Innate immune lectins kill bacteria expressing blood group antigen

Abstract

The expression of ABO(H) blood group antigens causes deletion of cells that generate self-specific antibodies to these antigens but this deletion limits adaptive immunity toward pathogens bearing cognate blood group antigens. To explore potential defense mechanisms against such pathogens, given these limitations in adaptive immunity, we screened for innate proteins that could recognize human blood group antigens. Here we report that two innate immune lectins, galectin-4 (Gal-4) and Gal-8, which are expressed in the intestinal tract, recognize and kill human blood group antigen-expressing Escherichia coli while failing to alter the viability of other E. coli strains or other Gram-negative or Gram-positive organisms both in vitro and in vivo. The killing activity of both Gal-4 and Gal-8 is mediated by their C-terminal domains, occurs rapidly and independently of complement and is accompanied by disruption of membrane integrity. These results demonstrate that innate defense lectins can provide immunity against pathogens that express blood group-like antigens on their surface.

Figure 1. Gal-3, Gal-4, and Gal-8 recognize blood group B positive E. coli

(a–d) Glycan microarray data obtained following incubation with (a) 0.2 μM Gal-1, (b) 0.2 μM Gal-3, (c) 0.5 μM Gal-4, and (d) 0.02 μM Gal-8. RFU = relative fluorescence units represented on the y-axis. Error bars = +/- 1 SEM. See Supplementary Table 1 for complete list of glycans represented on the x-axis. (e) Structure of E. coli O86 O antigen. (f-i) Flow cytometric analysis following incubation of E. coli O86 with (f) Gal-1, (g) Gal-3, (h) Gal-4, and (i) Gal-8 all tested at ~0.1 μM with or without inclusion of 20 mM lactose (Lac) where indicated.

Figure 2. Gal-4 and Gal-8 kill blood group B positive E. coli

E. coli O86 (BG B⁺ E. coli) were mixed with (a) 5 μM Gal-1, Gal-3, Gal-4, or Gal-8, (b) 5 μM Gal-4 with or without 20 mM lactose (Lact.) or 20 mM sucrose (Sucr.), (c) 5 μM Gal-8 with or without 20 mM lactose (Lact.) or 20 mM sucrose (Sucr.), or (d) the indicated concentrations of Gal-1, Gal-3, Gal-4, or Gal-8. Viable bacteria were quantified by dilution plating, n=3, 1 representative experiment in duplicate over 2 dilutions shown (a–c), error bars=SD. (e) Still-frame images from real-time video microscopy demonstrating bacterial mobility at 10-s intervals before and after addition of 5 μM Gal-8 as indicated (see Supplementary video 1). Arrows indicate one group of immobilized bacteria. Scale bars = 100 μm. (f) E. coli O86 (BG B⁺ E. coli) were grown to mid-log phase followed by addition of 5 μM Gal-8. Untreated and Gal-8 treated bacteria were stained with propidium iodide (red) and visualized by fluorescence microscopy. Scale bars = 100 μm. (g) Transmission electron microscopy images of E. coli O86 (BG B⁺ E. coli) following addition of PBS (NT) or 5 μM Gal-8. Lower panels show close up view of single bacterium. Scale bars = 500 nm. (h) Scanning electron microscopy images of E. coli O86 (BG B⁺ E. coli) followed by addition of PBS (NT) or Gal-8. Scale bars = 500 nm.

Figure 3. Gal-4 and Gal-8 kill BG B⁺ E. coli solely through the C-terminal domain

(a) 5 μM Gal-8, Gal-8R233H, or Gal-8R69H were added to mid-log phase E. coli O86 (BG B⁺ E. coli). Viable bacteria were quantified by dilution plating, n=3, 1 representative experiment in duplicate over 2 dilutions shown, error bars=SD. (b) Flow cytometric analysis following incubation of E. coli O86 (BG B⁺ E. coli) with Gal-8N or Gal-8C at ~0.1 μM with or without inclusion of 20 mM lactose (Lac) where indicated. (c) 5 μM Gal-8, Gal-8N, or Gal-8C were added to mid-log phase E. coli O86 (BG B⁺ E. coli). Viable bacteria were quantified by dilution plating, n=3, 1 representative experiment in duplicate over 2 dilutions shown, error bars=SD. (d) Flow cytometric analysis following incubation of E. coli O86 (BG B⁺ E. coli) with Gal-4N or Gal-4C at ~0.1 μM with or without inclusion of 20 mM lactose (Lac) where indicated. (e) 5 μM Gal-4, Gal-4N, or Gal-4C were added to mid-log phase E. coli O86 (BG B⁺ E. coli). Viable bacteria were quantified by dilution plating, n=3, 1 representative experiment in duplicate over 2 dilutions shown, error bars=SD.

Figure 4. Gal-4 and Gal-8 specifically kill blood group B positive E. coli

(a) Flow cytometric analysis of galectin binding after incubation of BG B⁺ E. coli and two different BG B⁻ E. coli reference strains obtained from a clinical laboratory with ∼0.1 μM Gal-8. (b–c) Incubation of (b) BG B⁺ E. coli or (c) BG B⁻ E. coli strain 1 with 5 μM Gal-1, Gal-3, Gal-4, or Gal-8 as indicated. Viable bacteria were quantified by dilution plating, n=3, representative experiment in duplicate over 2 dilutions shown, error bars=SD. (d) Flow cytometric analysis following incubation of BG B⁺ E. coli, K. pneumoniae, P. aeruginosa, and S. aureus with ∼0.1 μM Gal-8. (e–g) Incubation of (e) K. pneumoniae, (f) P. aeruginosa, or (g) S. aureus with 5 μM Gal-1, Gal-3, Gal-4, or Gal-8 as indicated. Viable bacteria were quantified by dilution plating, n=3, representative experiment in duplicate over 2 dilutions shown, error bars=SD. (h–i) Incubation with or without 5 μM Gal-8 with (h) GFP⁺ P. aeruginosa alone or (i) GFP⁺ P. aeruginosa mixed with BG B⁺ E. coli followed by determination of percent GFP⁺ P. aeruginosa by flow cytometric analysis in a mixing experiment. Gated values of GFP⁺ bacteria treated with PBS (blue) or Gal-8 (red) are shown. (j) Quantification of percent GFP⁺ bacteria utilizing flow cytometric analysis obtained following incubation of Gal-8 with either GFP⁺ P. aeruginosa alone (P.a.) or GFP⁺ P. aeruginosa mixed with BG B⁺ E. coli (P.a. + BG B⁺ E.c.).

Figure 5. Gal-4 and Gal-8 specifically recognize blood group B antigen on blood group B positive E. coli

(a) Schematic of O antigen structures on wild type (WT) BG B⁺ E. coli and mutants of BG B⁺ E. coli WaaL⁻ (Ligase-) and Wzy⁻ (Polymerase-) lacking a complete O antigen. (b) Flow cytometric analysis following incubation of BG B⁺ E. coli and mutants WaaL⁻ and Wzy⁻ with ∼0.1 μM Gal-8. (c–d) Incubation of (c) WT and WaaL⁻ mutant BG B⁺ E. coli or (d) WT and Wzy⁻ mutant BG B⁺ E. coli with 5 μM Gal-4 or Gal-8 as indicated. Viable bacteria were quantified by dilution plating, n=3, representative experiment in duplicate over 2 dilutions shown, error bars=SD.

Figure 6. Gal-4 and Gal-8 specifically kill blood group B positive E. coli in vivo

(a) Flow cytometric analysis following incubation of BG B⁺ E. coli with ∼0.1 μM mGal-4 with or without lactose. (b) Flow cytometric analysis following incubation of BG B⁺ E. coli and mutants WaaL⁻ with ∼0.1 μM mGal-4. (c) mGal-4 binding to the CFG glycan microarray at 20 μg/ml (0.5 μM). BGB= blood group B glycans, BGA= blood group A glycans (See Supplementary Table 1). (d) Quantification of WT BG B⁺ and ΔwaaL mutant E. coli after incubation with ~5 μM mGal-4. Viable bacteria were quantified by dilution plating; n = 3 experiments; one representative experiment in duplicate over two dilutions is shown; error bars represent means ± s.d. (e) Live antibiotic-treated mice were fed PBS, Wild type (WT), or WaaL⁻ mutant BG B+ E. coli. The number of viable bacteria in the intestine of mice sacrificed 24 h after feeding was quantified by dilution plating. * = p value 0.049. (f) Growth of WT and WaaL⁻ mutant BG B⁺ E. coli in the presence and absence of TDG. * = p value 0.008. (g) Schematic of O antigen structures on α-Gal E. coli. (h) Flow cytometric analysis of α-Gal expressing bacteria after incubation of α-Gal–expressing bacteria with ~0.1 μM human Gal-4 or Gal-8. (i) Bar graph showing the percentage of α-Gal–expressing bacteria and BG B⁻ bacteria remaining after incubation with 5 μM Gal-4 and Gal-8 as compared to PBS-treated control bacteria. Viable bacteria were quantified by dilution plating, n=3, representative experiment in duplicate over 2 dilutions shown, error bars=SD.