Bifidobacterium Breve Enhances Transforming Growth Factor β1 Signaling by Regulating Smad7 Expression in Preterm Infants

Abstract

Objectives: 

Transforming growth factor (TGF) β1 displays a broad spectrum of activities in mucosal regulation, including induction of oral tolerance, potent anti-inflammatory effects, mucosal IgA expression and effects on epithelial cell proliferation and differentiation. The present study examined the effect of probiotics on the immunologic system of preterm infants in relation to TGF-β signaling.

Methods: 

Subjects comprised 19 preterm infants divided into 2 groups: receiving Bifidobacterium breve supplementation (B. breve group) and without supplementation (controls). Blood samples were collected from both groups on days 0, 14 and 28 after birth. Serum cytokine levels were measured using enzyme-linked immunosorbent assay, and expression levels of the TGF-β signaling molecule, Smad, were examined using semiquantitative reverse transcriptase-polymerase chain reaction.

Results: 

Serum TGF-β1 level was elevated on day 14 and remained elevated on day 28 in the B. breve group. Level of messenger RNA expression was enhanced for Smad3 and reduced for Smad7 (antagonistic Smad) after B. breve administration relative to levels in controls on day 28.

Conclusions: 

These results demonstrated that the administration of B. breve to preterm infants can up-regulate TGF-β1 signaling and may possibly be beneficial in attenuating inflammatory and allergic reactions in these infants.

INTRODUCTION

Many physiological changes take place in newborn infants soon after delivery while adjusting to the new environment-one that is markedly different from the maternal environment. Because infants can be barraged by all kinds of antigens soon after delivery, dynamic changes are expected in the newborn immunologic system immediately after birth.

Th1 and Th2 cells are characterized by cytokine synthesis patterns. Th1 cells work predominantly for cell-mediated immunologic reactions and produce interleukin (IL) 2, interferon (IFN) γ, tumor necrosis factor (TNF) α and IL-12. Th2 cells work for antibody synthesis and produce IL-4, -5, -6 and -13. Th3 cells also exist, producing transforming growth factor (TGF) β and IL-10. TGF-β and IL-10 are important cytokines in preventing inappropriate immunologic reactions in infants because TGF-β and IL-10 display a broad spectrum of activities in mucosal regulation, including induction of oral tolerance and potent anti-inflammatory effects. In addition, TGF-β can up-regulate mucosal IgA expression and effects on epithelial cell proliferation and differentiation.(^1-3) TGF-β signals through a receptor serine kinase that then phosphorylates and thereby activates transcription factors Smad2 and Smad3. Smad4, a common mediator, then binds to these activated molecules and can send signals to the nucleus. The Smad7 molecule may also function as an antagonistic Smad and may inhibit TGF-β-mediated Smad3 phosphorylation to attenuate TGF-β signaling to the nucleus.

Probiotics have been shown to be useful in defense against and control of infection associated with stabilization of the intestinal microflora.(⁴) Li et al.(⁵) reported that very early administration of Bifidobacterium breve to low birth-weight infants helped to promote Bifidobacterium colonization, leading to the development of normal intestinal flora. A study by Kitajima et al.(⁶) using very low birth-weight infants found that B. breve colonized the immature bowel very effectively and was associated with fewer abnormal abdominal signs such as aspirated air volume and better weight gain than seen in the control group. Probiotics may also strengthen the intestinal mucosal immune system through secretory antibody production and T helper cell responses.(⁷) However, few thorough immunologic analyses have been performed for probiotics in preterm infants, particularly in early infancy.

The present study investigated the effects of probiotics in early infancy as reflected in serum cytokine levels and expression of TGF-β signaling Smad molecules, with particular emphasis on preterm infants.

MATERIALS AND METHODS

Subjects

This study was designed as a randomized controlled study. Subjects comprised 19 preterm infants admitted to our neonatal intensive care unit. Infants with complications, such as chromosomal or congenital anomalies or history of intrauterine infection or surgery, were excluded from this study. Infants who had received or whose mothers had received corticosteroid treatment were also excluded. Subjects were alternately allocated to 1 of 2 groups: either receiving B. breve supplementation (B. breve group); or receiving vehicle supplementation only (controls). The strain used for supplementation in this study was B. breve M-16V (Morinaga Milk Industry, Kanagawa, Japan), representing live but not viable bacteria. The B. breve group consisted of 11 infants (7 boys, 4 girls) with a mean birth weight of 1378 ± 365 g and a mean gestational age of 31.3 ± 3.16 weeks. In these infants, B. breve was administered by suspending 1.0 × 10⁹ cells in 0.5 mL of 5% glucose solution, starting several hours after birth. B. breve was administered using a nasogastric tube twice a day during the hospital stay. Controls consisted of 8 infants (5 boys, 3 girls), with a mean birth weight of 1496 ± 245 g and a mean gestational age of 31.2 ± 1.98 weeks. These subjects were administered a 5% glucose solution (without any B. breve) via a nasogastric tube, as above. Antibiotics were used primarily to prevent infants from infections due to tracheal intubation or use of a central venous catheter. Probiotics used in this study were resistant to commonly used antibiotics such as ampicillin, amikacin sulfate, arbekacin sulfate, imipenem/cilastatin sodium and cefotaxime sodium.

Capillary blood samples were collected on days 0, 14 and 28 from heel cuts into 1-mL tubes loaded with heparin. Soon after blood samples were taken, serum was separated for enzyme-linked immunosorbent assay (ELISA). All study protocols were approved by the Institutional Ethics Committee of Juntendo University Hospital and informed consent for participation was obtained from the parents of all infants before enrollment in the study.

Serum Cytokine Levels by ELISA

To determine serum cytokine levels in preterm, low birth-weight infants, ELISA kits for human IL-4 (R&D Systems Europe, Oxon, UK), IL-5 (R&D Systems Europe), IL-6 (Fujirebio, Tokyo, Japan), IFN-γ (Biosource International, Camarillo, CA), TNF-α (Japan Immunoresearch Laboratories, Gunma, Japan) and TGF-β1 (R&D System Europe) were used according to the instructions of the manufacturers. Each ELISA assay was performed in triplicate.

Expression of Smad messenger RNA by Reverse Transcriptase-Polymerase Chain Reaction

A random selection of 7 preterm infants from controls and 9 preterm infants from the B. breve group were selected for extraction of RNA from peripheral blood mononuclear cell using a monophasic solution of phenol and guanidine isothiocyanate (RNA STAT 60 reagent; Tel-Test, Friendswood, TX) and chloroform, followed by isopropanol precipitation. Synthesis of cDNA was performed using reverse transcriptase (RT) at 42°C for 30 minutes from 0.5 μg of total cellular RNA using the GeneAmp RNA polymerase chain reaction (PCR) kit (Applied Biosystems, Branchburg, NJ) according to the protocols of the manufacturer. Primer pairs for human Smad2 were as follows: sense, 5′-GAATTTGCTGCTCTTCTGGCTCAG-3′ and antisense, 5′-GCCATAGGGACCACAACACAATG-3′, resulting in a 322-bp PCR product. Primer pairs for human Smad3 were sense, 5′-GAGGGCAGGCTTGGGGAAAATG-3′, and antisense, 5′-GGGAGGGTGCCGGTGGTGTAATAC-3′, resulting in a 281-base pair (bp) PCR product. Primer pairs for human Smad4 were sense, 5′-AAAGGTGAAGGTGATGTTTGGGTC-3′, and antisense, 5′-CTGGAGCTATTCCACCTACTGATCC-3′, resulting in a 268-bp PCR product. Primer pairs for human Smad7 were sense, 5′-CATCACCTTAGCCGACTCTG-3′, and antisense, 5′-GTCTTCTCCTCCCAGTATGC-3′, resulting in a 227-bp PCR product.(⁸) Primer pairs for human β-actin, representing a house-keeping gene, were sense, 5′-ATCTGGCACCACACCTTCTACAATGAGCTGCG-3′, and antisense, 5′-CGTCATACTCCTGCTTGCTGATCCACATCTGC-3′, producing an 838-bp PCR product. The thermal cycle was programmed with a hot start at 95°C for 5 minutes followed by 35 cycles at 95°C for 1 minute and annealing at 60°C (Smad2), 62°C (Smad3), or 56°C (Smad4 and Smad7) for 1 minute, followed by extension at 72°C for 1 minute, as previously described.(⁹) RT-PCR products were electrophoresed in 1% agarose gel stained with ethidium bromide and visualized through ultraviolet light. A PCR mixture with distilled H₂O instead of RNA was used as a negative control. Band intensities were semiquantified by densitometry, and the intensity ratio of each reading compared with β-actin was also calculated.

Statistics

All data were analyzed using analysis of variance to determine individual changes at each time point within each group, the Mann-Whitney U test to evaluate the effects of B. breve at each time point between groups and Fisher exact test to determine differences between groups. Values of P < 0.05 were considered statistically significant.

RESULTS

Clinical Findings

No significant differences were noted between groups in mean gestational age, birth weight, duration of hospital stay, antibiotic use or central venous catheter use, infection or complications including frequency of respiratory distress syndrome, chronic lung disease and retinopathy of prematurity or asphyxia during or after delivery (Table 1). Neither was any differences noted in time until commencement of breast-feeding or proportion of breast-feeding to artificial feeding during the study period. Most infants were nutritionally maintained using a full-dose feeding via a minimum central venous catheter. No adverse effects were observed after B. breve supplementation.

TABLE 1: Profiles of infants in this study

Serum Cytokine Levels and Expression of TGF-β Signaling Molecules in Controls

Serum IL-4, IL-5, IL-6, IFN-γ, TNF-α and TGF-β1 levels were measured using ELISA. Only serum IL-4 and TGF-β1 levels were detectable during this study (Figs. 1, 2).

FIG. 1: Serum IL-4 levels as measured by ELISA. Although no significant difference was identified between groups, serum IL-4 level was elevated on day 14 but had returned to baseline by day 28 in Controls. Each assay was performed in triplicate.

FIG. 2: TGF-β levels as measured by ELISA. Serum TGF-β levels were significantly elevated on day 14 (P = 0.02) but returned to baseline by day 28 in controls. Significant elevations were confirmed on day 14 (P = 0.026) and day 28 (P = 0.029) compared with day 0 in the B. breve group. Serum TGF-β level was significantly elevated after B. breve administration on day 28 (P = 0.005). Each assay was performed in triplicate.

Regarding our analysis of cytokine synthesis, no significant differences were noted between serum IL-4 levels at day 0 (8.1 ± 4.3 pg/mL) and day 28 (7.0 ± 5.2 pg/mL) after birth (Fig. 1). Serum IL-4 levels were elevated on day 14 (10.7 ± 7.0 pg/mL) but had returned to baseline by day 28. A significant elevation of serum TGF-β1 level was confirmed from day 0 (23.7 ± 22.8 ng/mL) to day 14 (53.8 ± 31 ng/mL; P = 0.02) but had decreased to baseline by day 28 (33.1 ± 21.9 ng/mL) (Fig. 2).

To examine the development of Smad2, Smad3, Smad4 and Smad7 messenger RNA (mRNA) expression for these TGF-β signaling molecules, semiquantitative RT-PCR was performed, and levels were compared with levels of β-actin mRNA (Figs. 3-7. No significant differences in expression of Smad2 (Fig. 4) or Smad4 (Fig. 6) were noted between days 0, 14 and 28. Expression of Smad3 was significantly lower on day 28 than on day 14 (P = 0.04) (Fig. 5), whereas expression of Smad7 was significantly higher on day 28 than on day 14 (P = 0.037; Fig. 7).

FIG. 3: Expression of mRNA for Smad2, 3, 4 and 7 and β-actin by RT-PCR on days 0, 14 and 28. Expression was increased for Smad2, 3 and 4 mRNA and reduced for antagonistic Smad7 in early infancy. Representative data from each infant are presented.

FIG. 4: Expression of Smad2 mRNA was examined using semi-quantitative RT-PCR and was compared with expression of β-actin mRNA. In controls, no significant differences in expression of Smad2 were noted between days 0, 14 and 28. In the B. breve group, expression of Smad2 was significantly enhanced on day 28 compared with expression on day 14 (P = 0.041).

FIG. 5: Expression of Smad3 mRNA was examined by semi-quantitative RT-PCR and was compared with expression of β-actin mRNA. In controls, expression of Smad3 was significantly reduced on day 28 compared with expression on day 14 (P = 0.04). In the B. breve group, expression of Smad3 remained enhanced on day 28 and was significantly enhanced compared with expression levels in controls (P = 0.03).

FIG. 6: Expression of Smad4 mRNA was examined by semi-quantitative RT-PCR and was compared with expression of β-actin mRNA. In both the B. breve group and controls, no significant differences were seen in expression of Smad4 among days 0, 14 and 28.

FIG. 7: Expression of Smad7 mRNA was examined by semi-quantitative RT-PCR and was compared with expression of β-actin mRNA. In controls, expression of Smad7 was significantly higher on day 28 when compared with day 0 (P = 0.037). In the B. breve group, expression of Smad7 was not enhanced on day 28.

Serum Cytokine Levels and Expression of TGF-β Signaling Molecules in the Bifidobacterium breve Group

We then examined changes of cytokine synthesis at each time point within the B. breve group. Serum IL-4 level was practically unchanged on each day. Serum TGF-β1 level was elevated in the same way compared with controls from day 0 (35.6 ± 21.2 ng/mL) to day 14 (64.6 ± 28.2 ng/mL; P = 0.026) and was also significantly elevated on day 28 (68.5 ± 24.7 ng/mL) compared with day 0 (P = 0.029) (Figs. 1, 2.

Regarding Smad mRNA expression, no significant differences in Smad4 expression were noted between days 0, 14 and 28 (Fig. 6). Although no significant difference in Smad2 expression was found between controls and the B. breve group on days 0 and 14, Smad2 expression was significantly enhanced on day 28 compared with day 14 in the B. breve group (P = 0.041; Fig. 4). Expression of Smad3 remained enhanced on day 28 (Fig. 5). Expression of Smad7 thus remained low and was not elevated on day 28 compared with days 0 and 14 (Fig. 7).

Finally, when we compared each groups at each time point, no significance differences were identified in serum IL-4 level or expression of Smad2, Samd4 and Smad7. Serum TGF-β levels in the B. breve group were thus significantly elevated on day 28 (P = 0.005) (Fig. 2), and expression of Smad3 on day 28 was significantly enhanced compared with controls (P = 0.03; Fig. 5).

DISCUSSION

The present study examined the effect of probiotics on the immunologic system of preterm infants in relation to TGF-β signaling by measuring serum cytokine levels using ELISA and by measuring expression of Smad molecules using semiquantitative RT-PCR. Because our subjects were preterm infants, the study was limited to the analysis of peripheral samples rather than more direct approaches for studying elements of the mucosal immune system, and the current approach thus necessitated extrapolation to clinical situations. We identified significant elevation of serum TGF-β1 levels (Fig. 2) and enhanced mRNA expression of the TGF-β signaling molecule Smad3, whereas levels of the antagonistic Smad, Smad7, were not elevated after B. breve administration relative to controls on day 28 (Figs. 5, 7. Although pohosphorylation of Smads is key to the activation of these molecules, expression of Smad molecules seems to be established in early infancy because expression increases after birth. We thus considered that these molecules were activated. These findings strongly suggest that B. breve administration can enhance host TGF-β signals in preterm infants.

Transforming growth factor β1 is an important cytokine for preventing inappropriate immunologic reactions in infants because TGF-β1 displays a broad spectrum of activities in mucosal regulation, including induction of oral tolerance, potent anti-inflammatory effects, mucosal IgA expression and effects on epithelial cell proliferation and differentiation (¹). Continuous production of a certain amount of serum TGF-β1 and IL-4 in these infants suggests that natural antigen exposure is suitable for stimulating immunologic development biased toward Th2 responses in early infancy. Furthermore, higher expressions of Smad2, 3 and 4 mRNA and lower expression of antagonistic Smad7 may indicate the importance of TGF-β signaling in early infancy (Fig. 3).

These findings suggested being effective mainly for the production of IgA against natural antigens in early infancy. In fact, early elevation of serum TGF-β1 levels on day 14 might possibly represent an important signal for future Th2 development, including IgA production in preterm infants. Monteleone et al.(¹¹) reported that Smad7, as a Smad antagonist, was overexpressed in T cells in the mucosa of inflammatory bowel disease, indicating that TGF-β works as an important cytokine for regulating the mucosal immune system in inflammatory bowel disease.(^12-14) A Smad7 blockade of TGF-β signaling thus helps to regulate chronic production of proinflammatory cytokines that drive the inflammatory process in preterm infants.

We also confirmed up-regulation of Smad3 expression but no significant difference in Smad2 in the B. breve group compared with controls. Although the precise mechanisms remain unclear, Smad3 expression seems more sensitive to B. breve stimulation in early infancy.

Yang et al.(¹⁰) suggested that Smad3 plays an important role in TGF-β-mediated regulation of T-cell activation and mucosal immunity and that loss of these functions is responsible for chronic infection. Up-regulation of Smad3 expression after B. breve administration may thus prove helpful in preventing bacterial infection in preterm infants. Prescott et al.(^15,16) suggested that initial T-cell development is toward the Th2 cytokine profile. In the present study, production of IL-4 and TGF-β1 in preterm infants was relatively increased on day 14, suggesting that extramaternal environmental factors may stimulate and introduce a Th2 response in these infants soon after birth. However, an IFN-γ burst occurs at about 4 to 6 months after birth, regulating overexpression of Th2 cell response. Lack of this regulation would seem likely to increase the risk of allergic disease in the future in preterm infants.(¹⁶) Because TGF-β1 can regulate the effects of Th2 cell responses, administration of B. breve to preterm infants may prevent future allergic diseases.

Enterobacteriaceae are known to represent the main bacteria colonizing the gut lumen after birth, whereas the number of Bifidobacterium gradually increase by 3 to 5 days after birth.

Our previous study with Bifidobacterium supplementation in preterm infants demonstrated that a Bifidobacterium-predominant flora formed at an average of 2 to 4 weeks, earlier than that in the control group. In addition, the number of Enterobacteriaceae, one candidate for causing necrotizing enterocolitis, was lower in supplemented groups at 2 weeks after birth than in the control group (⁵). Other reports have also demonstrated that probiotics can prevent necrotizing enterocolitis and other infections by reducing chemical mediator release, including platelet-activating factor (¹⁷). We suggest that probiotics might establish an appropriate microbial environment at around 2 weeks after birth, and it takes another 2 weeks assisting development of appropriate immune responses, such as TGF-β1 production. The present findings suggest that the mechanism of anti-inflammatory effects of B. breve involves attenuation of proinflammatory signals through up-regulation of TGF-β signals.

In conclusion, administration of B. breve to preterm infants may be beneficial in attenuating inflammatory and allergic reactions, by enhancing both production and signaling of TGF-β1, particularly in early infancy.