Probing the transient interaction between the small heat-shock protein Hsp21 and a model substrate protein using crosslinking mass spectrometry

Abstract

Small heat-shock protein chaperones are important players in the protein quality control system of the cell, because they can immediately respond to partially unfolded proteins, thereby protecting the cell from harmful aggregates. The small heat-shock proteins can form large polydisperse oligomers that are exceptionally dynamic, which is implicated in their function of protecting substrate proteins from aggregation. Yet the mechanism of substrate recognition remains poorly understood, and little is known about what parts of the small heat-shock proteins interact with substrates and what parts of a partially unfolded substrate protein interact with the small heat-shock proteins. The transient nature of the interactions that prevent substrate aggregation rationalize probing this interaction by crosslinking mass spectrometry. Here, we used a workflow with lysine-specific crosslinking and offline nano-liquid chromatography matrix-assisted laser desorption/ionization tandem time-of-flight mass spectrometry to explore the interaction between the plant small heat-shock protein Hsp21 and a thermosensitive model substrate protein, malate dehydrogenase. The identified crosslinks point at an interaction between the disordered N-terminal region of Hsp21 and the C-terminal presumably unfolding part of the substrate protein.

Fig. 1

Hsp21 chaperone activity towards heat-treated MDH. The model substrate MDH was heat-treated by incubation at 45 °C in absence or presence of varying amounts of Hsp21. Thermal-induced precipitation of MDH was followed as light-scattering increase over time, and the suppression of light-scattering increase in presence of Hsp21 used to assess the chaperone activity. Samples assayed in duplicate or triplicate are marked with same color. The numbers of the ratios in the legend refer to concentrations (Hsp21/MDH) expressed in micromolars. a Hsp21/MSH molar ratios 1/1 and above, under different salt conditions (*): 1, low salt; 2, 0.1 M NaCl; 3, 0.2 M Na₂SO₄. b Hsp21/MDH molar ratios 1/1 and below, in 0.1 M NaCl

Fig. 2

MDH heat-treated in absence and presence of Hsp21 analyzed by Native-PAGE. Samples were pre-incubated at 25 °C or 45 °C as indicated and directly loaded on Blue-Native PAGE (3–12 % gradient, Bis–Tris buffer pH 7.6). MW, molecular weight marker; lanes 1 and 2, MDH and Hsp21, respectively, pre-incubated at 25 °C; lanes 3 and 4, MDH and Hsp21, respectively, pre-incubated at 45 °C (note that aggregated MDH does not enter the gel); lane 5, Hsp21 and MDH mixed at a 1/1 molar ratio and pre-incubated for 30 min at 25 °C; lane 6, Hsp21 and MDH mixed at a 1/1 molar ratio and pre-incubated for 30 min at 45 °C; lane 7, Hsp21 and MDH mixed at a 1/1 molar ratio and incubated for 180 min at 45 °C. The large stable complexes formed between Hsp21 and MDH pre-incubated for 180 min at 45 °C cannot enter the gel and remain in the loading well (lane 7). Comparison of lane 5 and 6 (marked by asterisks) shows that no stable complexes are formed by pre-incubation for 30 min at 45 °C, since Hsp21 dodecamers and MDH dimers run separately (lane 6), just as in the non-heated sample (lane 5). The sub-optimal resolution of the Hsp21 dodecamer in this Native-PAGE, compared with the better performance shown in Supplementary Figure S-1 and the appearance of the MDH dimer as a smear at an apparent mass approximately 700 kDa instead of a band at 66 kDa are due to technical limitations imposed by the high pI-value of MDH as further explained in the “Supplement”

Fig. 3

Crosslinked samples of Hsp21 and MDH analyzed by SDS-PAGE. Samples were pre-incubated at 20 °C or 45 °C M, as indicated, crosslinked, and then loaded to SDS-PAGE. MW, molecular weight marker; lane 1, Hsp21 + MDH, non-crosslinked reference sample; lane 2, Hsp21 + MDH, sample crosslinked after pre-incubation 30 min at 20 °C; lanes 3, 4, and 5 Hsp21 + MDH, triplicate samples crosslinked after pre-incubation 30 min at 45 °C; lane 6, Hsp21+MDH, sample crosslinked after pre-incubation 180 min at 45 °C. Extra bands are visible at 10 and 18 kDa which are degradation products of Hsp21 and at 42 kDa a dimer of Hsp21. The non-covalent bonds within the Hsp21 dimer are strong enough for a band to be visible in SDS-PAGE, more or less depending on the exact number of minutes and degrees in solubilization buffer. The Hsp21 dimer becomes stabilized by crosslinking and visible as an even stronger band (lanes 2–6)

Fig. 4

Residues involved in Hsp21–MDH crosslinks visualized in the 3D structures of Hsp21 and MDH. a Location of the three crosslinking Hsp21 residues M1, K27, and K173 in the Hsp21 dimeric subunit of the Hsp21 dodecamer. The crosslinking residues are only indicated on one of the two monomers shown. Residues M1 and K27 are within the predicted disordered N-terminal region (residues 1–84, dashed). b Location of the eight crosslinking MDH residues in the dimeric structure of MDH (PDB ID 1MLD). The dimer is made up of two monomers; one is colored dark gray, the other light gray. All eight residues crosslinking to Hsp21 are shown as sticks in both monomers. Residues 81, 215, 304, and 305 are only labelled in the dark gray monomer, and residues 1, 133, 273, and 277 are only labelled in the light gray monomer. a, b were prepared with PyMOL (www.pymol.org) (DeLano 2002)

Fig. 5

Abundance of Hsp21 residues involved in crosslinks. Observations of Hsp21 residues involved in the identified crosslinks were calculated and shown here as bars for each residue that can possibly get modified by the crosslinking reagent. One or several observations of the methionine-oxidized version of a crosslink were counted as one extra observation

Fig. 6

Abundance of MDH residues involved in crosslinks. Observations of MDH residues involved in the identified crosslinks were calculated and shown here as bars for each residue that can possibly get modified by the crosslinking reagent. One or several observations of the methionine-oxidized version of a crosslink were counted as one extra observation