Small heat shock proteins and α-crystallins: dynamic proteins with flexible functions

Abstract

The small heat shock proteins (sHSPs) and the related α-crystallins (αCs) are virtually ubiquitous proteins that are strongly induced by a variety of stresses, but that also function constitutively in multiple cell types in many organisms. Extensive research has demonstrated that a majority of sHSPs and αCs can act as ATP-independent molecular chaperones by binding denaturing proteins and thereby protecting cells from damage due to irreversible protein aggregation. As a result of their diverse evolutionary history, their connection to inherited human diseases, and their novel protein dynamics, sHSPs and αCs are of significant interest to many areas of biology and biochemistry. However, it is increasingly clear that no single model is sufficient to describe the structure, function or mechanism of action of sHSPs and αCs. In this review, we discuss recent data that provide insight into the variety of structures of these proteins, their dynamic behavior, how they recognize substrates, and their many possible cellular roles.

Figure 1

Amino acid sequence features and evolutionary relationships of sHSP/αCs. (a) Amino acid sequence alignment of sHSP/αCs for which structural data are available (Table 1). Residues comprising β-strands are in cyan background (note that the extent of B7 differs in different published structures). The ACD comprises β2 through β9 (red line). Note that the vertebrate proteins lack β6, but have an extended β7. The conserved I/L-X-I/L motif in the C-terminal extension is boxed. Phosphorylation sites in vertebrate proteins are highlighted in yellow. Sites of mutations in human HspB1 and HspB5 are indicated, and human HspB4 mutations are mapped on bovine HspB4 as follows: desmin or myofibrillar-related myopathy (olive), cataract (magenta), cardiomyopathy (red) and motor neuropathy or Charcot-Marie-Tooth disease (green) [8, 58]. The first and last residues in the PDB atomic structure file for each protein are underlined and highlighted in gray (Table 1). Abbreviations for organism names: ZFish, zebrafish (Danio rerio); T.a., Triticum aestivum (wheat); S.c., Saccharomyces cerevisiae; X.a., Xanthomonas axonopodis; M.t., Mycobacterium tuberculosis; M.j., Methanocaldococcus janaschii; S.t., Solfolobus tokodaii. Alignment was generated using ClustalW, and then the C-terminal region was adjusted to align the I/L-X-I/L motif for some sequences. (b) Phylogenetic relationships of sHSP/αCs for which structural data are available, along with the complete set of human proteins and conserved land plant paralogs. Only the ACD and C-termianl extension were used for the phylogenetic analysis. Plant lineage is in green and includes multiple orthologous sequences from Arabidopsis thaliana and Oryza sativa for those branches (gene families) identified in bold and named by subcellular localization. Thickness of the branch end is indicative of typical numbers of genes in that family. Single named genes in the plant lineage have no identified orthologs to date. The vertebrate lineage with all 10 human proteins is shown in red, selected invertebrate sequences are in orange, and microorganisms in black. Sequences were aligned with Promals3D, a multiple alignment program that uses protein structure to guide amino acid alignments [82]. Alignments were trimmed with Bioedit. The alignment was opened in the program MEGA4 [83] and neighbor-joining trees were constructed using the JTT matrix. Accession numbers: Plant Cytosolic I – Os17.6A,NP_001041951; At17.4I,AF410266_1; Os16.93I, Q943E9; At17.6AI, NP_176195; At17.6BI, NP_180511; At17.6CI,NP_175759; At18.1I,NP_200780; At17.8I, NP_172220;Os16.9I, NP_001041954; Os16.91I, NP_001041953;Os16.92I, NP_001041955; Os17.4AI, NP_001049660;Os17.4BI. NP_001049662; Os17.4DI, NP_001049657; Os17.4CI, NP_001049661;Triticum aestivum, 1GME_A; Plant ER - At22ER, NP_192763; OS22.3ER, NP_001052899. Plant Cytosol IV-At15.IV, NP_193918, Os18.8IV, NP_001059788. Plant Cytosol III – At17.4III, NP_175807; Os17.6BIII, NP_001048317. Plant Cytosol II – At17.6II, NP_196763; At17.7II, NP_196764; Os17.8II, NP_001042231; Os17.6CII< NP_001046302. Plant Peroxisome - At15.7PX, NP_198583, Os17.6PX, NP_001057300. Plant Cytosol IV – At15.4IV, NP_193918, Os18.8IIV, NP_001059788. Plant Mitochondria 2 – At26.5, NP_001117476, Os21.2, B7EZJ7. Plant Chloroplast – At25.6Cp, NP_1944497, Os26.oCp, NP_0001049541, Plant Mitochondria 1 – At23.6MI, NP_194250; At23.5MI, NP_199957; Os22MI, NP_0001048175, Os22.4MI, NP_001057162. Other plant proteins without identified orthologs – Os18, NP_001052899; Os18.2, NP_001045766; Os16.9C, Q0DY72; At14.2, NP_199571. Microbial proteins – Methanocaldococcus jannaschii, NP_247258; Synechocystis sp. 6803, NP_440316.1; Xanthomonas axonopodis, 3GT6_A; E. coli IbpA, ZP_04001711; E. coli IbpB, YP_543196; Saccharomyces cerevisiae Hsp26, CAA85016.1. Invertebrate proteins – Caenorhabditis elegans Hsp16, AAA28066.1; Drosophila melanogaster Hsp26, ABX80642.1; Hsp22, NP_001027114; Hsp23, NP_523999; Hsp27, NP_524000; Vertebrate proteins – Homo sapiens HspB1, AAA62175; HspB2, Q16082.2; HspB3, Q12988.2; HspB4, P02489.2; HspB5, P02511; HspB6, )14558; HspB7, Q9UBY9; HspB8, NP_055180.1HspB9, Q9BQS6.1;HspB10, Q14990.2; Danio rerio, 3N3E_A.

Figure 1

Amino acid sequence features and evolutionary relationships of sHSP/αCs. (a) Amino acid sequence alignment of sHSP/αCs for which structural data are available (Table 1). Residues comprising β-strands are in cyan background (note that the extent of B7 differs in different published structures). The ACD comprises β2 through β9 (red line). Note that the vertebrate proteins lack β6, but have an extended β7. The conserved I/L-X-I/L motif in the C-terminal extension is boxed. Phosphorylation sites in vertebrate proteins are highlighted in yellow. Sites of mutations in human HspB1 and HspB5 are indicated, and human HspB4 mutations are mapped on bovine HspB4 as follows: desmin or myofibrillar-related myopathy (olive), cataract (magenta), cardiomyopathy (red) and motor neuropathy or Charcot-Marie-Tooth disease (green) [8, 58]. The first and last residues in the PDB atomic structure file for each protein are underlined and highlighted in gray (Table 1). Abbreviations for organism names: ZFish, zebrafish (Danio rerio); T.a., Triticum aestivum (wheat); S.c., Saccharomyces cerevisiae; X.a., Xanthomonas axonopodis; M.t., Mycobacterium tuberculosis; M.j., Methanocaldococcus janaschii; S.t., Solfolobus tokodaii. Alignment was generated using ClustalW, and then the C-terminal region was adjusted to align the I/L-X-I/L motif for some sequences. (b) Phylogenetic relationships of sHSP/αCs for which structural data are available, along with the complete set of human proteins and conserved land plant paralogs. Only the ACD and C-termianl extension were used for the phylogenetic analysis. Plant lineage is in green and includes multiple orthologous sequences from Arabidopsis thaliana and Oryza sativa for those branches (gene families) identified in bold and named by subcellular localization. Thickness of the branch end is indicative of typical numbers of genes in that family. Single named genes in the plant lineage have no identified orthologs to date. The vertebrate lineage with all 10 human proteins is shown in red, selected invertebrate sequences are in orange, and microorganisms in black. Sequences were aligned with Promals3D, a multiple alignment program that uses protein structure to guide amino acid alignments [82]. Alignments were trimmed with Bioedit. The alignment was opened in the program MEGA4 [83] and neighbor-joining trees were constructed using the JTT matrix. Accession numbers: Plant Cytosolic I – Os17.6A,NP_001041951; At17.4I,AF410266_1; Os16.93I, Q943E9; At17.6AI, NP_176195; At17.6BI, NP_180511; At17.6CI,NP_175759; At18.1I,NP_200780; At17.8I, NP_172220;Os16.9I, NP_001041954; Os16.91I, NP_001041953;Os16.92I, NP_001041955; Os17.4AI, NP_001049660;Os17.4BI. NP_001049662; Os17.4DI, NP_001049657; Os17.4CI, NP_001049661;Triticum aestivum, 1GME_A; Plant ER - At22ER, NP_192763; OS22.3ER, NP_001052899. Plant Cytosol IV-At15.IV, NP_193918, Os18.8IV, NP_001059788. Plant Cytosol III – At17.4III, NP_175807; Os17.6BIII, NP_001048317. Plant Cytosol II – At17.6II, NP_196763; At17.7II, NP_196764; Os17.8II, NP_001042231; Os17.6CII< NP_001046302. Plant Peroxisome - At15.7PX, NP_198583, Os17.6PX, NP_001057300. Plant Cytosol IV – At15.4IV, NP_193918, Os18.8IIV, NP_001059788. Plant Mitochondria 2 – At26.5, NP_001117476, Os21.2, B7EZJ7. Plant Chloroplast – At25.6Cp, NP_1944497, Os26.oCp, NP_0001049541, Plant Mitochondria 1 – At23.6MI, NP_194250; At23.5MI, NP_199957; Os22MI, NP_0001048175, Os22.4MI, NP_001057162. Other plant proteins without identified orthologs – Os18, NP_001052899; Os18.2, NP_001045766; Os16.9C, Q0DY72; At14.2, NP_199571. Microbial proteins – Methanocaldococcus jannaschii, NP_247258; Synechocystis sp. 6803, NP_440316.1; Xanthomonas axonopodis, 3GT6_A; E. coli IbpA, ZP_04001711; E. coli IbpB, YP_543196; Saccharomyces cerevisiae Hsp26, CAA85016.1. Invertebrate proteins – Caenorhabditis elegans Hsp16, AAA28066.1; Drosophila melanogaster Hsp26, ABX80642.1; Hsp22, NP_001027114; Hsp23, NP_523999; Hsp27, NP_524000; Vertebrate proteins – Homo sapiens HspB1, AAA62175; HspB2, Q16082.2; HspB3, Q12988.2; HspB4, P02489.2; HspB5, P02511; HspB6, )14558; HspB7, Q9UBY9; HspB8, NP_055180.1HspB9, Q9BQS6.1;HspB10, Q14990.2; Danio rerio, 3N3E_A.

Figure 2

Different sHSP/αCs have different dimer structures but share a conserved contact for oligomer formation. (a) Dimer structure of wheat Hsp16.9 (1GME; left) and human HspB5 (2KLR; right). Individual monomers are colored red or blue. Regions outside the ACD in wheat Hsp16.9 are in gray, including the N-terminal arm of one monomer and the C-termini of both monomers; note that the HspB5 structure comprises only the ACD without β2. Differences in the dimer structures are discussed in the text. 2KLR is an NMR structure and represents only one of several structures available for related vertebrate proteins (Table 1). Although all the vertebrate structures show the same dimerization mode mediated by β7, details of the different vertebrate structures vary [28]. The most highly conserved residues (Fig. 1a) show different positions in the two dimer forms as highlighted for the conserved arginine in β7 (green stick) and the conserved G-X-L (green cartoon). (b) The conserved C-terminal I/V-X-I/V motif connects dimers in sHSP/αC oligomers by interacting with a hydrophobic groove formed by β4 and β8 at one edge of the ACD β-sandwich of another monomer. The model shown is derived from the wheat Hsp16.9 structure, but the same contact is observed in the Methanocaldococcus structure [17] and is proposed from recent NMR and X-ray data for vertebrate sHSP/αCs [21, 27, 30].

Figure 2

Different sHSP/αCs have different dimer structures but share a conserved contact for oligomer formation. (a) Dimer structure of wheat Hsp16.9 (1GME; left) and human HspB5 (2KLR; right). Individual monomers are colored red or blue. Regions outside the ACD in wheat Hsp16.9 are in gray, including the N-terminal arm of one monomer and the C-termini of both monomers; note that the HspB5 structure comprises only the ACD without β2. Differences in the dimer structures are discussed in the text. 2KLR is an NMR structure and represents only one of several structures available for related vertebrate proteins (Table 1). Although all the vertebrate structures show the same dimerization mode mediated by β7, details of the different vertebrate structures vary [28]. The most highly conserved residues (Fig. 1a) show different positions in the two dimer forms as highlighted for the conserved arginine in β7 (green stick) and the conserved G-X-L (green cartoon). (b) The conserved C-terminal I/V-X-I/V motif connects dimers in sHSP/αC oligomers by interacting with a hydrophobic groove formed by β4 and β8 at one edge of the ACD β-sandwich of another monomer. The model shown is derived from the wheat Hsp16.9 structure, but the same contact is observed in the Methanocaldococcus structure [17] and is proposed from recent NMR and X-ray data for vertebrate sHSP/αCs [21, 27, 30].

Figure 3

Diversity of sHSP oligomer architecture. sHSP oligomeric structures derived from X-ray and EM data (Table 1). Surface representations are shown in two orientations rotated by 90°. Ribbon diagrams of a single dimer structure are superimposed, although the wheat dimer (Hsp16.9) was used for the dimer structures of Mycobacterium tuberculosis and yeast. Bottom row shows geometric diagrams to illustrate the molecular symmetry of the molecules, with each solid line representing location of one dimer.

Figure 4

A model for the chaperone mechanism of sHSP/αC proteins. (i) sHSP/αC oligomers (here wheat Hsp16.9, 1GME) are dynamic structures that continually release and reassociate with their constituent subunits, either dimers (shown here) or monomers. (ii) Damaged or denaturing proteins (here malate dehydrogenase, 1MLD) expose hydrophobic surfaces prone to aggregation. (iii) The unfolding substrate binds to the sHSP to form large sHSP-substrate complexes with variable stoichiometries. Note that the form of the sHSP responsible for capturing denaturing substrate remains undefined and may be either oligomeric or a smaller species (such as the dimer). (iv) sHSP-substrate complexes can be acted on by Hsp70, co-chaperones and ATP, resulting in refolded active substrate. (v) In the absence of sufficient sHSP to capture denaturing substrate, protein aggregates are formed that are difficult for the cell to remove. (vi) sHSP-substrate complexes may also be acted on by cellular proteases, although this pathway is less well investigated than the refolding pathway. The specificities of different sHSP/αCs for different substrate proteins, as well as how different sHSP/αCs function together in one cell, remain to be defined. This model for chaperone activity may not explain all the different cellular roles of the sHSP/αCs.

Figure 5

In vitro assays for the study of the ATP-independent chaperone activity of sHSP/αCs. (a) Light scattering assay of suppression of aggregation. Relative light scattering measured at 360 nm can be used to monitor the aggregation of substrate proteins during denaturation induced by heat or by reduction. Scattering is suppressed with increasing amounts of chaperone. This is a rapid and simple assay. As shown here for reduction of insulin, maximal scattering (aggregation) is observed in the absence of the sHSP/αC. With increasing amounts of chaperone (here pea Hsp18.1), scattering is suppressed. The sHSP prevents insulin aggregation very efficiently at the molar ratio of 4 uM sHSP monomer to 45 uM insulin monomer, equivalent to a weight to weight ratio of 1:1. (b) Protection of solubility. A heat sensitive substrate (e.g. firefly luciferase, or Luc) is heated in the presence of a sHSP/αC (here pea Hsp18.1, or Ps18.1), then cooled and centrifuged. Soluble and pellet fractions are separated by SDS-PAGE and stained for protein. The protection efficiency can be estimated by varying the ratio of sHSP and substrate. In the absence of sHSP, all of the luciferase is in the pellet fraction, whereas in the presence of 3–6 µM sHSP monomers all the luciferase is found in the soluble fraction, representing protection of approximately an equal mass of substrate. (c) Visualizing sHSP-substrate complexes. Samples of sHSP plus substrate as shown in (b) can be separated by size exclusion chromatography to visualize sHSP-substrate interaction. Before heating, the substrate protein (Luc) and the sHSP elute separately at the expected position for their native molecular weight (dotted line). After heating (solid line), free substrate is no longer present, and the soluble sHSP-substrate complexes elute earlier than the free sHSP or substrate in the form of larger, heterogeneous mixtures. The complex size distribution changes with increasing ratio of sHSP to substrate, as shown by the differences in timing of elution of the complex peak in the different panels. Assays in (a–c) are non-equilibrium, end-point assays due to the essentially irreversible nature of complex formation. (d) A substrate refolding assay is used to examine how effective an sHsp/αC is in maintaining an enzymatic substrate in a conformation that can be refolded by ATP-dependent chaperones (such as Hsp70). Soluble complexes as in (b) are diluted into reticulocyte lysate (a rich source of ATP-dependent chaperones) plus ATP, or into a mixture of purified Hsp70/DnaK, co-chaperones and ATP, and then monitored for restoration of enzyme activity. Here, the endpoint of luciferase reactivation after one hour was plotted for different sHSP-luciferase ratios. Although luciferase is protected from aggregation at all ratios as in (b), it is obvious that different complexes facilitate a different extent of reactivation as in (c), which is consistent with differences in the size or organization of these complexes that may alter accessibility of substrate to the refolding machinery. (e) Equilibrium assay for binding of substrate to sHSP. Here, the binding of sHSP to thermodynamically destabilized variants of T4 lysozyme (T4L) is shown. Top: Characteristics of destabilized T4L mutants used as substrates. Although the T4L mutations do not change the structure of the native state (as demonstrated by superposition of mutant and wild-type T4L structures), they increase the free energy of unfolding as monitored by determining the midpoint value of guanidine-HCl concentration required to unfold the mutants. Thus, at equilibrium, a higher fraction of some T4L variants (eg. L99A) is in the unfolded state compared to other variants (e.g. L46A or the wild type). Using a bimane fluorescence label introduced at residue 151 of T4L, it is possible to monitor the binding of T4L to a HSP (here HspB5). Middle: The intensity and/or change in maximum intensity wavelength of fluorescence is used to monitor formation of a stable complex between T4L and sHSP. Bottom: Titration with increasing sHSP concentrations yields binding isotherms that can be fit to obtain apparent binding affinities and number of binding sites [7].