*3.5. Ordination Analysis*

The t-SNE plot allows embedding high-dimensional data into a two-dimensional map on the basis of each data point. Each species was categorized according to its prevalence (Figure 4A) or genus (Figure 4B). Through this graphical observation, highly prevalent species were conglomerated, and low-prevalence species were separated into distinct groups. Species were not separated on the basis of their genus. These results indicate that, on the species level, oral bacteria formed groups, and their coexistence was not regulated by their genus.

(**A**)

**Figure 4.** *Cont.*

**Figure 4.** Ordination analysis as a function of t-distributed stochastic neighbor embedding (t-SNE) plot categorized by (**A**) prevalence and (**B**) genus.

#### **4. Discussion**

In this study, we investigated 85 healthy people at the age of 90, focusing on analyzing their core oral microbiome. On the basis of the pyrosequencing analysis, 389,927 valid reads were obtained, and 1216 species were detected. Thirteen species were detected in all 85 subjects. Among them, seven belonged to the *Streptococcus* genus. These species represent potential candidates for the core human oral microbiome.

In this study, the most abundant phylum was Firmicutes, whereas *Streptococcus* constituted 45% of the genera present. The microbiome structure on the phylum level was categorized as follows: *Firmicutes* (62.7%), *Bacteroides* (9.6%), *Proteobacteria* (3.8%), *Fusobacteria* (2.7%), and *Actinobacteria* (20.3%). A previous study investigating subjects aged 60 years and older in China showed that the abundant phyla were *Firmicutes* (29.6%), *Bacteroides* (22.4%), Proteobacteria (20.4%), Fusobacteria (16.2%), and Actinobacteria (7.6%) [43]. Another study showed that the core oral microbiome predominantly comprises Firmicutes, followed by Proteobacteria and Bacteroides [44]. *Firmicutes*, *Proteobacteria*, *Bacteroides*, *Fusobacteria*, and *Actinobacteria* constitute more than 98% of the oral microbiome [45]. The oral-microbiome structure may change as a function of growth [46], food [47,48], oral diseases [49–51], or infection [51,52]. These results are consistent with the results of this study.

Few reports presented the core microbiome on the genus or species level. On the genus level, *Neisseria* (12.5%), *Leptotrichia* (11.1%), *Streptococcus* (10.7%), *Prevotella* (7.0%), *Veillonella* (6.9%), *Fusobacterium* (5.4%), *Capnocytophaga* (4.2%), *Prevotella* (4.1%), *Corynebacterium* (2.6%), *Saccharibacteria* (2.6%), *Actinomyces* (2.6%), *Haemophilus* (2.3%), and *Porphyromonas* (2.2%) were most prevalent [43]. Abundant genera according to another study were *Streptococcus* (26.1%), *Veillonella* (21.9%), *Neisseria* (16.9%), *Haemophilus* (10.7%), *Actinomycetes* (2.6%), *Rothia* (3.1%), and *Oribacterium* (1.7%) [53]. On the species level, a study carried out in Japan identified *Streptococcus salivarius* (9.5%), *Prevotella melaninogenica* (9.2%), *Rothia mucilaginosa* (8.8%), *Veillonella atypica* (6.0%), and *Neisseria flavescens* (5.8%) as species exhibiting > 5% abundance on the tongue surface [34]. When compared with our results, shown in Table 1, four of these species were detected in all our subjects, except for

*Neisseria flavescens*. *Neisseria* sp. is often detected in samples obtained from the dorsal surface of the tongue [54]. On the phylum level, the core microbiome observed in our study coincided with that in the literature. However, on the genus or species level, predominant bacteria varied across studies.

There are environmental and cultural differences, such as food consumption, that affect microbiome structure [49,55–57]; in our study, the proportions of Firmicutes and *Streptococcus* were higher than those found in other studies.

*Streptococcus* spp. are abundant in human milk, and they play an important role in the establishment of the oral-microbiome structure for breastfed infants [58,59]. Some *Streptococcus* spp. act as probiotic bacteria [60,61]. In contrast to these beneficial effects of *Streptococcus* for the human body, pathogenic *Streptococcus sinensis*, which is responsible for bacteremia [62] and endocarditis [63], and *Streptococcus pneumoniae* [64] were detected in all samples.

Two species of *Veillonella*, which are also classified as Firmicutes, were detected in all samples. *Veillonella* is known to be prevalently detected at various sites in the oral cavity, such as dental plaque, saliva [65], and oral mucosa [66]. *Veillonella parvula* is associated with the development of dental caries [67], endodontic infections [68], and periodontitis [69]. The most abundant species in our study was *Streptococcus sinensis*, followed by *Streptococcus pneumoniae*.

In this study, *Fusobacterium nucleatum* was detected in 94.1% of subjects. A previous study showed that Fusobacteria represent a predominant taxon in the oral microbiome [69]. Fusobacteria mediate the coaggregation of nonaggregating microbiota, and they are a structural element of plaque in both healthy and disease conditions [70]. They may contribute to the diversity of the oral microbiome.

A pioneering study of the oral microbiome using pyrosequencing suggested that the concept of a healthy core microbiome was supported by abundant oral taxa found in the oral cavity of healthy individuals [66,71]. Phylogenetic trees of 118 of the most predominant taxa identified at several sampling sites in the oral cavity were presented [71]. Among the 13 species detected in all subjects, five species were not included in this phylogenetic tree. These 13 species included pathogenic bacteria leading to human diseases. These bacteria were also detected in healthy older persons; however, they are not proposed as candidates for a healthy core oral microbiome.

It has been suggested that periodontal disease can be a major risk factor for some systemic diseases [72–77]. Recent advances in research on oral and general health have shown that there are protective host factors for periodontal-related systemic diseases [78–80]. The limitation of this study is its cross-sectional study. Further study is necessary to investigate the oral microbiome that can be the risk for mortality in combination with these host factors for older people.

In this study, we aimed to identify the core oral microbiome in healthy older people. However, on the basis of prevalence and abundance, pathogenic bacteria were also included. The human oral microbiome plays a crucial role in diseases and human health. Simple descriptive analysis as a function of prevalence and abundance may not be enough to define a healthy core microbiome. The effect of bacteria should be considered when defining a healthy human oral microbiome.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2076-3417/10/18/6450/s1. Figure S1. Heatmap of 1216 species detected in this study; Figure S2. Rarefaction curves of the 85 subjects; Figure S3. Composition bar plots for each subject at the phylum level; Figure S4. Core line plot; Table S1. Alpha diversity indices.

**Author Contributions:** Conceptualization, Y.N., A.Y., and N.H.; methodology, Y.N. and A.Y.; formal analysis, Y.N.; investigation, Y.N., E.K., and M.O.; resources, N.K.; writing—original-draft preparation, Y.N.; writing—review and editing, Y.N.; visualization, Y.N.; project administration, K.N.; funding acquisition, Y.N. and N.H. All authors read and agreed to the published version of the manuscript.

**Funding:** This study was supported by the JSPS KAKENHI (grant numbers 17K12030 and 20K10303) and the SECOM Science and Technology Foundation. The funders played no role in the design of the study, in data collection, in the analysis and interpretation of the results, or in the writing of the manuscript.

**Conflicts of Interest:** The authors declare no conflict of interest.
