450 - Maternal Lactoferrin Reduces Vertically-Acquired Escherichia coli Infection in the Offspring without Modifying Innate Immune Response in a Mouse Model

Publication Number: 2438.450

Mark Gamadia, Kansas City University, Kansas City, MO, United States; Joseph M. Varberg, Children's Mercy Hospitals and Clinics, Kansas City, MO, United States; Joshua L. Wheatley, Children's Mercy Hospitals and Clinics, Olathe, KS, United States; Sivani Badrivenkata, University of Kansas, Lawrence, KS, United States; Michael J.. Hageman, University of Kansas, Lawrence, KS, United States; Susana Chavez-Bueno, Children's Mercy Kansas City and University of Missouri Kansas City, Kansas City, MO, United States

Background: Escherichia coli is a leading pathogen in neonatal early-onset sepsis (EOS), a major cause of newborn mortality worldwide. Maternal administration of lactoferrin, an antimicrobial and immunomodulatory glycoprotein, may reduce EOS by protecting the offspring against vertical E. coli infection while modulating associated innate immune responses.

Objective: To determine if maternal vaginal administration of bovine lactoferrin (BLF) ameliorates invasive infection and modifies tissue CXCL2 production in the embryos of pregnant mice pretreated with BLF prior to vaginal E. coli inoculation.

Design/Methods: Pregnant C57BL/6 mice received BLF (75 mg/mL) or placebo vaginally in a novel oil/aqueous bi-gel on embryonic days (E) days 16 and 17, followed by vaginal inoculation with 105 colony forming units (CFU) of the neonatal E. coli strain RS218 on E17.
Embryos were collected on E18, and liver-spleen tissue homogenates were plated to quantify E. coli CFU per 100 mg tissue. CXCL2 (human IL-8 homolog) concentrations in homogenates were measured by ELISA. Four independent experiments were done, including 35 embryos in the BLF group, and 31 in the placebo group.
Bacterial load was modeled as a function of treatment group and CXCL2 concentrations using a generalized linear mixed model (GLMM) with a negative binomial distribution to address overdispersion and include random mouse effects. Spearman analyses were done to correlate CFU with CXCL2 levels. P < 0.05 was considered significant.

Results: Placebo treatment was associated with a 12.5-fold increase in bacterial load as compared to BLF treatment (incidence rate ratio (IRR) = 12.5; 95% CI 3.9-39.7, p < 0.001), supporting a substantial protective effect of BLF on infection severity (Fig. 1). In addition, bacterial load was associated with CXCL2 concentration (IRR = 4.3; 95% CI 2.6- 7.2; p < 0.001), indicating a strong positive relationship between bacterial infection and CXCL2 production. This was further supported by Spearman correlation analysis, which found positive correlations for both treatment groups (BLF: r = 0.49, p < 0.001; placebo: r = 0.61, p < 0.001) (Fig. 2). While increased bacterial loads correlated strongly with increased CXCL2 levels, we did not observe a significant difference in CXCL concentrations between treatment groups (Fig. 3).

Conclusion(s): Maternal vaginally administered BLF significantly reduced neonatal E. coli burden. CXCL2 levels positively correlated with bacterial loads but were not significantly modulated by treatment. Further studies are needed to elucidate lactoferrin's effects on innate host responses to neonatal E. coli infection.

Figure 1.
Generalized linear mixed model (GLMM) results summary table (top) and density plot (bottom) showing significantly lower E. coli CFU/100 mg of tissue in embryos from BLF-treated dams versus placebo.

Figure 2.
Spearman correlation plots showing a positive relationship between bacterial loads (CFU/100 mg tissue) and CXCL2 concentrations (pg/100 mg tissue) in BLF-treated and placebo-treated embryos. Higher bacterial burden correlated with increased CXCL2 levels in the BLF group (r = 0.49, p = 0.0028) and in the placebo group (r = 0.61, p = 0.00031).

Figure 3.
Log₂-transformed CXCL2 tissue concentrations measured with enzyme-linked immunosorbent assay (ELISA) by treatment group. CXCL2 levels are not significantly different between the groups (p = 0.51).