Dynamic interplay between the periplasmic chaperone SurA and the BAM complex in outer membrane protein folding

Abstract

Correct folding of outer membrane proteins (OMPs) into the outer membrane of Gram-negative bacteria depends on delivery of unfolded OMPs to the β-barrel assembly machinery (BAM). How unfolded substrates are presented to BAM remains elusive, but the major OMP chaperone SurA is proposed to play a key role. Here, we have used hydrogen deuterium exchange mass spectrometry (HDX-MS), crosslinking, in vitro folding and binding assays and computational modelling to show that the core domain of SurA and one of its two PPIase domains are key to the SurA-BAM interaction and are required for maximal catalysis of OMP folding. We reveal that binding causes changes in BAM and SurA conformation and/or dynamics distal to the sites of binding, including at the BamA β1-β16 seam. We propose a model for OMP biogenesis in which SurA plays a crucial role in OMP delivery and primes BAM to accept substrates for folding.

Fig. 1. Structure and interaction of SurA and BAM.

a Structure of BAM in the lateral-closed state (PDB: 5D0O). Subunits are coloured BamA (green), BamB (cyan), BamC (pink), BamD (yellow), BamE (light pink). b, c Structures of the BamA β-barrel domain and POTRA 5 (P5) from BAM structures in c the lateral-closed (‘inward-open’) state (PDB: 5D0O), and d the ‘lateral-open’ state (PDB: 5LJO). β-strands 1 and 16 of the BamA β-barrel, which form the lateral gate, are highlighted in magenta. The entrance to the BamA β-barrel lumen is accessible in the ‘lateral-closed’ conformation, but is occluded in the ‘lateral-open’ BAM conformation, as indicated in orange. d Crystal structure of E. coli SurA (PDB: 1M5Y). Regions are coloured grey (N-terminal region of the core domain), green (P1), yellow (P2) and orange (C-terminal region of the core domain). The colour scheme for BAM subunits and SurA domains is used throughout. e Domain architecture of E. coli SurA-WT and the SurA domain deletion variants used in this study. The signal sequence is not shown and was not present in any of the constructs used here, but the numbering used throughout reflects the gene numbering (including the signal peptide). Constructs were expressed with an N-terminal His₆-tag and TEV cleavage site (white box). f Microscale thermophoresis (MST) data for binding of SurA-WT to BAM. Samples contained 400 nM Alexa Fluor 488-labelled SurA, BAM (1.6 nM–52 μM), 0.02% (v/v) DDM, 150 mM NaCl, 20 mM Tris–HCl, pH 8, at 25 °C. Three independent replicates were performed and averaged prior to fitting. The mean for each BAM concentration is shown as open circles and the individual values for each replicate are shown as dots. The error bars represent the standard deviation between replicates. Data were fitted to a 1:1 quadratic binding model (see the “Methods” section). Source data are provided as a Source Data file (Supplementary Data 8).

Fig. 2. HDX-MS analysis of BAM in the presence of SurA reveals multiple chaperone binding sites and allosteric conformational changes.

a, b The structure of the BAM complex in the ‘lateral-open’ conformation (PDB: 5LJO) coloured by subunit. c, d Cartoon views and e, f surface views showing regions of HDX protection in the BAM complex upon binding SurA. Left panels show a side view of the complex, and right panels a view from the periplasmic face. Regions that are protected from hydrogen exchange in the presence of SurA are highlighted in blue. Regions in white show no change in deuterium uptake in the presence of SurA, while those in dark grey denote sequences for which peptides were not detected. Note that no regions showing deprotection from HDX upon SurA binding were observed. Patches of protection from HDX upon SurA binding in the BamA β-barrel domain (adjacent to the lateral gate), POTRAs 1 and 2, BamB, and BamE/POTRA 4 are ringed in orange. g–l Representative deuterium uptake plots for peptides from g, h BamA, i BamB, j BamC, k BamD, and l BamE. The extent of deuterium uptake (Da) in the absence (black) and presence of SurA (orange) is shown. Individual data points for each time point are shown as dots and error bars represent the standard deviation of three technical replicates. Source data are provided as a Source Data file (Supplementary Data 9).

Fig. 3. HDX-MS analysis of SurA in the presence of BAM reveals conformational changes in the chaperone upon binding.

a, b The crystal structure of SurA (PDB: 1M5Y) with the N-terminal and C-terminal regions of the core domain coloured in grey and orange, respectively. The SurA PPIase domains P1 and P2 are coloured in green and yellow, respectively. c, d Cartoon views and e, f surface views of differential HDX data for SurA upon binding BAM. Left panels show one view of the chaperone, and right panels a view rotated by 180° around the y-axis. Regions that are protected or deprotected from hydrogen exchange in the presence of BAM are highlighted in blue or red, respectively. Regions in white show no change in deuterium uptake in the presence of BAM, while those in dark grey denote sequences for which peptides were not detected. g, h Representative deuterium uptake plots for peptides spanning residues g 47–71 (deprotected) and h 100–111 (protected) from the core domain of SurA. The extent of deuterium uptake (Da) in the absence (black) and presence of BAM (orange) is shown. In the presence of BAM, these regions are deprotected and protected from exchange, respectively. Individual data points for each time point are shown as dots and error bars represent the standard deviation of three technical replicates. Source data are provided as a Source Data file (Supplementary Data 9).

Fig. 4. The presence of the SurA core and P2 domains are required for maximal rate enhancement of tOmpA folding on BAM.

a Fraction folded of tOmpA in the presence or absence of BAM and SurA variants at different time points as measured by cold SDS–PAGE. Points are the mean of two independent replicates and error bars represent the range of values covered by the replicates. For BAM-containing samples fits to a single exponential equation are shown. In the absence of BAM no folding is observed. b Observed rate constants (k_obs) from kinetic experiments in (a). Error bars represent the error on the fit to a single exponential. Example SDS–PAGE gels of the folding reactions are shown in Fig. S11. Samples contained 1 μM BAM (where included) in proteoliposomes containing E. coli polar lipid extract (see the “Methods” section), 2 μM tOmpA, 10 μM SurA variant (where included) in 20 mM Tris–HCl, 150 mM NaCl, pH 8.0, at 25 °C. Source data are provided as a Source Data file (Supplementary Data 10).

Fig. 5. AlphaFold-Multimer generated model of the BAM–SurA complex.

a, b The predicted structure of the BAM–SurA complex coloured by subunit. SurA is in grey. c, d Surface views of the BAM–SurA complex with regions of HDX protection in the BAM complex upon binding SurA highlighted, and SurA coloured by domain. Regions of BAM that are protected from HDX in the presence of SurA are highlighted in blue. Regions in white show no change in deuterium uptake in the presence of SurA, while those in dark grey denote sequences for which peptides were not detected. Patches of protection from HDX upon SurA binding in the BamA β-barrel domain (adjacent to the lateral gate), POTRAs 1 and 2, BamB, and BamE/POTRA 4 are ringed in magenta. The SurA core, P1 and P2 domains are coloured in orange, green and yellow, respectively. Left panels show a side view of the complex, and right panels a view from the periplasmic face.

Fig. 6. Proposed model of SurA-mediated delivery of OMPs to BAM for folding.

In the membrane, BAM is in dynamic equilibrium, populating a number of conformations including those in which the BamA barrel is in either a ‘lateral-open’ or ‘lateral-closed’ conformation (top). SurA is also dynamic, populating ‘Core-P1 open’ and ‘Core-P1 closed’ ensembles of states (bottom). The P2 domain is shown residing close to the core domain in solution (‘Core-P2 closed’). OMPs bind in a cradle formed by the Core and P1 domains, with binding sites primarily located on the core domain^,. OMP binding to SurA causes the OMP to populate expanded states^,, The OMP–SurA complex binds to BAM, to form an OMP–SurA–BAM ternary complex, resulting in conformational changes in both SurA (opening of the P1 and P2 domains) and BAM (favouring the lateral-closed state, which we propose to be the OMP acceptor state). The OMP binding cradle of SurA is oriented into the BAM periplasmic ring, allowing release of the unfolded OMP into a protective ‘chaperonin-like’ environment, and presentation of the OMP to the BamA β-barrel for folding in a C- to N-terminus direction via β-strand elongation from β1 of the BamA barrel^,,,,,. Here, a single β-hairpin of the substrate is depicted bound to β1 of the BamA barrel, however more extensive β-sheet structure in the substrate may begin to form in the periplasm (or at the water–membrane interface) prior to membrane integration^,,,, in a manner analogous to a seeded amyloid aggregation reaction^,,.