Active-State Structures of a Small Heat-Shock Protein Revealed a Molecular Switch for Chaperone Function

Article

Active-State Structures of a Small Heat-Shock Protein Revealed a Molecular Switch for Chaperone Function Highlights d

Small heat-shock protein structures with fully visible N-terminal domains

d

High-resolution structures presenting working state of the sHsp

d

A molecular switch model is proposed for chaperone function of this sHsp

Authors Liang Liu, Ji-Yun Chen, Bo Yang, Fang-Hua Wang, Yong-Hua Wang, Cai-Hong Yun

Correspondence [email protected] (C.-H.Y.), [email protected] (Y.-H.W.)

In Brief Liu et al. determined the crystal structures of wild-type and mutant SsHsp14.1 with fully visible N-terminal domains. The structural and functional data suggest a new model in which a molecular switch located in the N-terminal domain facilitates conformational changes for client protein binding.

Accession Numbers 4YL9 4YLB 4YLC

Liu et al., 2015, Structure 23, 1–10 November 3, 2015 ª2015 Elsevier Ltd All rights reserved http://dx.doi.org/10.1016/j.str.2015.08.015

Please cite this article in press as: Liu et al., Active-State Structures of a Small Heat-Shock Protein Revealed a Molecular Switch for Chaperone Function, Structure (2015), http://dx.doi.org/10.1016/j.str.2015.08.015

Structure

Article Active-State Structures of a Small Heat-Shock Protein Revealed a Molecular Switch for Chaperone Function Liang Liu,1,2,4 Ji-Yun Chen,1,4 Bo Yang,2 Fang-Hua Wang,3 Yong-Hua Wang,3,* and Cai-Hong Yun1,* 1Institute of Systems Biomedicine, Department of Biophysics, Beijing Key Laboratory of Tumor Systems Biology and Center for Molecular and Translational Medicine, School of Basic Medical Sciences, Peking University Health Science Center, Beijing 100191, P.R. China 2School of Bioscience and Bioengineering, South China University of Technology, Guangzhou 510006, P.R. China 3College of Light Industry and Food Sciences, South China University of Technology, Guangzhou 510641, P.R. China 4Co-first author *Correspondence: [email protected] (C.-H.Y.), [email protected] (Y.-H.W.) http://dx.doi.org/10.1016/j.str.2015.08.015

SUMMARY

Small heat-shock proteins (sHsps) maintain cellular homeostasis by binding to denatured client proteins to prevent aggregation. Numerous studies indicate that the N-terminal domain (NTD) of sHsps is responsible for binding to client proteins, but the binding mechanism and chaperone activity regulation remain elusive. Here, we report the crystal structures of the wild-type and mutants of an sHsp from Sulfolobus solfataricus representing the inactive and active state of this protein, respectively. All three structures reveal well-defined NTD, but their conformations are remarkably different. The mutant NTDs show disrupted helices presenting a reformed hydrophobic surface compatible with recognizing client proteins. Our functional data show that mutating key hydrophobic residues in this region drastically altered the chaperone activity of this sHsp. These data suggest a new model in which a molecular switch located in NTD facilitates conformational changes for client protein binding.

INTRODUCTION Living cells are subject to constant risk for protein misfolding and aggregation, and this risk becomes serious under stressed conditions (Bukau et al., 2006; Tyedmers et al., 2010). The maintenance of protein homeostasis is thus a fundamental biological process in cellular organisms (Hilton et al., 2013; Walter and Ron, 2011). Small heat-shock proteins (sHsps) are a superfamily of molecular chaperones existing in almost all living creatures and multiple biological tissues (Basha et al., 2012). Under stressed conditions, sHsps play essential roles through binding to denatured client proteins, preventing them from aggregating and precipitating (Haslbeck et al., 2005; Tyedmers et al., 2010). Loss of sHsp function sensitizes organism to heat stress (Zhong et al., 2013). In humans, malfunction of sHsps results in diseases such as cataract, certain myopathies, and neuropathies (Basha

et al., 2012). Although proteins of this superfamily are diverse in size (ranging from 12 to 43 kDa) and sequence, they share several common features. First, their chaperone activities are independent of ATP (Basha et al., 2012). Second, sHsps show diversity of oligomer architecture (Basha et al., 2012). Third, most sHsps remain inactive or partially active under physiological condition, and their chaperone activity can be induced by stressed conditions such as elevated temperature or the presence of denatured client proteins (Das and Surewicz, 1995; Haslbeck et al., 1999; Hilton et al., 2013). Most sHsps share a common domain structure consisting of the N-terminal domain (NTD), the a-crystalline domain (ACD), and the C-terminal extension (CTE) (Hochberg et al., 2014). The ACD contains two face-to-face placed b sheets formed by 7–8 b strands, and is the most structurally conserved part of sHsps. It forms the main body of the sHsp monomer and facilitates the formation of sHsp dimers. sHsp dimers are relatively stable in solution and act as the building blocks to form symmetric oligomers, usually containing 4–18 dimers or polydisperse oligomers (Basha et al., 2012; Kim et al., 1998). Oligomeric assembling can be regulated in stressed conditions to accommodate heterogeneous non-native client proteins (Braun et al., 2011; Hilton et al., 2013; McHaourab et al., 2012; Peschek et al., 2013; Shashidharamurthy et al., 2005; Stengel et al., 2010). CTEs of most sHsps bear a highly conserved three-residue Ile-X-Ile motif located at the C-terminal end, which plays a key role in oligomerization through hydrophobic binding to a hydrophobic groove located on a subunit of a neighboring dimer (Delbecq et al., 2012). The NTDs of sHsps are highly diverse in length and sequence between subfamilies of sHsps (Eifert et al., 2005; Sun and MacRae, 2005). According to previous structural and functional studies, the NTD is of high flexibility and has been proposed to be responsible for interacting with non-native client proteins (Jaya et al., 2009; Jehle et al., 2011; Koteiche et al., 2005; Stengel et al., 2010; White et al., 2006). However, it remains unclear how the NTD adapts to different kinds of denatured client proteins that are diverse in size, shape, surface hydrophobicity, and charge (Hilton et al., 2013). Since denatured proteins are heterogeneous without a defined three-dimensional structure, it has been very difficult (if not impossible) to study the structure of the complex formed by sHsps and non-native clients through

Structure 23, 1–10, November 3, 2015 ª2015 Elsevier Ltd All rights reserved 1

Please cite this article in press as: Liu et al., Active-State Structures of a Small Heat-Shock Protein Revealed a Molecular Switch for Chaperone Function, Structure (2015), http://dx.doi.org/10.1016/j.str.2015.08.015

Table 1. Crystallographic Data Collection and Refinement Statistics WT

Del-C4

A102D

4YL9

4YLC

4YLB

P21

I4

H32

65.7, 39.1, 119.7

189.1, 189.1, 88.5

132.8, 132.8, 213.1

a, b, g ( )

90.0, 91.6, 90.0

90.0, 90.0, 90.0

90.0, 90.0, 120.0

Resolution (A˚)

50–2.35 [2.43–2.35]

50–3.10 [3.34–3.10]

50–2.50 [2.63–2.50]

Rpimb

0.058 [0.382]

0.038 [0.330]

0.031 [0.395]

I/s

13.6 [2.1]

19.6 [2.5]

21.6 [2.5]

PDB ID Data Collectiona Space group Cell dimensions a, b, c (A˚)

Completeness (%)

99.3 [96.4]

99.7 [98.9]

96.6 [98.2]

Redundancy

3.6 [3.4]

3.7 [3.7]

5.7 [5.7]

Resolution (A˚)

39.88–2.35

44.57–3.10

42.59–2.50

No. of reflections

25,448

28,435

24,370

Rwork/Rfree

0.232/0.252

0.240/0.272

0.196/0.241

Protein

3,822

6,808

3,786

Solvent

218

93

203

Protein

57.4

98.8

66.7

Solvent

68.1

87.4

66.5

Bond lengths (A˚)

0.009

0.004

0.010

Bond angles ( )

0.931

0.854

1.179

Refinement

No. of atoms

B factors

RMSD

Ramachandran Plot (%) Favored region

97.7

98.5

97.1

Allowed region

2.3

1.5

2.9

Disallowed region

0.0

0.0

0.0

structure to 2.35-A˚ resolution with good statistics (R = 23.2%, Rfree = 25.2%; full statistics in Table 1). The electron density maps of this structure were of sufficient quality to enable tracing of nearly whole peptide chains including the NTD (residues 1–24) (Figure S1A), which has been partially or fully invisible in most of the previously reported sHsp crystal structures (Hanazono et al., 2013; Kim et al., 1998; van Montfort et al., 2001). Four SsHsp14.1 molecules of essentially the same conformation in the asymmetric unit form two homodimers (Figure 1A), in agreement with other sHsp structures (Kim et al., 1998; van Montfort et al., 2001). In this WT structure, the NTD adopts a long and straight a-helix conformation. The NTD helix appears to be amphipathic, with an array of hydrophobic residues facing away from the protein core to form a hydrophobic patch (Figures 2A and 2D). The N-terminal helices of the two molecules within a dimer stretch out from ACD toward opposite directions with only little interaction at the ‘‘roots’’ of them where they join the a-crystalline domain. Instead, the hydrophobic patch of the N-terminal helices interacts with its counterparts on the N-terminal helices from the other dimer through very limited contact (Figure 1A). The CTEs of the two molecules adopt a slightly different conformation, but both stretch out from the ACD dimer to interact with two neighbor dimers through hydrophobic interactions between the Ile-X-Ile motif and the b4-b8 hydrophobic groove, thus promoting oligomerization (Figure S2A). Functional mutagenesis has shown that NTD is responsible for client protein recognition by the sHsps (Braun et al., 2011; Jaya et al., 2009). However, it is difficult to understand how the rigid straight NTD helices can fit to and interact with a variety of non-native client proteins without defined three-dimensional structure. Therefore, this SsHsp14.1 WT structure with long straight NTD helices might represent a low-activity state of this protein. This is perhaps consistent with the observation that most sHsps remain inactive or partially active in physiological condition unless activated by stress conditions such as elevated temperature or the presence of denatured client proteins (Das and Surewicz, 1995; Hilton et al., 2013).

a

Numbers in brackets indicate values for the outermost-resolution shell. b For definition of Rpim see Evans (2006).

the usual structural biological approaches. Thus far, there has not been a structure of sHsp that provides a clear, high-resolution view of the NTD in the active state and in complex with client proteins. In this study, we solved crystal structures of an sHsp from Sulfolobus solfataricus P2 (SsHsp14.1) representing the conformation of this protein in the active and client binding state. We combined in vitro and in vivo functional studies with structural investigation to propose the molecular mechanism whereby the NTD of SsHsp14.1 changes its conformation and adapts to non-native clients. RESULTS Full-Length SsHsp14.1 Reveals a Straight Helix Conformation of the NTD We first determined the crystal structure of full-length wild-type (WT) SsHsp14.1 by molecular replacement, and refined this

SsHsp14.1 Mutant Structures Illustrate a Tightly Packed NTD To explore an active conformation of the NTD, we made great efforts to obtain crystals of SsHsp14.1 WT at high temperature and in complex with client proteins, but without success, likely because natural clients of sHsps are unfolded proteins that cannot be crystallized. A client protein suitable for the complex structure study would be the one that exposes hydrophobic residues to solvent to mimic the natural unfolded clients, but remains folded. We noted that the NTD of SsHsp14.1 itself may be a suitable candidate due to its highly exposed hydrophobic surface and defined structure. However, the SsHsp14.1 WT structure revealed only very limited interactions between the NTDs of the face-to-face positioned SsHsp14.1 dimers (Figure 1A). We hypothesized that intensive association between SsHsp14.1 NTDs requires that the two participating dimers can freely rotate and move so as to fit to the interaction, which is impossible in the case of WT SsHsp14.1 due to the oligomerization of the dimers through CTE and the b4-b8 hydrophobic groove interactions in the crystal. We then sought to break the high-order oligomerization by mutagenesis.

2 Structure 23, 1–10, November 3, 2015 ª2015 Elsevier Ltd All rights reserved

Please cite this article in press as: Liu et al., Active-State Structures of a Small Heat-Shock Protein Revealed a Molecular Switch for Chaperone Function, Structure (2015), http://dx.doi.org/10.1016/j.str.2015.08.015

Figure 1. Overall Structures and NTD-Mediated Interactions of the WT and Mutant SsHsp14.1 (A) Two dimers (colored lime/orange and deepsalmon/yellow, respectively) with essentially the same conformation are observed in the SsHsp14.1 WT asymmetric unit. Each dimer associates with two neighboring dimers through hydrophobic interactions mediated by the CTEs (see also Figure S2A). (B) Contents of two asymmetric units are shown for SsHsp14.1 A102D (two tetramers to form one octamer). One tetramer is formed by two dimers (colored slate/cyan and deep-salmon/yellow, respectively) that share essentially the same conformation in one asymmetric unit. This tetramer and its symmetry related tetramer in a neighboring asymmetric unit (gray) form an octamer, which is essentially the same as the octamer observed in the asymmetric unit of the SsHsp14.1 Del-C4 crystal structure (see Figure S2B). This tetramer and octamer seem to exist in solution (see Figure S2C). The NTD helices of the SsHsp14.1 WT dimers interact with each other only at the ‘‘tip’’ with limited contacts, while the NTDs of the mutant SsHsp14.1 dimers interact with each other through extensive hydrophobic interactions (see the insets to the right of the overall structures). The NTD hydrophobic residues mediating the dimer-dimer interactions are shown as yellow sticks in the magnified insets. Leu16 and Phe20 in SsHsp14.1 A102D/Del-C4 do not directly participate in dimerdimer interactions, but are important for promoting the dimer-dimer interactions, and thus are shown as red sticks.

To abolish oligomerization mediated by hydrophobic interaction between the Ile-X-Ile motif in the CTE and a hydrophobic groove between b4 and b8 strands on the neighboring dimer (Figure S2A), we designed a series of point mutations in the hydrophobic groove from a hydrophobic residue to a hydrophilic one, as well as deletions to shorten CTE containing the Ile-XIle motif. Extensive trials have been made to crystallize these mutants, but only two of them, A102D and Del-C4 (the last four residues including the Ile-X-Ile motif in CTE were removed), crystallized. The structures of Del-C4 and A102D were solved by molecular replacement and refined to 3.10 A˚ and 2.50 A˚, respectively (Table 1). A102D and Del-C4 crystal structures show the same assembly of four SsHsp14.1 dimers through extensive hydrophobic interactions via their NTDs (Figures 1B and S3), despite different crystallization conditions and space groups (H32 and I4, respectively). In both structures, two dimers closely associate with each other through hydrophobic interactions between the NTDs to form a tetramer, and two such tetramers associate with each other face to face through hydrophobic interactions among the NTDs to form an octamer. The octamer in Del-C4 (composing one asymmetric unit) and that in A102D (composing two asymmetric units) can be superimposed on each other very well with a root-mean-square deviation (RMSD) of 1.8 A˚ based on Ca atoms (Figure S2B). Moreover, the formation of the octamer depends solely on the same NTD-mediated hydrophobic interactions in both crystal structures (Figures 1B and S3).

Observation of the same molecular assembly modes (tetramer or octamer) with different mutants under different crystallization conditions and space groups strongly indicates that this molecular assembly is not an artifact of crystallization, and that it may exist in solution. Indeed, the size-exclusion elution volume indicates that SsHsp14.1 A102D and Del-C4 do form octamers and/or tetramers, but not higher-order oligomers in solution (Figure S2C). We therefore deduce that the extensive hydrophobic interactions among the NTDs dominate the interdimer interactions and that the A102D and Del-C4 structures present an example for sHsp/client interaction, in which the clients happen to be the NTDs from other mutant SsHsp14.1 dimers. SsHsp14.1 Structures Reveal Key Residues for NTD Plasticity and Substrate Binding Comparing the A102D and Del-C4 structures with the WT structure reveals striking conformational changes in the NTD (Figure 2). In contrast to the straight helix conformation of the NTDs in the WT dimer, the mutant NTD helices are severely bent. Moreover, the NTD conformations of the two molecules within each mutant dimer are different. In one conformation, the helix is broken at residues 16–17 to form an ‘‘L’’-like structure (Figure 2E). In another conformation, residues 10–14 lose helix conformation, and its flanking helical segments come together to form a hairpin-like structure (Figure 2F). It is worth noting that in the two different NTD conformations in the mutants, the helix in both cases breaks at residues 16–17.

Structure 23, 1–10, November 3, 2015 ª2015 Elsevier Ltd All rights reserved 3

Please cite this article in press as: Liu et al., Active-State Structures of a Small Heat-Shock Protein Revealed a Molecular Switch for Chaperone Function, Structure (2015), http://dx.doi.org/10.1016/j.str.2015.08.015

Figure 2. Different NTD Conformations of the WT and Mutant SsHsp14.1 (A) Side and top views of the SsHsp14.1 WT dimer. NTDs of the two molecules share the identical long and straight helical conformation. (B) Side and top views of the mutant SsHsp14.1 (A102D or Del-C4, figure drawn from the A102D structure) dimer. NTDs of the two molecules adopt clearly different conformations; one is an L-like bent helix (yellow) and the other is a hairpin-like structure (deep salmon) after further disruption of the helix. (C) Superimposition of the SsHsp14.1 WT (yellow), SsHsp14.1 A102D (slate), StHsp14.0 FKF Form I (magenta, PDB: 3AAB), and FKF Form II (cyan, PDB: 3AAC) dimers. (D–F) Hydrophobic patches (shown by transparent surface representation) on the NTDs of SsHsp14.1 WT (D) and SsHsp14.1 A102D/Del-C4 (E and F). Hydrophobic residues forming the hydrophobic patches are shown by stick representation. Red arrow indicates the gap between the two pieces of hydrophobic patches on SsHsp14.1 WT NTD surface.

The helix break at Leu16 leads to important structural consequences. N-terminal domain bears eight hydrophobic residues, namely Met2, Ile5, Met6, Ile9, Leu13, Leu16, Phe20, and Val24, and they are distributed differently in the WT and mutant NTD conformations. In the WT structure, these hydrophobic residues form two pieces of hydrophobic patches facing 90 away from each other, one containing residues 2, 5, 6, 9, 13, and 16, and the other containing residues 20 and 24 (Figure 2D), whereas in the mutant structures these hydrophobic residues are positioned on one side of the NTD to form a joint hydrophobic patch (Figures

2E and 2F). Since in aqueous solution hydrophobic side chains tend to cluster together, especially at higher temperature, we hypothesize that the hydrophobic interaction between Leu16 and Phe20 can become stronger at elevated temperature to induce unwinding of the helix at residues 16–17, thus acting as a driving force to trigger conformational change resulting in bending and disruption of the NTD long helix, i.e. to facilitate conformational changes of NTDs from the long, straight, low-activity helices to the more flexible high-activity L- or hairpin-like structures.

4 Structure 23, 1–10, November 3, 2015 ª2015 Elsevier Ltd All rights reserved

Please cite this article in press as: Liu et al., Active-State Structures of a Small Heat-Shock Protein Revealed a Molecular Switch for Chaperone Function, Structure (2015), http://dx.doi.org/10.1016/j.str.2015.08.015

Figure 3. Chaperone Activity of WT and Mutant SsHsp14.1 In Vitro and In Vivo (A) Chaperone activity of the N-terminal mutant variants (L16S, L16W, F20S) to inhibit aggregation of MDH in vitro. Aggregation of MDH (4.5 mM) incubated at 50 C was monitored by light absorbance at 360 nm in the absence or presence (108 mM) of SsHsp14.1 WT or N-terminal mutants. L16S and F20S did not co-precipitate with denatured MDH in this assay (see Figure S3). (B–E) Thermal activation of SsHsp14.1 WT (B), L16W (C), L16S (D), and F20S (E) in vitro. Aggregation of TEV after being denatured by 8 M urea and diluted into 20 mM Tris and 150 mM NaCl (pH 8.0) (the final concentration of TEV was 2 mM) was monitored in the absence or presence of WT or mutant SsHsp14.1 (the final concentration of SsHsp14.1 was 30 mM) by recording the light absorbance at 360 nm. The SsHsp14.1 proteins were pre-incubated in 20 mM Tris and 150 mM NaCl (pH 8.0) at 25 C, 37 C, 50 C, 75 C, or 95 C for 10 min and then cooled immediately to room temperature before mixing with TEV. (F and G) Colony formation assays to assess the chaperone activity of WT or mutant SsHsp14.1 in enhancing thermoresistance of E. coli (see also Figure S4). Viability data of E. coli after treatment at 50 C for 45 min are shown by bar-graph representation. The bacteria express no SsHsp14.1 (bearing empty pET23a), SsHsp14.1 WT, X-to-S mutants (F), or X-to-W mutants (G) in the NTD. The viability data are normalized to the strain expressing WT SsHsp14.1. Error bars indicate the SD.

Hydrophobicity of Residues 16 and 20 Is Essential for In Vitro Chaperone Activity As suggested by our structures, the hydrophobicity of Leu16 and Phe20 might be essential for turning on chaperone activity of SsHsp14.1 by triggering conformational changes favoring interactions with client proteins. To test this hypothesis, four mutants representing the substitution from hydrophobic to hydrophilic residues (L16S, F20S) or to more hydrophobic residues (L16W and F20W) were designed in the WT background. Attempts to obtain soluble F20W protein failed because it precipitated during purification. Assays for inhibition of heating induced protein aggregation were performed to test the diversities between the WT and mutant SsHsp14.1. The chaperone activity of SsHsp14.1 in preventing aggregation of denatured client protein was monitored

by the change in light absorbance (turbidity). SsHsp14.1 WT could almost completely inhibit the aggregation of mitochondrial malate dehydrogenase (MDH) when the molar ratio of MDH/WT reached 1:24, in agreement with previous study on similar sHsp (Usui et al., 2004). Under the same ratio, L16S and F20S exhibited nearly no chaperone activity, while L16W obtained an inhibition effect equal to that of the WT protein (Figure 3A). It is noteworthy that a turbidity assay may be misleading in assessing chaperone/client interaction without careful inspection of the protein behavior. Several crystallin mutants were found to bind the client protein strongly and to co-precipitate with it instead of inhibiting precipitation, leading to high turbidity readout (Koteiche and McHaourab, 2006). However, SsHsp14.1 L16S and F20S did not co-precipitate with MDH as was judged by SDS-PAGE (Figure S3). Hence, it is deduced

Structure 23, 1–10, November 3, 2015 ª2015 Elsevier Ltd All rights reserved 5

Please cite this article in press as: Liu et al., Active-State Structures of a Small Heat-Shock Protein Revealed a Molecular Switch for Chaperone Function, Structure (2015), http://dx.doi.org/10.1016/j.str.2015.08.015

that L16S and F20S lose chaperone activity due to lack of interaction with clients. Inhibition of aggregation assays were then employed to determine whether the activation of chaperone activity of SsHsp14.1 by elevated temperature is mediated by the Leu16-Phe20 hydrophobic interaction. The chemically denatured recombinant tobacco etch virus (TEV) protease was diluted 1:100 into buffers containing the same amount of WT or mutant SsHsp14.1 proteins pre-incubated at different temperatures. Without SsHsp14.1 the denatured TEV aggregates, resulting in turbidity and higher light absorbance, while in the presence of SsHsp14.1 with chaperone activity, the aggregation of denatured TEV is prevented. As shown in Figure 3B, activity of SsHsp14.1 WT showed a clear thermoactivation effect, suggesting the existence of some thermosensing regulation of this protein. After pre-incubation at 25 C or 37 C, chaperone activity of SsHsp14.1 WT toward denatured TEV was low, while after pre-incubation at temperatures above 50 C, the chaperone activity significantly increased. In agreement with the prediction based on the thermoswitch model by Leu16-Phe20 hydrophobic interaction, SsHsp14.1 L16W showed high chaperone activity even at temperatures as low as 25 C (Figure 3C), while even after pre-incubation at very high temperatures such as 95 C, the SsHsp14.1 L16S and F20S protein still remained inactive (Figures 3D and 3E). However, these data cannot solely prove that Leu16Phe20 act as a thermoswitch. The observed high activity of L16W and low activity of L16S and F20S could be explained by direct interaction between the side chains of these residues with the denatured client proteins. To test this possibility, we mutated a nearby residue, Leu13, to either Ser or Trp to test whether they show the same effect on chaperone activity. Unfortunately, we failed to obtain soluble L13S and L13W proteins because they precipitated during purification. We then turned to in vivo functional studies. The NTD Hydrophobic Surface Is Essential for Bacterial Heat Resistance We next carried out functional mutagenesis to study the hydrophobic residues in NTD in vivo by measuring the effect of mutating them to polar amino acids. Chaperone activity of sHsps can be estimated by colony formation assays (Liu and Hendrickson, 2007; Soto et al., 1999). The optimal temperature for E. coli proliferation is 37 C. Treating E. coli at 50 C for 45 min destroys protein homeostasis in the bacteria, resulting in death (Li et al., 2012; Soto et al., 1999). As expected, most bacteria of the negative control strain transformed with empty plasmid pET23a were killed after heating, showing poor thermoresistance, while most bacteria of the positive control strain expressing SsHsp14.1 WT survived, showing significant resistance to heating (Figure S4A). The N-terminal hydrophobic residues were mutated to Ser or Trp to test the effect of protecting bacteria in heat treatment, which is an indication of the chaperone activity of the sHsp. Compared with SsHsp14.1 WT, L16S and F20S led to a >100fold decrease in E. coli cell viability after treatment at 50 C, indicating the loss of SsHsp14.1 chaperone activity (Figure 3F). This observation is in agreement with our in vitro data above suggesting the essential role of L16 and F20 in chaperone activity. E. coli expressing SsHsp14.1 L16W and F20W remained at the same

level of cell viability as that expressing SsHsp14.1 WT (Figure 3G). However, these data do not solely validate the Leu16Phe20 thermoswitch model, since the observed higher viability of bacteria expressing WT, L16W, or F20W and lower viability of bacteria expressing L16S or F20S could be simply explained by direct interactions of these two residues with denatured client proteins. We therefore decided to study other hydrophobic residues located close to Leu16. Interestingly, mutants of hydrophobic residues at the extreme N-terminal end behaved differently from Leu16 and Phe20. They demonstrated an increasing preference toward hydrophilic residue for chaperone activity from the N-terminal residue Met2 to the residues in the middle of the NTD until Leu13. As shown in Figure 3F, by decreasing the hydrophobicity, M2S mutant leads to a 10-fold decrease in cell viability, whereas L13S leads to a 10fold increase (Figure 3F), and, on the other hand, mutation L13W to increase the hydrophobicity leads to a 20-fold decrease in cell viability (Figure 3G). It is worth noting that SsHsp14.1 expressed at a comparable level in all the mutants tested, which ruled out that the difference in cell viability is associated with low protein expression (Figure S4B). The distinct effects on chaperone activity of L13S/W compared with L16S/W and F20S/W is intriguing. The strikingly high chaperone activity of L13S indicates that this residue may not directly interact with denatured clients with exposed hydrophobic surfaces. Since Leu16 and Phe20 are located closer to the ACD and further away from the client than residue 13, they are even less likely to directly interact with clients. Therefore, we deduce that the major role of residues Leu16 and Phe20 should be promoting conformational changes to disrupt the NTD helix rather than binding clients directly. Why L13S largely increases the activity of SsHsp14.1 could also be explained by promoting the breakdown of the long straight NTD helix, since the continuous hydrophobic patch consisting of Met2, Ile5, Met6, Ile9, Leu13, and Leu16 would be expected to stabilize the amphipathic long helical NTD as was observed in our SsHsp14.1 WT crystal structure (Figure 2D). As for residues 2, 5, and 6, they stretch further out toward the client proteins and are hence expected to make direct contact with the clients. Thus, excess hydrophilicity (the X-to-Ser mutants, Figure 3F) would prevent sHsp-client interaction, while on the other hand excess hydrophobicity (the X-to-Trp mutants, Figure 3G) would prevent disruption of the NTD helices, both resulting in weakened chaperone activity. Together, these data suggest that disruption of the long and straight NTD helices may be the key event in the activation of the chaperone function of SsHsp14.1, and that this procedure is subtly regulated by the hydrophobic residues located in the NTD, among which the Leu16-Phe20 pair may act in response to elevated temperature to introduce a kink in the NTD helix, followed by the breakdown of more helix structure in the extreme N-terminal part of the NTD to facilitate the binding to denatured client proteins. DISCUSSION In this study, we determined three crystal structures of WT and mutant SsHsp14.1, which together provide snapshots of this sHsp utilizing its NTD to interact with a non-native client protein.

6 Structure 23, 1–10, November 3, 2015 ª2015 Elsevier Ltd All rights reserved

Please cite this article in press as: Liu et al., Active-State Structures of a Small Heat-Shock Protein Revealed a Molecular Switch for Chaperone Function, Structure (2015), http://dx.doi.org/10.1016/j.str.2015.08.015

Figure 4. The Thermoswitch Model for SsHsp14.1 Activation (A) NTD sequence alignment of several members of the Hsp14 subfamily. The NTDs of these proteins share the common pattern of hydrophobic residue distribution indicating the trend to form the long, straight, and amphipathic helix. Notably, an LXXXF motif is identified in these proteins (VXXXF in the case of Sulfolobus archaeon), indicating that the Leu-Phe thermoswitch model proposed for SsHsp14.1 may also apply to these proteins. (B) A schematic drawing to summarize the Leu-Phe thermoswitch model proposed in this report. The SsHsp14.1 dimers are shown as a pair of bars in light gray and dark gray, respectively. The Leu-Phe thermoswitch model proposed in this report is summarized in the inset. The solvent-exposed NTDs of the dimers or suboligomers may undergo fast conformational switches between the long straight low-activity helix conformation and the high-activity helix-broken conformation. Heating may promote the conformational changes toward the high-activity state through the Leu16-Phe20 interaction, which is followed by further disruption of the NTD to facilitate the binding of non-native clients.

Furthermore, by employing in vitro thermal unfolding and aggregation assays and an in vivo cell viability assay, we identified residues Leu16 and Phe20 as essential elements for the NTD conformational switch for client binding. Structure flexibility has been noted as the fundamental feature of sHsps and other molecular chaperones for chaperone activity, such as a-synuclein, a-casein, Hsp25, and GroEL (Bhattacharyya and Das, 1999; Kim et al., 2002; Lindner et al., 2000; Park et al., 2002; Sharma et al., 2008). In these systems, the flexible regions of the proteins are normally involved in interacting with the client proteins (Braig et al., 1994; Tompa and Csermely, 2004). Indeed, in most sHsps studied by crystallography, the NTDs were not observed (Basha et al., 2012; Hilton et al., 2013; Kim et al., 1998; van Montfort et al., 2001). Remarkably, in the crystal structures of the three SsHsp14.1 variants reported in this study, the NTDs are all well defined and reveal

very different conformations that are useful for elucidating the regulatory mechanism. A crystal structure of sHsp C-terminal IXI to FKF mutant from Sulfolobus tokodaii (StHsp14.0) with defined NTD structure was reported earlier (Takeda et al., 2011). Interestingly, StHsp14.0 NTD helix also bends at the residue position corresponding to Leu16 of SsHsp14.1. Superimposing these structures shows that the NTD helical structures are essentially the same from residues 25 to 17 and diverge drastically from residues 16 to 1 (Figure 2C). These NTD structures together provide consistent evidence that the NTDs of this subfamily of sHsps (sequence alignment shown in Figure 4A) have intrinsic conformational plasticity that enables them to bind a large array of heterogeneous client proteins, and that Leu16 might function as a hinge that allows the N-terminal segment to adopt different orientations to interact with different client proteins. In addition to recognizing unfolded client proteins, an important aspect of sHsps is their self-regulatory mechanism. Two trademarks of unfolded or misfolded client proteins are heterogeneous conformation and exposure of hydrophobic residues. Therefore, the NTD structural motifs developed by the sHsps to bind the diverse client proteins should also acquire the properties of being plastic and hydrophobic. These properties, however, could be a liability for cells under physiological conditions because the NTD self-association can lead to depletion of the sHsps. Therefore, the cell must evolve a mechanism to keep the sHsps in resting status to prevent this disastrous effect. Hsps are not highly expressed under normal conditions and can be induced to express under stressed conditions; this is the first mechanism to prevent aggregation and depletion of sHsps. Another proposed mechanism of aggregation prevention seems to be the formation of hollow spherical oligomers (Hochberg et al., 2014; Peschek et al., 2013; Shi et al., 2013), since this assembly could prevent the NTDs from contacting the clients. However, this does not mean that all oligomers of sHsps are in a resting state. According to previous studies, sHsps form hollow symmetric oligomers under physiological conditions (the resting oligomer). The oligomer undergoes dissociation in stressed conditions, and upon binding non-native clients the sHsp dimers reassemble to higher-order oligomers (the working oligomer). The client bound oligomer may vary in size and shape, and may be symmetric or asymmetric, probably depending on the properties of the non-native client proteins they bind (Stengel et al., 2010) (Figure 4B). Therefore, the spherical oligomer is not fixed; it is in equilibrium with suboligomers or sHsp dimers. However, suboligomers or sHsp dimers expose the NTD hydrophobic residues to solution, which could result in self-association between sHsp dimers. The cells should develop a mechanism to reduce these disastrous interactions. We propose that the long straight helical NTD (observed in the SsHsp14.1 WT structure) may act to reduce this risk, due to its rigid structure and consequently limited contacts between dimers (Figure 1A), hence representing a low-activity state of SsHsp14.1. Stressed conditions such as heat may trigger the Leu16-Phe20 switch to facilitate the conformational changes in the NTD, namely bending of the helix at residue 16, since hydrophobic interaction becomes stronger at elevated temperature. The bending, followed by further disruption of the extreme part of the NTD helix, leads to the formation of an expanded and flexible hydrophobic surface compatible

Structure 23, 1–10, November 3, 2015 ª2015 Elsevier Ltd All rights reserved 7

Please cite this article in press as: Liu et al., Active-State Structures of a Small Heat-Shock Protein Revealed a Molecular Switch for Chaperone Function, Structure (2015), http://dx.doi.org/10.1016/j.str.2015.08.015

with binding to heterogeneous client proteins, ultimately turns on the full activity of this sHsp (Figures 2D–2F and 4B). We were unable to obtain a structure of the complex between SsHsp14.1 and a bona fide client protein, since clients of sHsps are nonnative proteins that are heterogeneous in size, structure, and surface property, and are therefore not subject to crystallization. Nonetheless, since client proteins of sHsps are not specific (Delbecq and Klevit, 2013; Peschek et al., 2009), SsHsp14.1 itself could act both as chaperones and clients if they can bind to each other through the NTD hydrophobic surface. The two CTE hydrophobic interaction incompetent mutants (A102D and Del-C4) unleashed SsHsp14.1 dimers for free movement, leading to association between SsHsp14.1 dimers through extensive hydrophobic interactions among the NTDs (Figure 1B), thus resulting in the active-state-like structures. However, these mutant structures do not elucidate all possible functional working states of the NTD of SsHsp14.1, since the clients can be potentially any non-native proteins that vary in size, structure, and surface properties. Our structural studies thus only provided an example of how SsHsp14.1 can bind to a client that happens to be the unstructured NTD of a neighboring SsHsp14.1. sHsps constitute a large family highly diverse in the NTD. The mechanisms proposed here may not apply to other subfamilies of sHsps. For example, a thermosensor domain has been described for Hsp26, the principal sHsp of Saccharomyces cerevisiae (Franzmann et al., 2008). Nevertheless, the proposed mechanisms for temperature sensing, NTD structure plasticity, and self-regulation may be an inspiration for further studies on sHsps of other subfamilies and other systems that work with non-native client proteins. EXPERIMENTAL PROCEDURES Expression and Purification of WT and NTD Mutant SsHsp14.1 Proteins The full-length archaea SsHsp14.1 gene (encoding residues 1–124) was PCRamplified from S. solfataricus genomic DNA. SsHsp14.1 WT construct was cloned into pET-23a vector through the NdeI and XhoI sites. The protein was expressed in E. coli strain BL21 (DE3) (Novagen). The bacteria were grown to an OD600 of 0.6–0.8 and induced with 0.5 mM isopropyl b-D-1-thiogalactopyranoside (IPTG) at 37 C for 3 hr. The cells were harvested at 4 C by centrifugation (4,000 3 g for 10 min). Cells pellets were resuspended in lysis buffer containing 20 mM Tris (pH 8.0), 150 mM NaCl, 1% glycerol, 1 mM Tris(2-carboxyethyl)phosphine (TCEP) hydrochloride, and 1 mM PMSF, then lysed by sonication. The lysate was cleared by centrifugation (20,000 3 g at 4 C for 30 min). The supernatant was first incubated at 75 C for 30 min to denature host proteins and then centrifuged at 20,000 3 g at 4 C for 30 min to remove the denatured proteins. Subsequently, proteins were purified by ion-exchange chromatography on a Resource Q 1/6 column (GE Healthcare) equilibrated with 20 mM Tris-HCl buffer (pH 8.0) (with 1% glycerol and 1 mM TCEP). SsHsp14.1 WT was eluted with a linear gradient of 0–1,000 mM NaCl in 20 mM Tris-HCl buffer (pH 8.0) (with 1% glycerol and 1 mM TCEP). Target protein was collected and concentrated by ultrafiltration. Finally, the concentrated protein solution was loaded onto a Superdex 200 10/300 gel filtration column (GE Healthcare Life Sciences) equilibrated with 20 mM TrisHCl buffer (pH 8.0) (with 150 mM NaCl, 1% glycerol, and 1 mM TCEP). Fractions containing SsHsp14.1 WT were collected and concentrated to a final concentration of 20 mg/ml. Aliquots were flash-frozen in liquid nitrogen and stored at 80 C. N-Terminal mutant variants including L16S, L16W, and F20S were expressed and purified as SsHsp14.1 WT. However, we were unable to obtain soluble F20W protein because it precipitated during purification.

Expression and Purification of SsHsp14.1 A102D and Del-C4 SsHsp14.1 A102D and Del-C4 were cloned into a modified pET-23a vector containing an N-terminal human rhinovirus 3C protease-cleavable CBD (cellulose-binding domain) tag (Lan et al., 2012) through the KpnI and XhoI sites. The proteins were expressed in E. coli BL21 (DE3) (Novagen) as fusion proteins with N-terminal CBD tags. When the host cells were grown to an OD600 of 0.6–0.8, the proteins were induced with 0.5 mM IPTG at 37 C for 3 hr. The cells were harvested at 4 C by centrifugation (4,000 3 g for 10 min). Cell pellets were resuspended in lysis buffer containing 20 mM Tris (pH 8.0), 150 mM NaCl, 1% glycerol, 1 mM TCEP, and 1 mM PMSF, then lysed by sonication. After the lysate was centrifuged for 30 min at 4 C at 20,000 3 g, the supernatant was incubated with cellulose beads at 25 C for 30 min. The beads were washed with 20 column volumes of lysis buffer and wash buffer (20 mM Tris, 500 mM NaCl, 1% glycerol, 1 mM TCEP [pH 8.0]) and incubated overnight with recombinant CBD-3C protease to cleave the CBD tags. After cleavage, the SsHsp14.1 A102D and Del-C4 proteins were extracted from the cellulose beads by centrifugation (4,000 3 g for 10 min at 4 C). Subsequently, the supernatants were concentrated and applied to Superdex 200 10/300 gel filtration apparatus (GE Healthcare) in 20 mM Tris-HCl buffer (pH 8.0) (with 150 mM NaCl, 1% glycerol, and 1 mM TCEP). Fractions containing A102D and DelC4 proteins were collected and concentrated to final concentrations of 20 mg/ml. Aliquots were flash-frozen in liquid nitrogen and stored at 80 C. Crystallization and Data Collection SsHsp14.1 WT crystals were grown by hanging-drop vapor diffusion. Well solution (1 ml) containing 0.2 M calcium chloride dehydrate, 45% (v/v) (+/ )-2methyl-2,4-pentanediol, 0.1 M bis-tris (pH 6.5), and 5 mM EDTA disodium salt dihydrate was mixed with 1 ml of WT protein solution (12 mg/ml) and incubated at 293 K for up to 2 weeks. Crystals were cryoprotected by soaking in well solution supplemented with 20% ethylene glycol and flash-frozen in liquid nitrogen. Crystals of SsHsp14.1 A102D were obtained in 0.2 M magnesium sulfate heptahydrate and 20% (w/v) polyethylene glycol (PEG) 3350. The protein concentration was 5 mg/ml. Crystals were cryoprotected by soaking in well solution supplemented with 20% glycerol, and flash-frozen in liquid nitrogen. Crystals of SsHsp14.1 Del-C4 were obtained in 12% (w/v) PEG 3350 and 0.1 M sodium malonate (pH 7.0). The protein concentration was 12 mg/ml. Crystals were cryoprotected by soaking in well solution supplemented with 20% ethylene glycol, and flash-frozen in liquid nitrogen. Diffraction data were collected at the Shanghai SSRF BL17U1 beamline or at the Argonne National Laboratory APS ID19 beamline at 100 K. The data were processed with HKL3000 (Minor et al., 2006). Structure Determination and Refinement The WT structure was determined by molecular replacement using the StHsp14.0 structure (PDB: 3AAB) as the search model using Phaser (McCoy et al., 2007). A102D and Del-C4 structures were solved by molecular replacement using the WT structure as the search model using Phaser. To improve the model quality, simulated-annealing was performed using CNS (Brunger et al., 1998). Iterative cycles of manual refitting and crystallographic refinement were performed using Coot (Emsley and Cowtan, 2004) and PHENIX (Adams et al., 2002). All figures for the molecular models were prepared using the PyMOL program. Statistics of diffraction data processing and structure refinement are presented in Table 1. In Vitro Client Aggregation Assay: Turbidity Assay Aggregation of malate dehydrogenase (MDH, from porcine; Sigma) upon heating was monitored by measuring light absorbance at 360 nm with a UV-755B spectrophotometer (Shanghai Metash Instruments). For each measurement, MDH was diluted into 20 mM Tris-HCl (with 150 mM NaCl [pH 8.0]) buffer to a final concentration of 4.5 mM and incubated at 50 C in the presence (at different concentrations) or absence of 108 mM SsHsp14.1 WT or mutant variants. Thermal Activation Analysis with Chemically Denatured TEV Protease In vitro client aggregation assays were used to study the thermal activation behaviors of SsHsp14.1 WT and mutants. The SsHsp14.1 WT or mutants were diluted with 20 mM Tris-HCl (with 150 mM NaCl [pH 8.0]) buffer, treated

8 Structure 23, 1–10, November 3, 2015 ª2015 Elsevier Ltd All rights reserved

Please cite this article in press as: Liu et al., Active-State Structures of a Small Heat-Shock Protein Revealed a Molecular Switch for Chaperone Function, Structure (2015), http://dx.doi.org/10.1016/j.str.2015.08.015

at different temperatures (25 C, 37 C, 50 C, 75 C, or 95 C) for 10 min and cooled immediately to 25 C, then incubated with denatured TEV protease at 25 C. TEV protease was expressed in E. coli and purified as described previously (Fang et al., 2013). TEV protease was denatured in 8 M urea (with 20 mM Tris, 150 mM NaCl [pH 8.0]), and diluted 100-fold into 20 mM Tris-HCl (with 150 mM NaCl [pH 8.0]) buffer in the presence (30 mM) or absence of SsHsp14.1 WT or mutants to a final concentration of 2 mM. Colony Formation Assay Colony formation assays were conducted to study the thermotolerance of E. coli expressing WT or mutant SsHsp14.1. E. coli strain BL21 (DE3), with empty plasmid used as negative control. Cells were first cultured to mid-log phase (OD600 = 0.8) at 37 C and cooled immediately to 30 C; 1 mM IPTG was added to induce SsHsp14.1 expression, after incubating at 30 C for 2 hr. SDS-PAGE was then performed to test the protein expression level of SsHsp14.1. Meanwhile, the cultures were untreated or heat-shocked at 50 C for 45 min. The samples were diluted quickly and plated immediately on LB plates supplemented with 100 mg/ml ampicillin and 34 mg/ml chloramphenicol. For each test, 10-fold serial dilutions (10 1–10 5) of the culture were plated. Cell viability was estimated by counting the number of colonies formed after incubation of the plates overnight at 37 C. Size-Exclusion Chromatography Oligomeric states of SsHsp14.1 WT, A102D, and Del-C4 of were examined by size-exclusion chromatography, was performed using a Superdex 200 10/300 Gl column (GE Healthcare Life Sciences) at a flow rate of 0.5 ml/min. The column was first equilibrated with the assay buffer (20 mM Tris-HCl, 150 mM NaCl [pH 8.0]), then 200 ml of samples (1.0 mg/ml) was applied to the column. Standard proteins (ferritin: 440 kDa; BSA: 67 kDa; b-lactoglobulin: 35 kDa; RNase A: 13.7 kDa) were used as molecular mass markers for calibration. ACCESSION NUMBERS Coordinates of SsHsp14.1 WT, A102D, and Del-C4 have been deposited in the RCSB PDB under the accession code PDB: 4YL9, 4YLB, and 4YLC, respectively.

Basha, E., O’Neill, H., and Vierling, E. (2012). Small heat shock proteins and alpha-crystallins: dynamic proteins with flexible functions. Trends Biochem. Sci. 37, 106–117. Bhattacharyya, J., and Das, K.P. (1999). Molecular chaperone-like properties of an unfolded protein, alpha(s)-casein. J. Biol. Chem. 274, 15505–15509. Braig, K., Otwinowski, Z., Hegde, R., Boisvert, D.C., Joachimiak, A., Horwich, A.L., and Sigler, P.B. (1994). The crystal structure of the bacterial chaperonin GroEL at 2.8 A. Nature 371, 578–586. Braun, N., Zacharias, M., Peschek, J., Kastenmuller, A., Zou, J., Hanzlik, M., Haslbeck, M., Rappsilber, J., Buchner, J., and Weinkauf, S. (2011). Multiple molecular architectures of the eye lens chaperone alphaB-crystallin elucidated by a triple hybrid approach. Proc. Natl. Acad. Sci. USA 108, 20491–20496. Brunger, A.T., Adams, P.D., Clore, G.M., DeLano, W.L., Gros, P., GrosseKunstleve, R.W., Jiang, J.S., Kuszewski, J., Nilges, M., Pannu, N.S., et al. (1998). Crystallography & NMR system: a new software suite for macromolecular structure determination. Acta Crystallogr. D Biol. Crystallogr. 54, 905–921. Bukau, B., Weissman, J., and Horwich, A. (2006). Molecular chaperones and protein quality control. Cell 125, 443–451. Das, K.P., and Surewicz, W.K. (1995). Temperature-induced exposure of hydrophobic surfaces and its effect on the chaperone activity of alpha-crystallin. FEBS Lett. 369, 321–325. Delbecq, S.P., and Klevit, R.E. (2013). One size does not fit all: the oligomeric states of alphaB crystallin. FEBS Lett. 587, 1073–1080. Delbecq, S.P., Jehle, S., and Klevit, R. (2012). Binding determinants of the small heat shock protein, alphaB-crystallin: recognition of the ’IxI’ motif. EMBO J. 31, 4587–4594. Eifert, C., Burgio, M.R., Bennett, P.M., Salerno, J.C., and Koretz, J.F. (2005). N-terminal control of small heat shock protein oligomerization: changes in aggregate size and chaperone-like function. Biochim. Biophys. Acta 1748, 146–156. Emsley, P., and Cowtan, K. (2004). Coot: model-building tools for molecular graphics. Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132. Evans, P. (2006). Scaling and assessment of data quality. Acta Crystallogr. D Biol. Crystallogr. 62, 72–82. Fang, J., Chen, L., Cheng, B., and Fan, J. (2013). Engineering soluble tobacco etch virus protease accompanies the loss of stability. Protein Expr. Purif. 92, 29–35.

SUPPLEMENTAL INFORMATION Supplemental Information includes four figures and can be found with this article online at http://dx.doi.org/10.1016/j.str.2015.08.015. AUTHOR CONTRIBUTIONS L.L. and J.-Y.C. designed and performed the experiments. B.Y. and Y.-H.W. conceived the project. C.-H.Y. instructed the project and designed the experiments. All authors were involved in data analysis and manuscript writing.

Franzmann, T.M., Menhorn, P., Walter, S., and Buchner, J. (2008). Activation of the chaperone Hsp26 is controlled by the rearrangement of its thermosensor domain. Mol. Cell 29, 207–216. Hanazono, Y., Takeda, K., Oka, T., Abe, T., Tomonari, T., Akiyama, N., Aikawa, Y., Yohda, M., and Miki, K. (2013). Nonequivalence observed for the 16-meric structure of a small heat shock protein, SpHsp16.0, from Schizosaccharomyces pombe. Structure 21, 220–228. Haslbeck, M., Walke, S., Stromer, T., Ehrnsperger, M., White, H.E., Chen, S., Saibil, H.R., and Buchner, J. (1999). Hsp26: a temperature-regulated chaperone. EMBO J. 18, 6744–6751.

ACKNOWLEDGMENTS C.-H.Y. is funded by the National Basic Research Program of China (973 Program, No. 2012CB917202), the National Science Foundation of China (No. 31270769), and the Ministry of Science and Technology of China (NCET-12-0013). Y.-H.W. is funded by the National Science Funds for the Excellent Youth Scholars (No. 31222043), and fundamental Research Funds for the Central Universities (No. 2014ZM0062).

Haslbeck, M., Franzmann, T., Weinfurtner, D., and Buchner, J. (2005). Some like it hot: the structure and function of small heat-shock proteins. Nat. Struct. Mol. Biol. 12, 842–846. Hilton, G.R., Lioe, H., Stengel, F., Baldwin, A.J., and Benesch, J.L. (2013). Small heat-shock proteins: paramedics of the cell. Top. Curr. Chem. 328, 69–98. Hochberg, G.K., Ecroyd, H., Liu, C., Cox, D., Cascio, D., Sawaya, M.R., Collier, M.P., Stroud, J., Carver, J.A., Baldwin, A.J., et al. (2014). The structured core domain of alphaB-crystallin can prevent amyloid fibrillation and associated toxicity. Proc. Natl. Acad. Sci. USA 111, E1562–E1570.

Received: May 6, 2015 Revised: August 27, 2015 Accepted: August 29, 2015 Published: October 1, 2015 REFERENCES

Jaya, N., Garcia, V., and Vierling, E. (2009). Substrate binding site flexibility of the small heat shock protein molecular chaperones. Proc. Natl. Acad. Sci. USA 106, 15604–15609.

Adams, P.D., Grosse-Kunstleve, R.W., Hung, L.W., Ioerger, T.R., McCoy, A.J., Moriarty, N.W., Read, R.J., Sacchettini, J.C., Sauter, N.K., and Terwilliger, T.C. (2002). PHENIX: building new software for automated crystallographic structure determination. Acta Crystallogr. D Biol. Crystallogr. 58, 1948–1954.

Jehle, S., Vollmar, B.S., Bardiaux, B., Dove, K.K., Rajagopal, P., Gonen, T., Oschkinat, H., and Klevit, R.E. (2011). N-terminal domain of alphaB-crystallin provides a conformational switch for multimerization and structural heterogeneity. Proc. Natl. Acad. Sci. USA 108, 6409–6414.

Structure 23, 1–10, November 3, 2015 ª2015 Elsevier Ltd All rights reserved 9

Please cite this article in press as: Liu et al., Active-State Structures of a Small Heat-Shock Protein Revealed a Molecular Switch for Chaperone Function, Structure (2015), http://dx.doi.org/10.1016/j.str.2015.08.015

Kim, K.K., Kim, R., and Kim, S.H. (1998). Crystal structure of a small heatshock protein. Nature 394, 595–599. Kim, T.D., Paik, S.R., and Yang, C.H. (2002). Structural and functional implications of C-terminal regions of alpha-synuclein. Biochemistry 41, 13782–13790. Koteiche, H.A., and McHaourab, H.S. (2006). Mechanism of a hereditary cataract phenotype. Mutations in alphaA-crystallin activate substrate binding. J. Biol. Chem. 281, 14273–14279. Koteiche, H.A., Chiu, S., Majdoch, R.L., Stewart, P.L., and McHaourab, H.S. (2005). Atomic models by cryo-EM and site-directed spin labeling: application to the N-terminal region of Hsp16.5. Structure 13, 1165–1171. Lan, D., Tai, Y., Shen, Y., Wang, F., Yang, B., and Wang, Y. (2012). Efficient purification of native recombinant proteins using proteases immobilized on cellulose. J. Biosci. Bioeng. 113, 542–544. Li, D.C., Yang, F., Lu, B., Chen, D.F., and Yang, W.J. (2012). Thermotolerance and molecular chaperone function of the small heat shock protein HSP20 from hyperthermophilic archaeon, Sulfolobus solfataricus P2. Cell Stress Chaperones 17, 103–108. Lindner, R.A., Carver, J.A., Ehrnsperger, M., Buchner, J., Esposito, G., Behlke, J., Lutsch, G., Kotlyarov, A., and Gaestel, M. (2000). Mouse Hsp25, a small shock protein. The role of its C-terminal extension in oligomerization and chaperone action. Eur. J. Biochem. 267, 1923–1932.

Sharma, S., Chakraborty, K., Muller, B.K., Astola, N., Tang, Y.C., Lamb, D.C., Hayer-Hartl, M., and Hartl, F.U. (2008). Monitoring protein conformation along the pathway of chaperonin-assisted folding. Cell 133, 142–153. Shashidharamurthy, R., Koteiche, H.A., Dong, J., and McHaourab, H.S. (2005). Mechanism of chaperone function in small heat shock proteins: dissociation of the HSP27 oligomer is required for recognition and binding of destabilized T4 lysozyme. J. Biol. Chem. 280, 5281–5289. Shi, J., Koteiche, H.A., McDonald, E.T., Fox, T.L., Stewart, P.L., and McHaourab, H.S. (2013). Cryoelectron microscopy analysis of small heat shock protein 16.5 (Hsp16.5) complexes with T4 lysozyme reveals the structural basis of multimode binding. J. Biol. Chem. 288, 4819–4830. Soto, A., Allona, I., Collada, C., Guevara, M.A., Casado, R., Rodriguez-Cerezo, E., Aragoncillo, C., and Gomez, L. (1999). Heterologous expression of a plant small heat-shock protein enhances Escherichia coli viability under heat and cold stress. Plant Physiol. 120, 521–528. Stengel, F., Baldwin, A.J., Painter, A.J., Jaya, N., Basha, E., Kay, L.E., Vierling, E., Robinson, C.V., and Benesch, J.L. (2010). Quaternary dynamics and plasticity underlie small heat shock protein chaperone function. Proc. Natl. Acad. Sci. USA 107, 2007–2012. Sun, Y., and MacRae, T.H. (2005). Small heat shock proteins: molecular structure and chaperone function. Cell. Mol. Life Sci. 62, 2460–2476.

Liu, Q., and Hendrickson, W.A. (2007). Insights into Hsp70 chaperone activity from a crystal structure of the yeast Hsp110 Sse1. Cell 131, 106–120.

Takeda, K., Hayashi, T., Abe, T., Hirano, Y., Hanazono, Y., Yohda, M., and Miki, K. (2011). Dimer structure and conformational variability in the N-terminal region of an archaeal small heat shock protein, StHsp14.0. J. Struct. Biol. 174, 92–99.

McCoy, A.J., Grosse-Kunstleve, R.W., Adams, P.D., Winn, M.D., Storoni, L.C., and Read, R.J. (2007). Phaser crystallographic software. J. Appl. Crystallogr. 40, 658–674.

Tompa, P., and Csermely, P. (2004). The role of structural disorder in the function of RNA and protein chaperones. FASEB J. 18, 1169–1175.

McHaourab, H.S., Lin, Y.L., and Spiller, B.W. (2012). Crystal structure of an activated variant of small heat shock protein Hsp16.5. Biochemistry 51, 5105–5112. Minor, W., Cymborowski, M., Otwinowski, Z., and Chruszcz, M. (2006). HKL3000: the integration of data reduction and structure solution—from diffraction images to an initial model in minutes. Acta Crystallogr. D Biol. Crystallogr. 62, 859–866. Park, S.M., Jung, H.Y., Kim, T.D., Park, J.H., Yang, C.H., and Kim, J. (2002). Distinct roles of the N-terminal-binding domain and the C-terminal-solubilizing domain of alpha-synuclein, a molecular chaperone. J. Biol. Chem. 277, 28512– 28520. Peschek, J., Braun, N., Franzmann, T.M., Georgalis, Y., Haslbeck, M., Weinkauf, S., and Buchner, J. (2009). The eye lens chaperone alpha-crystallin forms defined globular assemblies. Proc. Natl. Acad. Sci. USA 106, 13272– 13277. Peschek, J., Braun, N., Rohrberg, J., Back, K.C., Kriehuber, T., Kastenmuller, A., Weinkauf, S., and Buchner, J. (2013). Regulated structural transitions unleash the chaperone activity of alphaB-crystallin. Proc. Natl. Acad. Sci. USA 110, E3780–E3789.

Tyedmers, J., Mogk, A., and Bukau, B. (2010). Cellular strategies for controlling protein aggregation. Nat. Rev. Mol. Cell Biol. 11, 777–788. Usui, K., Ishii, N., Kawarabayasi, Y., and Yohda, M. (2004). Expression and biochemical characterization of two small heat shock proteins from the thermoacidophilic crenarchaeon Sulfolobus tokodaii strain 7. Protein Sci. 13, 134–144. van Montfort, R.L., Basha, E., Friedrich, K.L., Slingsby, C., and Vierling, E. (2001). Crystal structure and assembly of a eukaryotic small heat shock protein. Nat. Struct. Biol. 8, 1025–1030. Walter, P., and Ron, D. (2011). The unfolded protein response: from stress pathway to homeostatic regulation. Science 334, 1081–1086. White, H.E., Orlova, E.V., Chen, S., Wang, L., Ignatiou, A., Gowen, B., Stromer, T., Franzmann, T.M., Haslbeck, M., Buchner, J., et al. (2006). Multiple distinct assemblies reveal conformational flexibility in the small heat shock protein Hsp26. Structure 14, 1197–1204. Zhong, L., Zhou, W., Wang, H., Ding, S., Lu, Q., Wen, X., Peng, L., Zhang, L., and Lu, C. (2013). Chloroplast small heat shock protein HSP21 interacts with plastid nucleoid protein pTAC5 and is essential for chloroplast development in Arabidopsis under heat stress. Plant Cell 25, 2925– 2943.

10 Structure 23, 1–10, November 3, 2015 ª2015 Elsevier Ltd All rights reserved