Outer membrane protein size and LPS O-antigen define protective antibody targeting to the Salmonella surface

Abstract

Lipopolysaccharide (LPS) O-antigen (O-Ag) is known to limit antibody binding to surface antigens, although the relationship between antibody, O-Ag and other outer-membrane antigens is poorly understood. Here we report, immunization with the trimeric porin OmpD from Salmonella Typhimurium (STmOmpD) protects against infection. Atomistic molecular dynamics simulations indicate this is because OmpD trimers generate footprints within the O-Ag layer sufficiently sized for a single IgG Fab to access. While STmOmpD differs from its orthologue in S. Enteritidis (SEn) by a single amino-acid residue, immunization with STmOmpD confers minimal protection to SEn. This is due to the OmpD-O-Ag interplay restricting IgG binding, with the pairing of OmpD with its native O-Ag being essential for optimal protection after immunization. Thus, both the chemical and physical structure of O-Ag are key for the presentation of specific epitopes within proteinaceous surface-antigens. This enhances combinatorial antigenic diversity in Gram-negative bacteria, while reducing associated fitness costs.

Fig. 1. STmOmpD induces antibody that can impair infection against STm.

a WT mice were primed and boosted i.p. with 20 μg of STmOmpD at days 0, 14, and 28, and sera obtained at day 35. Anti-STm (D23580 and SL1344) IgG was assessed by ELISA. Each point represents titers from an individual mouse, and is representative of three repeat experiments. b Bacterial numbers in the liver, spleen, and blood of nonimmunized (NI) and WT mice immunized twice (days 0 and 14) with 20 μg of STmOmpD and challenged with 1 × 10⁵ STm D23580 for 24 h. Each point represents bacterial numbers in one mouse. c Mice were infected i.p. with 5 × 10⁵ of either STm SL3261 or bacteria previously opsonized with STmOmpD antibodies. Bacteria numbers were quantified at day 3 from the spleen and liver. Data are representative of two independent experiments. d Serum bactericidal assays using STm SL1344 with sera from NI mice or mice immunized as in a, supplemented with human sera as a complement source with anti-STm-specific antibodies depleted. Normal human serum (NHS, blue) and heat-inactivated human serum (HiS, green) were used as controls. Negative values correspond with a decrease in viable STm compared with starting normals. Samples were taken at minutes 45, 90, and 180 using a starting concentration of 10⁶ CFU/ml. Sera from NI (yellow) and STmOmpD immune (red) WT mice were tested. Data are representative of two different experiments and represent the mean of 5–6 mice sera used. **P ≤ 0.01 assessed by two-tailed Mann–Whitney U test.

Fig. 2. Antibodies to STmOmpD offer limited cross-protection to SEn.

a Alignment of OmpD sequences from STm and SEn showing the site where an alanine-to-serine substitution is observed at amino acid 263 of the whole sequence. b Serum IgG titers to STm and SEn porins from WT NI mice and mice immunized with STmOmpD as in Fig. 1a. Each point represents the titers from one mouse. Data representative of two experiments. c NI mice and mice immunized twice with 20 μg of STmOmpD were challenged with 5 × 10⁵ aroA⁻ STm or SEn, and bacterial numbers per organ enumerated 3 days later. Each point represents the titers from one mouse. Data are representative of two individual experiments. *P ≤ 0.05, **P ≤ 0.01 assessed by two-tailed Mann–Whitney U test.

Fig. 3. Loss of O-Ag from SEn enhances protection by anti-STmOmpD antibodies.

a IgG titers to STm and SEn WT and O-Ag-deficient mutants in sera from NI mice and mice immunized with STmOmpD. b Serum bactericidal assays using STm D23580, SEn D24954, and STm and SEn wbaP⁻ mutants with sera from NI mice or mice immunized as in Fig. 1a, supplemented with human sera as a complement source with anti-STm-specific antibodies depleted. Normal human serum (NHS, blue) and heat-inactivated human serum (HiS, green) were used as controls. Negative values correspond with a decrease in viable STm compared with starting normals. Samples were taken at minutes 45, 90, and 180 using a starting concentration of 10⁶ CFU/ml. Sera from NI (yellow) and STmOmpD immune (red) WT mice were tested. c NI mice and mice immunized twice with 20 μg of STmOmpD were challenged with 1 × 10⁵ of STm or SEn (WT or wbaP⁻) for 24 h, and bacterial burdens determined from the spleen, liver, and blood. Each point represents the titers from one mouse. Data are from three different experiments. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001, NS = nonsignificant assessed by two-tailed Mann–Whitney U test.

Fig. 4. Acessibility of OmpD epitopes is determined by the LPS component.

OmpD structure and the location of the A263S polymorphism in the context of interaction with the LPS layer. a OmpD trimer (homology model) with the location of the A263S variation (blue spacefill representation) on the tip of the extracellular loop 6 (labeled L6, green in protomer 1). The other extracellular loops in this protomer are shown in different colors consistent with their representation in Supplementary Figs. S2 and S6. b Snapshot of the full atomistic MD simulation of the outer membrane showing an isolated OmpD trimer with a single LPS molecule. LPS elements are colored as follows: Lipid A—black; core sugars in red; O-Ag repeats in rainbow colors. c Size comparison of a murine IgG1 molecule (based on PDB 1IGY) to the OmpD trimer and OmpA. Top shows an IgG1 molecule. The heavy chains are colored in green and gold, while the light chains are in cyan and purple, respectively. Dimensions of various parts are indicated in Å. Bottom left, an OmpD trimer (protomers shown in yellow, blue, and red) and bottom right an OmpA monomer (red). d A wider context of the MD system is shown from the membrane plane (left) and from the top (right). OmpD protomers are colored in purple, red, and green, and LPS elements are colored as follows: Lipid A and inner leaflet in gray, core sugars in gold, and O-Ag in dark blue. The dense network provided by the LPS O-Ag prevents access to the surface epitopes on OmpD, which is especially clear on the spacefilling top-view representation on the right, showing a typical access channel dimension of 50 Å.

Fig. 5. O-Ag structure and variability influence access of antibody to OmpD.

a IgG titers to STm and SEn WT and LPS chimera strains were assessed using sera from NI mice and mice immunized with STmOmpD as in Fig. 1a. Each point represents the titers from one mouse. Representative of three experiments. b Serum bactericidal assays using STm D23580, SEn D24954, and STm O9 and SEn O4 chimera mutants with sera from nonimmunized mice (yellow) or mice immunized with STmOmpD (red) as in Fig. 1a, supplemented with human sera as a complement source with anti-STm-specific antibodies depleted. Normal human serum (NHS, blue) and heat-inactivated human serum (HiS, green) were used as controls. Negative values correspond with a decrease in viable STm compared with starting normals. Samples were taken at minutes 45, 90, and 180 using a starting concentration of 10⁶ CFU/ml. c Bacterial burdens from the spleens of NI and mice immunized twice with 20 μg of STmOmpD infected for 24 h with 1 × 10⁵ of STm or SEn WT and O-Ag chimeras. Data are representative of a minimum of two individual experiments. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, NS = nonsignificant assessed by two-tailed Mann–Whitney U test.

Fig. 6. Anti-OmpA antibodies are not protective against infection.

a A snapshot from the OmpA (red) after equilibration in the SEn outer membrane (membrane colored as in Fig. 4c) viewed from the membrane plane (left) and extracellular side (right) indicates that it will not be accessible to IgG due to LPS occlusion. b (Left) WT mice were challenged with 5 × 10⁵ STm SL3261 for 5 days, and the frequency of antibody-secreting cells (ASC) in the spleen assessed by ELISPOT. (Right) Anti-OmpA and anti-STm IgG titers of WT mice immunized twice with 20 μg of alum-precipitated OmpA were determined. Each point represents the titers from one mouse. Representative of two experiments. c Splenic bacterial burdens of WT NI mice and mice immunized with 20 μg of alum-precipitated OmpA mice challenged with 5 × 10⁵ STm for 4 days. Similar results seen at other time points up to 7 days post challenge. Each point represents the titers from one mouse. Data are representative of two independent experiments. *P ≤ 0.05, ***P ≤ 0.001, NS = nonsignificant assessed by two-tailed Mann–Whitney U test.