The composition of gut microbiota is thought to change during the aging process [1]; however, few reports have utilized molecular techniques to investigate the long-term, sequential changes in gut microbiota composition. Our results are in agreement with those of recent studies indicating clear differences in gut microbiota composition among infants, adults and the elderly [2, 23]. The present study using LEfSe method indicated that Actinobacteria, Clostridia and Bacteroidetes, Betaproteobacteria and Deltaproteobacteria were representative taxa in infant, adults and the elderly cluster. Additionally, our results revealed the sequential changes that occur with age from newborns to centenarians. Furthermore, our results showed that Japanese adults (21–69 years old) have a greater abundance of the genera Blautia and Bifidobacterium (interquartile ranges (IQR) of 18 (12–24) % and 7 (2–14) %, respectively) and a relatively lower abundance of genera related to Bacteroidetes (IQR 4 (1–10) %), compared with those reported in previous studies in other nations. For example, the estimated abundance of Blautia, Bifidobacterium and Bacteroidetes were??10 %, 2 % and??10 %, respectively, in US and Colombia [19], Korea, China and US [20] and Ireland [30]. These patterns might be characteristic of the gut microbiota composition of the Japanese population, although they may also reflect the DNA extraction method [16, 17] and the amplified region of the 16S rRNA gene [24].
Although there are differences among individuals, our analysis of the phylum composition of gut microbiota in each age group showed a significant shift in the relative abundance of Actinobacteria in infants from before to after weaning. The compositional pattern of gut microbiota during childhood has been thought to impact health later in life [11, 25], but children older than 2 years have not been sufficiently investigated. Our data show that some genera belonging to Bacteroides, Lachnospiraceae and Bifidobacterium CAGs and the alpha diversity of gut microbiota continued to change sequentially with age in subjects younger than twenty, reflecting the human gut microbiota maturation process. However, children younger than 20 years fell into both the infant and adult clusters, regardless of their age, when clustered based on the abundance of genus-like groups, thus illustrating the individual differences in the gut microbiota maturation process.
In the present study, the Wiggum plot showed negative relationships between the relative abundance of Enterobacteriaceae, which creates an greater endotoxin challenge for the weakened intestinal barrier and thus results in increased stimulation of the inflammatory response [26], and the abundance of Blautia and Lachnospiraceae;g (Fig. 4), which belonged to the adult-associated CAG (Lachnospiraceae;g CAG). Other butyrate-producing bacteria, such as Coprococcus, Roseburia and Faecalibacterium, were also clustered in the same CAG. Furusawa et al. reported that microbial-derived butyrate regulates Treg cell differentiation in vitro and in vivo [27]. Given the age-related reduction in the abundance of the genus Bifidobacterium, which down-regulates pro-inflammatory responses in the gut epithelium [28–30], our results suggested that the aging-related dysbiosis in elderly subjects may be a contributing factor to inflammatory responses that occur with advancing age.
We performed in vitro assays to investigate bacterial interactions, including those between Enterobacteriaceae and Blautia. These results were in accordance with the Wiggum plot results. However, these results might have some biases from the different of environmental condition between in human gut and in vitro assay. In addition, it is uncertain that all bacterial interactions are consistent with the relationships in the Wiggum plot, because genera with positive relationships in the Wiggum plot might grow well under the same environmental conditions without a mutualistic relationship. In contrast, some genera combinations have been reported to exhibit mutualistic relationships,although the absolute values of the correlation coefficients were below 0.3 (no visible relationship in the Wiggum plot). Pande et al. revealed that Acinetobacter baylyi and Escherichia coli reciprocally exchange essential amino acids [31]. It has also been reported that Bifidobacterium populations can be stimulated efficiently with a concomitant decrease in Enterobacteriaceae [32, 33]. Acetate, one of the main fermentation products of Bifidobacterium, was reported to promote the growth of butyrate-producing bacteria and the in vitro production of butyrate [34, 35]. Furthermore, Bifidobacterium longum has been reported to alter gut luminal metabolism via interactions with Bacteroides caccae and Eubacterium rectale [36]. Considering these reports, the correlations between gut microbiota members might be more complicated than shown in the Wiggum plot. Our computational analysis results must be interpreted cautiously because they are based on a limited data set. An advanced culture method is needed to clarify the relationships among gut microbiota.
A wide diversity of microorganisms is needed to utilize the many nutrients in adult diets [37]. In addition, a low gut microbiota diversity has been associated with an increasing number of conditions, such as autism [38], autoimmune disease [39] and obesity [40]. Maintaining sufficient bacterial richness and diversity is important for providing gut microbiota with functional redundancy, adaptability and thus systematic robustness against environmental challenges [41]. In this study, we observed an increase in gut microbiota diversity with aging until the centenarian stage. Claesson et al. reported that the alpha diversity of the gut microbiota in community dwellers was significantly higher than that of people in long-stay care [42]. Therefore, in the present study, the observed increase in microbiota diversity with age was likely due to the inclusion of community-dwelling elderly participants. In contrast, centenarians in a Chinese longevous village population had a more diverse gut microbiota than did younger elderly aged 85–99 years [43]. It is uncertain why the trend in diversity differed between Chinese and Japanese centenarians.
Biagi et al. [44] showed a significantly compromised gut microbiota in centenarians but not in elderly subjects aged approximately 70 years compared with a group of younger adults. The authors therefore suggested that a healthy gut microbiota community might be affected by aging-related physiological and behavioral changes that occur after 70 years of age, which was considered the threshold age for defining an individual as elderly. In agreement with this hypothesis, our clustering data suggested that most subjects over 70 years of age comprised two elderly-type clusters (Additional file 8). Nevertheless, the distribution of subjects was not clear from the classification based on subject age. This result suggested that other factors besides age contribute to the composition of gut microbiota communities.
Our study did not allow us to determine why gut microbiota changed with age because we lacked lifestyle and dietary habit information on the subjects. To begin to address this question, we performed PICRUSt analysis to predict the relative abundance of transporter genes because we hypothesized that bacterial transporters that incorporate nutrients from the gut environment would differ if dietary habit was a predominant contributor to gut microbiota composition. Clustering based on the relative abundance of the predicted transporters showed that subjects were divided into two age-related groups, the adult-enriched and infant/elderly-enriched clusters, implying that nutrients in the gut evoke the different gut microbiota compositions between adult and infant/elderly subjects. Not all transporters can explain the relationship with the dietary habits of subjects in each segmented age group, but, for example, the increase in the relative abundance of predicted xylose transporters in subjects after weaning seems to reflect the change from mother’s milk to an omnivorous diet. Interestingly, all the drug transporters based on KEGG Orthology groups were found in the infant/elderly-enriched cluster, perhaps due to the frequent antibiotic treatment of infants and the elderly compared with adults.
In addition, elderly people are known to have decreased intestinal function relative to younger people, which affects digestion, nutrient absorption and immune activity [45, 46] and may also impact the microbiota composition [47, 48]. We found that certain oral bacteria, such as Porphyromonas, Treponema, Fusobacterium and Pseudoramibacter, which have difficulty reaching the intestinal tract due to barriers such as gastric juice and bile acid, were enriched in the elderly-associated CAGs (Additional file 6). Therefore, the decline in gastrointestinal tract functionality in the elderly may also lead to significant changes in gut microbiota.