Peptides derived from

β-strands of OmpC selectively inhibit OMP assembly. A and C, 35S-labelled EspP was synthesized by rabbit reticulocyte lysate, then incubated with EMM, and samples prepared for analysis by SDS-PAGE and radio-imaging (see Methods). EspP assembly into EMM was performed in the presence of DMSO (D) or specific inhibitor peptides. The precursor (p) and mature (m) form of EspP are indicated. A, Peptides labelled 1 through 23 were derived from OmpC, while peptide 24 was taken from a mitochondrial β-barrel protein to serve as a negative control (see fig. S2). B, Sequence comparison of inhibitor peptides identified in the EspP assembly assays. The signal is comprised of key hydrophobic (Φ), polar (ζ), glycine (G), and 2 aromatic (Ω) or hydrophobic amino acids. Amino acids highlighted in blue conform to the consensus signal, while those highlighted in purple deviate from the consensus signal. C, Upper: sequence of the peptide 18 indicating residue codes for the mutated peptides e.g. at position “6”, the native residue Y was mutated to A in peptide “6A”. Position 0 designates the first hydrophobic residue and position 8 designates the final aromatic or hydrophobic residue. DMSO (D) was used for control. Middle: SDS-PAGE of EspP assembly assay. Lower: data for inhibition of EspP assembly into EMM assayed in the presence of each of the indicated peptides.

Mutations in β-signals inhibit OMP Assembly in vitro. A, Representation of OmpC, (upper panel) annotated to show the position of each b-strand (blue arrows) and the region mimicked in the indicated peptides. Sequence logos show the conservation of sequence for the residues in these regions (see Methods), where the height of letter corresponds to the degree of conservation across bacterial species, residues are color-coded: aromatic (orange), hydrophobic (green), basic (blue), acidic (red), polar (sky blue) and proline and glycine (black). In the lower panel, data from the EMM assembly assay of OmpC mutants is organized to show effects for key mutants: 35S-labelled OmpC wild-type or mutant proteins were incubated with EMM for 30 or 90 min, and analyzed by BN-PAGE and radio-imaging. B, Comparison of amino acids in the final and 5th to final (-5) β-strands of OmpC, OmpF, and LamB. All three OMPs contain the putative hydrophobic residue near the N-terminal region of the β-strands, highlight with a black box. A hydrophobic and an aromatic residue is also found to be highly conserved between these OMP β-strands at the C-terminal amino acids of the β-strands. C-D, Assembly of 35S-labelled OmpF (C) or LamB (D) with mutations to the -5 β-strand or final β-strand in to EMM. Samples were prepared as in A.

Modifications in the putative internal signal slow the rate of OMP assembly in vivo. A, Schematic model of FLAG-OmpC expression system in E. coli. The plasmid-borne copy of OmpC was mutated and the mutant proteins are selectively detected with antibodies to the FLAG epitope. B, Total cell lysates were prepared from the variant E. coli strains expressing FLAG-OmpC (WT), -5 strand mutants: F280A, Y286A, double mutant F280A/Y286A (FY), or final β-strand mutants: V359A, Y365A, double mutant V359A/Y365A (VY) with (+) or without (-) arabinose. The steady-state levels of the indicated proteins were assessed after SDS-PAGE by immunoblotting with anti-FLAG antibody. C, After growth with arabinose induction, EMM fraction was isolated from each of the FLAG-OmpC variant expressing E. coli strains to assess steady state levels of OmpC variants and BAM complex subunits: BamA, BamB, BamC, BamD, and BamE by SDS-PAGE and immunoblotting. D, EMM fractions were solubilized with 1.5% DDM and subjected to BN-PAGE and immunoblotting. The indicated proteins were detected by anti- FLAG (top) and anti-BamA antibodies (bottom), respectively.

Mutation that stalls OMP assembly in a high molecular weight assembly intermediate. A, Upper panel, assembly assay of 35S-OmpC wild-type (WT) and mutant Y286A. After incubation in the EMM assays for indicated time points, 35S-proteins were resolved via BN-PAGE. OmpC trimer and intermediate (int) are indicated. The lower panel shows a quantification of the intermediate form compared to total assembled protein (average of 3 independent experiments). Statistical significance was indicated by the following: N.S. (not significant), p<0.05; *, p<0.005; **. Exact p values of intermediate formed by Wt vs Y286A at each timepoint were as follows; 20 minutes: p = 0.03077, 40 minutes: p = 0.02402, 60 minutes: p = 0.00181, 80 minutes: p = 0.0545. B, 35S-labelled OmpC(Y286A) was incubated with EMM for 60 mins, then the EMM fraction was solubilized with 1.5% DDM and incubated with (+) or without (-) anti-BamC antibody for 40 min. These samples were then analyzed by BN-PAGE. The presumptive translocation intermediate (Int) and the gel-shift species (Int. +AB) are indicated. C, 35S-labelled OmpC was incubated with EMM and then analyzed by BN-PAGE and radio-imaging. EMM were then washed and solubilized in the presence (+) or absence (-) of 6 M urea, after which the membrane fraction was re-isolated via ultracentrifugation for 45 min. Isolated membranes were solubilized in 1.5% DDM and analyzed by BN-PAGE.

Proximity and dependence of BamD with OmpC(Y286A) revealed by Neutron Reflectometry analysis. A, Ni-NTA was used to affinity purify His6-BamA or His6-BamD. The protein-Ni-NTA bead complex were then incubated in the presence of DMSO or the indicated peptides resuspended in DMSO, and 35S-labelled OmpC was incubated with these beads. After washing, bound proteins were eluted with 400 mM imidazole buffer. The graph represents densitometry analysis of data from 3 independent experiments. B, NR profiles for BamA in membrane (1st measurement, square black line), after addition of BamD (2nd measurement, circle red line), and after addition of the OmpC(Y286A) substrate (3rd measurement, triangle blue line). The experimental data was fitted using a seven-layer model: chromium - gold - NTA - His8 - β-barrel - P3-5 - P1-2. C, Table summarizes of the thickness, roughness and volume fraction data of each layer from the NR analysis. The thickness refers to the depth of layered structures being studied as measured in Å. The roughness refers to the irregularities in the surface of the layered structures being studied as measured in Å. 1st, 2nd, and 3rd measurement displayed in black, magenta, and blue, respectively. P3-5: POTRA3, POTRA 4, and POTRA5, P1-2: POTRA1 and POTRA2, Lipid: POPC, Solution: D2O, gold match water (GMW) and H2O. Details were described in Table S8 to S10. D, Neutron reflectometry schematic of BamA (green), BamD (yellow) and OmpC(Y286A) (magenta). Numbers indicate corresponding POTRA domains of BamA. Note that the cartoon is a depiction of the results and not a scale drawing of the structures.

BamD acts as the receptor for internal β-signals. A, BamD-cys mutant proteins were purified and incubated with 35S-OmpC with cysteine mutations from the -5 to the final β-strand. After the addition of the oxidizing agent, CuSO4, samples were purified with Ni-NTA and analyzed by SDS-PAGE. B, BamD cysteine mutants, highlighted in magenta, formed disulfide bonds at distinct locations of OmpC. C, BPA containing BamD was expressed in the presence (+) or absence (-) of FLAG-OmpC overexpression. Pink arrows indicate BamD-FLAG-OmpC cross-link products. D, OMP substrate superimposed on BamD. BamD cross-linking positions corresponding to internal signal are marked in purple and substrate cross-linking regions are marked in orange.

Key residues in two structurally distinct regions of BamD promote β-strand formation and OMP assembly. A, Double cysteine residues were mutated into OmpC between anti-parallel β-strands to form artificial disulfide bonds B,35S-OmpC-cys variants were translated then incubated with (+) or without (-) BamD. The samples were analyzed by reducing or non-reducing condition. Disulfide bond specific bands (S-S) are indicated to the right of the gel. C, Sequence conservation analysis of BamD. Target residues, Y62 and R197, are indicated with dark pink spheres on the crystal structure of BamD, sequence logos represent conservation in the immediate region of these residues. Amino acids listed above logo sequence are from crystal structure. Sequence conservation of BamD (bottom). D, Schematic depiction of the BamD depletion strain of E. coli used to express mutant BamD proteins. E, Pull-down assay of the variants of BamD-His8. The EMMs as in (C) were solubilized with 1.5% DDM and then subjected to Ni-NTA. T- 5% total input, F- unbound fraction, W- wash fraction, E- Eluted fraction. Each fraction was analyzed by SDS-PAGE and immunoblotting against indicated antibodies. F, Assembly of OmpC was reduced in EMMs containing BamD mutated at Y62 and R197. 35S-labelled OmpC was incubated with EMM as in (C), and then analyzed by BN-PAGE and radio-imaging.

Substrate recognition by BamD is essential to maintain membrane integrity. A,

Strains of E. coli with a depletion of chromosomal bamD, complemented by expression of the indicated forms of BamD, were grown in restrictive media (LB supplemented with 0.5% glucose) for 4 hours, then diluted into fresh media and grown a further 4 hours, then diluted into fresh media and grown a further 4 hours. B, Quantitation of steady-state protein levels of OmpF and the indicated subunits of the BAM complex in WT, Y62A, R197A mutants were determined after SDS-PAGE and immunoblot analysis. Wedge indicates that in each case, two samples were assessed corresponding to either 4 μg or 12 μg of total protein. C, Quantitation of the level of trimeric OmpF formed in the indicated EMM assays. The immunoblot analysis is shown after BN-PAGE analysis and immunoblotting with antibodies recognizing OmpF (upper panel) or BamD (lower panel). D, E E coli bamD depletion cells expressing mutations at residues, Y62A and R197A, in the β-signal recognition regions of BamD were grown with of VCN. F, G, E coli cells expressing mutations to OmpC internal signal, as shown in Fig 3, grown in the presence of VCN. Mutations to two key residues of the internal signal were sensitive to the presence of VCN. H, Structure of BamD (top) and the final 5 β-strands of OmpC (bottom). Gray boxes indicate relative regions of interaction between BamD and substrates. I, OmpF complex assembly in vivo of cells expressing mutant OmpC. OmpF assembly was greatly reduced in cells expressing double mutations to the -5 internal β-signal.

Model for the recognition of internal and canonical β-signals by the BAM complex. A, BAM complex assisted-OMP assembly is initiated when substrate OMPs encounter the periplasmic domain of the BAM complex. The canonical β-signal (blue arrow) of the OMP engages with the C-terminal TPR motifs of BamD (pink), after which the internal β- signal (orange arrow) is marshalled by the N-terminal TPR motif. The internal signal interacts with BamD R49 and N60 (black stars) while the canonical β-signal interacts with BamD D204 (black star). The proximity of BamD-substrate to POTRA 5 and the organization of the C-terminal β-strands of OMPs can then stimulate the entrance to the lateral gate of BamA. Finally, barrel formation and release into the membrane follows as previously published28,29. B, Structure of substrate engaged BAM complex (PDBID: 6V05). C, Bottom-up view of periplasmic domain of complex as shown in (B). Arrows indicate the position of the residues on BamD that function in binding sites for -5 signal and the β signal. D, A cavity formed by POTRA 1-2 and POTRA 5 of BamA and BamD is sufficient to accommodate the C-terminal five β-strands of substrate OMP. Substrate -1 strand and -5 β-strand are indicated by light blue strand and orange strand, respectively. The signal binding region of BamD is emphasized by deep purple (PDBID; Bam complex: 7TT5 and 7TT2). E, Bottom-up view of complex as shown in (D).