BiP is the endoplasmic member of the Hsp70 family. BiP is regulated by several co-chaperones including the nucleotide-exchange factor (NEF) Bap (Sil1 in yeast). Bap is a two-domain protein. The interaction of the Bap C-terminal domain with the BiP ATPase domain is sufficient for its weak NEF activity. However, stimulation of the BiP ATPase activity requires full-length Bap, suggesting a complex interplay of these two factors. Here, single-molecule FRET experiments with mammalian proteins reveal that Bap affects the conformation of both BiP domains, including the lid subdomain, which is important for substrate binding. The largely unstructured Bap N-terminal domain promotes the substrate release from BiP. Thus, Bap is a conformational regulator affecting both nucleotide and substrate interactions. The preferential interaction with BiP in its ADP state places Bap at a late stage of the chaperone cycle, in which it coordinates release of substrate and ADP, thereby resetting BiP for ATP and substrate binding.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
We thank J. Lawatschek and K. Richter for help with experiments, L. Voith von Voithenberg for helpful and critical discussions regarding spFRET experiments as well as data analysis, W. Kügel for contributions to the spFRET analysis software, G. M. Feind for performing the HDX analysis and K. Buhr for help with the structural modeling of the BiP–peptide complexes. We gratefully acknowledge the financial support by the Deutsche Forschungsgemeinschaft to J.B. (DFG), D.C.L. (SFB1035, A11), I.A. (SFB1035, A10) and M.S. (IGSSE) and by the Ludwig-Maximilians-Universität through the Center for NanoScience (CeNS) and the BioImaging Network (BIN).
Integrated supplementary information
The sequence of human Bap with its signal sequence highlighted in red is depicted. Structural information from yeast Bap (PDB-ID: 3QML) was mapped onto the sequence of human Bap. Sequence regions resolved in the crystal are highlighted in purple. The missing N-terminal amino acids comprise ~25 % of the mature full-length protein.
(a) FUV CD spectra were measured for Bap, Bap-C and Bap-N. A comparison between the curves reveals an α-helical secondary structure of Bap and Bap-C and a random coil structure for Bap-N. (b) Thermal transitions of Bap (black) and Bap-C (red) at 220 nm and the corresponding sigmoidal fits showing a one-step thermal denaturation occurring at ~46 °C.
Analytical ultracentrifugation (AUC) measurements were performed in the presence of 1 mM nucleotide to detect the increase in size of the BiP-Bap complex depending on the Bap or Bap-C concentration. The presence of the Bip-Bap complex can be observed by a shift in the sedimentation coefficient with increasing Bap concentrations. (a-c) Titration performed with 1 mM AMP-PNP. (a) AUC sedimentation velocity (AUC-SV) curves for 0.4 μM ATTO 488-labeled BiP-167-638 (final concentration) titrated with different Bap concentrations. (b) AUC-SV curves for 0.4 μM ATTO 488-labeled BiP-167-638 (final concentration) titrated with different concentration of Bap-C. (c) AUC-SV curves for 0.5 μM ATTO 488-labelled BiP-NBD-167 (final concentration) titrated with Bap. The titrations are shown in Figure 2b. (d-g) AUC-SV curves for 0.4 μM ATTO 488-labeled BiP-167-638 (final concentration) titrated with different concentration of Bap (d, e) without nucleotide or (f, g) in presence of 1 mM ADP. (h-j) Interaction of BiP’s isolated NBD with Bap. Sedimentation velocity (SV) was determined by analytical ultracentrifugation (AUC). 0.5 μM Atto 488-labeled BiP NBD-167 was tested without Bap (black line) and in the presence of 5 μM Bap (red line). Measurements were performed (h) in the absence of a nucleotide (apo), (i) with 1 mM ATP or (j) with 1 mM ADP.
(a) Different BiP mutants and DnaK ATPase activity measured in an enzyme coupled assay. Data was normalized to the data of BiP wt. (b) Thin-layer chromatography was employed to separate [α-32P]-ATP from [α-32P]-ADP for the determination of single-turnover ATPase kinetics. 20 μM BiP were mixed with 5 μM ATP containing [α-32P]-ATP. Samples were quenched at defined time points and the band intensities were quantified. Simulation of ATPase in single turnover experiments of BiP and Bap (red) compared to BiP alone (black). The calculated rate constants from the fits are 0.4 min-1 and 1.3 min–1.
BiPPred prediction scores were determined for all heptamers within the BAP sequence. The score for each heptamer is shown in condensed form as a coloured box below the central residue Z of the heptameric peptide (XXXZXXX)
Supplementary Figure 6 Representative PIE-MFD analysis shown for mutant BiP (167-638) in the presence of 1 mM ADP and 10 μM Bap-N.
(a) A 2D histogram of stoichiometry versus FRET efficiency after filtering for double-labeled molecules. A stoichiometry of ~0.5 indicates a ratio of 1:1 labeling between donor and acceptor fluorophores. (b) A 2D histogram of FRET efficiency versus donor lifetime . The relationship between FRET efficiency and donor lifetime for a static population (the static FRET line) is shown in black. No deviation from the static FRET line is observed indicating an absence of conformational dynamics during the ~1 ms long observation time. (c,d) 2D histograms of the burstwise anisotropy of (c) the donor versus the fluorescence lifetime of the donor and (d) the acceptor versus the fluorescence lifetime of the acceptor fluorophore . The black lines are given by the Perrin equation , where is the fundamental anisotropy, is the fluorescence lifetime and is the rotational correlation time. is assumed to be 0.4. (e) The time-resolved anisotropy decay for the donor (blue) and acceptor (red) fluorophores are shown. Photons from all double-labelled molecules are pooled together to obtain the cumulative fluorescence decays, from which the anisotropy is determined. Fits to bi-exponential model functions are given by solid lines, accounting for the fast rotation of the fluorophore on the nanosecond timescale and the slow rotation of the protein on the timescale of tens of nanoseconds
(a) SpFRET efficiency distributions of the NBD-lid sensor (BiP-167-519) in the presence of different concentrations of Bap and 1 mM ATP were measured. Sample preparation was performed by incubating 1 μM BiP for 15 min at 37 °C with the final Bap concentrations and then BiP was diluted to spFRET concentrations (~20 pM) while keeping the Bap concentration constant and adding the nucleotide. (b) The spFRET distribution for a small peptide, HTFPVAL (70 μM), bound to BiP in the presence of 1 mM ATP. BiP and BiP with 10 μM Bap in 1 mM ATP are shown for comparison
SpFRET efficiency distributions of the SBD-lid sensor (BiP-519-638) were determined with different concentrations of ATP (a) in the absence of Bap, (b) in the presence of 10 μM Bap and (c) in the presence of 10 μM Bap-C. (d) The normalized area under the FRET efficiency histograms was summed up until a FRET efficiency value of 0.4 and plotted versus the ATP concentration. The K d of ATP is given by the Boltzmann fit as 4.7 ± 1.3 nM for BiP alone and as 412 ± 7 nM in the presence of Bap. The addition of Bap-C resulted in a K d of 4.1 ± 0.7 nM. Therefore, Bap decreases BiP’s affinity for ATP by a factor of 88 compared to BiP alone or with bound Bap-C. Sample preparation was performed by incubating 1 μM BiP with the final Bap concentration and then keeping the Bap and Bap-C concentrations constant and adding the nucleotide during the dilution of BiP to spFRET concentrations (~20 pM)
(a-c) SpFRET efficiency histograms were measured for the NBD-SBD BiP sensor (gray line), BiP-Δlid (black line), BiP-Δlid in the presence of 10 μM Bap (red line) or BiP-Δlid with 10 μM Bap-C (dark yellow line). Experiments were performed (a) in the absence of nucleotide (apo), (b) with 1 mM ATP or (c) 1 mM ADP included in the buffer. Sample preparation was performed by incubation of BiP (1 μM) with Bap, Bap-C and nucleotide at their final concentrations, which were kept constant during the dilution of BiP to spFRET concentrations (~20 pM). (d-f) SpFRET efficiency histograms for the SBD-lid BiP sensor (BiP 519 - 638), the NBD-SBD BiP sensor (BiP 167 - 519), and BiP- Δlid in 1 mM ADP without peptide or Bap (gray line), with 70 μM HTFPVAL (black line) and with 70 μM HTFPVAL and 10 μM Bap (red line)