OmpA controls order in the outer membrane and shares the mechanical load

Abstract

OmpA, a predominant outer membrane (OM) protein in Escherichia coli, affects virulence, adhesion, and bacterial OM integrity. However, despite more than 50 y of research, the molecular basis for the role of OmpA has remained elusive. In this study, we demonstrate that OmpA organizes the OM protein lattice and mechanically connects it to the cell wall (CW). Using gene fusions, atomic force microscopy, simulations, and microfluidics, we show that the β-barrel domain of OmpA is critical for maintaining the permeability barrier, but both the β-barrel and CW-binding domains are necessary to enhance the cell envelope's strength. OmpA integrates the compressive properties of the OM protein lattice with the tensile strength of the CW, forming a mechanically robust composite that increases overall integrity. This coupling likely underpins the ability of the entire envelope to function as a cohesive, resilient structure, critical for the survival of bacteria.

Keywords: atomic force microscopy; membrane biophysics; membrane organisation; outer membrane.

Fig. 1.

The efficiency of plating assays show that OmpA contributes to OM integrity. (A and B) Plating efficiency of MG1655 WT, ∆ompA, and ∆lpp cells on LB agar and LB agar supplemented with indicated antimicrobials. Dilutions from left to right are 10^{0 to} 10⁻⁵. (B) The similar structures of ampicillin and chloramphenicol are indicated to the right, with their similar molecular weights.

Fig. 2.

OmpA constructs and their impact on the efficiency of plating show that the β-barrel is particularly important for function, and AFM shows that removal of OmpA compromises OMP lattice order. (A) Schematic representations of the seven main strains used throughout the study. MG1655 has full-length OmpA, with an 8-stranded β-barrel domain, linker, and C-terminal CW binding domain. To investigate the importance of CW binding, this was reduced by mutating two residues important for binding (ompA^D262A/R277A) or binding was abolished by complete removal of the C-terminal domain (ompA^1-191). To investigate the importance of the β-barrel, the OmpA β-barrel was replaced with that of OmpX (ompX-ompA^192-346) or by the beginning of Lpp (lpp^1-35-ompA^195-346). To test the relative importance of CW binding in ompX-ompA^192-346, the C-terminal domain was also mutated (ompX-ompA^{192-346: D262A/R277A}). All constructs were inserted into the genome at the ompA locus and were well expressed, as detected by western blotting (SI Appendix, Fig. S2). (B and C) Plating efficiency of MG1655 WT and indicated strains on (B) LB agar and (C) LB agar supplemented with indicated antimicrobials. ompA^D262A/R277A, ompA^1-191, and ompX-ompA^192-346 survive well on bile and EDTA plates and have intermediate growth on cefsulodin and ampicillin. Like ∆ompA cells, lpp^1-35-ompA^195-346 and ompX-ompA^{192-346:D262A/R277A} grow poorly in the presence of bile, EDTA, cefsulodin, or ampicillin. Dilutions from left to right are 10^{0 to} 10⁻⁵. (D) Live cell AFM phase images of the WT, ∆ompA, and ∆lpp OMs show that the OMP lattice is abolished without OmpA. (Scale bar, 100 nm.) Color scales are 0.9, 0.8, and 0.6 deg. Color range is shown to the Right of panel A. (E) Quantification of lattice disruption shows that deletion of ompA significantly increases lattice disruption compared to the WT, whereas removing lpp has no effect. Center lines are means, and error bars are SD. ns = (P > 0.15) by the two-sample t test.

Fig. 3.

Simulations and AFM show that a CW–tethered OmpA can order surrounding OMPs. (A) Top–down view of the simulation setup. 29 copies of OmpF trimers (green), 17 copies of FhuA (red), and 44 OmpA (black) are arranged in an imperfect hexagonal lattice. ReLPS molecules are not shown, for clarity. (B) Lateral displacement of OmpF and FhuA proteins versus time, with corresponding slope. For each condition, data from three individual simulations were taken to obtain averages (shown as lines) and SD (shaded regions). In OmpA-NR conditions, surrounding proteins are significantly more mobile. (C) Density maps showing the lateral displacement of OMPs over the 2.75 µs-long production simulation when OmpA is restrained (OmpA-R) and not restrained (OmpA-NR). The increased mobility is seen as a blurring of proteins, but each protein stays approximately in its place in the lattice. (Scale bar, 20 nm.) Color density scale, shown on the right, is from 0 to 6. (D) Live cell AFM phase images of indicated strains show that the OMP lattice is better resolved in ompA^D262A/R277A, ompA^1-191, and ompX-ompA^192-346 strains than the deletion, but lpp^1-35-ompA^195-346 lattice disruption is not improved. (Scale bar, 100 nm.) Color scales are 0.25, 0.8, 0.4, and 0.9 deg. (E) Quantification of lattice disruption shows that ompA^D262A/R277A, ompA^1-191, and ompX-ompA^192-346 all have significantly lower lattice disruption than ∆ompA but are still significantly different from the WT. lpp^1-35-ompA^195-346 cells have lattice disruption equal to ∆ompA. Center lines are means, and error bars are SD. ns = (P > 0.08) by the two-sample t test.

Fig. 4.

Only full-length OmpA contributes to cell strength. OmpA orders the OMP lattice and connects the OM and CW to form a mechanically strong composite. (A) Osmotic force extension was used to calculate the relative stiffness of ∆ompA, ompA^1-191, and ompX-ompA^192-346 cell envelopes, normalized to the WT (MG1655). This shows that all three strains are significantly softer than the WT (P < 0.003 by the one-sided t test). Center lines are means, and error bars are SD. ns = (P > 0.15) by the two-sample t test. (B) Schematic of predicted OM organization. The hexagonal lattice of trimeric porins (purple) is embedded with other OMPs (orange and brown), leaving gaps that are filled by the abundant OmpA (blue) which interacts with surrounding OMPs (red regions) while remaining in place, maintaining an ordered lattice. (C) OmpA is uniquely able to couple the compressive strength of the OMP lattice and tensile strength of the CW via β-barrel interaction, a flexible linker, and CW binding, allowing the whole cell envelope to act as a mechanically strong composite.