Dynamic periplasmic chaperone reservoir facilitates biogenesis of outer membrane proteins

Abstract

Outer membrane protein (OMP) biogenesis is critical to bacterial physiology because the cellular envelope is vital to bacterial pathogenesis and antibiotic resistance. The process of OMP biogenesis has been studied in vivo, and each of its components has been studied in isolation in vitro. This work integrates parameters and observations from both in vivo and in vitro experiments into a holistic computational model termed "Outer Membrane Protein Biogenesis Model" (OMPBioM). We use OMPBioM to assess OMP biogenesis mathematically in a global manner. Using deterministic and stochastic methods, we are able to simulate OMP biogenesis under varying genetic conditions, each of which successfully replicates experimental observations. We observe that OMPs have a prolonged lifetime in the periplasm where an unfolded OMP makes, on average, hundreds of short-lived interactions with chaperones before folding into its native state. We find that some periplasmic chaperones function primarily as quality-control factors; this function complements the folding catalysis function of other chaperones. Additionally, the effective rate for the β-barrel assembly machinery complex necessary for physiological folding was found to be higher than has currently been observed in vitro. Overall, we find a finely tuned balance between thermodynamic and kinetic parameters maximizes OMP folding flux and minimizes aggregation and unnecessary degradation. In sum, OMPBioM provides a global view of OMP biogenesis that yields unique insights into this essential pathway.

Keywords: BAM; OMP biogenesis; kinetic model; membrane protein folding; β-barrel.

Fig. 1.

Diagram of mechanistic treatment used in OMPBioM. The downward vertical arrow at the top of the figure indicates uOMP synthesis and translocation. Nascent uOMP (U) can interact with itself through the aggregation pathway (purple), bind to chaperone (red) before folding into its native state (N), or be degraded (orange). Chaperones enter the system as monomers before undergoing oligomerization into a binding-competent oligomerization state indicated by subscript (blue). All species are subject to a rate of dilution (gray). Chaperones are regenerated upon uOMP folding or unbinding. Folding pathways that are assumed to be accelerated by BAM are shown (green). For more information of this mechanistic treatment see SI Appendix.

Fig. 2.

Trends observed in the simulated and experimentally observed phenotypes agree. The steady-state concentrations of each OMP species in the simulations for the WT and indicated chaperone single- and double-null mutants are shown. Species include fOMP (dark green), free monomeric uOMP (blue), aggregated uOMP (gray), bound uOMP (light green), and degraded uOMP (hatched segments). Bound uOMP is the sum of uOMP bound to all chaperones, including SurA, Skp, FkpA, and DegP. The x axis indicates simulated phenotypes. Simulations indicate that minimal populations of free and aggregated uOMP are present under WT and mild phenotype conditions. Simulated σ^E responses are included for ∆surA, ∆surA∆degP, and ∆surA∆skp. Data are provided in tabular form in SI Appendix, Table S6.

Fig. 3.

A folding rate enhancement provided by SurA is required to recapitulate phenotypes. The rate enhancement provided by SurA is defined as the folding rate constant for SurA–BAM divided by the folding rate constant for BAM-only. Shown are contour lines for OMP periplasmic lifetime (green) and the summed concentration of free and aggregated uOMP (red). The concentration of fOMP in the ∆surA simulation and WT simulation are related; the ∆surA simulation is expected to return a concentration of fOMP ∼10% of that in the WT simulation (13). The limits defining where the fOMP concentration in the ∆surA simulation is 5% or 15% of the WT simulation are shown as gray contour lines. If this value is less than 5%, SurA is more essential than expected; if the value is greater than 15%, SurA is more inconsequential than expected. The cyan area indicates the parameter space that results in lifetimes, free and aggregated uOMP concentrations, and fOMP concentrations that agree with experimental observations; a mild rate enhancement of 3–100 fold is required to recapitulate experimental observations.

Fig. 4.

OMPBioM allows the assessment of in vivo BAM folding rates. Shown are the contour lines for OMP periplasmic lifetimes (green) and copy number per cell (CN) (blue) as a function of covariation of the periplasmic input rate (k_in) and the effective BAM folding rate (k_fold). The parameter space allowed by the known values for OMP lifetimes and CN is shown in cyan. The dashed red line indicates the boundary where the concentration of free and aggregated uOMP in the periplasm equals 1 µM; this is a viable parameter space. The solid red line is the boundary at which uOMP + Aggregate = 10 µM; this concentration would be expected to induce the envelope stress response. The increasing red shading in the bottom right corner indicates the increasing accumulation of uOMP in the periplasm; these levels would be expected to lead to cell death.

Fig. 5.

OMP biogenesis is highly dynamic, with many binding events occurring between uOMP synthesis and folding. (A) The number of binding events for each synthesized uOMP under WT conditions. Fitting these data to an exponential decay results in 348 binding events on average. (B) A representative trajectory (600,000 steps) of binding events for a single (representative) uOMP over its periplasmic lifetime; between every binding event, the OMP is released to form free uOMP before it is bound by another chaperone. This particular OMP has a periplasmic lifetime of 54 s. SI Appendix, Table S4 shows the number of binding events and lifetimes for simulated phenotypes.