Insights into the structure and assembly of a bacterial cellulose secretion system

Abstract

Secreted exopolysaccharides present important determinants for bacterial biofilm formation, survival, and virulence. Cellulose secretion typically requires the concerted action of a c-di-GMP-responsive inner membrane synthase (BcsA), an accessory membrane-anchored protein (BcsB), and several additional Bcs components. Although the BcsAB catalytic duo has been studied in great detail, its interplay with co-expressed subunits remains enigmatic. Here we show that E. coli Bcs proteins partake in a complex protein interaction network. Electron microscopy reveals a stable, megadalton-sized macromolecular assembly, which encompasses most of the inner membrane and cytosolic Bcs components and features a previously unobserved asymmetric architecture. Heterologous reconstitution and mutational analyses point toward a structure-function model, where accessory proteins regulate secretion by affecting both the assembly and stability of the system. Altogether, these results lay the foundation for more comprehensive models of synthase-dependent exopolysaccharide secretion in biofilms and add a sophisticated secretory nanomachine to the diverse bacterial arsenal for virulence and adaptation.

Fig. 1

E. coli-like cellulose secretion systems—components and macrocomplex detection. a E. coli bcs operon organization (top) and examples of organisms sharing E. coli-like cellulose secretion system conservation. b Predicted localization and function of the encoded Bcs components. Domain architectures for all Bcs proteins are detailed in Supplementary Fig. 1. c Chromosomal epitope-tagging of the BcsQ and BcsA subunits and functional validation of the recombinant proteins. The sequence of the respective N-terminal and C-terminal epitope tags is shown on the left; calcofluor-binding and fluorescence of the recombinant E. coli 1094 strains is shown on the right. d Thumbnail summary of co-purifying Bcs partners as determined by anti-FLAG affinity purification and mass-spectrometry analyses

Fig. 2

Electron microscopy analysis of the Bcs macrocomplex. a Purification of the Bcs macrocomplex expressed in the overexpression E. coli 1094 bcsA ^HA-FLAG 2K7 strain. SDS-PAGE of the affinity-gradient and density gradient-purified Bcs macrocomplex stained with Coomassie blue. Band labels are as identified by mass-spectrometry and immunoblotting. Asterisks denote presence of consistently identified contaminants discussed in Supplementary Fig. 2. b A representative micrograph of the negatively stained Bcs macrocomplex from the dataset used for image processing. c Representative views (class averages of 2D projections; not to scale) of the negatively stained Bcs macrocomplex. d Structure of the Bcs macrocomplex at 16.7 Å resolution. Different views and the relative rotational angles are shown. The characteristic layers and crown repeats are indicated in the bottom. The overall dimensions of the complex are indicated by size bars (top left and middle). The approximate molecular weight of the complex calculated from volumetric analyses of the reconstruction is indicated in megadaltons (MDa). Owing to the nature of the negatively stained sample, this estimation accounts only for the low-resolution envelope reconstruction and not intrinsic to the proteins electron density

Fig. 3

Map interpretation and electron microscopy studies of an inner membrane (IM) subcomplex. a Representative fitting of a BcsB homology model in a crown repeat volume and proposed orientation of the Bcs macrocomplex relative to the inner membrane. b Co-expressed subunits for recombinant overexpression and Bcs complex reconstitution in BL21 (DE3) cells (macrocomplex: top; IM subcomplex: bottom). c Structural and functional analyses of the recombinantly expressed Bcs macrocomplex. Representative 2D class averages are shown on the top (not to scale); the 3D structure reconstruction is shown on the bottom. End-point de novo cellulose synthesis by the purified complex is shown on the bottom; results are representative of two biological replicates. The data is plotted relative to cellulose detection in a wild-type E. coli 1094 cell culture, with bcs operon-deleted strain serving as negative control. d Structural analyses of the IM subcomplex. A coomassie-stained SDS-PAGE gel, 2D class averages (not to scale) and different views of the 3D structure reconstruction are shown from left to right. The corresponding molecular weight is indicated in megadaltons and the dimensions of the subcomplex are indicated by size bars. e. Fitting of the IM subcomplex into the Bcs macrocomplex volume and proposed orientation relative to the inner membrane

Fig. 4

Bcs protein interactions and functional model of cellulose microfibril secretion. a A summary of the direct protein–protein interactions as observed by bacterial two-hybrid screen. b Proposed microfibril formation at the periplasmic or cell surface level. The hexameric assembly is proposed to secure a cell pole-localized nanoarray for cellulose biogenesis and microfibril formation, the open architecture is proposed to secure access for the hydrolizing endo-1,4-β-d-glucanase BcsZ necessary for maximal cellulose production. The BcsC TPR-rich periplasmic motifs are shown in red, whereas the C-terminal outer membrane porin domain is shown in yellow. Microfibril formation at the periplasmic level would require the assembly of a wider composite porin for outer membrane secretion (left)