Pathways of allosteric regulation in Hsp70 chaperones

Abstract

Central to the protein folding activity of Hsp70 chaperones is their ability to interact with protein substrates in an ATP-controlled manner, which relies on allosteric regulation between their nucleotide-binding (NBD) and substrate-binding domains (SBD). Here we dissect this mechanism by analysing mutant variants of the Escherichia coli Hsp70 DnaK blocked at distinct steps of allosteric communication. We show that the SBD inhibits ATPase activity by interacting with the NBD through a highly conserved hydrogen bond network, and define the signal transduction pathway that allows bound substrates to trigger ATP hydrolysis. We identify variants deficient in only one direction of allosteric control and demonstrate that ATP-induced substrate release is more important for chaperone activity than substrate-stimulated ATP hydrolysis. These findings provide evidence of an unexpected dichotomic allostery mechanism in Hsp70 chaperones and provide the basis for a comprehensive mechanical model of allostery in Hsp70s.

Figure 1. Asp481 is essential for interdomain communication in DnaK.

(a) Hsp70s alternate between closed and open conformations controlled by nucleotides and substrates through an allosteric mechanism. Cartoon representation of DnaK in the ADP-bound closed conformation (left, PDB ID 2KHO) and in the ATP-bound open conformation (right, 4B9Q15). Lobes of the nucleotide-binding domain are coloured in dark blue (I) and light blue (II), SBDβ in dark red and SBDα in orange. The highly conserved interdomain linker is shown in purple. Indicated are residues important for NBD–SBDβ docking. All structure figures were prepared in PyMOL (The PyMOL Molecular Graphics System, Version 1.7.4 Schrödinger, LLC). (b) Zoom into the interface between NBD and SBDβ of the ATP-bound open conformation (PDB ID 4B9Q15), rotated by 180° as compared with the right panel in a. NBD lobes are shown in dark blue and light blue, and SBDβ in dark red and interdomain linker in purple. Residues bridging the domain interface by hydrogen bonds are shown in sticks and coloured according to the atom with carbon in the colour of the respective domain. Putative hydrogen bonds (distances between proton donor and acceptor ≤3.5 Å) are shown as dashed lines. (c) ATP-induced blueshift of the emission maximum of typtophane fluorescence. (d,e) Peptide dissociation rates measured in the absence of nucleotides (d) or presence of ATP (e). (f) Single-turnover ATPase rates in the absence of DnaJ and substrate (basal rate), in the presence of DnaJ, the protein substrate σ³² or both DnaJ and σ³², as indicated. (g) Comparison of basal single-turnover ATPase rates measured in a quenched-flow apparatus for DnaK-D481A, DnaK-D481K and DnaK-K414I. Error bars represent s.e.m. of at least three independent measurements.

Figure 2. Pathway of substrate stimulation of the ATPase rate.

(a) Alignment of the SBDβ of the co-crystal structure of DnaK-SBD (cyan, PDB ID 1DKX60) with a substrate peptide (green) with the SBDβ (brown) of the crystal structure of DnaK in the ATP-bound open conformation (4B9Q15). β1 to β8 indicate the individual β-strands of the two-layered sandwich. Substrate peptide (NRLLLTG) is shown in sticks with the central leucine, which inserts into the hydrophobic substrate-binding pocket of DnaK, in space-filling representation. Residues Val440 and Leu484 in both structures are shown in space-filling representation. Residues that link Leu484 to the ATPase catalytic centre are shown as sticks in atom colours with carbon in dark blue. (b) Single-turnover ATPase rates of wild-type and mutant DnaK in the absence of any other protein and in the presence of DnaJ, σ³² or both as indicated. (c) Dissociation equilibrium constant of wild-type and mutant DnaK for binding of a HiLyte Fluor 488-labelled model peptide (σ³²-Q132-Q144-C). (d) Basal peptide dissociation rates. (e) ATP-induced substrate dissociation rates. (f) ATP-induced blueshift of the emission maximum of typtophane fluorescence in the absence of a substrate peptide (grey bars) or the presence of 130 μM peptide (σ³²-Q132-Q144-C). Error bars represent s.e.m. of at least three independent experiments.

Figure 3. Replacement of Phe146 by alanine compromises signal transduction from ATP to the SBD but not from substrate to the NBD.

(a) Zoom into the NBD–SBDβ interface of the ATP-bound open conformation of DnaK (PDB ID 4B9Q15) with NBD in blue and SBDβ in dark red. ATP and residues involved in catalysis (K70, E171) or allostery (P143, Y145, F146, D148, R151, R167, L484) are shown in sticks in atom colours with carbon in blue (NBD residues) or dark red (SBDβ residues). (b) ATP-induced blueshift of the emission maximum of tryptophane fluorescence. (c,d) Substrate dissociation rates in the absence of nucleotides (c) or presence of ATP (d). (e) Single-turnover ATPase rates of wild-type and mutant DnaK in the absence of any other protein and in the presence of DnaJ, σ³² or both as indicated. Error bars represent s.e.m. of at least three independent experiments.

Figure 4. All DnaK variants completely or partially impaired in allosteric regulation had residual chaperone activity in vitro.

(a–c) Refolding of chemically denatured luciferase (80 nM) by DnaK (800 nM), DnaJ (160 nM) and GrpE (400 nM). Luciferase was denatured in 6 M guanidinium HCl and diluted 125-fold into refolding buffer containing the respective DnaK variant, DnaJ and GrpE. Shown is the activity of luciferase relative to the not denatured control. Error bars represent s.e.m. of at least three independent experiments. (d) Aggregation prevention assay. Guanidinium-denatured luciferase (80 nM) was diluted into a solution containing DnaK (800 nM) and DnaJ (40 nM), and light scattering at 600 nm was followed over time.

Figure 5. Model of allosteric regulation in Hsp70s.

(a–c) Surface representation of the NBD of DnaK in the ATP-bound open conformation (PDB ID 4B9Q15) with lobe I in dark grey, lobe II in light grey and residues interacting with SBDβ coloured according to effects on basal ATPase activity (a), on substrate and DnaJ stimulated ATPase activity (b), and on ATP stimulated substrate release (c), if corresponding residue is replaced by alanine (Y145A, F146A, D148A, R151A, K155A, R167A) or if interacting residue is replaced (I168[D481A], D326[K414I]). (a) Fold increase in basal ATPase rate relative to DnaKwt; (b) stimulation of ATPase rate by DnaJ plus σ³² relative to the basal rate; (c) stimulation of substrate release by ATP relative to basal release rates; colouring as indicated. (d) Model of allosteric regulation (see main text). For clarity, the SBDα was left away.