αB-crystallin is a sensor for assembly intermediates and for the subunit topology of desmin intermediate filaments
- 646 Downloads
Mutations in the small heat shock protein chaperone CRYAB (αB-crystallin/HSPB5) and the intermediate filament protein desmin, phenocopy each other causing cardiomyopathies. Whilst the binding sites for desmin on CRYAB have been determined, desmin epitopes responsible for CRYAB binding and also the parameters that determine CRYAB binding to desmin filaments are unknown. Using a combination of co-sedimentation centrifugation, viscometric assays and electron microscopy of negatively stained filaments to analyse the in vitro assembly of desmin filaments, we show that the binding of CRYAB to desmin is subject to its assembly status, to the subunit organization within filaments formed and to the integrity of the C-terminal tail domain of desmin. Our in vitro studies using a rapid assembly protocol, C-terminally truncated desmin and two disease-causing mutants (I451M and R454W) suggest that CRYAB is a sensor for the surface topology of the desmin filament. Our data also suggest that CRYAB performs an assembly chaperone role because the assembling filaments have different CRYAB-binding properties during the maturation process. We suggest that the capability of CRYAB to distinguish between filaments with different surface topologies due either to mutation (R454W) or assembly protocol is important to understanding the pathomechanism(s) of desmin-CRYAB myopathies.
KeywordsChaperone CRYAB Intermediate filaments Desmin Desminopathy Cardiomyopathy
Coomassie Brilliant Blue G-250
The α-crystallins are small heat shock proteins (sHSPs); (Carra et al. 2013; Kappe et al. 2003) that are part of the molecular chaperone family (Brandvold and Morimoto 2015; Kim et al. 2013; Treweek et al. 2014). Protein chaperones assist the folding of nascent protein chains, bind to and prevent the aggregation of misfolded and stress-denatured proteins (Brandvold and Morimoto 2015; Strauch and Haslbeck 2016), as well as assisting oligomer assembly and the assembly of protein complexes (Ellis 2013). Their role as assembly chaperones is often overshadowed by their folding pathway(s) function (Ellis 2013). For sHSPs and particularly for αB-crystallin (CRYAB, HSPB5), its role as an assembly chaperone is critical to its many functions in the cell that involve multiprotein complexes (Quinlan and Ellis 2013). Mutations in sHSPs cause many different diseases (Carra et al. 2013; Treweek et al. 2014) typified in their histopathology by protein aggregates of intermediate filaments (IFs) and associated proteins (Perng et al. 2016; Quinlan and Van Den Ijssel 1999). The α-crystallin complex (CRYAA and CRYAB) in the eye lens was first found to modulate the assembly of lenticular IFs (Nicholl and Quinlan 1994). Subsequently, mutations in both CRYAB (Vicart et al. 1998) and the muscle-specific IF protein desmin (Goldfarb et al. 1998) were found to phenocopy each other, causing cardiomyopathy. This first disease-causing mutation in CRYAB, R120G (Vicart et al. 1998), decreased the dissociation constant by half for desmin, causing the assembled desmin filaments to aggregate in both in vitro assembly assays and in transfected cells (Perng et al. 1999b; Perng et al. 2004). Cytosolic heterogeneous multiprotein aggregation and mitochondrial abnormalities characterize myopathies caused by human mutations in desmin or CRYAB (reviewed in (Capetanaki et al. 2015; Kley et al. 2016). Thus, the phenocopying by each highlights the importance of exploring in more detail the desmin-CRYAB interaction.
Desmin is part of a large multigene family of IF cytoskeletal proteins responsible for the mechanical and stress-coping resilience of cells (Guo et al. 2013) and no more so than in muscle (Palmisano et al. 2014) where the targeted deletion of desmin from the mouse genome causes cardiomyopathy, fibrosis and heart failure (Milner et al. 1996). The desmin network is recognized as a key mechanical element (Kiss et al. 2006; Li et al. 2013; Li et al. 1997; Palmisano et al. 2015; Wojtowicz et al. 2015) that can become progressively dysfunctional through myopathy-causing desmin mutations (Kreplak and Bar 2009).
Desminopathy is a rare neuromuscular disorder, belonging to the so-called myofibrillar myopathies, caused by inherited mutations in DES (Clemen et al. 2013; Palmio and Udd 2016). A major histopathological feature of desminopathy is the accumulation of insoluble desmin and partner proteins including CRYAB into aggregates (Maerkens et al. 2013). The optimal biomechanical properties for muscle sarcomeres depend heavily upon desmin (Conover et al. 2009; Diokmetzidou et al. 2016; Li et al. 2013; Palmisano et al. 2014) and CRYAB (Diokmetzidou et al. 2016; Wojtowicz et al. 2015). Thus far, it is only the desmin-binding sites on CRYAB that are known (Houck et al. 2011), but those domains in desmin that determine CRYAB binding are unknown. According to pin array studies, three surface exposed peptides within two beta strands of CRYAB and a C-terminal peptide had the strongest binding sites to desmin in a temperature-dependent manner (Ghosh et al. 2007).
The direct interaction of sHSPs and IFs has been proposed to have important cytoprotective roles during normal muscle physiology (Capetanaki et al. 2015; Wettstein et al. 2012). Deeper molecular insights are needed to understand how the assembly of desmin filaments and its network is impacted by CRYAB. There is precedence for desmin-binding partners to affect desmin assembly as recent studies show a delay of a mutant desmin assembly occurs when it binds to nebulin, a giant actin-binding skeletal muscle protein (Baker et al. 2013), or to nebulette, an actin-binding cardiac muscle protein (Hernandez et al. 2016). These studies help explain the molecular basis for the pathology found in desminopathy patients that carry the filament-forming mutant desmin E245D. Furthermore, it is known that other disease-associated desmin mutations alter filament morphology, whilst others dramatically halt filament assembly at distinct stages (Bar et al. 2007b; Bar et al. 2005; Bar et al. 2010). As a major chaperone in muscle (Kato et al. 1991), CRYAB has been suggested to modulate filament assembly and network formation (Perng et al. 1999b; Perng and Quinlan 2015; Perng et al. 2004). CRYAB is a major component in the desmin aggregates taken from muscle biopsies of patients with desminopathies (Maerkens et al. 2013). Thus far, there are no data to suggest that filament morphology and subunit topology affect CRYAB binding. Here, we have investigated the chaperone function of CRYAB in desmin assembly in vitro, using different buffer conditions to manipulate the assembly pathway and alter filament morphology (Herrmann et al. 1999; Herrmann et al. 1996) and find that this influences CRYAB to desmin filaments. We show that the non-α-helical C-terminal domain (“tail”) of desmin contributes to CRYAB binding, and we provide evidence that the mutations I451M and R454W, which map to this domain, alter CRYAB binding. Our data support a sensor role for CRYAB (McHaourab et al. 2009) with respect to desmin filaments and their surface topology.
Materials and methods
Cloning, mutagenesis and recombinant protein expression
Point mutations were introduced into the full-length clone of human desmin cDNA by site-directed mutagenesis (Quickchange, Stratagene) using the listed primers (supplemental materials). For protein expression, cDNAs of WT desmin, point mutations and deletion constructs were subcloned into the T5 promoter-driven prokaryotic expression vector pDS5, modified to contain a proper Shine-Dalgarno sequence followed by a NcoI restriction site CCATGG, the ATG sequence of which was used as translation start (Herrmann et al. 1993). The identity of the clones was verified by sequencing. Plasmids were expressed in Escherichia coli strains TG1 (Amersham Biosciences) or JM109 (Novagen) for protein purification from inclusion bodies followed by cation and anion exchange chromatography, as described (Herrmann et al. 1999). Protein size and purity was verified by SDS PAGE and colloidal Coomassie Brilliant Blue G-250 (CBB; data not shown) staining of the gel. Human recombinant CRYAB was purified as described (Perng et al. 1999a).
Protein dialysis, desmin assembly, co-sedimentation and viscometry assays
Desmin filament morphology was changed using three different assembly buffers taken from previously published protocols (Herrmann et al. 1996; Mucke et al. 2004; Perng et al. 1999a). Very low ionic strength buffers (preassembly buffers, see below) maintained desmin in a soluble form, such as tetramers or their pre-unit-length filament oligomers (Mucke et al. 2004), prior to initiating assembly by the addition of an equal volume of “assembly buffer” (“Phosphate-buffer”—2 mM sodium phosphate 100 mM NaCl pH 7.5 (Mucke et al. 2004); “Tris-buffer”—100 mM NaCl, 40 mM Tris-HCl, pH 7.0 (Herrmann et al. 1996); “Imidazole-buffer”—200 mM imidazole-HCl, 2 mM DTT (Perng et al. 1999a)) at the indicated temperatures and times. Corresponding low ionic strength preassembly buffers were “low-phosphate”—2 mM Na2PO4, 1 mM DTT pH 7.5; “low-Tris” —5 mM Tris-HCl, 1 mM EDTA, 0.1 mM EGTA, 1 mM DTT, pH 8.4; “low-imidazole”—10 mM Tris-HCl, pH 8.0, 1 mM DTT, 0.2 mM PMSF. The final pH after mixing of equal volumes of preassembly and assembly buffers for the three different buffer compositions was pH 7.5. To monitor the binding of CRYAB to desmin filaments, equimolar ratios of CRYAB and desmin (0.3 μg/μl desmin and 0.12 μg/μl CRYAB) were mixed together in the corresponding low salt buffer prior to adding the assembly buffer. CRYAB was directly dialyzed into the corresponding low salt buffer used to investigate desmin binding.
Buffer pH influences the staging of IF assembly in a tetramer-dependent fashion (Mucke et al. 2004; Wickert et al. 2005). For desmin, reducing the pH from 8.4 to 7.5 arrests the assembly process at an intermediate stage between tetramer complex and unit-length filament (ULF; (Wickert et al. 2005)). Samples were dialyzed overnight into “low-Tris” preassembly buffers at two different pHs: 5 mM Tris-HCl, 1 mM EDTA, 0.1 mM EGTA, 1 mM DTT, pH 8.4 or pH 7.4. When mixed with an equal volume of assembly-inducing “Tris-buffer”, the final pHs of the assembly mixture were 7.5 and 7.1, respectively. To optimise CRYAB binding to desmin still further, the following preassembly buffers, all adjusted to pH 7.4, were tested: “5.0 mM low Tris” (5.0 mM Tris-HCl, 1 mM EDTA, 0.1 mM EGTA, 1 mM DTT), “2.5 mM low Tris” (2.5 mM Tris-HCl, 0.5 mM EDTA, 0.05 mM EGTA, 1 mM DTT) and “1.0 mM low Tris” (1.0 mM Tris-HCl, 0.2 mM EDTA, 0.02 mM EGTA, 1 mM DTT). Assembly was performed at 37 °C for 1 h by adding an equal volume of the assembly “Tris-buffer”, leading to a final pH of 7.1 and final Tris-HCl concentrations of 22.50, 21.25 and 20.50 mM, respectively.
Samples were centrifuged at 30,000 rpm for 30 min at 20 °C in a Beckman centrifuge with swing-out TLS 55 rotor (Beckman Coulter). Viscometry measurements (Ostwald viscometer; Cannon-Nanning) were made as described (Schopferer et al. 2009). For diameter measurements, protein samples were dialyzed and assembled the same day. Data are reported as mean ± SD. A two-sided Student’s t test is used to determine the significant differences (p value <0.05).
Band densitometry quantification
Individual protein bands from colloidal CBB-stained and subsequently scanned gels were quantified with ImageJ (http://rsb.info.nih.gov/ij), and the values obtained for each fraction were graphed using Prism 5.0. Experiments were performed in triplicate to make statistical evaluations.
Proteins were fixed in 0.2% (v/v) glutaraldehyde and mounted on glow-discharged formvar carbon-coated 200-mesh copper grids (SPI Supplies, USA) or on self-coated grids. Samples were negatively stained with 0.2% (w/v) uranyl acetate for 20 s, and images were acquired on a CCD camera in a Zeiss 900 transmission electron microscope (Carl Zeiss, Germany) at various magnifications between 21,000× and 112,000× at 80 kV. Images were processed for presentation using Adobe CS6. For the measurement of filament diameters, EM images were processed using ImageJ 1.32j. At least 100 measurements were carried out per sample, the mean and SEM calculated, significance determined using Student’s t test. All reported p values were two-sided and considered to be statistically significant at p < 0.05.
The morphology of the desmin filaments influences CRYAB binding
Optimisation of the binding of CRYAB to desmin.
Our data consistently show that more CRYAB (~16%) was recovered when a final pH of 7.1 was realized rather than pH 7.5, (~7%; Fig. 3a) in “Tris-buffer”. EM analyses indicated that the desmin filament diameter was similar when the assembly reaction was completed in pH 7.1 and 7.5 by the addition of an equal volume of assembly “Tris-buffer” (data not shown). Additionally, our data show that CRYAB binding to desmin could be enhanced further by reducing the Tris concentration of the “low-Tris” buffer from 5 to 1 mM (Fig. 3b), whilst maintaining the pH of the assembly conditions at pH 7.1. Our results demonstrate that varying the composition of the preassembly buffer by reducing the Tris concentration increased filament diameters from 11.3 (22.5 mM Tris-HCl), 11.7 (21.25 mM Tris-HCl) and 12.4 nm (20.50 mM Tris-HCl). In summary, our systematic analysis revealed that a “1 mM low-Tris” buffer at pH 7.1 when added to an equal volume of pH 7.5 “Tris-buffer” produced desmin filaments that most efficiently bound CRYAB. These analyses established the experimental platform on which to optimize the binding of CRYAB to desmin filaments assembled in vitro.
Previous studies had noted that for desmin, in “5 mM low-Tris” buffer pH 7.4, the addition of chelators shifted the s value from 5.5 to 13 S (Wickert et al. 2005); thus, the buffer composition did alter the starting point in the assembly pathway by driving tetramer association (Lopez et al. 2016).
Chaperone role for CRYAB in desmin assembly.
Next, we compared the viscosity profiles of the desmin filament networks formed during the coassembly and sequential assembly regimes as described in Fig. 4a. When CRYAB was coassembled with desmin, then a 50% drop in the viscosity was observed. If CRYAB was added 45 min after the initiation of filament assembly, then no decrease in viscosity was seen (Fig. 4d). Visualization of the desmin and CRYAB by EM confirmed that by coassembly, CRYAB coated the desmin filaments (Fig. 4e (i)). In contrast, when CRYAB was added after 45 min, this association was not readily seen (Fig. 4e (ii)). These data are in close agreement with our co-sedimentation and viscosity data (Fig. 4b–d) and furthermore suggest that CRYAB is able to sense differences in the assembling desmin filaments.
The tail domain of desmin harbours critical CRYAB-binding sites
The C-terminal domain of vimentin and desmin is recognized for its role in modulating filament width control and filament-filament interactions (Herrmann et al. 1996, Bar et al. 2007a, Bar et al. 2010). Cardiomyopathy disease-causing desmin tail mutations do not coassemble correctly with WT desmin in vitro and in transient transfection of myoblasts and have a notably decreased network viscosity, suggesting that the tail domain of desmin is important for biomechanical force transmission between myocytes (Bar et al. 2007a; Bar et al. 2010). Passive rheology measurements indicated that CRYAB likely modulates the desmin network viscosity (Elliott et al. 2013). These findings are consistent with our bulk network viscosity measurements (Fig. 4d. Therefore, based on these findings, we sought to investigate in further detail how the C-terminal domain of desmin might modulate CRYAB binding.
CRYAB has opposing binding defects for desminopathy-linked tail mutants
CRYAB plays a central role in stress tolerance in muscle cells. As such, more CRYAB binds to desmin at elevated temperatures (Elliott et al. 2013; Perng et al. 2004), and temperature elevation increases sHSP subunit exchange and potential chaperone activity (Aquilina et al. 2013; Datskevich and Gusev 2014). In this study, we compared the ability of CRYAB to function as a desmin sensor in physiological temperatures and raised temperatures (Fig. 6b, c). Both I451M and R454W desmin were not as efficient as wild-type desmin in their assembly as measured by the pelletable protein (Fig. 4b), but these differences were only significant for I451M for which 74.6, 82.3 and 86.3% protein was pelleted at 22, 37 and 44 °C, respectively (cf 83.8, 93.3 and 94.3% at 22, 37 and 44 °C, respectively, for wild-type desmin). Our EM data were consistent with the biochemical analyses as the percentages of CRYAB recovered in the pellets after in vitro assembly increased in a temperature-dependent manner (22, 37 and 44 °C, Fig. 6c). Our data show that significantly more CRYAB bound to mutant desmin R454W filaments as compared to WT desmin at 37 and 44 °C. Our data also revealed significantly lower amounts of CRYAB bound to mutant desmin I451M at all three temperatures investigated (Fig. 6c), more than could be explained by the small reduction in its sedimentation properties (halving in CRYAB levels for a 10–15% decrease in sedimentable desmin). These data indicate that CRYAB detects differences in filament surface topologies of disease-causing desmin mutants. We conclude that these findings suggest a novel sensing function for CRYAB with respect to detecting and surveying desmin filament subunit topology during its assembly.
The R454W mutation in desmin drives the association of CRYAB
In this study, we report that desmin filament morphology influences CRYAB binding suggesting that it can act as a sensor detecting changes in the surface topology of IFs. It is well established that the assembly protocol, i.e. rapid versus dialysis (Herrmann et al. 1999) as well as the monovalent or divalent cation concentration (Brennich et al. 2014) all influence the diameter of the desmin filaments produced (Stromer et al. 1987). In the case of vimentin, a structurally related IF protein to desmin, the MPL values of filaments formed by dialysis exhibited a homogeneous filament population rather than the polymorphism seen by rapid assembly protocols (Herrmann et al. 1996). We suggest this indicates desmin subunit topology is an important aspect of filament structure and therefore its function. CRYAB binding is dramatically influenced by desmin filament morphology (Figs. 1 and 2). Here, we deliberately used two in vitro assembly protocols to produce assembled filaments of different diameter, in order to manipulate the mass-per-length (MPL; (Herrmann et al. 1999)), and the data we have presented here suggest that the resulting topological differences are detected by differences in the extent of CRYAB binding and forming a stable complex with the IFs.
The prevalent model for IF assembly is largely based on length measurements of vimentin IFs at different time points (reviewed in (Herrmann and Aebi 2016), since vimentin and desmin belong to the same assembly group and can co-polymerize (Quinlan and Franke 1982) and therefore share a common assembly pathway (Herrmann and Aebi 2016). Filament assembly begins as tetramers associate laterally to form ULFs (phase 1), next ULFs anneal longitudinally into filaments (phase 2) and then filaments undergo a radial compaction event (phase 3; (Herrmann and Aebi 2016)). This process occurs very fast, since during the first 10 s of assembly, ~85% of all subunits are assembled into ULFs, and subsequently ULFs are consumed within the 10 min of assembly (Lopez et al. 2016; Sokolova et al. 2006). In the assembly regimes utilized here, we show that CRYAB rapidly binds desmin subunits within the first 3 min of assembly (Fig. 4b, c). This finding is consistent with the ideas that chaperones can modulate IF assembly (Landsbury et al. 2010; Nicholl and Quinlan 1994; Perng and Quinlan 2015). Accordingly, it has been demonstrated that HSP27 (HSPB1) regulates the early assembly and bundling dynamics of keratin K8/K18 IF networks (Kayser et al. 2013), with the well-documented role of protein chaperones in assisting protein assemblies (Ellis 2013). We also demonstrate that the C-terminal domain of desmin is important for CRYAB binding, consistent with previous reports that implicate this domain of desmin in high-order filament organization (Bar et al. 2007a; Bar et al. 2010).
Once the compaction stage has started during the in vitro assembly of desmin, then little CRYAB binding will occur (Fig. 4b, c). If, however, desmin filaments are prepared by serial dialysis, CRYAB can still bind (Nicholl and Quinlan 1994) arguing that CRYAB-binding sites remain accessible in some assembly regimes, but not in others as we evidence here (Figs. 1 and 2). The inclusion of CRYAB at the start of the assembly process ensures that the desmin filaments retain their CRYAB binding, reducing their solution viscosity (Fig. 4d; (Elliott et al. 2013)). Such effects are lost if CRYAB is added once filament compaction has occurred (Fig. 4b, c, e (ii)). We suggest that even though morphologically there appears to be little difference, CRYAB can distinguish the stage of desmin assembly and filament end products. Our data are consistent with the kinetic trapping mechanisms reported for reconstituted actin and keratin cytoskeletal networks (Kayser et al. 2013). Such regulation is a recognized feature of other self-assembly systems (Yan et al. 2016) from viral capsids (Cardone et al. 2014) to individual proteins (Barducci et al. 2013) and suits the emerging role for sHSPs as sensors (McHaourab et al. 2009) capable of using the energy landscape of protein assembly and particularly IFs.
The C-terminal domain of desmin is required for CRYAB binding
Our optimization strategy investigated the effects of pH and ionic strength using Tris buffers at different molarities (Fig. 3b). We used these to show that the C-terminal tail domain of desmin was important for CRYAB binding. CRYAB also detects changes to filament topology as result of the incorporation of desmin with mutations in the C-terminal tail domain (Figs. 5, 6 and 7). This domain regulates desmin filament bundling (Kaufmann et al. 1985) as corroborated by the unique viscometric profiles of the truncated proteins (Bar et al. 2007a; Bar et al. 2010; Herrmann et al. 1996). The complete removal of the C-terminal domain does not alter the in vitro assembly of desmin (Rogers et al. 1995), but it could influence the subunit conformation in the filaments as well as affecting the availability of docking sites for IF-associated proteins (Herrmann and Aebi 2016; Rogers et al. 1995). In transfected cells, the C-terminal tail of desmin affects the higher order organization of filaments (Bar et al. 2007a; Herrmann et al. 2003). Disease-associated tail mutations in desmin change the CRYAB-binding properties (Figs. 6 and 7) and also help explain the previously described altered filament morphology (R454W) and aberrations in filament bundling (I451R) desmin structures (Bar et al. 2010). The dominant nature of the R454W mutation in CRYAB binding to assembled filaments was particularly striking and resonates with the histopathological feature of desminopathies. Together, our data suggest that the surface topology of desmin filaments strongly influences the binding of CRYAB, expanding the role of this chaperone from merely modulating filament assembly to also sensing changes in the filament surface topology.
Role of CRYAB in desminopathy
A hallmark of biopsied muscles in desminopathy is the presence of granulofilamentous aggregates (Clemen et al. 2013; van Spaendonck-Zwarts et al. 2011). Typically, the accumulation of IFs, chaperones and sarcomere proteins in insoluble amorphous protein aggregates disrupts the alignment of serially connected Z-discs and the extra-sarcomeric linkages of muscle myofibers to the intercalated discs (Maerkens et al. 2013). Desminopathy caused by R454W desmin mutation is associated with disintegrated Z-discs and severe misalignment of myofibrils (Bar et al. 2007a; Mavroidis et al. 2008). One major role of CRYAB is to prevent the incorrect self-association of desmin filaments that may produce aggregates in cells (Houck et al. 2011), given that in vitro, the CRYAB alleviates desmin aggregation (Elliott et al. 2013).
Overexpression studies of mutant desmin I451M and R454W in C2C12 myoblasts show that both mutants can be readily incorporated into the endogenous IF networks; however, in humans, they cause severe cardiomyopathy and skeletal muscle disease (Bar et al. 2007b). Structurally, it is known that desmin R454W produced unravelled open filamentous structures, whilst mixtures of mutant desmin I451M and WT desmin yielded abnormal bundles (Bar et al. 2007a). These previous observations are consistent with our findings that both of these mutant desmin exhibited abnormal filament topology, and this is reflected in the altered CRYAB binding reported here (Figs. 6 and 7). We interpret this as evidence for the ability of CRYAB to detect changes to the surface topology of the in vitro assembled desmin filaments as a result of the incorporation of these mutant desmins.
The assembly chaperone and sensor properties of CRYAB toward desmin
Although the precise structural details for filament surface topology sensor function for CRYAB or nebulin are as yet undetermined, our data support a model in which the degree of CRYAB interaction with desmin varies in response to the filament surface dissimilarities (Fig. 8). In other words, CRYAB is a sensor for intermediate filament topology. At the same time, the filament surface topology can be a consequence of the involvement of CRYAB in the assembly pathway, linking the assembly chaperone and sensor aspects of CRYAB with respect to IFs. In cells and tissues, the formation of mechanically active IF cytoskeletal networks is highly regulated (Capetanaki et al. 2015). Thus, alterations in the binding of CRYAB to mutant desmin lead to either building highly dense or loosely compacted IF networks that will also impact CRYAB function in IF assembly and in muscle itself (Diokmetzidou et al. 2016).
IFs are dynamic structures both in vitro and in vivo (Herrmann and Aebi 2016). In support of our model, it was reported that muscle contraction drives the association of CRYAB to desmin-rich compartments in myocytes (Frankenberg et al. 2014), and that the concomitant pH reduction that accompanies contraction favours CRYAB association to desmin filaments (Elliott et al. 2013). In cells, IF subunits exchange along the length of the filament, modulated by kinases, phosphatases, ATP levels and other post-translational modifiers (Chang et al. 2009; Colakoglu and Brown 2009; Vikstrom et al. 1991). In vitro experiments, as seen ex vivo experiments filaments can fragment and anneal end-to-end (Noding et al. 2014; Winheim et al. 2011). Therefore, a changing surface filament topology is a novel feature of IF structure and function and an aspect of their physiology that clearly involves CRYAB and other sHSPs (Landsbury et al. 2010; Perng and Quinlan 2015) that remains to be further examined in the context of beating myocytes.
We thank Helga Kleiner and Tatjana Wedig for excellent technical assistance. We also thank Norbert Mücke and Andrew Landsbury for helpful discussions during the preparation of this manuscript. This work was supported by grants from the German Research Foundation (DFG; HE 1853/9-2 to Harald Herrmann) and the American Heart Association (09SDG2110057 to Gloria M. Conover), BBSRC (Studentship to Jayne Elliot; BB/D52617X/1), Royal Society (IE140736) and Leverhulme Trust (RPG-2012-554) to RAQ.
Compliance with ethical standards
Conflict of interest
The authors declare that they have no conflict of interest.
- Elliott JL, Der Perng M, Prescott AR, Jansen KA, Koenderink GH, Quinlan RA (2013) The specificity of the interaction between alphaB-crystallin and desmin filaments and its impact on filament aggregation and cell viability. Philos Trans R Soc Lond Ser B Biol Sci 368:20120375. doi: 10.1098/rstb.2012.0375 CrossRefGoogle Scholar
- Frankenberg NT, Lamb GD, Overgaard K, Murphy RM, Vissing K (2014) Small heat shock proteins translocate to the cytoskeleton in human skeletal muscle following eccentric exercise independently of phosphorylation. J Appl Physiol (1985) 116:1463–1472. doi: 10.1152/japplphysiol.01026.2013 CrossRefGoogle Scholar
- Hernandez DA, Bennett CM, Dunina-Barkovskaya L, Wedig T, Capetanaki Y, Herrmann H, Conover GM (2016) Nebulette is a powerful cytolinker organizing desmin and actin in mouse hearts. Mol Biol Cell 27(24):3869-3882. doi: 10.1091/mbc.E16-04-0237
- Herrmann H, Aebi U (2016) Intermediate filaments: structure and assembly. Cold Spring Harb Perspect Biol 8(11). doi: 10.1101/cshperspect.a018242
- Landsbury A, Perng MD, Pohl E, Quinlan RA (2010) Functional symbiosis between the intermediate filament cytoskeleton and small heat shock proteins. In: Simon S, Arrigo AP (eds) Small stress proteins and human diseases. Nova Science Publishers Inc., Hauppauge, pp. 55–87Google Scholar
- Mavroidis M, Panagopoulou P, Kostavasili I, Weisleder N, Capetanaki Y (2008) A missense mutation in desmin tail domain linked to human dilated cardiomyopathy promotes cleavage of the head domain and abolishes its Z-disc localization. FASEB J 22:3318–3327. doi: 10.1096/fj.07-088724 CrossRefPubMedPubMedCentralGoogle Scholar
- Perng MD, Muchowski PJ, van den IJssel P, GJS W, Clark JI, Quinlan RA (1999b) The cardiomyopathy and lens cataract mutation in aB-crystallin compromises secondary, tertiary and quaternary protein structure and reduces in vitro chaperone activity. J Biol Chem 274:33235-33243CrossRefPubMedGoogle Scholar
- Perng MD, Quinlan RA (2015) The dynamic duo of small heat proteins and IFs maintain cell homeostasis, resist cellular stress and enable evolution in cells and tissues. In: Tanguay RM, Hightower LE (eds) The Big Book of Small Heat Shock Proteins, vol 8. Heat Shock Proteins. Springer International Publishing AG pp 401–434. doi: 10.1007/978-3-319-16077-1
- Quinlan R, Van Den Ijssel P (1999) Fatal attraction: when chaperone turns harlot. Nat Med 5(1):25–26Google Scholar
- Treweek TM, Meehan S, Ecroyd H, Carver JA (2015) Small heat-shock proteins: important players in regulating cellular proteostasis. Cell Mol Life Sci 2(3):429–451 doi: 10.1007/s00018-014-1754-5
- Wickert U, Mucke N, Wedig T, Muller SA, Aebi U, Herrmann H (2005) Characterization of the in vitro co-assembly process of the intermediate filament proteins vimentin and desmin: mixed polymers at all stages of assembly. Eur J Cell Biol 84:379–391. doi: 10.1016/j.ejcb.2005.01.004 CrossRefPubMedGoogle Scholar
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.