Fig 1.
Genes encoding the l-rhamnose biosynthesis pathway are distributed in listeriae and other bacterial species.
Comparison of the genomic organization of the l-rhamnose pathway genes in the genus Listeria and other bacteria. The corresponding species and strains are indicated on the left (Lmo, Listeria monocytogenes; Lin, Listeria innocua; Lse, Listeria seeligeri; Liv, Listeria ivanovii; Lwe, Listeria welshimeri; Smu, Streptococcus mutans; Mtu, Mycobacterium tuberculosis; Sen, Salmonella enterica serovar Typhimurium; Sfl, Shigella flexneri; Pae, Pseudomonas aeruginosa) and listerial serotypes are indicated on the right. Genes are represented by boxed arrows and their names are provided for strain EGD-e. Operons are underlined by dashed arrows and homologs of the rml genes are shown with identical colors. Numbered gaps indicate the genetic distance (Mb, mega base pairs) between rml genes located far apart in the chromosome. Bacterial genomic sequences were obtained from NCBI database and chromosomal alignments assembled using Microbial Genomic context Viewer and Adobe Illustrator.
Fig 2.
A functional rml operon is required for glycosylation of Lm WTAs with l-rhamnose.
(A) Alcian blue-stained 20% polyacrylamide gel containing WTA extracts from logarithmic-phase cultures of different Lm strains. (B–D) HPAEC-PAD analyses of the sugar composition of the (B) WTA, (C) peptidoglycan and (D) cytoplasmic fractions isolated from the indicated Lm strains. Samples were hydrolyzed in 3 M HCl (2 h, 95°C), diluted with water and lyophilized before injection into the HPLC equipment. Standards for ribitol (Rib), l-rhamnose (Rha), glucosamine (GlcN), and muramic acid (Mur) were eluted under identical conditions to allow peak identification.
Fig 3.
WTA l-rhamnosylation promotes Lm resistance against AMPs.
(A) Growth of Lm strains in BHI broth supplemented with 5% NaCl. A growth curve of wild type EGD-e in the absence of 5% NaCl was included as a control for optimal growth. (B) Growth of mid-exponential-phase Lm strains untreated (black symbols) or challenged with 50 μg/ml (gray symbols) or 1 mg/ml (white symbols) of lysozyme. Optical density of the shaking cultures was monitored spectrophotometrically at 600 nm. (C) Quantification of viable bacteria after treatment of mid-exponential-phase Lm strains (2 h, 37°C) with gallidermin (1 μg/ml), CRAMP or LL-37 (5 μg/ml). Averaged replicate values from AMP-treated samples were normalized to untreated control samples and the transformed data expressed as the percentage of surviving bacteria relative to wild type Lm (set at 100). Data represent mean±SD of three independent experiments. *, p≤0.05; ***, p≤0.001.
Fig 4.
WTA l-rhamnosylation interferes with the Lm cell wall crossing by AMPs.
(A and B) Flow cytometry analysis of Lm surface-exposed CRAMP levels in mid-exponential-phase Lm strains, following incubation (5 min) in a 5-μg/ml solution of the peptide and immunolabeling with anti-CRAMP and Alexa Fluor 488-conjugated antibodies. (A) Representative experiment showing overlaid histograms of CRAMP-treated (solid line) and untreated (dashed line) samples, with mean fluorescence intensity (MFI) values from treated samples indicated by vertical dashed lines. (B) Mean±SD of the MFI values of CRAMP-treated samples from three independent experiments. (C) Cell surface charge analysis of Lm strains deficient for WTA l-rhamnosylation as determined by cytochrome c binding assays. Mid-exponential-phase bacteria were incubated with equine cytochrome c (0.5 mg/ml), centrifuged and the supernatant was recovered for spectrophotometric quantification of the unbound protein fraction. Values from Lm-containing samples are expressed as the percentage of unbound cytochrome c relative to control samples lacking bacteria. Data represent the mean±SD of three independent experiments. (D and E) Flow cytometry analysis of total Lm-associated CRAMP levels in mid-exponential-phase Lm strains, following incubation (5 min) with a 5-μg/ml solution of fluorescently labeled peptide (5-FAM-CRAMP). (D) Representative experiment showing overlaid histograms of FAM-CRAMP-treated (solid line) and untreated (dashed line) samples, with MFI values from treated samples indicated by vertical dashed lines. (E) Mean±SD of the MFI values of 5-FAM-CRAMP-treated samples from three independent experiments. (F) Fluorometric quantification of the unbound CRAMP fraction in the supernatant of suspensions of mid-exponential-phase Lm strains, following incubation (5 min) with a 5-μg/ml solution of 5-FAM-CRAMP. Data are expressed as the percentage of unbound fluorescent peptide relative to control samples lacking bacteria, and represent the mean±SD of three independent experiments performed in triplicates. ns = not significant, p>0.05; **, p≤0.01; ***, p≤0.001.
Fig 5.
WTA l-rhamnosylation delays AMP interaction with the Lm plasma membrane.
(A) Depolarization rate of Lm strains in response to CRAMP. Mid-exponential-phase bacteria pre-stained (15 min) with 30 μM DiOC2(3) were challenged with 50 μg/ml CRAMP and changes in the membrane potential, expressed as the ratio of CRAMP-treated versus untreated samples, were monitored during 30 min. Data represent the mean±SD of three independent experiments. (B) SYTOX Green uptake kinetics of Lm strains in response to CRAMP-mediated membrane permeabilization. Exponential-phase bacteria were incubated (37°C) with PBS (white symbols) or 50 μg/ml CRAMP (black symbols), in the presence of 1 μM SYTOX Green, and the increase in green fluorescence emission was recorded over time. (C and D) Transmission electron microscopy analysis of the subcellular distribution of CRAMP in immunogold-labeled sections of mid-exponential-phase wild type and ΔrmlACBD Lm strains treated with 50 μg/ml CRAMP (15 min, 37°C). (C) Representative images of contrasted sections of Lm cells showing CRAMP-specific gold labeling (10-nm black dots). Scale bar: 0.2 μm. (D) Quantification of the subcellular partition of CRAMP labeling in wild type and ΔrmlACBD Lm strains, for two independent assays. The percentages of cell envelope- and cytoplasm-associated gold dots per bacterium were quantified (at least 90 cells per strain) and the results expressed for each strain as mean±SD. (E and F) Western blot analysis of levels of CRAMP bound to purified cell wall of different Lm strains. Purified cell wall (100 μg) was incubated with CRAMP (5 min), washed and digested overnight with mutanolysin. (E) Supernatants from mutanolysin-treated samples were resolved in 16% Tris-tricine SDS-PAGE and immunoblotted for CRAMP. The Lm cell wall-anchored protein InlA was used as loading control. (F) Quantification of the relative CRAMP levels represented as the mean±SD of four independent blots. *, p≤0.05; **, p≤0.01.
Fig 6.
WTA l-rhamnosylation is necessary for AMP resistance in vivo and Lm virulence.
(A–D) Quantification of viable bacteria in the spleen and liver recovered from BALB/c mice (n = 5), three days after (A and B) oral or (C and D) intravenous infection with sub-lethal doses of indicated Lm strains. Data are presented as scatter plots, with each animal indicated by a dot and the mean indicated by a horizontal line. (E and F) Quantification of the fecal shedding of wild type or ΔrmlACBD Lm strains after oral infection of (E) wild type (WT, cramp+/+) and (F) CRAMP knockout (KO, cramp-/-) 129/SvJ mice (n = 5). Total feces produced by each animal at specific time points were collected and processed for bacterial enumeration in Listeria-selective agar media. Data are expressed as mean±SD. (G and H) Quantification of viable bacteria in spleens and livers recovered from (G) wild type (WT, cramp+/+) and (H) CRAMP knockout (KO, cramp-/-) 129/Sv mice (n = 5), three days after intravenous infection with sub-lethal doses of wild type or ΔrmlACBD Lm strains. Data are presented as scatter plots, with each animal represented by a dot and the mean indicated by a horizontal line. *, p≤0.05; **, p≤0.01; ***, p≤0.001.
Table 1.
Plasmids and bacterial strains.