Adaptive Synthesis of a Rough Lipopolysaccharide in Geobacter sulfurreducens for Metal Reduction and Detoxification

Abstract

The ability of some metal-reducing bacteria to produce a rough (no O-antigen) lipopolysaccharide (LPS) could facilitate surface interactions with minerals and metal reduction. Consistent with this, the laboratory model metal reducer Geobacter sulfurreducens PCA produced two rough LPS isoforms (with or without a terminal methyl-quinovosamine sugar) when growing with the soluble electron acceptor fumarate but expressed only the shorter and more hydrophilic variant when reducing iron oxides. We reconstructed from genomic data conserved pathways for the synthesis of the rough LPS and generated heptosyltransferase mutants with partial (ΔrfaQ) or complete (ΔrfaC) truncations in the core oligosaccharide. The stepwise removal of the LPS core sugars reduced the hydrophilicity of the cell and increased outer membrane vesiculation. These changes in surface charge and remodeling did not substantially impact planktonic growth but disrupted the developmental stages and structure of electroactive biofilms. Furthermore, the mutants assembled conductive pili for extracellular mineralization of the toxic uranyl cation but were unable to prevent permeation and mineralization of the radionuclide in the cell envelope. Hence, not only does the rough LPS promote cell-cell and cell-mineral interactions critical to biofilm formation and metal respiration but it also functions as a permeability barrier to toxic metal cations. In doing so, the rough LPS maximizes the extracellular reduction of soluble and insoluble metals and preserves cell envelope functions critical to the environmental survival of Geobacter bacteria in metal-rich environments and their performance in bioremediation and bioenergy applications. IMPORTANCE Some metal-reducing bacteria produce an LPS without the repeating sugars (O-antigen) that decorate the surface of most Gram-negative bacteria, but the biological significance of this adaptive feature was not previously investigated. Using the model representative Geobacter sulfurreducens strain PCA and mutants carrying stepwise truncations in the LPS core sugars, we demonstrate the importance of the rough LPS in the control of cell surface chemistry during the respiration of iron minerals and the formation of electroactive biofilms. Importantly, we describe hitherto overlooked roles for the rough LPS in metal sequestration and outer membrane vesiculation that are critical for the extracellular reduction and detoxification of toxic metals and radionuclides. These results are of interest for the optimization of bioremediation schemes and electricity-harvesting platforms using these bacteria.

Keywords: biofilms; cell envelope; electromicrobiology; extracellular electron transfer; metal reduction; outer membrane vesicles; type IV pili.

FIG 1

LPS biosynthesis in G. sulfurreducens. (A) Illustration of the structure of the rough LPS variants (with or without the methyl-quinovosamine sugar [Q]), including the lipid A, kdo₂ disaccharide, and core oligosaccharide sugar chains attached to heptoses HepI, HepII, and HepIII. (B) Isolation and SDS-PAGE (12%) separation of the two LPS isoforms with (Q⁺) and without (Q⁻) methyl-quinovosamine from nonpiliated (30°C) or piliated (25°C) cells grown in NBAFYE medium. Piliated cells from acetate-Fe(III) oxide cultures (FWAFeOx medium, 30°C) produced only the short LPS variant without methyl-quinovosamine. (C and D) Genomic reconstruction of the kdo₂-lipid A biosynthetic pathway (C) and identification of a genomic cluster with genes involved in the synthesis of the core oligosaccharide (D). The cluster shown in panel D includes the three heptosyltransferases, namely, HepI (rfaC), HepII (rfaF), and HepIII (rfaQ). Overlapping genes are shown in a lower row. Color coding: blue, lipid A biosynthesis; gray, kdo attachment; black, heptosyltransferases.

FIG 2

Effect of LPS truncation on cell surface topology. Topographic images of the WT (A), ΔrfaQ (B), and ΔrfaC (C) cell surface collected with an AFM operated in tapping mode at 0.3 Hz. The net negative charge and polymeric nature of the LPS core oligosaccharide softened interactions between the WT cell surface and the AFM tip and resolved a smooth surface topography (A). The stepwise truncation of the LPS core in the ΔrfaQ (B) and ΔrfaC (C) mutants reduced the LPS polymeric cushion and net negative charge of the cell, promoting surface-tip interactions that show as topographic scratches on the cell surface.

FIG 3

Effect of LPS truncation on growth and OMV production. (A) Biochemical characterization of the LPS isolated from WT, ΔrfaQ, and ΔrfaC cells by SDS-PAGE in a 4 to 20% polyacrylamide gel and silver staining (the ruler on the right marks the approximate migration of protein standards but cannot be used for LPS mass determination). The two bands in the WT and ΔrfaQ samples are expected for LPS variants with (Q⁺) and without (Q⁻) the terminal methyl-quinovosamine sugar (Q). The complete truncation of the core oligosaccharide in the ΔrfaC mutant produced a single band containing the lipid A-kdo₂ moiety. (B) Average growth (and standard deviation) of the WT strain and the LPS mutants in triplicate wells of a microtiter plate incubated inside an anaerobic chamber. (C) OMV production by mid-log-phase cells of the WT strain and LPS mutants. The images show representative AFM scans (tapping mode, with 0.3-Hz scan rates) of mid-log-phase culture samples (OD₆₀₀ of ∼0.5), showing cells (∼500-nm diameter) and the much smaller OMVs after chemical fixation on a highly oriented pyrolytic graphite stage. We used ImageJ to estimate the OMV counts per field and to calculate the average number of OMVs per square micrometer in triplicate viewing fields for each strain. *, P < 0.05, two-tailed t test. All cultures were grown in anoxic NBAFYE medium at 30°C except for those in panel C, which used an optimal temperature for growth (35°C) that minimized vesiculation in the WT strain.

FIG 4

Effect of LPS truncations on surface hydrophobicity and biofilm formation. (A) Relative surface hydrophobicity (SHb) of WT and LPS mutant cells, measured as cell aggregation in the presence of hydrocarbons (xylene, toluene, and hexane). Shown are averages and standard deviations of four measurements, corresponding to two independent cultures each assayed in two technical replicates per strain. *, P < 0.05; **, P < 0.005; ***, P < 0.0005, two-tailed t test. (B and C) Biofilm formation in the WT strain and the LPS mutants after 24, 48, and 72 h of incubation at 30°C. Each strain was grown in triplicate wells of a 6-well plate. Two of the wells were used for crystal violet staining of the biofilm biomass (B), and the third was used for fluorescence staining with the viable dye SYTO 9 and top and side biofilm image reconstructions (scale bar, 20 μm) with a confocal scanning laser microscope (C).

FIG 5

Uranium immobilization by the WT strain and LPS mutants of G. sulfurreducens. (Left) Transmission electron micrographs of unstained resting cells treated with uranyl cation for 6 h. The images show the reductive precipitation of the uranyl cation along the conductive pili (black arrows) in all of the strains and increased electron density (darker cells) in the LPS mutant cells, which is consistent with the periplasmic mineralization of uranium. The white arrows show examples of cytoplasmic blebbing and L-forms in the LPS mutants but not in the WT cells. (Right) Outer membrane production in mid-log-phase NBAFYE cultures (OD₆₀₀ of 0.4 to 0.6) grown at 30°C in the absence (0 mM) or presence (1 mM) of uranyl acetate. Culture aliquots (10 μl) were deposited and chemically fixed on the surface of a highly oriented pyrolytic graphite stage before AFM imaging in tapping mode with a 0.3-Hz scan rate. *, P < 0.01.