Impact of Diet in Shaping Gut Microbiota Revealed by a Comparative Study in Infants During the First Six Months of Life

Abstract

The development of the gut is controlled and modulated by different interacting mechanisms, such as genetic endowment, intrinsic biological regulatory functions, environment influences and last but no least, the diet influence. In this work, we compared the fecal microbiota of breast-fed (BF), formula-fed (FF), and mixed-fed (MF) infants from Hebei Province, China. By using high-throughput 16S rDNA sequencing analyses, we found some differences in gut microbiota in the three groups. Firmicutes and Proteobacteria were the dominant bacteria at the phylum level in the three groups, where FF infants showed a significant depletion in Bacteroidetes (p < 0.001) and Actinobacteria (p < 0.05). Enterobacteriaceae was the dominant bacteria at the family level in the three groups, but FF infants showed higher Enterobacteriaceae enrichment than BF and MF infants (p < 0.05); the abundance of the Bifidobacteriaceae was only 8.16% in the feces of BF infants, but higher than in MF and FF infants (p < 0.05). The number of genera detected (abundance >0.01%) in BF, MF, and FF infants was only 15, 16, and 13, respectively. This study could provide more accurate and scientific data for the future study of infant intestinal flora.

Introduction

The gastrointestinal microbiota plays a crucial role in health since it is involved in nutrition, pathogenesis, and immunology [42]. Infancy is a critical period of colonization of the intestinal flora, which affects the adult intestinal flora and the future of the health [26]. Infant diet could regulate the diversity of human gut microbiota [3]. Microbial imbalance has been linked to several functional gut disorders, such as inflammatory bowel disease [33, 34], irritable bowel syndrome [19], stomach cancer [29], mucosa-associated lymphoid tissue lymphoma [22], obesity [11, 38], and necrotizing enterocolitis [10]. Therefore, a study of the interaction between the infant diet and intestinal flora has important implications. Previous studies in intestinal ecology had highlighted the field of the human gut microbiota [8, 12, 39], but the studies of the intestinal flora are appearing only gradually. The main obstacle of these studies is the detection technique, as the cultivate-independent techniques have some considerable limitations, and they could only detect the cultivate-based microorganisms, and their proportion was only 1%-10% [18]. At the same time, some non-cultivate techniques were used for the study of intestinal flora, such as DGGE [31], real-time PCR [29], FISH [20], etc. These techniques provided some important information about the intestinal flora, but they also had some shortcomings, especially as they only could detect a certain abundance of flora or a specific microorganism. In recent years, high-throughput sequencing technology has been introduced. In one finished study, high-throughput sequencing technology has been applied in the study of intestinal flora [30], where it could detect almost all the flora in the feces, and enables us to have a more comprehensive understanding of the composition of infant intestinal flora and have a huge role in promoting the study of intestinal pathology. Therefore, we performed multiplex pyrosequencing of the V6 hypervariable regions of the 16S rRNA gene with Illumina high-throughput sequencing.

The aim of this study was to compare the gut microbiota of infants aged 1-6 months; the fecal samples were from 24 infants. The target of the study was to establishing a complete picture of the differences between breast-fed, mixed-fed, and formula-fed infants with regard to the composition of the healthy fecal microbiota following the type of feeding. We could illustrate the diversity partitioned within and between the three populations studied and the possible correlation between diversity and diet.

 

Materials and Methods

Infant Enrollment

We enrolled 24 healthy infants from a rural village in Hebei Province, China (Table 1). The infants were divided into three groups: breast-fed (BF) group (twelve), mixed-fed (MF) group (eight), and formula-fed (FF) group (four). The 24 infants were individually numbered: the breast-fed group (1-12), the mixedfed group (15-22), and the formula-fed group (23-26). All infants were 1-6 months of age; exclusive breast-feeding, formula-feeding, and mixed-feeding were established throughout the research period; no infant received any additional foods. In addition to the feeding pattern, there were no significant differences in gestational age and birth weight; and there were no use of antibiotics for infants in t he three groups. The infants d id n ot have a ny f ood allergies or any gastrointestinal diseases. Infants No. 1, 2, 6, and 11 were born by caesarean section.

Table 1.Characteristics of the studied samples.

Fecal Sample Collection and DNA Extraction

In all populations, fecal samples were collected by physicians from each individual in the morning, 1-2 h after the first meal, and preserved in RNAlater (Qiagen) at 4℃ for the first 48 h, and then kept at -80℃ until extraction. DNA was extracted according to the protocol described by the QIAamp DNA Stool Mini Kit (Qiagen). In order to improve efficiency of cell wall lysis of grampositive bacteria in the feces, the sample was incubated at 95℃ for 5 min. The extracted DNA was stored at -20℃ until processed.

PCR Amplification of V6 Region of Bacterial 16S rRNA Gene and Pyrosequencing

For each sample, we amplified 16S rRNA genes using a primer set corresponding to primers 976F and 1046R: CAACGCGAAGAACCTTACC, CGACAGCCATGCANCACCT [13]. These PCR primers target the V6 hypervariable 16S rRNA region. For each sample, a PCR mix of 50 μl was prepared containing 10× PCR buffer, 5 U of FastStart High Fidelity polymerase blend and dNTPs from the FastStart High Fidelity PCR system (Roche), 200 nM primers (Sangon), and 100 ng of genomic DNA. Thermal cycling consisted of initial denaturation at 94℃ for 4 min, followed by 25 cycles of denaturation at 94℃ for 30 sec, annealing at 55℃ for 30 sec, and extension at 72℃ for 1 min, with a final extension of 5 min at 72℃. Amplicons were visualized on 2.0% agarose gels using SYBR Safe DNA gel stain in 0.5× TBE (Invitrogen).

The PCR products derived from amplification of the specific 16 SrRNA-V6 hypervariable regions were sent to be purified and sequenced by Illumina technology (Illumina HiSeq 2000; BGIShenzhen, China).

Taxonomic Assignment to 16S Reads and Comparative Analysis of Multiple Samples

RDP classifier software [41] was used to classify the sequences, maintained at the Ribosomal Database Project. Based on the relative content of each species at the best taxonomic level, PCA (principal component analysis) was operated and the species with the biggest contribution were found. Then cluster analysis was used to determine the similarity of the species at the best taxonomic level. Two p-values, AU (Approximately Unbiased) and BP (Bootstrap Probability), were provided to evaluate the accuracy of the cluster analysis.

Richness and Diversity Index

A rarefaction curve was drawn to evaluate whether the sequencing was sufficient to cover all groups, and reflect the richness of species indirectly. Richness and biodiversity indices were obtained with the Mothur software package [34]. For richness estimation, related to the number of observed operational taxonomic units (OTUs), we used the Chao1 index. Biodiversity, which depends on how uniformly the sequences were spread into the different observed OTUs, was instead estimated with the nonparametric Shannon formula [7]. Both indexes were evaluated at an OTU distance unit cutoff of 0.03 to test different selectivity in the definition of OTUs.

Ethical Consideration

This study was approved by the Ethics Committee of Hebei, Province, China. The mothers of all of the infants enrolled in the research signed an informed consent form.

Statistical Analyses

Differences between populations had been analyzed using parametric (ANOVA) and nonparametric statistical methods. All results are presented as the mean value (± SE). Differences between groups were declared significant at p < 0.05.

 

Results

In this study, we characterized the fecal microbiota of 24 healthy infants from Hebei Province, China (Table 1). Twelve healthy infants were selected as representative consumers of the traditional breast-fed diet, 8 healthy infants were fed a mixed diet mixing the breast milk and formula milk powder (Table 2), and 4 healthy infants were selected to obtain the formula-fed diet (Table 2).

Table 2.All nutrients are expressed per 100 g except for energy.

Diversity of Intestinal Microbiota in Different Patterns of Feeding Infants

To characterize the bacterial lineages present in the fecal microbiota of these 24 infants, we performed multiplex pyrosequencing of the V6 hypervariable regions of the 16S rRNA gene with Illumina high-throughput sequencing. We generated a dataset consisting of 1,760,707 filtered, taxonomic 16S rRNA gene sequences (Supplemental Fig. S1 and Tables S1-S3) with a mean average (± SE) of 73,360 ± 486 sequences per sample (Table 3). The size of most tag sequences was about 60 bp, which was consistent with the length of most bacterial V6 regions. Nineteen bacterial phyla were detected in all samples by using Illumina sequencing of the 16S rRNA-V6 region, and more than 98.9% of the sequences in all samples were found to belong to the four most populated bacterial phyla; namely, Actinobacteria, Bacteroidetes, Firmicutes, and Proteobacteria. Proteobacteria and Firmicutes were the predominant bacterial phyla in the three groups, and relevant differences were found in the proportions of four phyla: Proteobacteria was more abundant in FF and MF infants than in BF infants (53.8%, 42.4% versus 29.85%), whereas Firmicutes were more represented in BF than in FF and MF infants (44.21% versus 42.78%, 34.37%). Actinobacteria and Bacteroidetes belonged to the secondary bacterial phyla in the three groups, where their abundance in BF, MF, and FF infants was 10.88%, 8.96%, 3.06% and 14.96%, 8.30%, 0.11%, respectively; therefore, the relative abundance of these two bacterial phyla in the BF group was highest, and their relative abundance in FF infants was lowest (Fig. 1). Statistical analysis indicated that Actinobacteria (p = 0.023) differentiated MF from FF infants, and it significantly differentiated BF from FF infants (p = 0.0078), whereas Bacteroidetes (p < 0.001) differentiated MF from FF infants, and it (p < 0.001) significantly differentiated BF from FF infants (Fig. 1).

Fig. 1.Relative abundance of the intestinal flora at the phylum level in infants (±SE). Asterisks indicate significant differences (*p < 0.05; **p < 0.01; ***p < 0.001).

Table 3.Sequences summary in each sample.

Fig. 2.Relative abundance (>1%) of the intestinal flora at the family level in infants (percent).

Fig. 3.Fig. 3. Statistical analysis of microbial diversity of different samples at the family level. Relative abundance ≥1% of the bacteria in samples are shown, and less than 1% of those were counted as others.

Fig. 4.Relative abundance of the intestinal flora at the genus level in infants (percent).

The obtained OTUs in all samples showed that 90% of the tag sequences could be annotated at the family level, but the proportion of annotated tag at the genus level was only 5.06% (Supplemental Tables S4-S6), so we selected the family level as the best taxonomic level. The number of detected bacteria (abundance 1%) at the family level in all infant fecal samples was 13; the other bacteria (abundance <1%) were ranged to others (Figs. 2 and 3). Enterobacteriaceae (Proteobacteria) was the dominant bacteria in BF, MF, and FF infants (28.66%, 41.44%, and 50.82%), Veillonellaceae (Firmicutes) was more abundant in BF than in MF and FF infants (21.42% versus 3.65%, 3.25%), but Streptococcaceae (Firmicutes) was less abundant in BF than in MF and FF infants (8.67% versus 24.01%, 15.14%). Bifidobacteriaceae (Actinobacteria) and Bacteroidaceae (Bacteroidetes) were more represented in BF infants than MF and FF infants (8.16% versus 6.16%, 1.48%; 12.82% versus 4.32%, 0.04%). The intestinal bacteria of FF infants were more complex, and their species diversity was higher than in BF and MF infants.

The detected bacteria were few at the genus level, where the number of genera (abundance >0.01%) in BF, MF, and FF infants was only 15, 16, and 13 respectively (Fig. 4). The abundance of Bifidobacterium (19.15%) and Bacteroides (30.08%) in BF infants was higher than in MF (12.37%, 10.76%) and FF infants (4.04%, 0.11%), but Streptococcus and Klebsiella were less represented in BF infants than in MF and FF infants. Interestingly, Parabacteroides and Collinsella were nearly not found in FF infants.

16S rRNA Gene Surveys Reveal Hierarchical Separation of the Three Populations

We further assessed differences in the total bacterial community at the single sample level by clustering BF, MF, and FF samples according to their bacterial genera as found by the RDP classifier. Complete linkage hierarchical clustering produced a net separation of BF, MF, and FF populations (Fig. 5). It was noteworthy that the samples were divided into the two major clusters, the left cluster being mainly FF and MF groups (1-3 months), and the right cluster being mainly BF and MF groups (4-6 months). The resemblance between the left and the right cluster was very low, which indicated that the intestinal microbiota in FF and BF infants were different. It was prominent that in these three groups, we observed abundant Proteobacteria (Fig. 6), mainly represented by Enterobacteriaceae family (Fig. 7). We observed abundant Actinobacteria, mainly represented by Bifidobacterium genus, where it was found in all subjects (Fig. 3) and that was known to be strictly related to breastfeeding in infants. This result provided a clear indication of the dominant role of diet over the variables mentioned above in shaping the microbial composition of the gut.

Fig. 5.Cluster analysis of all samples.

Fig. 6.Relative abundance (percentage of sequences) of the four most abundant bacterial phyla in each individual among the BF, MF, and FF infants.

Fig. 7.Principal component analysis of different samples at the family level. (A) Figures represent the sample ID. Red (BF), black (MF), blue (FF). (B) Arrows indicate the contributive species for the diversity.

Microbial Richness and Biodiversity

Most of the sample dilution curves tendde to flatten (Fig. 8); it meant that most of bacterial species were detected. We then compared the microbial richness, estimated by the Chao1 index, and the biodiversity, assessed by a nonparametric Shannon index for the three groups. In our calculations, we took into account the OTU distance unit cutoff of 0.03 (Table 4). We found a high microbial diversity in the infant feces and detected 461 species at most, but the diversity of the infant intestinal flora with different feeding patterns was different; the abundance of FF infant fecal flora was higher than for BF infants. Using the parametric test for comparisons, we could not find significant differences (p > 0.05) in both richness and biodiversity between the BF, MF, and FF samples at the OTU cutoff of 0.03.

Fig. 8.Dilution curves of all samples.

Table 4.Richness and diversity indexes relative to each fecal sample: Number of observed OTUs (Obs. OTU), the Chao1 index (Chao), and the nonparametric Shannon index (Np Shannon) at OTU cutoff of 0.03.

 

Discussion

The establishment and evolution of the infant intestinal flora are a complex process. Microbial colonization of the gastrointestinal tract begins immediately after birth and is carried out by microbes that are derived from the mother and the environment. The course of microbial colonization of the newborn gut is determined by a complex interaction of factors, including the mode of delivery, feeding practice, and the bacterial load in the environment. When the baby is born, the intestinal tract is sterile, and some vaginal and dermatic flora start to plant in the infant intestine, and they reach an initial balance after 1 to 2 weeks, and then the feeding patterns may become the main factors [7, 21, 25]. The intestinal microbiotas of infants are different from the adult; similarities appear around 1 year of age and converge towards a more commonly shared adult-like microbiota [25].

In our study, we used Illumina sequencing technology to study the composition of the infant intestinal flora under different feeding patterns, and found a high microbial diversity in the infant feces. The results showed that more than 98.9% of genes belonged to the Actinobacteria, Bacteroidetes, Firmicutes, and Proteobacteria phyla, in agreement with the previous study describing such phyla as those contributing to the majority of infant gut microbiota [2]. The previous research pointed out that the human dominant intestinal flora was mainly Bacteroidetes and Firmicutes bacterial phyla [9], but the study showed that the infant dominant intestinal microflora were Firmicutes and Proteobacteria bacterial phyla. By comparing the microbial richness, the abundance of FF infant fecal flora was higher than BF infants, which was consistent with the previous study [4].

The impact of feeding patterns for the infant intestinal flora is mainly concentrated in the initial two or three months after birth, and previous studies have shown that the dominant bacteria in the breast-fed infant intestine were Bifidobacteriaceae and Lactobacillaceae, but artificial feeding infant intestinal flora were multiple [14, 33]. This study found that the dominant bacteria in BF infant intestine were Enterobacteriaceae, Veillonellaceae, and Bacteroidaceae during 1 to 6 months. The average abundance of Bifidobacteriaceae was only about 8%, which was not the dominant bacteria, but significantly higher than in the FF group. The abundance of Bifidobacterium (Bifidobacteriaceae) of in BF group (19.15%) was significantly higher than in the FF group (4.04%); the first feasible reason was that the human breast milk is an important source of oligosaccharides, which have a strong prebiotic effect for the eonate’s developing microbiota, suggesting an important role of breast milk as a delivering system for probiotic bacteria [15], and these human milk oligosaccharides may upregulate the expression of several pathways in these bacteria [17]; the second possible cause was that the breast milk contained these microbes such as bifidobacteria [23], and thus may be a direct source of the Bifidobacterium that became established in the infant gut [24, 35]. Lactobacillaceae in the three groups was little; the main reason was the elevatory fecal pH, so the environment was not suitable for Lactobacillaceae growth. The FF infant fecal flora were complex and the dominant bacteria were Enterobacteriaceae and Streptococcaceae. The abudance of other bacteria such as Clostridiaceae, Peptostreptococcaceae, Enterococcaceae and Erysipelotrichaceae was little, which indicated that the role of these bacteria in the gut may be relatively small [40]. In addition, this study found that Enterobacteriaceae, Veillonellaceae, Bacteroidaceae, and Bifidobacteriaceae were the main bacteria that caused the diversity of the infant intestinal flora. Enterobacteriaceae was the dominant bacteria in the three groups but the abundance in the BF group was significantly lower than is the FF and MF groups; the abundance of Veillonellaceae in the BF group was significantly more than in the FF and MF groups. The abundance of Veillonellaceae, Bacteroidaceae, and Bifidobacteriaceae in the BF group was higher than FF group; it may be these bacteria were well suited for the utilization of breast milk.

In addition, this study also found that there was large diversity in each group, which was caused by the infant individual differences. First of all, the establishment of the infant intestinal flora was a dynamic evolution; the sampling period was the possible reason for the diversity. Second, the mode of delivery and environment also affected the composition of infant intestinal flora. Infants 1, 2, 6, and 11 were caesarean children in this study, and the dominant bacteria were Veillonellaceae and Enterobacteriaceae, and the abundance of Bacteroidaceae and Bifidobacteriaceae was very little, and thus their intestinal flora were different from other samples. Some previous studies have reported that the abundance of the intestinal flora of caesarean children was very low, and the colonization of the intestinal flora lagged behind full-term infants. It was noteworthy that the initial colonization of Bifidobacterium and Bacteroides was not normal, which would affect the intestinal flora evolution, and the influence might last a long time [5, 6, 27].

Although it is now apparent that the feeding pattern is not the sole determinant of the levels of bacteria in the infant gut, it is clear that feeding does have a crucial impact. Breastfeeding elevates the levels of bifidobacteria and lactobacilli in the infant gut, and the consumption of breast milk could significantly reduce the risk of necrotizing enterocolitis in infants relative to those who were formula fed [1], so breastfeeding is accepted as being highly beneficial to infants. Breast milk is a nutritious food for the newborn, which contains the appropriate nutrients for the growing infant, and it can also have a significant impact on the gut microbial composition by virtue of being a source of prebiotics, which beneficially effect the infant by selectively stimulating the growth of one or a limited number of bacteria in the gut [16]. From the nutritional point of view, breastfeeding is considered the sublimate feeding pattern; its sustainable protective effect could reach to two years after the birth.

In conclusion, there were some differences in the gut microbiota in the three groups; the study provides a more comprehensive understanding of the structure and diversity of the intestinal flora diversity in BF, MF, and FF infants, and since the MF infant intestinal flora were less studied in the past, our results could provide a theoretical and technical support for future studies of the infant intestine. In comparison with high levels of breastfeeding in the past, increasing countries have much lower rates of breastfeeding. In this context, the research has supplied plentiful data on changes in the composition of the intestinal microbiota during early weaning and how the feeding patterns influence the structure of the infant intestinal microbiota before the introduction of complementary foods, and the data could be applied to develop new infant products that closely resemble breast milk. Next, we should closely investigate the exact dynamic changes in the gut microbial composition of breast-fed infants before and after adding complementary foods. More investigations are also required to determine the impact of formula feeding on the infant health both short-term or long-term, to illustrate the correlation between the impact on infant health and diseases, and explore what we can do to treat the diseases.