Global landscape of cell envelope protein complexes in Escherichia coli

Abstract

Bacterial cell envelope protein (CEP) complexes mediate a range of processes, including membrane assembly, antibiotic resistance and metabolic coordination. However, only limited characterization of relevant macromolecules has been reported to date. Here we present a proteomic survey of 1,347 CEPs encompassing 90% inner- and outer-membrane and periplasmic proteins of Escherichia coli. After extraction with non-denaturing detergents, we affinity-purified 785 endogenously tagged CEPs and identified stably associated polypeptides by precision mass spectrometry. The resulting high-quality physical interaction network, comprising 77% of targeted CEPs, revealed many previously uncharacterized heteromeric complexes. We found that the secretion of autotransporters requires translocation and the assembly module TamB to nucleate proper folding from periplasm to cell surface through a cooperative mechanism involving the β-barrel assembly machinery. We also establish that an ABC transporter of unknown function, YadH, together with the Mla system preserves outer membrane lipid asymmetry. This E. coli CEP 'interactome' provides insights into the functional landscape governing CE systems essential to bacterial growth, metabolism and drug resistance.

Figure 1

Purification and benchmarking of E. coli CEP assemblies. (a) Schematic of successfully tagged CEPs targeted in each envelope compartment (number in parentheses). PG, peptidoglycan; PL, phospholipid; LPI and LPO, IM- and OM-lipoproteins, respectively. (b) Average correlation (protein spectral counts) and variance of replicate AP/MS analyses. (c) Performance (true-positive rate, TPR, vs. false-positive rate, FPR) of cePPI scoring algorithms (top) and cumulative log-likelihood score (LLS; bottom). Threshold based on literature-curated interactions; Σ LLS refers to LLS HGSCore + LLS CompPASS (Supplementary Note 1). (d) Overlap of unique CEPs (left) and cePPIs (right) in the filtered (high + medium confidence) network compared to large^,, and small-scale (curated in IntAct or EcoCyc databases) PPI studies. (e) Evidence supporting AP/MS-derived cePPIs based on all available experimental, computational or functional association criteria. Number in parenthesis for redundant PPIs indicate prey-prey associations with at least one CEP. (f) Interaction coverage as compared to previous large^,, or small-scale studies as measured against reference cePPIs from EcoCyc.

Figure 2

Global organization and conservation of E. coli MPCs. Schematic showing putative E. coli CEP (purple circles) or cytosolic (brown circles) assemblies. CEP complex membership indicated by circle size, while edges (lines) indicate previously known (red) and novel (brown) interactions; hexagonal nodes represent putative novel components of a known complex or CE system. Topology of the MPC on the periphery indicates if the complex is based on a matrix (M) or socio-affinity (SA) model. Zoom-ins of representative highlighted complexes are shown at the periphery. Inset plot (right) shows evolutionary conservation of cePPIs based on the fraction co-occurrence of orthologs across bacterial classes for interacting CEP pairs observed within and between E. coli MPCs.

Figure 3

CEP interactions and MPCs underlie antibiotic susceptibility. (a) CEP complexes conferring drug resistance; node (complex ID number) size is proportional to the number of interacting components, while edge thickness is proportional to statistical enrichment (Supplementary Note 1). Zoom-in of representative complexes (dotted circles; characterized as socio-affinity model) that confer hypersensitivity to small-molecule inhibitors indicating bait (brown) and prey (green) subunits. (b) Graphs showing AcrA-MdtF-dependent export. Plots indicate the enhanced retention (fluorescence intensity) of ethidium bromide (EtBr, 5 μg/mL) and sensitivity (distance of the diffused killing zones) to carbenicillin or levofloxacin (100 mg/mL) of strains lacking acrAB mdtEF or expressing acrA⁺ or mdtF⁺ alone relative to acrA⁺mdtF⁺ or mdtEF⁺ (+ve; positive control) together; WT, wild type; *P-values calculated using Student’s t-tests; error bars, mean ± s.d. from triplicate biological measurements.

Figure 4

Bam-Tam interconnections mediate autotransporter biogenesis. (a) TamA/B interacting proteins; node color indicates bait proteins involved in translocation and assembly (Tam), autotransporter (Ag43) display, and β-barrel assembly (Bam). (b) Synthetic growth defect of tamA/B-bamA double vs. single mutants and unrelated (ompG, ychM) strains; asterisk indicates hypomorphic allele. (c) Hyper-agglutination of E. coli mutants relative to wild-type (WT) cells expressing the indicated autotransporter; data represented as the mean ± s.d. from triplicate biological measurements. (d) Endpoint settling assay of WT and tam mutants expressing His-tagged Ag43 or a parental vector. (e) Representative flow cytometry profiles (from three biological replicates) using fluorescently labeled anti-His antibodies to quantify autotransporter passenger domain accumulation on the outer surface under normal or thermal denaturation (60 °C, 5 min) conditions; strains transformed with empty plasmid or untagged Ag43 serve as negative controls. (f) Export of Ag43 passenger domain, as measured by mean accumulated cell fluorescence intensity (mean ± s.d.) by flow cytometry. (g) Immunoblot (anti-His) time-course analysis of proteinase-K susceptibility of His-tagged Ag43 at the surface of WT cells and tamB mutants, along with loading controls. (h) Secretion model for type Va autotransporters (Supplementary Note 2).

Figure 5

YadH assists Mla-mediated phospholipid trafficking. (a) Immunoblot (anti-FLAG) analysis of sucrose density gradient CE fractions (lanes 1 and 3 are soluble fractions) from E. coli expressing FLAG-tagged YadH or controls (LptB and MlaA); molecular masses (kDa) are indicated. (b) YadH interactions with Mla pathway members captured by core-attachment algorithm (complex ID: 145; Supplementary Table 4); PPIs captured among Mla transporters, between Mla and surface-exposed phospholipid (PldA), or as detected by repeat AP/MS or bacterial two-hybrid (B2H) showing corresponding bait (brown) and prey (green) proteins. Zoom-in shows β-gal (i) produced by colonies (dashed boxes; EL designates empty lane). Edges (dotted lines) connecting YadH-MlaABDF reflect B2H (i) or repeat AP/MS (ii) confirmations. (c) Growth sensitivity (OD₆₀₀ after 24-h culture at 32 °C) of single and double mutants vs. wild-type (WT) cells to 0.5% SDS/1.1 mM EDTA, and suppression of yadH hyper-permeability upon pldA overexpression. *P < 0.05, by Student’s t-test. Error bars, mean ± s.d. (d) Immunoblots of LPS and OMP marker levels in the indicated strains; periplasmic maltose-binding protein (MBP) serves as a loading control (LC). (e) Cardiolipin (CL) species quantified in single and double mutants; values indicate mean ± s.d. in ten biological replicates. *P-value ≤ 0.05, computed using Student’s t-test, refers to double vs. single mutants. (f) Model illustrating the role of YadH in Mla-mediated maintenance of OM lipid asymmetry.