Human milk oligosaccharides (HMOs) are highly abundant in human breast milk, but not in infant formula. In addition to serving as prebiotics that are selectively metabolized by specific bacteria and shape the infant's intestinal microbiome, HMOs may provide specific anti-infective functions by interfering with bacteria–host interactions (1). One liter of mature human milk contains 5 to 15 g of unbound oligosaccharides (in addition to lactose), which often exceeds the amount of total milk proteins. The building blocks of milk oligosaccharides are the 5 monosaccharides D-glucose (Glc), D-galactose (Gal), N-acetylglucosamine (GlcNAc), L-fucose (Fuc), and sialic acid (Sia; N-acetyl neuraminic acid [Neu5Ac] in humans and both Neu5Ac and N-glycolyl neuraminic acid [Neu5Gc] in most other species). Enteropathogenic Escherichia coli (EPEC) can contaminate food and water supplies and cause serious diarrheal illness with potentially high mortality in infants, especially in developing countries (2). EPEC is classified as an attaching/effacing pathogen because of its ability to adhere to intestinal epithelium, efface microvilli, and induce characteristic, actin-filled membranous pedestals (3). Initial attachment leads to the formation of distinct microcolonies and is mediated by the plasmid-encoded type IV bundle-forming pili (4). Intimate adherence is a key feature of a number of important enteric pathogens (5), including enterohemorrhagic E coli and Campylobacter jejuni. Attachment of C jejuni to human intestinal mucosa ex vivo can be inhibited by HMOs, particularly fucosylated oligosaccharides (6). The goal of the present study was to determine whether HMOs block the attachment of EPEC to intestinal epithelial cells in culture and protects against EPEC infection in neonatal mice.
Bacteria and Reagents
EPEC strain 2348/69 (serotype O127:H6) was used for these studies (7). HMOs were isolated and purified from pooled human milk from 41 different donors as previously described (8).
Cell Culture and Attachment Assays
EPEC was grown overnight at 37°C in Luria-Bertani broth, diluted 1:20, and grown for an additional 2 hours. After washing with phosphate-buffered saline (PBS), bacterial density was estimated by optical density 600, and bacteria were resuspended in Dulbecco modified eagle medium (DMEM) containing different amounts of HMOs or galactooligosaccharides (GOS) and incubated at 37°C for different times before addition to epithelial cells. The human epithelial cell lines HeLa (ATCC CCL-2) and HEp-2 (CCL-23) cells were cultured in DMEM supplemented with 10% FCS and 1% penicillin-streptomycin. For infection experiments, cells were seeded into 6-well plates and grown to confluence in antibiotic-free medium. The human colon epithelial cell line T84 (ATCC CCL-248) was seeded onto 42-mm2 permeable filter supports (EMD Millipore, Billerica, MA) in 6-well plates and grown to confluence for 10 to 14 days to reach a minimum transepithelial resistance of 1000/Ωcm2. One day before infection, the medium was changed to serum-free DMEM. Subsequently, medium was replaced by medium containing bacteria preincubated with HMO, GOS, or neither, and plates were placed onto a rocking shaker (Boekel Scientific, Feasterville, PA; 15 degree angle, 18 rpm) for 1 hour at 37°C. To assay attached bacteria, cells were washed 4 times with warm PBS to remove nonadherent bacteria, and incubated with 0.1% Triton X-100 in PBS for 5 minutes. Cells were scraped off and lysates were transferred to 1.5-mL microtubes. After vigorous vortexing, serial dilutions of the lysates were plated onto tryptic soy agar plates. Colony-forming units (CFUs) were counted after overnight incubation. All cell culture experiments were repeated at least 3 times.
Mice and Infection Protocols
Wild-type C57BL/6J mice (6–8 week old) were obtained from Jackson Laboratory (Bar Harbor, ME) and bred in our animal facility. The resulting newborn mice (7 days) were infected by oral gavage with 105 CFU/mouse of EPEC strain 2348/69 in PBS, with or without preincubation with HMO or GOS (15 mg/mL) for 1 hour before infection. For some experiments, pups were treated with HMO (15 mg/day), GOS (15 mg/day), or vehicle (PBS) 3 times daily on the day before and after infection. To ensure that preinucbation with HMO or GOS did not alter EPEC viability, aliquots were plated onto MacConkey agar without showing any differences in EPEC growth.
To determine bacterial numbers in the intestine, mice were sacrificed on indicated time points and small intestine and colon were collected, weighed, and homogenized in 5-mL PBS. Homogenates were plated onto MacConkey agar and CFUs were counted after overnight incubation. The detection limit of the CFU assay was 103 CFU/g organ. To confirm identity of single colonies, polymerase chain reaction analysis was done for EPEC gene EspB(7). All experiments were repeated at least 3 times; overall, 90 mice were used. All of the animal studies were reviewed and approved by the University of California, San Diego Institutional Animal Care and Use Committee.
CFU counts from in vivo experiments were log10 transformed, and means and standard error of the mean of the mean were calculated from the log values. Mice without detectable bacteria in small intestine or colon were assigned a log10 value equivalent to half of the detection limit of the CFU assay (103 CFU/g). Results from males and females were combined because no significant differences were observed between the sexes in bacterial colonization or mucosal responses after infection. Data from bacterial attachment assays in vitro are expressed as mean ± standard error of the mean. CFU data are either shown as percentage of initial inoculum or counts were log10 transformed (see above). Differences between groups were evaluated by 1-sample t test or Wilcoxon rank sum test, with P < 0.05 considered significant.
HMOs Block EPEC Attachment to Epithelial Cells
As a first step to determine whether HMO can block EPEC attachment to epithelial cells, we used in vitro assays. To better mimic the dynamic conditions in the intestine, where intestinal motility and bulk fluid movements compromise attachment opportunities, we developed a new infection model that mimics the situation by rocking the culture plates, reducing the bacterial inoculum, and shortening the infection times. EPEC (3 × 106 CFU) were preincubated with HMO (10 mg/mL) or medium (DMEM) alone for 1 hour, added to different epithelial cell monolayers (HeLa, HEp-2, and T84 cells), and incubated for 1 hour to allow attachment. HMO preincubation significantly reduced EPEC attachment in all 3 cell lines (Fig. 1A). The effect was dose dependent (Fig. 1B) and even more prominent when EPEC were preincubated with HMO for 2 hours (Fig. 1C). In contrast, GOS, which are structurally distinct from HMO and are added to some infant formula, were used as control and did not reduce EPEC attachment to any of the epithelial cell lines (Fig. 1B, C). Subsequently, we first preincubated epithelial cells with HMO and then added EPEC to the cells. Under these conditions, HMO were no longer able to interfere with EPEC attachment (Fig. 1D), suggesting that HMOs bind to critical bacterial adhesion molecules to block attachment rather than act directly on epithelial cells.
Oral HMO Administration to Suckling Mice Attenuates EPEC Infection
Next, we sought to determine the physiological importance of the attachment-blocking effects of HMOs. Adult mice are not readily infectable with human EPEC strains (9) and were less relevant for the present study. Instead, we established a new model of EPEC infection in newborn mice. Testing of different bacterial inocula indicated that an inoculum of 105 CFU/mouse was optimal for achieving stable infection for at least 3 to 4 days (Fig. 2A). After an initial decline in bacterial numbers in the colon, numbers rebounded and reached stable levels after 18 to 24 hours. Compared with the low infection rates in adult mice, newborn mice displayed 103- to 105-fold greater bacterial colonization in the small intestine and colon after 3 days (Fig. 2B). Together, these data show that newborn mice were actively and selectively colonized by EPEC, making this model suitable for evaluating the physiological effect of HMOs on infection. HMOs (15 mg/day) were orally administered to newborn mice on the day of oral infection with EPEC, and administration was continued throughout the infection period. HMO-treated pups showed significantly lower EPEC colonization than PBS-treated animals. By comparison, administration of GOS to pups had no significant effect on EPEC colonization (Fig. 2C).
In this study, we show that HMOs, the third most abundant solid component of human milk next to lipids and protein, reduce EPEC attachment to human epithelial cells and significantly protect against EPEC infection in newborn mice. Intimate adherence to the intestinal epithelium is important for EPEC pathogenicity (10), so interference in this essential step by HMOs is likely to attenuate infection and disease induction. It has been reported that the bundling protein of EPEC has the properties of an N-acetyllactosamine-specific lectin (11), which may be responsible for EPEC binding to intestinal epithelial host cells. Because N-acetyllactosamine forms the backbone of most HMOs (1), we speculate that HMOs prevent the lectin from binding to glycans on the intestinal epithelial surface. Future studies must reveal which of the >150 different HMOs are responsible for the observed effects. Our data also show that HMOs attenuate intestinal EPEC colonization of infant mice. Our murine model uses the human pathogen EPEC instead of the mouse-specific attaching/effacing pathogen Citrobacter rodentium(12), making it likely that the present insights can be applied to humans. Thus, we predict that HMOs will protect breast-fed infants from EPEC infection, thus providing an explanation for the lower incidence of EPEC infections in breast-fed compared with formula-fed infants (13).
1. Bode L. Human milk oligosaccharides
: every baby needs a sugar mama. Glycobiology
2. Bryce J, Boschi-Pinto C, Shibuya K, et al. WHO estimates of the causes of death in children. Lancet
3. Frankel G, Phillips AD, Rosenshine I, et al. Enteropathogenic and enterohaemorrhagic Escherichia coli: more subversive elements. Mol Microbiol
4. Donnenberg MS, Kaper JB. Enteropathogenic Escherichia coli. Infect Immun
5. Bardiau M, Szalo M, Mainil JG. Initial adherence of EPEC, EHEC and VTEC to host cells. Vet Res
6. Ruiz-Palacios GM, Cervantes LE, Ramos P, et al. Campylobacter jejuni binds intestinal H(O) antigen (Fuc alpha 1, 2Gal beta 1, 4GlcNAc), and fucosyloligosaccharides of human milk inhibit its binding and infection. J Biol Chem
7. Savkovic SD, Villanueva J, Turner JR, et al. Mouse model of enteropathogenic Escherichia coli infection. Infect Immun
8. Jantscher-Krenn E, Zherebtsov M, Nissan C, et al. The human milk oligosaccharide disialyllacto-N-tetraose prevents necrotising enterocolitis in neonatal rats. Gut
9. Mundy R, Girard F, FitzGerald AJ, et al. Comparison of colonization dynamics and pathology of mice infected with enteropathogenic Escherichia coli, enterohaemorrhagic E. coli and Citrobacter rodentium. FEMS Microbiol Lett
10. Wong AR, Pearson JS, Bright MD, et al. Enteropathogenic and enterohaemorrhagic Escherichia coli: even more subversive elements. Mol Microbiol
11. Hyland RM, Sun J, Griener TP, et al. The bundlin pilin protein of enteropathogenic Escherichia coli is an N-acetyllactosamine-specific lectin. Cell Microbiol
12. Frankel G, Phillips AD, Novakova M, et al. Intimin from enteropathogenic Escherichia coli restores murine virulence to a Citrobacter rodentium eaeA mutant: induction of an immunoglobulin A response to intimin and EspB. Infect Immun
13. Creek TL, Kim A, Lu L, et al. Hospitalization and mortality among primarily nonbreastfed children during a large outbreak of diarrhea and malnutrition in Botswana, 2006. J Acquir Immune Defic Syndr