Secondary Logo

Journal Logo

Original Articles: Gastroenterology

Role of Postnatal Acquisition of the Intestinal Microbiome in the Early Development of Immune Function

Dimmitt, Reed A; Staley, Elizabeth M*; Chuang, Gin§; Tanner, Scott M§; Soltau, Thomas D; Lorenz, Robin G*

Author Information
Journal of Pediatric Gastroenterology and Nutrition: September 2010 - Volume 51 - Issue 3 - p 262-273
doi: 10.1097/MPG.0b013e3181e1a114
  • Free


Following birth, the sterile newborn intestine is colonized by environmental microorganisms that expand exponentially to numbers that exceed the total number of host mammalian cells (1). The current paradigm holds that a state of mutualism is essential for overall health, growth, and development (2). Acquiring intestinal microbiota is an active process in which host and microbial factors mediate the intestinal bacterial composition (3).

The systemic and mucosal immune functions of newborns are markedly different than adults (4). The fetal immune system has evolved to avoid maternal rejection rather than to process potential pathogens (5). The neonatal immune system undergoes extensive postnatal development (6), and the acquisition of intestinal microbiota is a major determinant of early immune development (7); for example, the T-cell population of the fetus favors a naïve or T-helper cell 2 (TH2) phenotype that undergoes a switch to a T-helper 1 (TH1) or T-helper 17 (TH17) phenotype upon exposure to intestinal microbiota to facilitate protection against potential injurious microorganisms (8). The classic TH1, TH2, and TH17 phenotypes are largely defined by their associated cytokine production: TH1 (interferon-γ [IFN-γ]), TH2 (interleukin-4 [IL-4]), and TH17 (IL-17). The first step in this phenotype commitment is believed to be the interaction between commensal bacteria and the mucosal immune system. The microbial sensors known as Toll-like receptors (TLRs) sense specific bacterial products, that is, peptidoglycan (TLR2), lipopolysaccharide (LPS) (TLR4), and flagellin (TLR5), with a resultant innate immune response mediated in large part by chemokine production (9).

Extremely premature human infants have a different postnatal experience than term infants. They often receive broad-spectrum antimicrobials and experience delayed feedings. Thus, these infants acquire a skewed commensal microbiota that, we hypothesize, could adversely affect the developing mucosal and systemic immune function. This phenomenon may, in part, explain a high incidence of necrotizing enterocolitis (NEC) in this patient population (10). Recent reports of probiotic therapy reducing the incidence of NEC in premature infants (11,12) give credence to the notion that acquiring intestinal commensal bacteria in the early postnatal period is of great importance. In addition, recent studies from our laboratory have demonstrated that 2-week-old mice lacking TLR2, TLR4, or both TLR2 and 4, as well as microbial-reduced (MR) mice, have increased intestinal injury in a model of NEC (12a). These results support the idea that the presence of commensal bacteria and the sensing of those bacteria through TLRs are important in preventing NEC.

Previous work (13) examining the role of commensal microbiota in the development of immune function have used adult germ-free mice or animals colonized with a defined microbiota (gnotobiotic). These studies (14,15) have indicated that in the absence of a commensal microbiota, mice have decreased immunoglobulin A, reduced intraepithelial lymphocytes, reduced Peyer's patch and mesenteric lymph node (MLN) size and cellularity, decreased responses to T-cell mitogens, and altered susceptibility to pathogenic organisms. Because the neonatal immune system is not fully developed during normal acquisition of commensal microbiota, these studies on fully developed adults may not accurately reflect the outcome of this crucial interaction. Previous studies in suckling rats have demonstrated that antibiotic-induced microbial restriction during early neonatal development results in decreased Paneth cell products (16). Additionally, it has been reported that in C57BL/6 mice exposure to microbial products after vaginal delivery transiently increases chemokine expression by intestinal epithelial cells, a result that is not seen after a sterile cesarean delivery (17). However, the effects of these colonizing microorganisms and their microbial products on the initial establishment of mucosal and systemic immunity have not been evaluated. Because some components of the developing postnatal murine immune system are similar to those of premature human infants, it is a useful model for the investigation of the establishment of this ecosystem (18). The aim of the present study was to characterize early postnatal mucosal and systemic immune function during the neonatal period when mice are acquiring their intestinal microbiota. We then manipulated exposure to these microbiota during this period by antibiotic treatment of pups (MR) to recapitulate the clinical course of premature infants. We hypothesized that newborn mice that have altered acquisition of normal microbiota would demonstrate significant changes in postnatal immune development.



Adult C57BL/6 mice were obtained from Jackson Laboratories (Bar Harbor, ME), raised under specific pathogen-free (SPF) conditions, and acclimatized to our facility 2 weeks before mating. SPF mice were chosen because they are known to be free of bacterial, viral, and parasitic mouse pathogens, as opposed to “conventional” mice, which are not known to be free of pathogens. SPF conditions, as defined by the University of Alabama at Birmingham (UAB) Animal Resources Program, can be found at The pups were born and housed under SPF or experimental conditions (as noted), and nursed until time of sacrifice. All of the experimental protocols were approved by the Institutional Animal Care and Use Committee of the University of Alabama at Birmingham.

RNA Isolation and Gene Expression

RNA isolation and gene expression by quantitative real-time reverse-transcriptase-polymerase chain reaction (RT-PCR) was performed as previously described (19,38,49). Briefly, small intestine (SI) and colon were harvested for analysis of the expression of microbial sensors and inflammatory molecules. RNA was isolated, genomic DNA contamination was removed, and then RNA was transcribed into cDNA. RT-PCR was performed using Applied Biosystems (Carlsbad, CA) primer probe sets. Specific gene RNA levels were determined using crossing thresholds read by the RT-cycler MX3000P (Stratagene, La Jolla, CA). Fluorescence thresholds were averaged generating a gene-specific numeral, which then could be normalized to the average expression of the 18S housekeeping gene and further stratified by particular strain and experimental condition. Gene expression was calculated as an average fold change when compared with that of control strain values, and shown on a log 2 scale as fold changes from the control baseline (=1). The protocol for this data analysis format is provided in the Applied Biosystems manufacturer's instructions (4371095 Rev A, PE Applied Biosystems).

Epithelial Versus Lamina Propria TLR Expression

SI and colonic epithelium were removed to determine the mucosal location of TLR expression at 2 weeks of age. Briefly, SI and colon from 2-week-old animals were removed and opened along the antimesenteric border. The tissue was then placed in Hank's balanced salt solution-plus (HBBS+) (HBSS 500 mL, 25 mmol/L 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid, 1% fetal bovine serum (HyClone Laboratories, Logan, UT), 100 U/mL penicillin, and 100 μg/mL streptomycin). The tissue was gently inverted several times to remove the luminal contents. To remove the epithelial layer, the tissue was then transferred to HBSS without calcium and magnesium + 5 mmol/L ethylenediaminetetraacetic acid (EDTA) and agitated at 100 rpm for 15 minutes at 37°C. The tissue was removed and the supernatant, which contained the epithelium, was saved on ice. The process was repeated and the final volume of supernatant underwent centrifugation at 1500 rpm for 5 minutes at 4°C. The pellet was resuspended in Trizol for RNA isolation (described as above). Adequate removal of epithelium was confirmed by hematoxylin and eosin staining of histological sections of the remaining tissue.

Microbial Reduction

Antibiotics were added to the drinking water based on a modification by Hans et al (20). MR animals were maintained by the addition of streptomycin sulfate (4.8 mg/mL), ampicillin (1.2 mg/mL), metronidazole (1.2 mg/mL), and vancomycin (0.6 mg/mL) (Sigma, St Louis, MO) to sterile drinking water ad libitum and changed biweekly. Antibiotic treatment of female C57BL/6 mice was initiated 5 days following mating and maintained throughout the experimental period, including the nursing period, until sacrifice of the pups. The choice of these antibiotics was based on an approximation of the early postnatal experience in extremely premature human infants in the neonatal intensive care unit.

Analysis of Fecal Samples by PCR-denaturing Gel Electrophoresis

DNA extraction of fecal samples for PCR-denaturing gradient gel electrophoresis (DGGE) analysis was performed as described (21). Primers were used for PCR amplification targeting different variable regions of the 16S rDNA. The forward primer sequence was ACTACGTGCCAGCAGCCCGCCGGGCCGCGGCCCGCCCGCCCGCGGGGGCACGGGGGACTACGTGCCAGCAGCC, and the reverse sequence was GGACTACCAGGG TATCTAATCC. Analysis of PCR products was performed using a gradient denaturant gel, which was scanned using the GS-710 Calibrated Imaging Densitometer (Bio-Rad Inc, Hercules, CA). Selected bands were excised from the gel, and DNA was extracted by a crush soak process. The DNA was used for an additional PCR using the above primers joined to M13 vector sequences as previously described (22) and the sequence analyzed using the online Ribosomal Database Project provided by the Center for Microbial Ecology, Michigan State University (

Flow Cytometry

To conduct lymphocyte population analysis, spleens and MLNs underwent mechanical disruption, erythrocyte lysis, and preparation for flow cytometry as previously described (50). Lymphocytes were labeled for the lymphocyte markers CD4, CD45, CD8, CD3, CD19, CD44, CD45, CD62L, or CD69 (BD Biosciences, San Diego, CA). Flow cytometry was performed using the Becton Dickinson FACS Caliber (Franklin Lakes, NJ) using the Cell Quest analysis software (San Diego, CA) and analyzed with FlowJo software (TreeStar, Ashland, OR).

Bacterial Translocation and Barrier Permeability Experiments

Green fluorescent protein (GFP)–producing Escherichia coli (pET 21d/eTS GFP #4) were a gift from Dr Casey Weaver (University of Alabama at Birmingham) and cultured as described (23). GFP E coli was inoculated into liquid Luria-Bertani broth with 100 mg/L ampicillin (Sigma-Aldrich, St Louis, MO) and incubated at 37°C overnight with agitation. A final dilution of 5 × 1010 colony-forming units (CFU)/mL was prepared in sterile media, as determined by optical density at 450 nm (OD450/3 × 109 = CFU/mL).

Two-week-old SPF and MR pups and their dams were orally inoculated with 1 × 108 CFU/mL of GFP E coli on days 1, 2, 3, and 4. On day 1, antibiotic therapy was discontinued in the MR group before bacterial challenge. On day 6, the pups were sacrificed and individual MLNs, spleen, and liver were harvested using sterile technique. Cecal contents were obtained after the organ harvest. The organs and cecal contents were used to inoculate Luria-Bertani broth agar plates containing ampicillin and incubated overnight at 37°C. GFP-producing colonies were confirmed by fluorescence. The presence of ≥1 GFP E coli colony was defined as being positive for bacterial translocation for that organ.

Barrier permeability was assessed by analyzing serum concentrations of fluorescein isothiocyanate (FITC)-dextran (MW 4000, Sigma-Aldrich, St Louis, MO) following oral gavage as previously described (24). Briefly, mice were orally gavaged 4 hours before sacrifice using an 80 mg/mL stock of FITC-dextran, for a total dose of 60 mg/100 g of body weight. Total blood volume was collected at sacrifice via cardiac puncture. Blood samples were coagulated in the dark at 4°C for 1 hour before centrifugation and serum harvest. Samples were analyzed using the Synergy HT Multi-Mode Microplate Reader (Bio-Tek Technologies, Winooski, VT) and standardized to FITC-loaded control serum; FITC was detectable between 25 and 0.195 ng/mL.

Splenocyte Culture and Cytokine Analysis by Enzyme-linked Immunosorbent Assay

Spleens were harvested from SPF and MR neonates following sacrifice. A single cell suspension was filtered through a 70 μmol/L Nitex membrane (Tetko, Elmsford, NY). Erythrocytes were lysed using ammonium chloride, and the splenocytes were washed twice in HBSS+. The cells were resuspended in R10 medium (RPMI 1640, 10% FBS, 1% GlutaMAX-I (Invitrogen, Carlsbad, CA) and 50 μg/mL gentamicin sulfate) and inoculated at 2 × 105 cells per well onto 96-well culture plates, coated with hamster αCD3 (clone 145-2C11) or hamster immunoglobulin G control (clone UC8-4B3) at a concentration of 1 μg/mL (BD Biosciences). Cells were incubated at 37°C in a humidified atmosphere containing 5%CO2/95% air for 96 hours. Supernatants were analyzed for expression of IFN-γ, IL-17, IL-10, IL-4, and IL-2 with BioLegend cytokine enzyme-linked immunosorbent assay reagents and read at 450 nm with a VERSA-Max microplate reader (San Diego, CA) and analyzed using Softmax Pro software (Molecular Devices, Sunnyvale, CA).

Statistical Analysis

Sigma Stat 2.03 (SPSS Inc, Chicago, IL) software was used for data analysis. Student t test and 1-way analysis of variance were used for continuous variables, and Fisher exact test was used for categorical variables. Nonparametric analysis was performed with a Mann-Whitney test. Gene expression differences were analyzed by using Student t test to compare ΔΔCT and standard deviation between the experimental and control animals. A P value <0.05 was regarded as significant.


Bacterial Sensing Molecules During Early Development

Previous work has shown that immature intestinal cells express a basal level of the LPS signaling molecule TLR4 (25). Additionally, several investigators have demonstrated that the expression and responsiveness of TLR2 and TLR4 is regulated by exposure to bacterial product ligand (26–28). To determine the early postnatal response to commensal microbial colonization with regard to TLR expression, cohorts of SPF C57BL/6 mice were sacrificed at multiple developmental time points. RNA was harvested from the SI and colon after which gene expression was determined using RT-PCR analysis and values were normalized to E20.5 levels. This was done based on the assumption that prenatal organs at this stage would be fully developed but sterile due to a lack of exposure to a commensal microbiota. Relative to the E20.5 expression, TLR2, TLR4, and TLR5 were significantly elevated in the 2-week-old mouse colon (Fig. 1B, D, and F). In addition, SI expression of TLR4 and TLR5 was also significantly greater in the 2-week-old cohorts (Fig. 1C and E). Although not achieving statistical significance, the same trend was observed with regard to SI TLR2 expression (Fig. 1A). These expression peaks were attenuated in the 3-week-old animals and remained diminished into adulthood. Interestingly, the increased level of expression at 2 weeks coincides with the period of microbial expansion in the postnatal murine gastrointestinal tract (29,30). In addition, the relative organ-specific expression of TLRs was consistent with increased microbial colonization in the colon compared with the SI. Thus, the induction of TLR expression at 2 weeks is greatest in the colon, the organ with the greatest number of commensal microbiota.

Quantitative reverse-transcriptase polymerase chain reaction analysis of toll-like receptor gene expression in specific pathogen-free mice at various postnatal ages. A, Small intestinal TLR2; B, colonic TLR2; C, small intestinal TLR4; D, colonic TLR4; E, small intestinal TLR5; and F, colonic TLR5. Data are expressed on a log2 scale and normalized to 18S rRNA expression, with fold change calculated by comparison with the level of expression in E20.5 mice (E20.5, n = 7; day 3, n = 8; week 1, n = 10; week 2, n = 13; week 3, n = 9; month 6, n = 7). The bar graph to the right represents increased expression and that to the left decreased expression.

Epithelial Versus Lamina Propria TLR Expression at 2-week Postnatal Development

To determine the mucosal compartment of the increased TLR expression, separate cohorts of 2-week-old pups raised in the SPF environment were studied. After isolation of the epithelium, the relative expression of epithelial-derived TLR2, TLR4, and TLR5 were compared with that of the total SI or colon. Relative to that of the total SI or colon, the epithelial expression of TLR2, TLR4, and TLR5 was significantly less, indicating that the majority of the increased expression was a result of induction of TLRs in the cells of the lamina propria (Fig. 2A, B). The removal of the epithelium by EDTA treatment was confirmed in each sample by histological examination (Fig. 2C, D, E, and F).

Quantitative reverse-transcriptase-polymerase chain reaction analysis of Toll-like receptor (TLR) gene expression in specific pathogen-free mice, comparing epithelial expression to total small intestinal or colonic expression. A, Small intestinal TLR2, -4, and -5 expression. B, Colonic TLR2, -4, and -5 expression. C, Hematoxylin and eosin of small intestinal section without and (D) with ethylenediamintetaacetic acid digestion. E and F are similar sections of colon without and with ethylenediamintetaacetic acid digestion. Reverse-transcriptase-polymerase chain reaction analysis data are expressed in log2 scale and normalized to 18S rRNA expression with fold change calculated comparing epithelial expression to total intestinal or colonic expression, n = 6. The bar graph to the right represents increased expression and that to the left decreased expression.

Commensal Intestinal Microbial Reduction Using Antibiotics

The developmental experiments described above identified the 2-week period as a potential checkpoint of microbial–host interaction. Therefore, we set out to reproduce a simulated MR environment during this period by using antibiotics added to recently bred dam drinking water. This replicated the broad-spectrum antibiotic treatment that is observed in the early postnatal period of premature human neonates.

To determine the alteration in maternal microbiota induced by the broad-spectrum antibiotic treatment, feces from the 2-week-old MR and SPF mice were analyzed for the presence and diversity of bacteria using amplification of 16S rRNA. Following PCR with conserved bacterial 16S rRNA primers, a DGGE analysis revealed a relatively diverse microbial population in the SPF feces (Fig. 3, lanes 1, 2) compared with that of MR mice (Fig. 3, lanes 3–9). A consistent band representing single bacterial species was observed in the MR mice (Fig. 3, arrow). This band was also noted in the SPF mice. Sequence analysis revealed that this band corresponded with Pseudomonas aeruginosa with 93% sequence similarity, as found in the Ribosomal Database Project. Thus, the mice reared by dams receiving antimicrobial therapy had a marked reduction in microbial diversity with a selection for a potentially pathogenic bacterium.

Polymerase chain reaction-denaturing gradient gel electrophoresis analysis of feces from 2-week-old specific pathogen-free and microbial reduced mice. Lanes 1, 2 specific pathogen-free; lanes 3 to 9 MR. The black arrow indicates persistent banding pattern observed in the microbial reduced mice.

Microbial Reduction Results in Alteration of Bacterial Sensor and Immune Function Molecule Expression

Because there was an increased expression of TLR2, TLR4, and TLR5 in 2-week-old mice compared with the other cohorts, we hypothesized that MR mice would have less expression of these microbial-sensing molecules. The SI expression of TLR5 in the MR mice was significantly less than that in the SPF controls, with no difference in TLR2 and TLR4 expression (Fig. 4A). In contrast, there was a marked reduction in TLR2 and TLR5 in the MR colons with a trend toward less TLR4 expression (Fig. 4B). These results support the notion that commensal microbiota are critical for the induced TLR expression observed at 2 weeks postpartum.

Quantitative reverse-transcriptase polymerase chain reaction analysis of toll-like receptor (TLR) gene expression in microbial reduced (MR) mice compared with specific pathogen-free (SPF). Reverse-transcriptase polymerase chain reaction data are expressed in log2 scale and normalized to 18S rRNA expression with fold change calculated comparing MR with SPF (MR, n = 10; SPF, n = 10). The bar graph to the right represents increased expression and that to the left decreased expression.

Bacterial signaling through TLRs results in an initial innate immune response, including the production of neutrophil-attracting chemokines (27). Therefore, we speculated that the 2-week-old MR mice would have diminished chemokines bearing the glutamic acid-leucine-arginine and cysteine-amino acid-cysteine (ELR+ CXC) motif. These chemokines are the murine equivalent of human IL-8. As seen in Figure 5, there was decreased expression of 3 ELR+ CXC chemokines in the MR mice compared with SPF controls. Specifically, there was significantly less SI and colonic lipopolysaccharide-inducible CXC chemokine (CXCL5) expression (Fig. 5A, B). In addition, there was significantly less colonic macrophage inflammatory protein-2 (CXCL2) (Fig. 5B). Although there was a trend toward decreased expression of SI and colonic keratinocyte-derived chemokine (CXCL1), this reduction did not achieve statistical significance. As was previously seen with regard to intestinal versus colonic differences, we observed a greater relative reduction in ELR+ CXC expression in the colon, which has orders of magnitude more microbiota.

Quantitative reverse-transcriptase-polymerase chain reaction (RT-PCR) analysis of cytokine and chemokine expression in microbial reduced (MR) mice compared with specific pathogen-free (SPF). A, Small intestinal chemokine expression; B, colonic chemokine expression; C, small intestinal cytokine expression; and D, colonic cytokine expression. RT-PCR data are expressed in log2 scale and normalized to 18S rRNA expression with fold change calculated comparing MR with SPF (MR, n = 10; SPF, n = 10). The bar graph to the right represents increased expression and that to the left decreased expression. MR = microbial reduced; SPF = specific pathogen-free.

We next set out to determine the role of microbiota in T-cell phenotype–associated adaptive immune development. In general, there was no difference in mucosal IL-4, IL-17, or IFN-γ expression between the 2-week-old MR and SPR animals (Fig. 5C, D). There was higher IL-10 expression in the MR SI, but no difference in colonic expression. In summary, taken together, these data demonstrate that the presence of intestinal and colonic commensal bacteria induce their own signaling molecules, trigger an innate mucosal immune response, but do not seem to affect the early postnatal adaptive mucosal response.

Role of Microbiota in MLN T-Cell Activation

The MLNs from 2-week-old MR and SPF animals were analyzed for evidence of T-cell activation. Overall, we saw no statistically significant difference in the percentage of T cells expressing the activation markers CD44, CD62L, CD69, or CD25 (data not shown). There was a trend toward less CD44 expression (Fig. 6A) and greater CD62L and CD69 (Fig. 6C, E), whereas there was similar CD25 expression (Fig. 6G) in the SPF mice compared with the MR mice. However, the mean fluorescence intensity in the groups for each activation marker was not significantly different.

Flow cytometry analysis of mesenteric lymph nodes–derived T-cell activation markers comparing microbial reduced (MR) to specific pathogen-free (SPF) mice. A, C, E, and G represent histograms of T-cell surface activation markers CD44, CD62L, CD69, and CD25, respectively. Dotted line = SPF; solid line = MR; gray filled = unlabeled SPF cells. B, D, F, and H represent absolute numbers of CD4+ cells that are CD44, CD62L, CD69, and CD25 positive. White bars = SPF; striped bars = MR. Results are 3 separate experiments in which the mesenteric lymph nodes from 3 to 5 mice were pooled for analysis. * P < 0.05.

Next we analyzed the number of T cells expressing the selected activation markers. Although the percentage of activated CD4+ cells was not different in the SPF and MR cohorts, the mean absolute number of CD4+ cells per MLN was significantly decreased in the MR mice (8.6 ± 2.4 × 106 (SPF) vs 1.7 ± 0.42 × 106 (MR); P = 0.0012). This alteration in total number of cells resulted in significantly fewer CD44+, CD62L+, CD69+, and CD25+ T cells in the MR mice compared with the SPF group (Fig. 6B, D, F, and H). Taken together, these results suggest that in the early postnatal period, the presence of commensal bacteria induces the expansion of activated T cells but does not alter the intensity of surface expression of markers of activation.

Alterations in Microbiota Affect Intestinal Barrier Function

Although the mechanisms by which broad-spectrum antibiotics and probiotics alter the incidence of NEC in premature infants are unknown, 1 hypothesis is that these bacteria protect the host from intestinal bacterial translocation. To test the hypothesis that mice with an altered commensal microbiota secondary to early antibiotic exposure would have a relative loss of barrier function, cohorts of 2-week-old MR and SPF pups were orally inoculated with GFP-labeled E coli. Six days after inoculation, pups were sacrificed under sterile conditions, and cecal contents and peripheral lymphoid organs were cultured and analyzed for the presence of GFP E coli. Twenty-four hours following initial culture, GFP-producing colonies were detected in the cecal contents from all of the innoculated neonates, indicating adequate colonization. Colonies were detected in the MLN, spleen, and liver of MR animals, whereas organs from SPF neonates were demonstrably free of GFP-labeled organisms (Fig. 7A).

A, Bacterial translocation as a function of intestinal barrier permeability. The columns represent the percentage of animals with positive green florescent protein-producing Escherichia coli colonies cultured from mesenteric lymph nodes, spleen, and liver from SPF (first 3 columns) and MR (last 3 columns) neonates. The median number of colony forming units per whole organ culture is expressed below the x-axis. * P < 0.05, (microbial reduced, n = 12; specific pathogen-free, n = 11). B, Epithelial barrier function assay. The columns represent serum fluorescein isothiocyanate-dextran concentrations 4 hours after oral gavage. The concentration is expressed as mean and standard deviation (microbial reduced n = 9, specific pathogen-free = 8). ND = none detected.

To test whether bacterial translocation was a result of increased intestinal permeability, MR and SPF mice were orally gavaged with FITC-dextran and sera harvested after 4 hours. The mean serum FITC-dextran concentration of the MR and SPF mice was not significantly different (Fig. 7B), arguing that paracellular barrier permeability and epithelial tight junctions were intact.

Role of Microbiota in the Development of Systemic T- and B-cell Populations

The preceding experiments demonstrated that the acquisition of commensal bacteria in the first 2 weeks of life is a major determinant of postnatal innate mucosal immune development. We next wanted to determine whether MR pups would have altered numbers and proportions of peripheral lymphocytes. Splenocytes were labeled with fluorescently tagged lymphocyte markers and analyzed using flow cytometry. Overall, there were significantly more lymphocytes in the spleen in SPF mice compared with the MR group, similar to what was observed in the MLN experiments. There was no difference in the absolute number or percentage of CD45+CD3+ lymphocytes, indicating that early in postnatal life, microbiota have little effect on systemic T-cell numbers (Table 1). In contrast, there were significantly more CD45+CD19+ lymphocytes, thus showing that commensal bacteria influence early postnatal B-cell numbers.

Spleen lymphocyte populations analyzed by flow cytometry

Role of Microbiota in the Development of Systemic T-Cell Phenotype

Although the overall percentage of T cells was similar in the SPF and MR groups, we next set out to test the hypothesis that acquisition of commensal intestinal microbiota influences systemic T-helper cell phenotypic commitment. Following a 96-hour culture, the supernatants from splenocyte cultures obtained from 2-week-old SPF and MR mice were analyzed for cytokine production. When compared with MR conditions, there was a significantly greater concentration of IFN-γ and IL-17 in the SPF splenocyte cultures (Fig. 8C, D). In contrast, the MR cultures had a significantly greater amount of IL-4 and a trend toward more IL-10 production (Fig. 8A, B). These results demonstrate that in the absence of intestinal microbiota, newborns remain committed to a systemic TH2 phenotype and fail to switch to the TH1 and TH17 pathways, which are essential for defending against microbial infections.

Cytokine enzyme-linked immunosorbent assay results from microbial reduced and specific pathogen-free splenocyte cultures after αCD3 activation. A, IL-4; B, IL-10; C, IL-17; and D, IFN-γ. * P < 0.05 (MR, n = 10; SPF, n = 10). IFN = interferon; IL = interleukin; MR = microbial reduced; SPF = specific pathogen-free.


The results of these experiments demonstrate that early acquisition of commensal microbiota significantly influences neonatal mucosal and systemic immune development and function. From a mechanistic prospective, these experiments suggest that during early postnatal mucosal immune development, the colonization of the intestine with microbiota results in a self-induction process of bacterial sensors with subsequent activation of innate mucosal immune processes. This activation process is temporal and rapidly attenuated as the host ages. The presence of commensal bacteria also promotes intestinal barrier function against potential pathogens and stimulates the number of activated T cells in the MLN. In concert, the acquisition of commensal intestinal bacteria induces expansion of systemic B cells, and thus also shapes early humoral immunity. Finally, this colonization process induces the switch from the fetal TH2 state that promotes intrauterine fetal–maternal coexistence to a TH1/TH17 state that prevents invasion of potential pathogens.

Our results agree with other work (31) demonstrating that intestinal TLR2 and TLR4 are expressed in the sterile fetal environment. Constitutive expression of baseline TLRs would be necessary if increased expression is inducible. The increased expression of TLRs at 2 weeks postnatal development reflects a relative increase, not an absolute value. The downregulation of TLRs occurs in the same time frame as TLR tolerance has been reported to occur, indicating that both functional and physical mechanisms are induced by exogenous stimuli to control intestinal responses to TLR ligands (32).

These results also tease out the specific roles of commensal microbiota and breast milk in the induction of TLR expression (33). TLR expression in human adult and fetal epithelial cells was different when treated with breast milk supplement medium (34). In our study, both the SPF and MR pups received breast milk; thus, the difference in microbial sensor expression was specifically related to the presence or absence of commensal microbiota.

Other investigators have demonstrated that germ-free animals lacking a commensal microbiota are at risk for invasive bacteria (14,35,36). Our findings demonstrate that this barrier function is established early in the postnatal period. Another difference in our results is that although MR mice were in an MR environment, the dam was not from a gnotobiotic environment and as such was able to confer passive immunity by both transplacental or breast milk pathways. Even though the MR pups were nursed by immunocompetent mothers, the lack of normal commensal microbiota created an environment allowing bacterial translocation. Interestingly, the FITC-dextran experiments revealed similar serum levels in the MR and SPF pups, indicating no differences in the paracellular mucosal permeability in the early postnatal period. Thus, the increased bacterial translocation is not a result of a “leaky” intestinal epithelial barrier, but due to a more global impairment of immune function. Because chemokines are critical for the appropriate trafficking of innate and adaptive immune cells to the intestine, the decreased expression of these molecules may explain the absence of appropriate immune mechanisms to clear the administered GFP E coli.

However, some commensal microbial products have been shown to alter mucosal epithelial integrity and barrier function. An in vitro study of human intestinal epithelial cells has demonstrated that the tight junction, ZO-1, is increased in the presence of TLR2 ligands (37). In addition, our laboratory has shown that adult mice lacking TLR2 have greater injury in an ischemia/reperfusion model of NEC (38), and that 2-week-old MR mice or mice lacking TLR2 or TLR4 also have increased ischemia/reperfusion injury (12a). Thus, intestinal sensing of commensal microbiota is also essential to impart protection against acute intestinal injury.

Of additional interest is that the MR mice have reduced bacteria but are not germ-free. The PCR-DGGE results show that broad-spectrum antibiotics in this model reduce not only the number of bacteria but more important the bacterial diversity. It is unlikely that any 1 commensal bacterium is the major determinant of postnatal development, but rather it is the interplay between several species. Our findings are in agreement with previous studies looking at the introduction of defined bacterial microbiota into gnotobiotic mice, which show that a diverse population of bacteria is necessary (39,40). Of interest was the predominance of Pseudomonas in the MR mice. This finding has also been reported in a case-control study in very-low-birth-weight infants reporting an association with prolonged use of antibiotics with Pseudomonas and NEC (41). Thus, our findings seem to recapitulate the early postnatal experience of premature infants in the neonatal intensive care unit receiving broad-spectrum antibiotics.

Our findings of activation of innate immunity in response to microbiota have been previously shown in adult germ-free mice (15). A unique finding of our study is the relative increase in IL-10 gene expression in the 2-week-old MR mice. Because IL-10 is considered a regulatory cytokine, it is unclear why mice with a reduced microbiota would have greater expression of the gene. It is possible that this increase reflects the earlier response to food antigens and the development of tolerance in an environment with markedly less bacteria (42).

The finding of increased B cells in SPF animals is consistent with the observation in humans that Peyer patches and other mucosa-associated lymphoid tissue do not contain secondary follicles with germinal centers and B-2 B cells at birth. These structures develop several weeks later, reflecting that this activation needs exogenous stimuli (43). However, this early postnatal increase in B cells could also reflect expansion of B-1 B progenitor cells (44). In addition, antibiotic therapy early in life has been shown to impair humoral immunity and response to oral antigens, thus demonstrating the role for microbiota to direct postnatal immune development (45).

The temporal and functional nature of TH2 and TH1/TH17 commitment as it relates to health and disease and the need for microbiota is well documented (46). Our study, again, focuses on the timely commitment during the early postnatal period. Several studies have shown that progesterone induces the production of TH2 cytokines, presumably to prevent allograft (fetus) rejection (47). It appears that evolution has selected for a switch from a TH2 to TH1/17 preponderance early in development as a means of protection against potential pathogens. Our study shows that neonatal animals that fail to acquire commensal microbiota remain in a TH2 state and are at risk for pathogenic infections.

Finally, our study offers a potential mechanism for how the commensal microbiota is involved in the susceptibility to NEC. The alteration and/or delay in establishing a normal commensal microbiota has been theorized as a possible mechanism for NEC (48). Our data show that the interplay of microbiota and immune function during the early postnatal period is critical and that the treatment of premature infants with probiotic medication may help to mimic the establishment of a normal commensal microbiota, and therefore establish a normal mucosal and systemic immune system.

In conclusion, we have demonstrated that the acquisition of a commensal microbiota is necessary for the early postnatal development of immune function. In this model, newly arriving commensal bacteria are sensed by cells, both in the epithelium and in the lamina propria that express a basal level of TLRs. These bacteria induce expression of host TLR and an accompanying innate response. This early immune response is necessary for activation of the mucosal immune function, prevention of bacterial translocation, and as a checkpoint for the switch from the in utero TH2 predominant phenotype to a mature TH1/TH17 T-cell commitment (Fig. 9). Determining the specific bacterium and/or bacterial products that are essential for this process could provide insight into the mechanism of both disease and protection in neonatal sepsis and NEC.

Proposed mechanism for commensal bacteria acquisition and early postnatal mucosal and immune development.


We thank Dr Vance McCracken (Southern Illinois University at Edwardsville), Jamie McNaught, and the Cell and Molecular Pathology Core of the Digestive Disease Developmental Center at UAB for assistance with the PCR-DGGE experiments and histology, Peggie McKie-Bell for animal husbandry assistance, and Dr Suzanne M. Michalek for the use of her flow cytometer.


1. Bäckhed F, Ley R, Sonnenburg J, et al. Host-bacterial mutualism in the human intestine. Science 2005; 307:1915–1920.
2. Hooper L, Midtvedt T, Gordon J. How host-microbial interactions shape the nutrient environment of the mammalian intestine. Annu Rev Nutr 2002; 22:283–307.
3. Ley R, Bäckhed F, Turnbaugh P, et al. Obesity alters gut microbial ecology. Proc Natl Acad Sci U S A 2005; 102:11070–11075.
4. Levy O. Innate immunity of the newborn: basic mechanisms and clinical correlates. Nat Rev Immunol 2007; 7:379–390.
5. Trowsdale J, Betz A. Mother's little helpers: mechanisms of maternal-fetal tolerance. Nat Immunol 2006; 7:241–246.
6. Schroeder HJ, Zhang L, Philips JR. Slow, programmed maturation of the immunoglobulin HCDR3 repertoire during the third trimester of fetal life. Blood 2001; 98:2745–51.
7. Salzman N, Underwood M, Bevins C. Paneth cells, defensins, and the commensal microbiota: a hypothesis on intimate interplay at the intestinal mucosa. Semin Immunol 2007; 19:70–83.
8. Morein B, Blomqvist G, Hu K. Immune responsiveness in the neonatal period. J Comp Pathol 2007; 137(Suppl 1):S27–31.
9. Parker L, Whyte M, Dower S, et al. The expression and roles of Toll-like receptors in the biology of the human neutrophil. J Leukoc Biol 2005; 77:886–892.
10. Lemons J, Bauer C, Oh W, et al. Very low birth weight outcomes of the National Institute of Child Health and Human Development Neonatal Research Network, January 1995 through December 1996. NICHD Neonatal Research Network. Pediatrics 2001; 107:E1.
11. Barclay A, Stenson B, Simpson J, et al. Probiotics for necrotizing enterocolitis: a systematic review. J Pediatr Gastroenterol Nutr 2007; 45:569–576.
12. Alfaleh K, Bassler D. Probiotics for prevention of necrotizing enterocolitis in preterm infants. Cochrane Database Syst Rev 2008(1):CD005496.
12a. Tatum PM, Harmon CM, Lorenz, RG, et al. Toll-like receptors 2 and 4 are protective against injury in a neonatal murine model of NEC. J Pediatr Surg. In press.
13. Williams A, Probert C, Stepankova R, et al. Effects of microflora on the neonatal development of gut mucosal T cells and myeloid cells in the mouse. Immunology 2006; 119:470–478.
14. Macpherson A, Uhr T. Compartmentalization of the mucosal immune responses to commensal intestinal bacteria. Ann N Y Acad Sci 2004; 1029:36–43.
15. Macpherson A, Hapfelmeier S, McCoy K. The armed truce between the intestinal microflora and host mucosal immunity. Semin Immunol 2007; 19:57–58.
16. Schumann A, Nutten S, Donnicola D, et al. Neonatal antibiotic treatment alters gastrointestinal tract developmental gene expression and intestinal barrier transcriptome. Physiol Genomics 2005; 23:235–245.
17. Masumoto J, Yang K, Varambally S, et al. Nod1 acts as an intracellular receptor to stimulate chemokine production and neutrophil recruitment in vivo. J Exp Med 2006; 203:203–213.
18. McCracken V, Lorenz R. The gastrointestinal ecosystem: a precarious alliance among epithelium, immunity and microbiota. Cell Microbiol 2001; 3:1–11.
19. Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 1987; 162:156–159.
20. Hans W, Schölmerich J, Gross V, et al. The role of the resident intestinal flora in acute and chronic dextran sulfate sodium-induced colitis in mice. Eur J Gastroenterol Hepatol 2000; 12:267–273.
21. Simpson J, Martineau B, Jones W, et al. Characterization of fecal bacterial populations in canines: effects of age, breed and dietary fiber. Microb Ecol 2002; 44:186–197.
22. O'Sullivan L, Webster G, Fry J, et al. Modified linker-PCR primers facilitate complete sequencing of DGGE DNA fragments. J Microbiol Methods 2008; 75:579–581.
23. Schultz C, Edrington T, Schroeder S, et al. Effect of the thyroid on faecal shedding of E. coli O157:H7 and Escherichia coli in naturally infected yearling beef cattle. J Appl Microbiol 2005; 99:1176–1180.
24. Cario E, Gerken G, Podolsky D. Toll-like receptor 2 controls mucosal inflammation by regulating epithelial barrier function. Gastroenterology 2007; 132:1359–1374.
25. Savidge T, Newman P, Pan W, et al. Lipopolysaccharide-induced human enterocyte tolerance to cytokine-mediated interleukin-8 production may occur independently of TLR-4/MD-2 signaling. Pediatr Res 2006; 59:89–95.
26. Fan J, Frey R, Malik A. TLR4 signaling induces TLR2 expression in endothelial cells via neutrophil NADPH oxidase. J Clin Invest 2003; 112:1234–1243.
27. Fan J, Malik A. Toll-like receptor-4 (TLR4) signaling augments chemokine-induced neutrophil migration by modulating cell surface expression of chemokine receptors. Nat Med 2003; 9:315–321.
28. Zhou X, Gao X, Fan J, et al. LPS activation of Toll-like receptor 4 signals CD11b/CD18 expression in neutrophils. Am J Physiol Lung Cell Mol Physiol 2005; 288:L655–L662.
29. Makhoul I, Sujov P, Ardekian L, et al. Factors influencing oral colonization in premature infants. Isr Med Assoc J 2002; 4:98–102.
30. Fanaro S, Chierici R, Guerrini P, et al. Intestinal microflora in early infancy: composition and development. Acta Paediatr Suppl 2003; 91:48–55.
31. Fusunyan R, Nanthakumar N, Baldeon M, et al. Evidence for an innate immune response in the immature human intestine: Toll-like receptors on fetal enterocytes. Pediatr Res 2001; 49:589–593.
32. Lotz M, Gütle D, Walther S, et al. Postnatal acquisition of endotoxin tolerance in intestinal epithelial cells. J Exp Med 2006; 203:973–984.
33. Newburg D, Walker W. Protection of the neonate by the innate immune system of developing gut and of human milk. Pediatr Res 2007; 61:2–8.
34. LeBouder E, Rey-Nores J, Raby A, et al. Modulation of neonatal microbial recognition: TLR-mediated innate immune responses are specifically and differentially modulated by human milk. J Immunol 2006; 176:3742–3752.
35. Berg R, Garlington A. Translocation of certain indigenous bacteria from the gastrointestinal tract to the mesenteric lymph nodes and other organs in a gnotobiotic mouse model. Infect Immun 1979; 23:403–411.
36. Lecuit M, Sonnenburg J, Cossart P, et al. Functional genomic studies of the intestinal response to a foodborne enteropathogen in a humanized gnotobiotic mouse model. J Biol Chem 2007; 282:15065–15072.
37. Cario E, Gerken G, Podolsky D. Toll-like receptor 2 enhances ZO-1-associated intestinal epithelial barrier integrity via protein kinase C. Gastroenterology 2004; 127:224–238.
38. Aprahamian C, Lorenz R, Harmon C, et al. Toll-like receptor 2 is protective of ischemia-reperfusion-mediated small-bowel injury in a murine model. Pediatr Crit Care Med 2008; 9:105–109.
39. Martin F, Wang Y, Sprenger N, et al. Probiotic modulation of symbiotic gut microbial-host metabolic interactions in a humanized microbiome mouse model. Mol Syst Biol 2008; 4:157.
40. Che C, Pang X, Hua X, et al. Effects of human fecal flora on intestinal morphology and mucosal immunity in human flora-associated piglet. Scand J Immunol 2009; 69:223–233.
41. Leigh L, Stoll B, Rahman M, et al. Pseudomonas aeruginosa infection in very low birth weight infants: a case-control study. Pediatr Infect Dis J 1995; 14:367–371.
42. Allam J, Peng W, Appel T, et al. Toll-like receptor 4 ligation enforces tolerogenic properties of oral mucosal Langerhans cells. J Allergy Clin Immunol 2008; 121:368–374, e361.
43. Bauer E, Williams B, Smidt H, et al. Influence of the gastrointestinal microbiota on development of the immune system in young animals. Curr Issues Intest Microbiol 2006; 7:35–51.
44. Herzenberg L, Tung J. B cell lineages: documented at last! Nat Immunol 2006; 7:225–226.
45. Nutten S, Schumann A, Donnicola D, et al. Antibiotic administration early in life impairs specific humoral responses to an oral antigen and increases intestinal mast cell numbers and mediator concentrations. Clin Vaccine Immunol 2007; 14:190–197.
46. Forchielli M, Walker W. The role of gut-associated lymphoid tissues and mucosal defence. Br J Nutr 2005; 93(Suppl 1):S41–S48.
47. Piccinni M, Scaletti C, Mavilia C, et al. Production of IL-4 and leukemia inhibitory factor by T cells of the cumulus oophorus: a favorable microenvironment for pre-implantation embryo development. Eur J Immunol 2001; 31:2431–2437.
48. Neu J, Chen M, Beierle E. Intestinal innate immunity: how does it relate to the pathogenesis of necrotizing enterocolitis. Semin Pediatr Surg 2005; 14:137–144.
49. Staley E, Schoeb T, Lorenz R. Differential susceptibility of p-glycoprotein deficient mice to colitis induction by environmental insults. Inflamm Bowel Dis 2009; 15:684–696.
50. Newberry R, McDonough J, McDonald K, et al. Postgestational lymphotoxin/lymphotoxin beta receptor interactions are essential for the presence of intestinal B lymphocytes. J Immunol 2002; 168:4988–4997.

antibiotics; commensal microbiota; necrotizing enterocolitis; probiotics; Toll-like receptors

Copyright 2010 by ESPGHAN and NASPGHAN