Skip to main content
NCBI home page
Protein Science : A Publication of the Protein Society logoLink to Protein Science : A Publication of the Protein Society
. 2023 Jan 1;32(1):e4549. doi: 10.1002/pro.4549

Effect of evolution of the C‐terminal region on chaperone activity of Hsp70

Hong Zhang 1,2,3,✉, Huimin Hu 1,2, Si Wu 1,2, Sarah Perrett 1,2,✉
PMCID: PMC9798248  PMID: 36533311

Abstract

Dynamic interdomain interactions within the Hsp70 protein is the basis for the allosteric and functional properties of Hsp70s. While Hsp70s are generally conserved in terms of structure, allosteric behavior, and some overlapping functions, Hsp70s also contain variable sequence regions which are related to distinct functions. In the Hsp70 sequence, the part with the greatest sequence variation is the C‐terminal α‐helical lid subdomain of substrate‐binding domain (SBDα) together with the intrinsically disordered region. Dynamic interactions between the SBDα and β‐sandwich substrate‐binding subdomain (SBDβ) contribute to the chaperone functions of Hsp70s by tuning kinetics of substrate binding. To investigate how the C‐terminal region of Hsp70 has evolved from prokaryotic to eukaryotic organisms, we tested whether this region can be exchanged among different Hsp70 members to support basic chaperone functions. We found that this region from eukaryotic Hsp70 members cannot substitute for the same region in Escherichia coli DnaK to facilitate normal chaperone activity of DnaK. In contrast, this region from E. coli DnaK and Saccharomyces cerevisiae Hsp70 (Ssa1 and Ssa4) can partially support some roles of human stress inducible Hsp70 (hHsp70) and human cognate Hsp70 (hHsc70). Our results indicate that the C‐terminal region from eukaryotic Hsp70 members cannot properly support SBDα–SBDβ interactions in DnaK, but this region from DnaK/Ssa1/Ssa4 can still form some SBDα–SBDβ interactions in hHsp70 or hHsc70, which suggests that the mode for SBDα–SBDβ interactions is different in prokaryotic and eukaryotic Hsp70 members. This study provides new insight in the divergency among different Hsp70 homologs and the evolution of Hsp70s.

Keywords: ATPase activity, chaperone activity, evolution, Hsp70, substrate binding

1. INTRODUCTION

Hsp70s are vital multifunctional proteins which are involved in diverse cellular processes including protein folding and degradation, autophagy, and apoptosis (Rosenzweig et al., 2019; Roufayel & Kadry, 2019; Stricher et al., 2013). As the central hub of the protein quality control network, Hsp70s facilitate the activity of numerous clients and form associations with other chaperones, including Hsp40s and Hsp90s, and with protein degradation systems, contributing to proteostasis (Fernández‐Fernández & Valpuesta, 2018).

Hsp70 homologs have conserved structure including two major domains: an N‐terminal nucleotide‐binding domain (NBD) and a C‐terminal substrate binding domain (SBD) which are joined by a conserved linker (Figure 1a; Bertelsen et al., 2009). The NBD consists of two lobes (I and II) which can be further subdivided into four subdomains (IA, IB, IIA, and IIB) (Figure 1a; Zhang & Zuiderweg, 2004). ATP or ADP binds to the deep cleft between the two lobes and hydrolysis of ATP in the NBD is often very slow without stimulation of co‐chaperones or substrates (Zhuravleva & Gierasch, 2011). The SBD consists of a β‐sandwich SBD (SBDβ), an α‐helical lid subdomain of SBD (SBDα) including four to five α‐helices, and the C‐terminal intrinsically disordered region (C‐IDR; Figure 1a; Zhang et al., 2014; Zhu et al., 1996). The C‐terminal part of the SBDα is an α‐helical bundle and the SBDα forms a lid to cover the hydrophobic cleft for substrate binding in SBDβ (Zhang et al., 2014; Zhu et al., 1996). For eukaryotic Hsp70 members, a conserved tetratricopeptide‐repeat‐domain‐interaction EEVD motif at the C‐terminus facilitates cooperation between Hsp70 and co‐chaperones (Jiang et al., 2019; Matsumura et al., 2013).

FIGURE 1.

FIGURE 1

Open in a new tab

Conservative structure of Hsp70s and nonconservative sequence of the α‐helical subdomain of substrate‐binding domain (SBDα) and C‐terminal intrinsic disordered region (C‐IDR) in Hsp70s. (a) the canonical Hsp70 structure is represented by the NMR structure of ADP‐bound DnaK (PDB code 2KHO, left) and the crystal structure of ATP‐bound DnaK (PDB code 4JNE, right). The NBD and a small part of the linker (residues 1–383) is displayed in blue. The β‐sheet subdomain of the substrate binding domain (SBDβ) and the remaining part of the linker (residues 384–507) is displayed in dark pink. The αA of SBDα (residues 508–524), αB‐E of SBDα (residues 525–601) and a schematic drawing of C‐IDR (residues 602–637) are displayed in lime, yellow and purple, respectively. In the structure of ADP‐bound DnaK the four loops (L4,5, L1,2, L3,4, and L5,6) that are orientated towards SBDα in SBDβ and contribute to SBDα–SBDβ interactions are labeled. (b) Comparison of the SBDα–SBDβ interface in Escherichia coli DnaK (PDB code 1DKX, beige) and in human stress inducible Hsp70 (hHsp70, PDB code 4PO2, green). The side chains of the residues that participate in formation of hydrogen bonds and a salt bridge in the SBDα–SBDβ interface are displayed as sticks and the three hydrogen bonds in DnaK are labeled with numbers and circles. (c) Alignment of the SBD and C‐IDR of hHsp70, hHsc70, Ssa1, Ssa4, and DnaK. The residues that participate in formation of hydrogen bonds and the salt bridge in SBDα–SBDβ interactions in DnaK and their corresponding residues in the other Hsp70 members are indicated by dark blue boxes, and the related residues in DnaK are indicated under the boxes. The loop region between the SBDα and SBDβ (Lα,β) is indicated by a light blue box, and the yellow line is the boundary.

Both conformation and function of Hsp70 are allosterically regulated by nucleotide and substrate binding (Zuiderweg et al., 2013). ATP binding induces rotation of the two lobes in the NBD and docking between the NBD and SBD (Kityk et al., 2012; Qi et al., 2013; Figure 1a). The hydrolysis of ATP to ADP leads to undocking of the NBD and SBD (Zhuravleva et al., 2012). The SBDβ and SBDα combine to act as a clamp and permit association and dissociation of exposed hydrophobic segments in non‐native polypeptides (Buczynski et al., 2001; Schlecht et al., 2011; Zhang et al., 2014). The structure of the isolated SBDβ is highly dynamic which can be stabilized by SBDα as well as substrate binding (Pellecchia et al., 2000; Swain et al., 2006). The docking between SBDβ and SBDα involving hydrogen bonds and salt bridges between side chains in the four upward protruding loops of SBDβ and the αB helix of SBDα (Figure 1b) leads to high affinity and slow association and dissociation of substrates (Fernández‐Sáiz et al., 2006; Mayer et al., 2000; Schlecht et al., 2011). In contrast, undocking between SBDβ and SBDα results in rapid association and dissociation of substrates (Buczynski et al., 2001; Mayer et al., 2000; Slepenkov & Witt, 2002). In the ATP‐bound state, the interactions between the NBD and SBDβ cause conformational changes in SBDβ and result in undocking between SBDβ and SBDα, leading to low substrate affinity (Kityk et al., 2012; Mayer et al., 2000; Qi et al., 2013; Wu et al., 2020; Zhuravleva et al., 2012). In the ADP‐bound state, weaker interactions between the NBD and SBDβ facilitate docking between SBDβ and SBDα, resulting in high substrate affinity (Bertelsen et al., 2009; Mayer et al., 2000; Wu et al., 2020; Zhuravleva et al., 2012). In both the ATP‐bound and ADP‐bound states, the equilibrium between docked and undocked states results in conformational heterogeneity among the different domains of Hsp70, although there is a predominant conformation for each state (Wang et al., 2021; Wu et al., 2020; Zhuravleva & Gierasch, 2015). High concentration of environmental ATP facilitates ATP‐ADP exchange in the NBD and helps Hsp70 to recover to the ATP‐bound state, releasing the bound substrate (Mayer et al., 2000; Wu et al., 2020). The above processes form the functional cycle of Hsp70, and different cochaperone Hsp40s and nucleotide exchange factors (NEFs) accelerate and finely tune this cycle to assist protein folding and participate in other cellular activities (Kityk et al., 2018; Mayer, 2021; Rauch & Gestwicki, 2014; Rosenzweig et al., 2019).

In the above functional cycle of Hsp70, SBDα is thought to contribute to the following aspects: (1) SBDα modulates the kinetics of substrate binding by forming different interactions with SBDβ in different nucleotide‐bound states, which can facilitate foldase activity of Hsp70 (Fernández‐Sáiz et al., 2006; Schlecht et al., 2011). (2) SBDα interacts with substrate to enhance substrate binding, especially for folding intermediates, native protein, or protein aggregates (Gong et al., 2018; Marcinowski et al., 2011; Mashaghi et al., 2016; Schlecht et al., 2011). (3) SBDα interacts with Hsp40 or NEFs to facilitate cooperation between Hsp70 and its cochaperones (Banerjee et al., 2016; Zhang et al., 2015). (4) Truncation or post‐translational modification of SBDα leads to unfolding of SBDα, blocking of SBDβ by the unfolded SBDα and switching off the chaperone activity (Morshauser et al., 1999; Wang et al., 1998; Yang et al., 2020).

HspA1A and HspA8 are the best‐studied human cytosolic Hsp70s. Human HspA1A (hHsp70) is stress inducible. Human HspA8 (hHsc70), also termed 71 kDa heat shock cognate, is the housekeeping Hsp70. The two Hsp70s have both overlapping and distinct functions and properties, and they also have a nonoverlapping range of substrates (Daugaard et al., 2007; Ryu et al., 2020). Ssa1 and Ssa4 in Saccharomyces cerevisiae probably correspond to hHsc70 and hHsp70 in humans, while the function of DnaK in Escherichia coli amounts to the function of both of hHsc70 and hHsp70. With evolution the sequence of Hsp70s shows some variation, for example, the sequence identity between DnaK and Ssa1 is only 48%, while the identity between Ssa1 and hHsc70 is 75%. The sequence identity decreases from the N‐terminal domain to the C‐terminal domain, and the C‐IDR has the highest variability in sequence among different Hsp70 members (Figure 1c). Why this region shows more divergence than the other regions is still a mystery. To address the functional significance of the sequence variation in this region along the path of evolution we examined whether this region from different Hsp70 members can be substituted with each other and the effect on the basic chaperone function, including intrinsic and cochaperone/peptide‐promoted ATPase activity, substrate binding affinity and kinetics, and foldase activity. Based on these results we analyzed the interactions between SBDα and SBDβ from different Hsp70 members and the possible evolution of SBDα–SBDβ interactions. We found the mode for SBDα–SBDβ interactions is different in prokaryotic DnaK and in eukaryotic Hsp70 members.

2. RESULTS

2.1. Truncation or substitution of the SBDα and C‐IDR disrupts the chaperone activity of DnaK

It was previously reported that DnaK truncation mutants lacking the SBDα and C‐IDR retain allosteric behavior, although the truncation of the SBDα lid domain dramatically affects the kinetics of substrate binding and release (Buczynski et al., 2001; Slepenkov & Witt, 2002). In this study, we constructed two different C‐terminal truncation mutants of DnaK: DnaK(1–507) without the SBDα lid and C‐IDR, and DnaK(1–524) without αB‐αE helices of the SBDα lid and C‐IDR (Figure 2a). ATPase activity of DnaK(1–507) can still be promoted by co‐chaperones (DnaJ and GrpE) and peptide (Figure 2b). DnaK(1–507) also showed a comparable equilibrium dissociation constant (K D) value with wild‐type (WT) DnaK upon binding to fluorescently labeled peptide (FITC‐ALLLSAPRR peptide [FAR]; Figures 2c, S1, and Table 1). However, k off and k on of peptide binding to DnaK(1–507) in the apo state and ADP‐bound state were higher than for WT DnaK (Table 1). When compared with WT DnaK, DnaK(1–507) cannot cooperate with co‐chaperones (DnaJ and GrpE) to facilitate luciferase refolding (Figure 2d). This indicates that the SBDα and C‐IDR region of DnaK is essential for the normal foldase activity of the DnaK machinery. The αA helix of SBDα interacts with loop (L4,5) in SBDβ by hydrophobic contacts (Zhu et al., 1996). DnaK(1–524) containing αA showed similar K D and kinetics of substrate binding and release to DnaK(1–507; Figure 2c and Table 1), which indicates that αB–αE of the lid domain slows down the rate of substrate association and dissociation.

FIGURE 2.

FIGURE 2

Open in a new tab

Contribution of the α‐helical subdomain of substrate‐binding domain (SBDα) and C‐terminal intrinsic disordered region (C‐IDR) from different Hsp70 homologs to chaperone activity of DnaK. (a) Schematic diagram of DnaK, the chimeric proteins with substitution of the SBDα and C‐IDR (DA, DP, and DC) and DnaK truncation mutants (DnaK(1–524) and DnaK(1–507)) are shown. (b) The stimulatory effects of co‐chaperones DnaJ (1 μM) and GrpE (1 μM) on ATPase activity of DnaK, DA, DP, DC, and DnaK(1–507) (0.5 μM) were compared. (c) the equilibrium dissociation constants (K D) for binding of DnaK, DA, DP, DC, DnaK(1–507), and DnaK(1–524) to FITC‐ALLLSAPRR peptide in apo, ADP and ATP states were compared. The K D values for peptide binding to DnaK, DA, DP, DC, DnaK(1–507), and DnaK(1–524) are just an estimation because the highest attainable protein concentrations were not sufficient for the peptide binding to reach the plateau of the binding curve as shown in Figure S1c. (d) Effect of substitution or truncation of the SBDα and C‐IDR on luciferase refolding activity of DnaK was detected. Refolding of chemically denatured firefly luciferase by DnaK, DA, DP, DC, and DnaK(1–507) in the absence and presence of DnaJ and/or GrpE was compared. The data shown are the mean of three individual experiments and the error bars represent the standard error of the mean

TABLE 1.

Effect of substitution or truncation of the SBDα and C‐IDR on peptide substrate binding to DnaK.

K D  (μM) k off (×10−3 s−1) k on (×10−3 μM−1 s−1) a
DnaK 0.07 ± 0.01 0.78 ± 0.11 11.93 ± 1.70
DnaK ADP 0.12 ± 0.01 1.77 ± 0.35 14.45 ± 2.88
DA 0.42 ± 0.07 2.50 ± 0.40 5.90 ± 0.94
DA ADP 0.61 ± 0.09 4.15 ± 0.35 6.85 ± 0.58
DP 0.48 ± 0.05 16.45 ± 2.65 34.43 ± 5.55
DP ADP 0.59 ± 0.05 37.35 ± 4.55 63.40 ± 7.72
DC 1.07 ± 0.11 10.00 ± 5.10 9.31 ± 4.75
DC ADP 1.21 ± 0.09 24.60 ± 1.50 20.26 ± 1.24
DnaK(1–524) 0.20 ± 0.04 8.41 ± 1.71 41.41 ± 8.43
DnaK(1–524) ADP 0.19 ± 0.03 7.24 ± 1.23 38.91 ± 6.53
DnaK(1–507) 0.17 ± 0.03 3.20 ± 1.00 19.14 ± 0.60
DnaK(1–507) ADP 0.09 ± 0.02 3.20 ± 0.50 36.51 ± 5.70
Open in a new tab

Abbreviations: C‐IDR, C‐terminal intrinsically disordered region; SBD, substrate‐binding domain; SBDα, α‐helical lid subdomain of SBD.

a

Where k on values were calculated from K D and k off values.

We then tested whether the SBDα and C‐IDR from yeast or human Hsp70 can replace the SBDα and C‐IDR in DnaK and constructed chimeric proteins DA containing SBDα and C‐IDR from Ssa1, DP containing SBDα and C‐IDR from hHsp70, and DC containing SBDα and C‐IDR from hHsc70 (Figure 2a). We observed that the ATPase activity of DA, DP, and DC can also be stimulated by co‐chaperones (DnaJ and GrpE) and peptide, similar to DnaK(1–507) and WT DnaK (Figure 2b). However DA, DP, and DC had slightly lower affinity for peptide substrate than WT DnaK, DnaK(1–524) and DnaK(1–507), and in the apo state and ADP‐bound state the k off of peptide binding to DA, DP, and DC increased significantly (Figure 2c, Figure S1 and Table 1). We also noticed that the k on of peptide binding to DA, DP, and DC were different and only the k on value of DP increases significantly (Table 1). This indicates that the SBDα from other Hsp70 homologs may have different effects on k off and k on of substrate binding. Similar to C‐terminal truncation mutants, DA, DP, and DC still retained allosteric behavior upon nucleotide binding, showing lower substrate binding affinity in the ATP‐bound state than in the apo or ADP‐bound state (Figure 2c, Figure S1, and Table 1; In the presence of ATP the highest attainable protein concentrations were not sufficient for the peptide binding to reach the plateau of the binding curve.) Moreover, DA, DP, and DC were also unable to cooperate with co‐chaperones in assisting luciferase refolding (Figure 2d). Thus, the SBDα and C‐IDR from yeast or human Hsp70s cannot support corresponding roles in DnaK.

Previous studies have shown that in DnaK the first hydrogen bond between D526 in SBDα and R445 in L4,5 of SBDβ is crucial for slowing down the association and dissociation rates of substrate binding (Figure 1b,c; Fernández‐Sáiz et al., 2006; Schlecht et al., 2011). DnaK(D526A) or DnaK(D526C/R445C) under reducing conditions which lose the interaction between D526 and R445 showed dramatically increased k off and k on values for peptide binding (Fernández‐Sáiz et al., 2006; Schlecht et al., 2011). This residue pair is conserved in DnaK, Ssa1 (D526 and R444), Ssa4 (D526 and R444), hHsc70 (D529 and R447), and hHsp70 (D529 and R447; Figure 1b,c). Additionally, the second hydrogen bond between D540 in SBDα and R467 in L5,6 of SBDβ, the third hydrogen bond between H544 in SBDα and D431 in L3,4 of SBDβ, and the salt bridge between K548 in SBDα and D431 are also important for interactions between SBDα and SBDβ, and form a latch to control the opening and closing of the remote part of the SBDα lid (Figure 1b,c; Zhu et al., 1996). However, the residues participating in the formation of the latch are not fully conserved between prokaryotic Hsp70 DnaK and eukaryotic Hsp70 members (Figure 1c; Zhang et al., 2014). The hydrogen bond between E543 and R469 in eukaryotic Hsp70 members is equivalent to the hydrogen bond between D540 and R467 in DnaK (Figure 1b,c). The third hydrogen bond and the salt bridge are missing in Ssa1, Ssa4, hHsp70, and hHsc70 (Figure 1b,c). A previous study also found disruption of the second hydrogen bond or the salt bridge render DnaK unable to assist protein folding, although it only slightly affects substrate association and dissociation rates (Fernández‐Sáiz et al., 2006). Thus, from the finding that substrate dissociates more rapidly from DA, DP, and DC and luciferase refolding was not assisted by these chimeras, we can deduce that the SBDα from Ssa1, hHsp70, and hHsc70 cannot form sufficiently strong interactions with the SBDβ of DnaK in the apo or ADP‐bound states. The dramatically increased k off values for peptide binding of DP and DC is possibly due to failing to form the first hydrogen bond. Different from DP and DC, DA only showed a modest increase in k off and a decreased k on values for peptide binding (Table 1), which indicates that the SBDα from Ssa1 could form more interactions with the SBDβ of DnaK, compared with the SBDα from hHsp70 or hHsc70.

2.2. Truncation of the SBDα and C‐IDR disrupts the chaperone activity of hHsp70 and hHsc70

To investigate the contribution of the SBDα and C‐IDR to chaperone activity of human Hsp70, we constructed hHsp70(1–511) and hHsc70(1–511) without the SBDα and C‐IDR, and hHsp70(1–527) and hHsc70(1–527) without αB‐αE helices of the SBDα lid and C‐IDR (Figure 3a). Similar to the C‐terminal truncation mutants of DnaK, ATPase activity of the four truncation mutants of human Hsp70 can still be stimulated by co‐chaperones (Figure 3b). This indicates that both the NBD and SBDβ (but not the SBDα and C‐IDR) are enough for stimulation of the ATPase activity of Hsp70 by cochaperones. The docking between the NBD and SBD in the ATP‐bound state restricts movement of the NBD and the hydrolysis of ATP, while undocking between the NBD and SBD facilitates the hydrolysis of ATP by Hsp70 (Kityk et al., 2015; Zhuravleva et al., 2012) (Figure 1a). We noticed that C‐terminal truncation of hHsp70 resulted in increased intrinsic ATPase activity (Figure 3b). The intrinsic ATPase activity of hHsp70 (1–527) was higher than hHsp70 and lower than hHsp70(1–511; Figure 3b), suggesting the SBDα and C‐IDR of hHsp70 contribute to the docking between the NBD and SBD, and deletion of the SBDα and C‐IDR in hHsp70 increases the proportion of the undocked population. This is in contrast to the finding that C‐terminal truncation of hHsc70 had no obvious effect on intrinsic ATPase activity (Figure 3b).

FIGURE 3.

FIGURE 3

Open in a new tab

Effect of deletion of the α‐helical subdomain of substrate‐binding domain (SBDα) and C‐terminal intrinsic disordered region (C‐IDR) on the chaperone activity of hHsp70 and hHsc70. (a) Schematic diagram of hHsp70, hHsp70 truncation mutants (hHsp70(1–527) and hHsp70(1–511)), hHsc70 and hHsc70 truncation mutants (hHsc70(1–527) and hHsc70(1–511)) are shown. (b) The stimulatory effects of co‐chaperones Hdj1 (1 μM) and Bag1 (0.5 μM) on ATPase activity of hHsp70, hHsp70(1–527), hHsp70(1–511), hHsc70, hHsc70(1–527) and hHsc70(1–511) (1 μM) were compared. (c) The stimulatory effects of different concentrations of peptide (NRLLLTG) as indicated on ATPase activity of hHsp70, hHsp70(1–527), hHsp70(1–511), hHsc70, hHsc70(1–527) and hHsc70(1–511) (1 μM) were compared. (d) The equilibrium dissociation constants (K D) for binding of hHsp70, hHsp70(1–527), hHsp70(1–511), hHsc70, hHsc70(1–527) and hHsc70(1–511) to FITC‐ALLLSAPRR peptide in apo, ADP and ATP states were compared. The K D values for peptide binding to hHsp70, hHsp70(1–527), hHsp70(1–511), hHsc70, hHsc70(1–527) and hHsc70(1–511) are only an estimation because the highest attainable protein concentrations were not sufficient for the peptide binding to reach the plateau of the binding curve as shown in Figure S2e,f. We performed Student's t‐test to compare the K D values of WT hHsc70 and hHsc70 mutants with C‐terminal truncation in apo and ADP‐bound states. There was a statistically significant difference in the following comparisons: Apo WT hHsc70 versus apo hHsc70(1–527) (p = 0.006), apo WT hHsc70 versus apo hHsc70(1–511) (p < 0.001) and ADP‐bound hHsc70 versus ADP‐bound hHsc70(1–511) (p = 0.005). (e) Effect of substitution or truncation of the SBDα and C‐IDR on luciferase refolding activity of hHsp70 was detected. Refolding of chemically denatured firefly luciferase by hHsp70, hHsp70(1–527), hHsp70(1–511), hHsc70, hHsc70(1–527), and hHsc70(1–511) in the absence and presence of Hdj1 and/or Bag1 was compared. The data shown are the mean of three individual experiments and the error bars represent the standard error of the mean.

In terms of peptide‐promoted ATPase activity, peptide had a weaker stimulatory effect on the ATPase activity of hHsp70(1–511), hHsp70(1–527), and hHsc70(1–527) compared with WT hHsp70 and WT hHsc70 (Figure 3c). It is interesting that peptide stimulation of the ATPase activity of hHsc70(1–511) was much stronger than for WT hHsc70 (Figure 3c). We then measured peptide binding for these truncation mutants and found C‐terminal truncation of hHsp70 had no obvious effect on the K D of peptide binding (Figures S2, 3d, and Table 2), while C‐terminal truncation of hHsc70 increased the affinity of peptide binding (Figures S2, 3d, and Table 2), which is consistent with the recent published work regarding C‐terminal truncation of hHsc70 (Taylor et al., 2018). Similar to C‐terminal truncation of DnaK, C‐terminal truncation of hHsp70 or hHsc70 increased the k off and k on of peptide binding in the apo and ADP‐bound states (Table 2). Compared with C‐terminal truncation of hHsp70, C‐terminal truncation of hHsc70 had a very obvious promotion effect on the k on of peptide binding, while the effect of C‐terminal truncation on k off was similar for hHsp70 and hHsc70 (Table 2). In assisting luciferase refolding, C‐terminal truncation of hHsp70 or hHsc70 resulted in de‐coordination between the Hsp70 protein and cochaperones (Figure 3e), which is similar to C‐terminal truncation of DnaK. This suggests that the role of the SBDα and C‐IDR in slowing down the association and dissociation rate of substrates and in cooperating with cochaperones to assist protein folding may be conserved from prokaryotes to eukaryotes.

TABLE 2.

Effect of truncation of the SBDα and C‐IDR on peptide substrate binding to hHsp70 and hHsc70.

K D  (μM) k off (×10−3 s−1) k on (×10−3 μM−1 s−1) a
hHsp70 1.09 ± 0.07 0.28 ± 0.01 0.25 ± 0.01
hHsp70 ADP 0.96 ± 0.05 0.29 ± 0.01 0.30 ± 0.01
hHsp70(1–527) 1.75 ± 0.23 0.79 ± 0.16 0.45 ± 0.09
hHsp70(1–527) ADP 1.53 ± 0.26 0.96 ± 0.36 0.63 ± 0.23
hHsp70(1–511) 1.65 ± 0.09 0.72 ± 0.13 0.44 ± 0.08
hHsp70(1–511) ADP 1.40 ± 0.07 0.67 ± 0.22 0.48 ± 0.16
Hsc70 1.08 ± 0.08 0.20 ± 0.02 0.18 ± 0.02
Hsc70 ADP 0.59 ± 0.04 0.29 ± 0.02 0.50 ± 0.03
hHsc70(1–527) 0.36 ± 0.11 0.86 ± 0.16 2.37 ± 0.44
hHsc70(1–527) ADP 0.48 ± 0.06 0.98 ± 0.29 2.02 ± 0.60
hHsc70(1–511) 0.22 ± 0.02 1.10 ± 0.23 4.99 ± 1.06
hHsc70(1–511) ADP 0.35 ± 0.02 0.91 ± 0.16 2.65 ± 0.47
Open in a new tab

Abbreviations: C‐IDR, C‐terminal intrinsically disordered region; SBD, substrate‐binding domain; SBDα, α‐helical lid subdomain of SBD.

a

Where k on values were calculated from K D and k off values.

2.3. Exchange of the SBDα and C‐IDR between hHsp70 and hHsc70 does not affect their chaperone activity

hHsp70 and hHsc70 have many distinct functions and clients although they share 86% identity (Ryu et al., 2020). Sequence alignment shows that the most distinct sequence between the two different Hsp70 homologs is the SBDα and C‐IDR region (Figure 1c). We exchanged the SBDα and C‐IDR between hHsp70 and hHsc70 and constructed chimeric proteins PC (in which the NBD and SBDβ are from hHsp70 and the SBDα and C‐IDR are from hHsc70) and CP (in which the NBD and SBDβ are from hHsc70 and the SBDα and C‐IDR are from hHsp70) as indicated in Figure 4a. The stimulation of ATPase activity of hHsp70, hHsc70, and the chimeric proteins PC and CP by co‐chaperones (Hdj1 and Bag1) were similar (Figure 4b). We then measured the peptide stimulated ATPase activity and it was stronger for hHsp70 than hHsc70 (Figure 4c). It is interesting that the stimulation of ATPase activity from peptide is not dependent on the exchange of SBDα (behavior of PC is similar to hHsp70, and behavior of CP is also similar to hHsc70; Figure 4c).

FIGURE 4.

FIGURE 4

Open in a new tab

Effect of exchanging the α‐helical subdomain of substrate‐binding domain (SBDα) and C‐terminal intrinsic disordered region (C‐IDR) between hHsp70 and hHsc70 on their chaperone activity. (a) Schematic diagram of hHsp70, hHsc70, and the chimeric proteins with exchange of the SBDα and C‐IDR region (PC and CP) are shown. (b) The stimulatory effects of co‐chaperones Hdj1 (1 μM) and Bag1 (0.5 μM) on ATPase activity of hHsp70, PC, hHsc70, and CP (1 μM) were compared. (c) The stimulatory effects of different concentrations of peptide (NRLLLTG) as indicated on ATPase activity of hHsp70, PC, hHsc70, and CP (1 μM) were compared. (d) The equilibrium dissociation constants (K D) for binding of hHsp70, PC, hHsc70, and CP to FITC‐ALLLSAPRR peptide in apo, ADP and ATP states were compared. The K D values for peptide binding to hHsp70, PC, hHsc70, and CP are only an estimation because the highest attainable protein concentrations were not sufficient for the peptide binding to reach the plateau of the binding curve as shown in Figure S2e,f. (E) Kinetics of 20 nM peptide binding to 2 μM hHsp70, PC, hHsc70, and CP in the apo state were compared. (f) Kinetics of 20 nM peptide binding to 2 μM SBD truncation mutants from hHsp70, PC, hHsc70, and CP were compared. (g) Kinetics of 20 nM peptide binding to 2 μM hHsp70, PC, hHsc70, and CP in the ADP‐bound state were compared. (h) Effect of exchanging the SBDα and C‐IDR between hHsp70 and hHsc70 on luciferase refolding activity of hHsp70 and hHsc70 was detected. Refolding of chemically denatured firefly luciferase by hHsp70, PC, hHsc70, and CP in the absence and presence of Hdj1 and/or Bag1 was compared. The data shown are the mean of three individual experiments and the error bars represent the standard error of the mean.

We then compared substrate binding to these proteins and found they had similar values of K D (Figures 4D, S2, and Table 3). When the kinetics of peptide binding were compared, we found they also fell into two distinct groups. In the apo state, peptide binding to hHsp70 or PC was much more rapid than peptide binding to hHsc70 or CP (Figure 4e). A similar phenomenon was observed when measuring the peptide binding to the SBD of these proteins (Figure 4f). However, in the ADP‐bound state the kinetics of peptide binding to these four proteins was similar (Figure 4g). Consistent with this we found that the k off of peptide binding to hHsp70, PC, hHsc70, and CP were similar in the ADP‐bound state but distinct in the apo state (Table 3). In the apo state the k off and k on of peptide binding to hHsp70 or PC was slightly higher than for hHsc70 or CP (Table 3). The k off of peptide binding to all four ADP‐bound proteins and the apo‐state of WT hHsp70 and PC is 0.3 × 10−3 s−1, while it is 0.2 × 10−3 s−1 for the apo‐state of hHsc70 and CP (Table 3). For SBD constructs, hHsp70 SBD and PC SBD also had higher k off values than hHsc70 SBD and CP SBD although the four SBD constructs had similar values of K D (Table 3 and Figure S2g). This indicates that the different substrate binding kinetics in the apo state between hHsp70 and hHsc70 is decided by SBDβ, not SBDα. Under some extreme conditions including severe oxidative stress, ATP becomes depleted and the fraction of apo‐state Hsp70 increases (Winter et al., 2005). Under such conditions hHsp70 may largely retain its holdase activity and may bind to substrates with higher efficiency than hHsc70.

TABLE 3.

Effect of exchanging the SBDα and C‐IDR on peptide substrate binding to hHsp70 and hHsc70.

K D  (μM) k off (* 10−3 s−1) k on (* 10−3 μM−1 s−1) a
hHsp70 1.09 ± 0.07 0.28 ± 0.01 0.25 ± 0.01
hHsp70 ADP 0.96 ± 0.05 0.29 ± 0.01 0.30 ± 0.01
PC 0.93 ± 0.05 0.27 ± 0.02 0.29 ± 0.02
PC ADP 0.96 ± 0.05 0.30 ± 0.01 0.31 ± 0.01
hHsc70 1.08 ± 0.08 0.20 ± 0.02 0.18 ± 0.02
hHsc70 ADP 0.59 ± 0.04 0.29 ± 0.02 0.50 ± 0.03
CP 1.60 ± 0.13 0.19 ± 0.02 0.12 ± 0.01
CP ADP 1.30 ± 0.08 0.29 ± 0.01 0.22 ± 0.01
hHsp70 SBD 0.87 ± 0.08 1.25 ± 0.07 1.44 ± 0.08
PC SBD 0.59 ± 0.03 1.51 ± 0.13 2.56 ± 0.12
hHsc70 SBD 0.68 ± 0.02 0.32 ± 0.02 0.47 ± 0.02
CP SBD 0.65 ± 0.03 0.32 ± 0.01 0.49 ± 0.01
Open in a new tab

Abbreviations: C‐IDR, C‐terminal intrinsically disordered region; SBD, substrate‐binding domain; SBDα, α‐helical lid subdomain of SBD.

a

Where k on values were calculated from K D and k off values.

In assisting luciferase refolding, WT hHsp70, PC, WT hHsc70, and CP can cooperate with co‐chaperones (Hdj1 and Bag1) to nearly the same extent (Figure 4h). In conclusion, the SBDα and C‐IDR from hHsp70 and hHsc70 can substitute for each other to support the NBD and SBDβ to perform basic chaperone activity, and will not change the original characteristics of substrate binding for the parent protein. It also indicates that the modes for SBDα–SBDβ interaction in hHsp70 and hHsc70 are almost the same and the residue differences between hHsp70 and hHsc70 do not have a major effect on SBDα–SBDβ interactions.

2.4. The SBDα and C‐IDR from E. coli or yeast Hsp70 can replace the same region in human Hsp70 to partially retain some functions

We then investigated whether the SBDα and C‐IDR from E. coli or yeast Hsp70 can be substitution for the SBDα and C‐IDR from human Hsp70 by constructing chimeric proteins PDnaK, PSsa1, PSsa4, CDnaK, CSsa1, and CSsa4 as indicated in Figure 5a. As expected the ATPase activity of PDnaK, PSsa1, and PSsa4 can still be promoted by cochaperones Hdj1 and Bag1 (Figure 5b). When we measured the peptide stimulated ATPase activity, we found the behavior of PSsa1 and PSsa4 were more similar to WT hHsp70 than hHsp70(1–527) or hHsp70(1–511), and the behavior of PDnaK was intermediate between WT hHsp70 and hHsp70 truncation mutants (Figure 5c). Next, we measured their affinity and kinetics of peptide binding and found PDnaK, PSsa1, and PSsa4 behaved like WT hHsp70 but not hHsp70 truncation mutants (Figure 5d, Tables 2 and 4). The affinity of peptide binding to PDnaK, PSsa1, and PSsa4 was slightly increased compared with WT hHsp70 (Table 4). We noticed that the k off of peptide binding to PDnaK, PSsa1 and PSsa4 in the apo or ADP‐bound states was almost the same as WT hHsp70 and their slightly decreased K D of peptide binding resulted from the increased k on for peptide binding (Table 4). So, the SBDα and C‐IDR from E. coli and yeast Hsp70 can slow down the dissociation of substrate from hHsp70 SBDβ but not the association process, and they can partially compensate for C‐terminal truncation of hHsp70. We speculate that SBDα from DnaK, Ssa1, and Ssa4 can still interact with SBDβ from hHsp70 in apo or ADP‐bound states at least by forming the first hydrogen bond which is essential for slowing down the dissociation rate of substrate.

FIGURE 5.

FIGURE 5

Open in a new tab

Contribution of the α‐helical subdomain of substrate‐binding domain (SBDα) and C‐terminal intrinsic disordered region (C‐IDR) region from different Hsp70 homologs to chaperone activity of hHsp70 and hHsc70. (a) Schematic diagram of hHsp70, hHsc70, and the chimeric proteins with substitution of the SBDα and C‐IDR (PDnaK, PSsa1, PSsa4, CDnaK, CSsa1, and CSsa4) are shown. (b) The stimulatory effects of co‐chaperones Hdj1 (1 μM) and Bag1 (0.5 μM) on ATPase activity of hHsp70, PC, PDnaK, PSsa1, PSsa4, hHsp70(1–527), hHsp70(1–511), and hHsp70 NBD (1 μM) were compared. (c) The stimulatory effects of different concentrations of peptide (NRLLLTG) as indicated on ATPase activity of hHsp70, PDnaK, PSsa1, Pssa4, hHsp70(1–527), and hHsp70(1–511) (1 μM) were compared. (d) the equilibrium dissociation constants (K D) for binding of hHsp70, PDnaK, PSsa1, PSsa4, hHsp70(1–527), and hHsp70(1–511) to FITC‐ALLLSAPRR peptide in apo, ADP and ATP states were compared. The K D values for peptide binding to hHsp70, PDnaK, PSsa1, PSsa4, hHsp70(1–527), and hHsp70(1–511) are just an estimation because the highest attainable protein concentrations were not sufficient for the peptide binding to reach the plateau of the binding curve as shown in Figure S2e. (e) Effect of substitution of the SBDα and C‐IDR, adding a C‐terminal EEVD motif into PDnaK or mutation mimicking phosphorylation on luciferase refolding activity of hHsp70 was detected. Refolding of chemically denatured firefly luciferase by hHsp70, PSsa1, PSsa4, PDnaK, PDnaK(EEVD), hHsp70(T636E), hHsp70(1–527), and hHsp70(1–511) in the absence and presence of Hdj1 and Bag1 was compared. (f) Effect of substitution of the SBDα and C‐IDR, adding a C‐terminal EEVD motif into CDnaK or mutation mimicking phosphorylation on luciferase refolding activity of hHsc70 was detected. Refolding of chemically denatured firefly luciferase by hHsc70, CSsa1, Cssa4, CDnaK, CDnaK(EEVD), hHsc70(T641E), hHsc70(1–527), and hHsc70(1–511) in the absence and presence of Hdj1 and Bag1 was compared. The data shown are the mean of three individual experiments and the error bars represent the standard error of the mean.

TABLE 4.

Effect of substitution of the SBDα and C‐IDR from Escherichia coli or yeast Hsp70 on peptide substrate binding ability of hHsp70

K D  (μM) k off (* 10−3 s−1) k on (* 10−3 μM−1 s−1) a
hHsp70 1.09 ± 0.07 0.28 ± 0.01 0.25 ± 0.01
hHsp70 ADP 0.96 ± 0.05 0.29 ± 0.01 0.30 ± 0.01
PDnaK 0.54 ± 0.02 0.29 ± 0.02 0.53 ± 0.03
PDnaK ADP 0.62 ± 0.01 0.31 ± 0.01 0.50 ± 0.01
PSsa1 0.35 ± 0.02 0.30 ± 0.01 0.86 ± 0.04
PSsa1 ADP 0.31 ± 0.01 0.32 ± 0.02 1.02 ± 0.05
PSsa4 0.50 ± 0.03 0.28 ± 0.01 0.56 ± 0.01
PSsa4 ADP 0.58 ± 0.02 0.31 ± 0.01 0.54 ± 0.01
Open in a new tab

Abbreviations: C‐IDR, C‐terminal intrinsically disordered region; SBD, substrate‐binding domain; SBDα, α‐helical lid subdomain of SBD.

a

Where k on values were calculated from K D and k off values.

In term of assisting luciferase refolding, we found PSsa1 and PSsa4 but not PDnaK can work with the same efficiency as WT hHsp70 (Figure 5e). Considering that the C‐IDR of DnaK lacks the EEVD motif which is essential for interaction with eukaryotic Hsp40 in the foldase activity of eukaryotic Hsp70, we mutated the last eight residues in PDnaK from “FEEVKDKK” to “GPTIEEVD” (Figure 1c) and generated a new mutant PDnaK(EEVD). However, PDnaK(EEVD) acted like PDnaK in assisting luciferase refolding and still lacked ability to cooperate with Hdj1 and Bag1 (Figure 5e). For hHsc70 we found the consequences of SBDα and C‐IDR substitution is very similar to hHsp70: CSsa1 and Cssa4 can cooperate with Hdj1 and Bag1 to assist luciferase refolding like WT hHsc70, but not CDnaK(EEVD) or CDnaK (i.e., with or without the EEVD motif; Figure 5f). Therefore, besides the EEVD motif the other parts of SBDα and C‐IDR are also important for interaction between human Hsp70 and cochaperones Hdj1 and Bag1 to assist protein folding. There are multiple phosphorylation sites in the SBDα and C‐IDR region and it is reported that T636 in hHsp70 and T641 in hHsc70 which are near the EEVD motif can affect interaction between Hsp70 and cochaperones (Muller et al., 2013). We constructed hHsp70(T636E) and hHsc70(T641E) to mimic phosphorylation. However, we found the mutations only slightly decreased the foldase activity and this indicates that phosphorylation adjacent to the EEVD motif in hHsp70 and hHsc70 only has a limited effect on interaction between Hdj1 and the EEVD motif (Figure 5e,f).

3. DISCUSSION

Dynamic changes of domain–domain interactions caused by nucleotide and/or substrate binding are the foundation for allostery of Hsp70 (Bertelsen et al., 2009; Kityk et al., 2012; Kityk et al., 2015; Zhuravleva et al., 2012). Although Hsp70 truncation mutants without SBDα still retain some allosteric effects and holdase activity, SBDα was found to contribute to controlling the kinetics of substrate binding and efficient foldase activity of Hsp70 by its interactions with SBDβ (Buczynski et al., 2001; Schlecht et al., 2011; Taylor et al., 2018; Xu et al., 2013). To explore the evolution of the C‐terminal region of Hsp70, we tested whether the C‐terminal region (SBDα and C‐IDR) can be exchanged between different Hsp70 homologs and still preserve chaperone activity in chimeras. It is interesting to find that the C‐terminal region from eukaryotic Hsp70 members cannot execute functions in E. coli DnaK, but the C‐terminal region from DnaK/Ssa1/Ssa4 can execute some functions in human cytosolic Hsp70. Our results indicate that SBDβ from prokaryotic Hsp70 cannot properly interact with SBDα from eukaryotic Hsp70 members. In contrast, SBDβ from eukaryotic Hsp70 members can still form efficient interactions with SBDα from prokaryotic Hsp70 to modulate kinetics of substrate binding, and SBDβ from a eukaryotic Hsp70 member can form sufficient interactions with SBDα from another eukaryotic Hsp70 member to support foldase activity. It is clear that the mode of interaction of SBDα‐SBDβ has changed significantly between prokaryotes and eukaryotes. When the crystal structure of hHsp70 SBD with NR peptide bound was compared with that of DnaK SBD also with NR peptide bound, it was found that DnaK shows a more extensive network of interactions between SBDα and SBDβ than hHsp70 (Zhang et al., 2014). This network of interactions between the N‐terminal end of the αB helix of SBDα and the inner loops of SBDβ serve to rigidly link this region of αB as well as the preceding αA to SBDβ in different Hsp70 homologs, but the C‐terminal part of αB and associated helix bundle region of SBDα show more divergence in interaction with SBDβ in hHsp70 and DnaK (Zhang et al., 2014). It seems that stronger interactions at the C‐terminal part of αB and outer loops of SBDβ are needed for DnaK compared with hHsp70 in order to maintain sufficient interaction between SBDα and SBDβ for chaperone activity, for example, hHsp70 lacks the hydrogen bond between H544 and D431 and the salt bridge between K548 and D431 in DnaK (Figure 1b,c).

Comparison of the structure of DnaK SBD with NR peptide binding in Types 1 and 2 crystals reveals that the interactions between the C‐terminal part of αB and the outer loops (L3,4 and L5,6) are weaker in Type 2 crystals than in Type 1 crystals although the interactions between the N‐terminal part of αB and the inner loops (L4,5 and L1,2) are comparable in both crystal types (Zhu et al., 1996). Consistent with this, FRET study for DnaK found that in the ADP‐only state about 50% of DnaK molecules were in the lid‐open state (Banerjee et al., 2016). Binding of peptide substrates to ADP‐bound DnaK promoted closing of the SBDα lid (60%–70% lid‐closed) and binding of native protein substrates boosted opening of the SBDα lid (80%–90% lid‐open; Banerjee et al., 2016). FRET study of the ER‐located Hsp70 homolog BiP revealed that in the presence of the peptide and ADP, as in the ADP‐only state, about 60% of BiP was in the lid‐open state, and upon binding of protein substrate about 74% BiP was in the lid‐open state (Marcinowski et al., 2011). FRET study of hHsp70 found that in the ADP‐only state only a small fraction of hHsp70 (about 15%) was in the lid‐open state (Wu et al., 2020). The equilibrium of opening and closing of the SBDα lid is primarily decided by the interactions between SBDα and SBDβ, and affects substrate binding and release. The change in k on and k off for substrate binding can be used to evaluate the change in the interactions between SBDα and SBDβ. In this study, k on and k off values of C‐terminal truncation mutants of Hsp70s were higher than WT Hsp70 members, and the value of DnaK chimeras with the C‐terminal region from Ssa1, hHsp70, or hHsc70 were also higher than for WT DnaK, reflecting reduced interaction between SBDα and SBDβ. Although ATPase activity of these mutants can still be promoted by cochaperones, luciferase refolding assisted by Hsp70 proteins is not promoted by cochaperones. This suggests that both the kinetics of substrate binding to Hsp70 controlled by SBDα–SBDβ interaction and the cooperation between Hsp70 and Hsp40 mediated by the C‐terminal region of Hsp70 are important for promoting luciferase refolding with high efficiency. Although the chimera PDnaK had comparable k on and k off values with WT hHsp70, PSsa1, PSsa4, and PC (Table 4), it still cannot efficiently promote luciferase refolding by cooperation with human cochaperone homologs (Figure 5e), indicating that the C‐terminal region from DnaK cannot properly interact with human cochaperones. So the evolution of SBDα and C‐IDR appears to have been driven not only by evolution of SBDβ to form proper SBDα‐SBDβ interactions, but also by evolution of cochaperones to allow expanded roles of Hsp70.

In this study we found that for binding of the same peptide substrate, the k on and k off values for hHsp70, hHsc70, or their truncation mutants were smaller than for DnaK or its truncation mutants, and the k on and k off values for PDnaK which contains SBDβ from hHsp70 and SBDα from DnaK were similar to hHsp70 but not DnaK, suggesting that k on and k off values are also affected by SBDβ. When the SBDα and C‐IDR was exchanged between hHsp70 and hHsc70, or the C‐terminal region of hHsp70 was replaced by the C‐terminal region from DnaK, Ssa1, or Ssa4, the characteristics for substrate binding were still primarily decided by the SBDβ part and the SBDα from the other Hsp70 members cooperated with SBDβ to help it work well. From this it suggests that DnaK has a more stringent requirement for SBDα to cooperate with its SBDβ than in eukaryotic Hsp70 homologs, thus the SBDα from eukaryotic Hsp70 homologs cannot replace the role of SBDα in DnaK while the reverse can be accommodated. The dependence of SBDβ on SBDα probably decreases with evolution so that the further evolved SBDα and C‐IDR can satisfy broad interactions with other parts of Hsp70 as well as other proteins including evolved substrates and an expanded range of cochaperones for diverse functions of Hsp70s, since Hsp70s could coevolve together with their substrates and cochaperones (Rebeaud et al., 2021). As SBDα and C‐IDR often interact with folding intermediates, native protein and protein aggregates (Marcinowski et al., 2011; Mashaghi et al., 2016; Schlecht et al., 2011; Smock et al., 2011), the relationship between the variation in the SBDα and C‐IDR and the evolution of substrates of Hsp70s still needs to be interpreted. Further in vivo study will also be needed to clarify how the evolution of the SBDα and C‐IDR facilitates Hsp70 to adapt to changes in the cellular environment.

4. EXPERIMENTAL PROCEDURES

4.1. Protein expression and purification

The dnaK, dnaJ, and grpE genes were cloned from the genomic DNA of E. coli C41 cells. The SSA1 gene was provided by Prof. Susan Lindquist (available from Adgene as plasmid 1231) and the SSA4 genes was cloned from genomic DNA of S. cerevisiae cells. Human HSPA1A gene (UniProtKB code: P0DMV8), HSPA8 gene (UniProtKB code: P11142), and HDJ1 gene (UniProtKB code: P25685) were kindly provided by Prof. Richard Morimoto (Northwestern University). The human BAG1S (referred to as BAG1) gene was cloned from cDNA of HEK293 cells. HDJ1, dnaJ, and grpE genes were ligated into the mini pRSETa expression plasmid for expression of protein with a noncleavable His6 tag. HSPA1A, HSPA8, SSA1, SSA4, dnaK, and BAG1 genes were subcloned into the pET‐28a‐His6‐SMT3 vector for expression of protein with a cleavable His6‐SMT3 tag (Mossessova & Lima, 2000). All plasmids for expression of DnaK, hHsp70, and hHsc70 mutants mentioned in this study (Figures 2a, 3a, 4a, and 5a,e,f) were derived from pET‐28a‐His6‐SMT3‐dnaK, pET‐28a‐His6‐SMT3‐HSPA1A, pET‐28a‐His6‐SMT3‐HSPA8, pET‐28a‐His6‐SMT3‐SSA1, and pET‐28a‐His6‐SMT3‐SSA4 plasmids through an In‐Fusion cloning method using a commercial kit (NEB).

Expression and purification of DnaK, hHsp70, hHsc70, their mutants and Bag1 without His6 tag was performed as described previously (Wu et al., 2020; Yang et al., 2020; Zhang et al., 2016). Expression and purification of DnaJ, GrpE, and Hdj1 with His6 tag was performed as described previously (Zhang et al., 2016). All protein concentrations are given in terms of monomer and were determined using a bicinchoninic acid assay kit (Pierce). The purity of all of the purified proteins was verified by 10% sodium dodecyl sulfate‐polyacrylamide gel electrophoresis (SDS‐PAGE) as shown in Figure S3.

4.2. ATPase assay (malachite green)

Colorimetric determination of inorganic phosphate produced by ATP hydrolysis was performed using the malachite green reagent, prepared as described (Chang et al., 2008; Zhang et al., 2009). ATP hydrolysis was measured in a 20‐μl volume of Buffer B (50 mM Tris‐HCl buffer, pH 7.5, containing 100 mM KCl and 5 mM MgCl2) with 1 mM ATP in the presence or absence of chaperones in a 96‐well plate. The plate was incubated for 4 h at 37 °C. An 80‐μl volume of malachite green and 10 μl of 34% sodium citrate was added sequentially. The samples were mixed thoroughly and incubated at 37°C for 30 min before measuring the OD620 on a SpectraMax M3e plate reader (Molecular Devices, USA). The rate of chaperone‐catalyzed ATP hydrolysis was deduced by subtracting the signal from the ATP hydrolysis in the absence of chaperone.

In the assay of DnaK and its mutants, if DnaJ, GrpE, or peptide was added, their final concentrations were 0.5 μM for DnaK and its mutants, 1 μM for DnaJ, 1 μM for GrpE, and 100 μM for NRLLLTG peptide. In the assay of hHsp70, hHsc70, and their mutants, if Hdj1, Bag1, or peptide was added, their final concentrations were 1 μM for hHsp70, hHsc70 and their mutants, 1 μM for Hdj1, 0.5 μM for Bag1, and 0–2000 μM for NRLLLTG peptide.

4.3. Peptide binding assay

Peptide binding assays based on fluorescence polarization (FP) were performed as described previously (Ricci & Williams, 2008; Yang et al., 2020). Steady‐state FP measurements were performed at room temperature (RT) with 30–300 min incubation in Buffer B containing 0.01% Tween 20 to give the equilibrium dissociation constant (K D). In the presence of ATP, the incubation time was 30 min. In the absence of nucleotide or in the presence of ADP, the incubation time for DnaK or its mutants, hHsp70 or its mutants, and hHsc70 or its mutants was 60, 120, and 300 min, respectively. Binding was assessed by incubating increasing concentrations of Hsp70 proteins with a fixed concentration (20 nM) of fluorescently labeled substrate (FAR) in the absence or presence of 1 mM ADP or ATP and FP values were measured. FP measurements were performed on a Fluostar microplate reader (BMG Labtech) using the FP filter set (emission at 485 nm and excitation at 520 nm). FP values are expressed in millipolarization units. All statistical analyses were performed with Origin 9 software. Binding data were analyzed using nonlinear regression analysis (single site binding model) to get the value of K D. Kinetic FP measurements were performed by monitoring the time course of peptide binding at RT. After rapid mixing of 20 nM FAR and 1 or 2 μM Hsp70 proteins in the absence or in the presence of 1 mM ADP or ATP, FP was recorded against time.

The dissociation rate constants (k off) for peptides were determined by preincubating an equimolar mixture (0.5, 1, or 2 μM) of FAR and Hsp70 (0.5 or 1 μM was used for DnaK and its mutants, while 1 or 2 μM was used for hHsp70 / hHsc70 and their mutants) in the absence or presence of 1 mM ADP for 60–300 min for each of the different Hsp70 proteins as above. After addition of unlabeled ALLLSAPRR (55, 110, or 220 μM, to guarantee the concentration of unlabeled peptide is at least 100 times the concentration of labeled peptide) and rapid mixing, FP was recorded against time. The dissociation curves were fitted using a single‐exponential decay equation (f = y0 + a × exp[−b × x]) where b is the value of k off. The association rate constants (k on) were calculated by dividing k off by K D.

4.4. Luciferase refolding assay

Hsp70‐assisted luciferase refolding assays were performed as described previously (Wisén & Gestwicki, 2008) with slight modifications. The refolding of guanidine hydrochloride (GuHCl) denatured firefly luciferase (Promega) was performed in Buffer B containing 2 mM ATP and 2.2 mM DTT at 37 °C in the presence or absence of chaperones. DnaK or its mutants, DnaJ, and GrpE were added into the refolding system at final concentrations of 800, 160, and 400 nM respectively. hHsp70 or its mutants, Hdj1, and Bag1 were added into the refolding system at final concentrations of 2.0, 1.0, and 0.2 μM respectively. hHsc70 or its mutants, Hdj1, and Bag1 were added into the refolding system at final concentrations of 1.0, 0.5, and 0.5 μM respectively. The endpoints of luciferase refolding were measured after incubation for 2 h at 37 °C. Each reaction was performed in triplicate, and 5 μl of the refolding mixture was removed and added to a white flat‐bottomed 96‐well plate (JET Biofil) that was preloaded with 10 μl of SteadyGlo (Promega). After mixing, the luminescence was measured on a SpectraMax M3e multimode plate reader (Molecular Devices, USA) using a 500‐ms integration time.

AUTHOR CONTRIBUTIONS

Hong Zhang: Conceptualization (lead); data curation (lead); formal analysis (lead); funding acquisition (equal); investigation (lead); methodology (lead); project administration (lead); validation (equal); visualization (lead); writing – original draft (lead); writing – review & editing (equal). Huimin Hu: Data curation (supporting); investigation (supporting); validation (supporting). Si Wu: Funding acquisition (equal); validation (equal); writing – review and editing (equal). Sarah Perrett: Conceptualization (equal); funding acquisition (equal); project administration (equal); supervision (lead); writing – review & editing (lead).

FUNDING INFORMATION

This work was supported by Chinese Ministry of Science and Technology (2017YFA0504000), National Natural Science Foundation of China (31770829, 31920103011, and 32171443), the National Laboratory of Biomacromolecules, and the CAS Center of Excellence in Biomacromolecules.

CONFLICT OF INTEREST

The authors declare there is no conflict of interest.

Supporting information

FIGURE S1. The binding of FITC‐labeled ALLLSAPRR peptide to different concentrations of DnaK or its mutants. Fluorescence polarization at 520 nm after excitation at 485 nm was used to monitor the binding of 20 nM FITC‐labeled ALLLSAPRR peptide to different concentrations of DnaK or its mutants, in the absence of nucleotide (a), or in the presence of 1 mM ADP (b) or ATP (c), as indicated. The data shown are the mean of three individual experiments and the error bars represent the standard error of the mean and were fitted as described in Section 4.

FIGURE S2. The binding of FITC‐labeled ALLLSAPRR peptide to different concentrations of hHsp70, hHsc70, or their mutants. Fluorescence polarization at 520 nm after excitation at 485 nm was used to monitor the binding of 20 nM FITC‐labeled ALLLSAPRR peptide to different concentrations of hHsp70, hHsc70 or their mutants, in the absence of nucleotide (a,b,g), or in the presence of 1 mM ADP (c,d) or ATP (e,f), as indicated. The data shown are the mean of three individual experiments and the error bars represent the standard error of the mean and were fitted as described in Section 4.

FIGURE S3. SDS‐PAGE (10%) of all the proteins used in this study with molecular mass standards as indicated (a–d).

Click here for additional data file. (1.7MB, pdf)

ACKNOWLEDGMENT

We thank Dr. Weibin Gong for helpful discussion and assistance in drawing figures related to the structure of Hsp70s.

Zhang H, Hu H, Wu S, Perrett S. Effect of evolution of the C‐terminal region on chaperone activity of Hsp70. Protein Science. 2023;32(1):e4549. 10.1002/pro.4549

Review Editor: John Kuriyan

Funding information Ministry of Science and Technology of the People's Republic of China, Grant/Award Number: 2017YFA0504000; National Natural Science Foundation of China, Grant/Award Numbers: 31770829, 31920103011, 32171443; National Laboratory of Biomacromolecules; CAS Center of Excellence in Biomacromolecules

Contributor Information

Hong Zhang, Email: zhangh@moon.ibp.ac.cn.

Sarah Perrett, Email: sarah.perrett@cantab.net.

DATA AVAILABILITY STATEMENT

All data are contained within the manuscript.

REFERENCES

  1. Banerjee R, Jayaraj GG, Peter JJ, Kumar V, Mapa K. Monitoring conformational heterogeneity of the lid of DnaK substrate‐binding domain during its chaperone cycle. FEBS J. 2016;283(15):2853–68. [DOI] [PubMed] [Google Scholar]
  2. Bertelsen EB, Chang L, Gestwicki JE, Zuiderweg ER. Solution conformation of wild‐type E. coli Hsp70 (DnaK) chaperone complexed with ADP and substrate. Proc Natl Acad Sci U S A. 2009;106(21):8471–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Buczynski G, Slepenkov SV, Sehorn MG, Witt SN. Characterization of a lidless form of the molecular chaperone DnaK: deletion of the lid increases peptide on‐ and off‐rate constants. J Biol Chem. 2001;276(29):27231–6. [DOI] [PubMed] [Google Scholar]
  4. Chang L, Bertelsen EB, Wisen S, Larsen EM, Zuiderweg ER, Gestwicki JE. High‐throughput screen for small molecules that modulate the ATPase activity of the molecular chaperone DnaK. Anal Biochem. 2008;372(2):167–76. [DOI] [PubMed] [Google Scholar]
  5. Daugaard M, Rohde M, Jäättelä M. The heat shock protein 70 family: highly homologous proteins with overlapping and distinct functions. FEBS Lett. 2007;581(19):3702–10. [DOI] [PubMed] [Google Scholar]
  6. Fernández‐Fernández MR, Valpuesta JM. Hsp70 chaperone: a master player in protein homeostasis. F1000Res. 2018;7:7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Fernández‐Sáiz V, Moro F, Arizmendi JM, Acebrón SP, Muga A. Ionic contacts at DnaK substrate binding domain involved in the allosteric regulation of lid dynamics. J Biol Chem. 2006;281(11):7479–88. [DOI] [PubMed] [Google Scholar]
  8. Gong W, Hu W, Xu L, Wu H, Wu S, Zhang H, et al. The C‐terminal GGAP motif of Hsp70 mediates substrate recognition and stress response in yeast. J Biol Chem. 2018;293(46):17663–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Jiang Y, Rossi P, Kalodimos CG. Structural basis for client recognition and activity of Hsp40 chaperones. Science. 2019;365(6459):1313–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Kityk R, Kopp J, Mayer MP. Molecular mechanism of J‐domain‐triggered ATP hydrolysis by Hsp70 chaperones. Mol Cell. 2018;69(2):227–237.e4. [DOI] [PubMed] [Google Scholar]
  11. Kityk R, Kopp J, Sinning I, Mayer MP. Structure and dynamics of the ATP‐bound open conformation of Hsp70 chaperones. Mol Cell. 2012;48(6):863–74. [DOI] [PubMed] [Google Scholar]
  12. Kityk R, Vogel M, Schlecht R, Bukau B, Mayer MP. Pathways of allosteric regulation in Hsp70 chaperones. Nat Commun. 2015;6:8308. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Marcinowski M, Höller M, Feige MJ, Baerend D, Lamb DC, Buchner J. Substrate discrimination of the chaperone BiP by autonomous and cochaperone‐regulated conformational transitions. Nat Struct Mol Biol. 2011;18(2):150–8. [DOI] [PubMed] [Google Scholar]
  14. Mashaghi A, Bezrukavnikov S, Minde DP, Wentink AS, Kityk R, Zachmann‐Brand B, et al. Alternative modes of client binding enable functional plasticity of Hsp70. Nature. 2016;539(7629):448–51. [DOI] [PubMed] [Google Scholar]
  15. Matsumura Y, Sakai J, Skach WR. Endoplasmic reticulum protein quality control is determined by cooperative interactions between Hsp/c70 protein and the CHIP E3 ligase. J Biol Chem. 2013;288(43):31069–79. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Mayer MP. The Hsp70‐chaperone Machines in Bacteria. Front Mol Biosci. 2021;8:694012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Mayer MP, Schröder H, Rüdiger S, Paal K, Laufen T, Bukau B. Multistep mechanism of substrate binding determines chaperone activity of Hsp70. Nat Struct Biol. 2000;7(7):586–93. [DOI] [PubMed] [Google Scholar]
  18. Morshauser RC, Hu W, Wang H, Pang Y, Flynn GC, Zuiderweg ER. High‐resolution solution structure of the 18 kDa substrate‐binding domain of the mammalian chaperone protein Hsc70. J Mol Biol. 1999;289(5):1387–403. [DOI] [PubMed] [Google Scholar]
  19. Mossessova E, Lima CD. Ulp1‐SUMO crystal structure and genetic analysis reveal conserved interactions and a regulatory element essential for cell growth in yeast. Mol Cell. 2000;5(5):865–76. [DOI] [PubMed] [Google Scholar]
  20. Muller P, Ruckova E, Halada P, Coates PJ, Hrstka R, Lane DP, et al. C‐terminal phosphorylation of Hsp70 and Hsp90 regulates alternate binding to co‐chaperones CHIP and HOP to determine cellular protein folding/degradation balances. Oncogene. 2013;32(25):3101–10. [DOI] [PubMed] [Google Scholar]
  21. Pellecchia M, Montgomery DL, Stevens SY, Vander Kooi CW, Feng HP, Gierasch LM, et al. Structural insights into substrate binding by the molecular chaperone DnaK. Nat Struct Biol. 2000;7(4):298–303. [DOI] [PubMed] [Google Scholar]
  22. Qi R, Sarbeng EB, Liu Q, Le KQ, Xu X, Xu H, et al. Allosteric opening of the polypeptide‐binding site when an Hsp70 binds ATP. Nat Struct Mol Biol. 2013;20(7):900–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Rauch JN, Gestwicki JE. Binding of human nucleotide exchange factors to heat shock protein 70 (Hsp70) generates functionally distinct complexes in vitro. J Biol Chem. 2014;289(3):1402–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Rebeaud ME, Mallik S, Goloubinoff P, Tawfik DS. On the evolution of chaperones and cochaperones and the expansion of proteomes across the tree of life. Proc Natl Acad Sci U S A. 2021;118(21):e2020885118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Ricci L, Williams KP. Development of fluorescence polarization assays for the molecular chaperone Hsp70 family members: Hsp72 and DnaK. Curr Chem Genomics. 2008;2:90–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Rosenzweig R, Nillegoda NB, Mayer MP, Bukau B. The Hsp70 chaperone network. Nat Rev Mol Cell Biol. 2019;20(11):665–80. [DOI] [PubMed] [Google Scholar]
  27. Roufayel R, Kadry S. Molecular chaperone HSP70 and key regulators of apoptosis ‐ a review. Curr Mol Med. 2019;19(5):315–25. [DOI] [PubMed] [Google Scholar]
  28. Ryu SW, Stewart R, Pectol DC, Ender NA, Wimalarathne O, Lee JH, et al. Proteome‐wide identification of HSP70/HSC70 chaperone clients in human cells. PLoS Biol. 2020;18(7):e3000606. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Schlecht R, Erbse AH, Bukau B, Mayer MP. Mechanics of Hsp70 chaperones enables differential interaction with client proteins. Nat Struct Mol Biol. 2011;18(3):345–51. [DOI] [PubMed] [Google Scholar]
  30. Slepenkov SV, Witt SN. Kinetic analysis of interdomain coupling in a lidless variant of the molecular chaperone DnaK: DnaK's lid inhibits transition to the low affinity state. Biochemistry. 2002;41(40):12224–35. [DOI] [PubMed] [Google Scholar]
  31. Smock RG, Blackburn ME, Gierasch LM. Conserved, disordered C terminus of DnaK enhances cellular survival upon stress and DnaK in vitro chaperone activity. J Biol Chem. 2011;286(36):31821–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Stricher F, Macri C, Ruff M, Muller S. HSPA8/HSC70 chaperone protein: structure, function, and chemical targeting. Autophagy. 2013;9(12):1937–54. [DOI] [PubMed] [Google Scholar]
  33. Swain JF, Schulz EG, Gierasch LM. Direct comparison of a stable isolated Hsp70 substrate‐binding domain in the empty and substrate‐bound states. J Biol Chem. 2006;281(3):1605–11. [DOI] [PubMed] [Google Scholar]
  34. Taylor IR, Ahmad A, Wu T, Nordhues BA, Bhullar A, Gestwicki JE, et al. The disorderly conduct of Hsc70 and its interaction with the Alzheimer's‐related tau protein. J Biol Chem. 2018;293(27):10796–809. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Wang H, Kurochkin AV, Pang Y, Hu W, Flynn GC, Zuiderweg ER. NMR solution structure of the 21 kDa chaperone protein DnaK substrate binding domain: a preview of chaperone‐protein interaction. Biochemistry. 1998;37(22):7929–40. [DOI] [PubMed] [Google Scholar]
  36. Wang W, Liu Q, Liu Q, Hendrickson WA. Conformational equilibria in allosteric control of Hsp70 chaperones. Mol Cell. 2021;81(19):3919–3933.e7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Winter J, Linke K, Jatzek A, Jakob U. Severe oxidative stress causes inactivation of DnaK and activation of the redox‐regulated chaperone Hsp33. Mol Cell. 2005;17(3):381–92. [DOI] [PubMed] [Google Scholar]
  38. Wisén S, Gestwicki JE. Identification of small molecules that modify the protein folding activity of heat shock protein 70. Anal Biochem. 2008;374(2):371–7. [DOI] [PubMed] [Google Scholar]
  39. Wu S, Hong L, Wang Y, Yu J, Yang J, Yang J, et al. Kinetics of the conformational cycle of Hsp70 reveals the importance of the dynamic and heterogeneous nature of Hsp70 for its function. Proc Natl Acad Sci U S A. 2020;117(14):7814–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Xu LQ, Wu S, Buell AK, Cohen SI, Chen LJ, Hu WH, et al. Influence of specific HSP70 domains on fibril formation of the yeast prion protein Ure2. Philos Trans R Soc Lond B Biol Sci. 2013;368(1617):20110410. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Yang J, Zhang H, Gong W, Liu Z, Wu H, Hu W, et al. S‐Glutathionylation of human inducible Hsp70 reveals a regulatory mechanism involving the C‐terminal α‐helical lid. J Biol Chem. 2020;295(24):8302–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Zhang H, Amick J, Chakravarti R, Santarriaga S, Schlanger S, McGlone C, et al. A bipartite interaction between Hsp70 and CHIP regulates ubiquitination of chaperoned client proteins. Structure. 2015;23(3):472–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Zhang H, Loovers HM, Xu LQ, Wang M, Rowling PJ, Itzhaki LS, et al. Alcohol oxidase (AOX1) from Pichia pastoris is a novel inhibitor of prion propagation and a potential ATPase. Mol Microbiol. 2009;71(3):702–16. [DOI] [PubMed] [Google Scholar]
  44. Zhang H, Yang J, Wu S, Gong W, Chen C, Perrett S. Glutathionylation of the bacterial Hsp70 chaperone DnaK provides a link between oxidative stress and the heat shock response. J Biol Chem. 2016;291(13):6967–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Zhang P, Leu JI, Murphy ME, George DL, Marmorstein R. Crystal structure of the stress‐inducible human heat shock protein 70 substrate‐binding domain in complex with peptide substrate. PLoS One. 2014;9(7):e103518. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Zhang Y, Zuiderweg ER. The 70‐kDa heat shock protein chaperone nucleotide‐binding domain in solution unveiled as a molecular machine that can reorient its functional subdomains. Proc Natl Acad Sci U S A. 2004;101(28):10272–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Zhu X, Zhao X, Burkholder WF, Gragerov A, Ogata CM, Gottesman ME, et al. Structural analysis of substrate binding by the molecular chaperone DnaK. Science. 1996;272(5268):1606–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Zhuravleva A, Clerico EM, Gierasch LM. An interdomain energetic tug‐of‐war creates the allosterically active state in Hsp70 molecular chaperones. Cell. 2012;151(6):1296–307. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Zhuravleva A, Gierasch LM. Allosteric signal transmission in the nucleotide‐binding domain of 70‐kDa heat shock protein (Hsp70) molecular chaperones. Proc Natl Acad Sci U S A. 2011;108(17):6987–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Zhuravleva A, Gierasch LM. Substrate‐binding domain conformational dynamics mediate Hsp70 allostery. Proc Natl Acad Sci U S A. 2015;112(22):E2865–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Zuiderweg ER, Bertelsen EB, Rousaki A, Mayer MP, Gestwicki JE, Ahmad A. Allostery in the Hsp70 chaperone proteins. Top Curr Chem. 2013;328:99–153. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

FIGURE S1. The binding of FITC‐labeled ALLLSAPRR peptide to different concentrations of DnaK or its mutants. Fluorescence polarization at 520 nm after excitation at 485 nm was used to monitor the binding of 20 nM FITC‐labeled ALLLSAPRR peptide to different concentrations of DnaK or its mutants, in the absence of nucleotide (a), or in the presence of 1 mM ADP (b) or ATP (c), as indicated. The data shown are the mean of three individual experiments and the error bars represent the standard error of the mean and were fitted as described in Section 4.

FIGURE S2. The binding of FITC‐labeled ALLLSAPRR peptide to different concentrations of hHsp70, hHsc70, or their mutants. Fluorescence polarization at 520 nm after excitation at 485 nm was used to monitor the binding of 20 nM FITC‐labeled ALLLSAPRR peptide to different concentrations of hHsp70, hHsc70 or their mutants, in the absence of nucleotide (a,b,g), or in the presence of 1 mM ADP (c,d) or ATP (e,f), as indicated. The data shown are the mean of three individual experiments and the error bars represent the standard error of the mean and were fitted as described in Section 4.

FIGURE S3. SDS‐PAGE (10%) of all the proteins used in this study with molecular mass standards as indicated (a–d).

Click here for additional data file. (1.7MB, pdf)

Data Availability Statement

All data are contained within the manuscript.

Articles from Protein Science : A Publication of the Protein Society are provided here courtesy of The Protein Society

RESOURCES