Maternal Immunization Confers Protection to the Offspring against an Attaching and Effacing Pathogen through Delivery of IgG in Breast Milk

Abstract

Owing to immature immune systems and impaired colonization resistance mediated by the microbiota, infants are more susceptible to enteric infections. Maternal antibodies can provide immunity, with maternal vaccination offering a protective strategy. We find that oral infection of adult females with the enteric pathogen Citrobacter rodentium protects dams and offspring against oral challenge. Parenteral immunization of dams with heat-inactivated C. rodentium reduces pathogen loads and mortality in offspring but not mothers. IgG, but not IgA or IgM, transferred through breast milk to the intestinal lumen of suckling offspring, coats the pathogen and reduces intestinal colonization. Protective IgG largely recognizes virulence factors encoded within the locus of enterocyte effacement (LEE) pathogenicity island, including the adhesin Intimin and T3SS filament EspA, which are major antigens conferring protection. Thus, pathogen-specific IgG in breast milk induced during maternal infection or immunization protects neonates against infection with an attaching and effacing pathogen.

Keywords: C. rodentium; Citrobacter rodentium; EPEC; IgG; LEE-virulence factors; breast milk; enteric infection; maternal vaccination; mucosal immunity; neonatal immunity.

Figure 1.. Parenteral immunization protects the offspring, but not the dam, against pathogen oral challenge

(A) Adult C57BL/6 mice were orally- or IP-inoculated with alive or heat-inactivated C. rodentium, respectively. After one month, mice were challenged orally with 5×10⁸ CFU and bacterial counts in feces were determined on the indicated days after infection. Naïve, untreated animals. (B) Serum α-C. rodentium IgG subtypes in orally infected or IP immunized adult mice determined by ELISA two weeks after inoculation (n= 4 mice per group). (C) Immunoblotting of C. rodentium WT and Δler bacterial lysates using serum from orally infected or IP immunized adult mice. Proteins recognized by antibodies in serum were revealed with HRP-anti-mouse IgG. SDS-PAGE gel stained with Coomassie blue is shown to reveal total amounts of cell lysates loaded. (D) Fecal α-C. rodentium IgG subtypes in orally infected or IP immunized adult mice determined by ELISA two weeks after inoculation (n= 4 mice per group). (E) Adult mice (6-week old) and pups (18-day old) were orally infected with 5×10⁸, 5×10⁷ or 5×10⁶ CFU and mouse survival was determined over the indicated time after infection. (F) 18-day old pups from orally infected or IP immunized mothers were challenged orally with 5×10⁸ CFU of C. rodentium and survival rates were determined. Naïve, pups from untreated dams. Representative results from two independent experiments are shown. Data are represented as mean ± SD (A-B, and D). *p < 0.05, **p < 0.01, ***, p<0.001 by Mann-Whitney U test (B and D) or log rank test (E and F, compared to control groups indicated in red). Only statistically significant differences are shown. See also Figure S1.

Figure 2.. Parenteral maternal immunization reduces pathogen loads and intestinal inflammation in the offspring after oral challenge

(A) 18-day old pups from naïve or HK-Cr immunized mothers were challenged orally with 5×10⁸ CFU of C. rodentium and mouse survival was monitored at the indicated days post-infection (dpi). (B-D) Pathogen load in feces at 5 dpi (B), liver (C) and spleen (D) (7 dpi), in pups from naïve and HK-Cr immunized mothers after C. rodentium challenge. (E-F) Representative H&E staining sections (E) and pathology scores (F) of colonic tissue harvested from pathogen challenged pups depicted in B-D. Arrows indicate areas with marked epithelial damage and inflammation and arrowheads pathogen foci in the mucosa. Scale bars, 400 μm. Data show representative (E) or combined (A-D, and F) results from at least two independent experiments. Each dot represents one mouse (B-D, and F). Dotted line indicates the limit of detection. *p < 0.05, **p < 0.01, ***, p<0.001 by log rank test (A), unpaired two-tailed T- test (B) or Mann-Whitney U test (C-D, and F). See also Figure S2.

Figure 3.. Passive immunity is transferred from the mother to the offspring mainly through breast milk.

(A) Experimental design. Newborn pups from naïve and HK-Cr immunized dams were cross-fostered and nursed by the indicated recipient mother. Pups were then challenged with 5×10⁸ CFU of C. rodentium when they were 18-day old. (B) Survival rates of the challenged pups over the indicated days post-infection (dpi). Prenatal, pups from immunized mothers nursed by naïve dams; Postnatal, pups from naïve mothers nursed by immunized dams; Naïve, non-fostered pups from untreated mothers. (C-E) Pathogen load in feces at 5 dpi (C), liver (D) and spleen (E) (7 dpi) of the pathogen challenged pups. (F) Pathology scores of cecal and colonic tissues from pups depicted in B-E. (G) Immunoblotting of C. rodentium WT and Δler bacterial lysates with breast milk from naïve or HK-Cr immunized mothers. Proteins recognized by antibodies in milk were revealed with HRP-anti-mouse IgG. SDS-PAGE gel stained with Coomassie blue is shown to reveal total amounts of cell lysates loaded. Data show representative (G) or combined (B-F) results from at least two independent experiments. Each dot represents one mouse (C-F). Dotted line indicates the limit of detection. *p < 0.05, **p < 0.01, ***, p<0.001 by log rank test (B), one-way ANOVA (C) or Kruskal-Wallis test (D-F). Only statistically significant differences are indicated.

Figure 4.. Maternal IgG, but not IgA or IgM, is required for protection of the offspring.

(A) 18-day old pups from naïve or HK-Cr immunized IgA-deficient (Igha^−/−) mothers were challenged with 5×10⁸ CFU of C. rodentium and survival rates were determined over the indicated days post-infection (dpi). (B-D) Pathogen load in feces at 5 dpi (B), liver (C) and spleen (D) (7 dpi) in pups from naïve and HK-Cr immunized Igha^−/− mothers after pathogen challenge. (E) Pathology scores of cecal and colonic tissue harvested from infected pups depicted in B-D. (F) Pups from naïve or HK-Cr immunized pIgR-deficient (Pigr^−/−) mothers were challenged and mouse survival recorded as in A). (G-I) Pathogen load in feces at 5 dpi (G), liver (H) and spleen (I) (7 dpi), in pups from naïve and HK-Cr immunized Pigr^−/− mothers after C. rodentium challenge. (J) Pathology scores of cecal and colonic tissue harvested from infected pups depicted in G-I. (K) Pups from naïve or HK-Cr immunized FcRn^−/−-deficient (Fcgrt^−/−) mothers were challenged and mouse survival recorded as in A). (L-N) Pathogen load in feces at 5 dpi (L), liver (M) and spleen (N) (7 dpi), in pups from naïve and HK-Cr immunized Fcgrt^−/− mothers after C. rodentium challenge. (O) Pathology scores of cecal and colonic tissue harvested from infected pups depicted in L-N. Data represent combined results from two independent experiments. Each dot represents one mouse (B-E, G-J, L-N). Dotted line indicates the limit of detection. *p < 0.05, **p < 0.01, ***, p<0.001 by log rank test (A, F and K) and Mann-Whitney U test (BE, G-J, L-O). ns, not significant. See also Figures S3–5.

Figure 5.. Maternal IgG coats C. rodentium in the intestine of the offspring, promotes phagocytosis of virulent bacteria, and reduces pathogen attachment.

(A) IgG coating of C. rodentium-GFP incubated in vitro with breast milk from naïve or HK-Cr immunized mothers. Left panel, representative flow cytometry profiles; right panel, percentage of IgG-coated C. rodentium-GFP (naïve, n= 4 mice; immunized, n= 5 mice). (B) IgG coating of C. rodentium-GFP in intestinal contents of infected pups from naïve or HKCr immunized mothers. Left panel, representative flow cytometry profiles; right panel, percentage of IgG-coated C. rodentium-GFP (naïve, n= 4 mice; immunized, n= 5 mice). (C) Neutrophil phagocytosis assays in vitro. WT/Δler bacterial mixtures (1:1) were treated with breast milk from naïve or HK-Cr immunized dams and incubated with neutrophils harvested from the peritoneal cavity of naïve mice. Ratios (WT/Δler) of intracellular bacteria in gentamicin-treated neutrophils are shown (naïve, n= 5 mice; immunized, n= 6 mice). (D) Neutrophil phagocytosis assays in vivo. WT/Δler mixtures (1:1) were IP-injected into the neutrophil-rich peritoneal cavity of pups from naïve or HK-Cr immunized mothers. Ratios (WT/ler) of intracellular bacteria in gentamicin-treated neutrophils are shown (naïve, n= 6 mice; immunized, n= 6 mice). (E-F) Levels of C. rodentium attached to intestinal epithelium. Pups from naïve or HK-Cr immunized mothers were orally inoculated with 5×10⁸ CFU of a C. rodentium-lux reporter strain. Cecal and colonic tissues were harvested, washed to remove non-adherent bacteria, and used to assess bacterial bioluminescence (E) and total bacteria in tissue homogenates (F). Data show representative (A and B, left panels; and E) or combined (A and B, right panels; C, D and F) results from two independent experiments. Data are represented as mean ± SD (A-B, right panels; C and D). Each dot represents one mouse (F). *p < 0.05, **p < 0.01, ***, p<0.001 by Mann-Whitney U test (A to D) or unpaired two-tailed T-test (F).

Figure 6.. LEE virulence factors are required to confer passive immunity against C. rodentium

(A) 18-day old pups from mothers immunized with WT or Δler HK-Cr were challenged with 5×10⁸ CFU of C. rodentium and mouse survival was determined over the indicated days post-infection (dpi). Naïve, pups from untreated dams. (B-D) Pathogen load in feces at 5 dpi (B), liver (C) and spleen (D) (7 dpi) in pups from naïve and HK-Cr immunized mothers after C. rodentium challenge. (E) Pathology scores of cecal and colonic tissue harvested from infected pups depicted in B-D. Data represent combined results from two independent experiments. Each dot represents one mouse (B-E). Dotted line indicates the limit of detection. *p < 0.05, **p < 0.01, ***, p<0.001 by log rank test (A, compared to naïve group), one-way ANOVA (B) or Kruskal-Wallis test (C-E). Only statistically significant differences are indicated.

Figure 7.. Maternal vaccination with purified LEE antigens confers protection to the offspring against C. rodentium.

(A) Immunoblotting of purified Intimin and EspA using serum or breast milk from mothers immunized with WT or Δler HK-Cr. Proteins recognized by antibodies in serum and milk were revealed with HRP-anti-mouse IgG. SDS-PAGE gel stained with Coomassie blue is shown to reveal total amounts of protein loaded. (B) Antigen-specific IgG subtypes determined by ELISA using serum from dams immunized with the indicated proteins. OVA, Ovalbumin. (C) 19-days old pups from dams immunized with the indicated proteins were challenged with 5×10⁷ CFU of C. rodentium and survival rates were determined over the indicated days post-infection (dpi). (D-F) Pathogen load in feces at 5 dpi (D), liver (E) and spleen (F) (7 dpi) in the pathogen challenged pups. (G) Pathology scores of cecal and colonic tissue harvested from the infected pups depicted in D-F. Data show representative (A and B) or combined (C-G) results from two independent experiments. Each dot represents one mouse (D-G). Dotted line indicates the limit of detection. *p < 0.05, **p < 0.01, ***, p<0.001 by log rank test (C, compared to OVA group), one-way ANOVA (D and G) or Kruskal-Wallis test (E and F). Only statistically significant differences are indicated. See also Figures S6–7.