香京julia种子在线播放

The microorganisms that live in and on the human body are estimated to outnumber human cells by at least an order of magnitude. Human beings are regarded as “superorganisms” whose metabolism represents an amalgamation of microbial and human attributes (Gill et al., 2006). Recent findings suggest that the microbial community is indispensable for maintaining human health, yet is also capable of eliciting disease. As one of the largest and most complex microbial habitats, the oral cavity harbors hundreds of bacteria that play vital roles in maintaining oral homeostasis and influencing the development of oral diseases, mainly dental caries and periodontal disease (Chen and Jiang, 2014). In addition, oral pathogens may be important players in systemic diseases, such as diabetes (Moodley et al., 2013), infective endocarditis (Beck and Offenbacher, 2005), bacterial pneumonia (Awano et al., 2008), adverse pregnancy outcomes (Han, 2011), inflammatory bowel disease (Ismail et al., 2012), pancreatic cancer (Farrell et al., 2012), and colon cancer (Castellarin et al., 2012). Despite its substantial impact on overall human health, the complexities of the oral bacterial community and how its composition changes when shifting between the healthy and diseased states are far from being fully understood.

Dental caries, otherwise known as tooth decay, is one of the most prevalent chronic infectious diseases, affecting 80–90% of the worldwide population (Selwitz et al., 2007). Individuals are susceptible to this disease throughout their lifetime, and it is the primary cause of tooth loss, oral infection, and pain (Featherstone, 2000). Dental caries is a multifactorial disease that initiates from bacterial shifts within the supragingival dental plaque, a complex biofilm formed on the tooth surfaces, by the aggregation of a group of bacteria in a well-organized manner (Marsh, 2006). A long history of debate is associated with cariogenic theories of dental plaques, which have mainly focused on four hypotheses: the non-specific plaque hypothesis (Schultz-Haudt et al., 1954), the specific plaque hypothesis (Loesche, 1979), the ecological plaque hypothesis (Marsh, 1994), and, recently, the polymicrobial synergy and dysbiosis hypothesis (Hajishengallis and Lamont, 2012).

Traditional cultivation methods have been used to isolate and characterize over 280 oral bacterial species (Aas et al., 2005). Streptococcus mutans was the first species isolated from caries lesions and has been considered the main cariogenic bacteria for decades (Clarke, 1924; Loesche et al., 1975). However, a large proportion of oral bacteria cannot be cultivated in vitro, which impedes the thorough and in-depth understanding of the natural microbial community residing in dental plaques (Paster et al., 2001). The use of molecular biological techniques, such as denaturing gradient gel electrophoresis (Muyzer et al., 1993), the quantitative real-time polymerase chain reaction (Ciric et al., 2010), microarray chips (Preza et al., 2009), checkerboard hybridization (Wall-Manning et al., 2002), and cloning sequencing of DNA (Kanasi et al., 2010), enables the identification and classification of uncultured oral bacteria. However, these approaches have remarkable biases and drawbacks that do not facilitate the comprehensive study of bacterial diversity, as only the predominant bacteria can be identified while many low-abundance species cannot be detected (Schulze-Schweifing et al., 2014). In recent years, next-generation sequencing technologies have dramatically improved sequencing capabilities (Metzker, 2010) and have been successfully applied for oral microbial analysis in a high-throughput manner (Zaura, 2012; Nyvad et al., 2013; Wade, 2013; Camelo-Castillo et al., 2015), including five major platforms for microbiome studies (namely the 454, Illumina, SOLiD, Ion Torrent, and PacBio platforms) (Hodkinson and Grice, 2015). To date, the overall diversity and structure of supragingival plaque microbiota in adult dental health and caries have not been completely elucidated.

Thus, we used amplicon pyrosequencing of the 16S rDNA V1–V3 hypervariable regions to comprehensively define the bacterial composition, abundance, and structure of supragingival plaques in a relatively large number of adult subjects with dental health and caries. Unlike previous studies, the cases of dental caries were grouped into different categories of severity according to the decayed missing filled tooth (DMFT) index. Furthermore, we aimed to identify bacterial shifts related with dental health and caries, and to investigate the presence of a core plaque microbiome. Our findings expand the understanding of bacterial ecology in dental caries and provide directions for prevention and treatment strategies of dental caries.

A comprehensive and thorough investigation of the bacterial diversity of plaque microbiota is essential for understanding their etiologies and for developing effective prevention and treatment strategies of dental caries. The introduction of high-throughput pyrosequencing has provided new insights into the compositions and structures of microbial communities. We used 454 FLX+ pyrosequencing to explore the bacterial diversity and community structure of 160 supragingival plaque samples by sequencing the 16S rDNA hypervariable V1–V3 region in samples from 131 adults with dental caries and 29 volunteer with good oral health. Different from previous studies using basic molecular techniques such as DGGE, qPCR and microarray chips (Muyzer et al., 1993; Preza et al., 2009), we not only identified the predominant taxa but also a very large number of low-abundance or rare taxa such as Ochrobactrum, Centipeda, Lachnoanaerobaculum, GN02_[G-1], Moryella, Bacteroides, Megasphaera, Rhizobium, Filifactor, and Dietzia. The V1–V3 region has been used in several studies (Kumar et al., 2011; Xu et al., 2014; Yoo et al., 2016) because it can provide greater phylogenetic resolution and better analysis of diversity and richness. In addition, this region has more sequences deposited into databases, which greatly facilitates diversity analysis (Kim et al., 2011).

We obtained 2,261,700 high-quality sequences with an average of 14,136 sequences per sample (Table S3), which was much higher than reported in previous studies (Keijser et al., 2008; Ling et al., 2010). For pyrosequencing studies of 16S rRNA, a depth of coverage of approximately 1000 sequences per sample provided a good balance between sample number and sampling depth, and to enabled detection of species at 1% abundance with reasonable accuracy (Hamady and Knight, 2009). Thus, the sequencing depth in our study (14,136 sequences per sample) was reasonable and large enough to enable detection of the vast majority of bacterial species in plaque samples. Furthermore, a Good's coverage estimator of >95% also suggested that our sequencing depth was sufficient to reflect the full bacterial diversity of plaque microbiota (Table 1), indicating that some extremely low abundance of rare species could be detected and might be related to dental caries (Schloss et al., 2011).

At 97% identity, 10,365 OTUs belonging to 453 species, 122 genera, 66 families, 34 orders, 21 classes, and 12 phyla were obtained after filtering out the low-credibility OTUs and BLAST searching against the HOMD database for the taxonomic assignment. In agreement with the results of the present study, many investigators have used 97% identity to classify microorganisms at the species level (Acinas et al., 2004; Yarza et al., 2008; De Gannes et al., 2013; Lindeque et al., 2013). Other studies have explored the classification of oral microbiota. For instance, Pushalkar et al. (2011) identified 860 species, 52 genera, and 8 phyla; Ling et al. (2010) identified 203 genera and 10 phyla; Keijser et al. (2008) identified 318 genera and 22 phyla; and Siqueira and Rôças (2009) identified 460 species, 100 genera, and 9 phyla. The differences in microbial taxonomy observed among these studies may be due to differences related to the selection process for subjects, the geographic location, oral sample collection sites, target sequencing regions of 16S rDNA, databases, and the data-analysis methods.

Compared with Xu's study of supragingival plaque microbiota diversity in children younger than 30 months ranging from 125 to 255 OTUs per sample (Xu et al., 2014), we observed a high diversity in adults aging 20–50 years with dental health or caries ranging from 339 to 2559 OTUs per sample. This suggested that the diversity of oral microbiota increased with age and developmental status, such as primary and permanent tooth growth (Papaioannou et al., 2009). Tsai et al recently identified an average of 774 classified phylotypes per sample and a total of 6 phyla across all subgingival samples in individuals with severe chronic periodontitis (Tsai et al., 2016). In contrast, our study found high diversity with an average of 1156 OTUs per sample and 12 phyla across all supragingival samples in adults with dental caries. We have confirmed that sites in the oral cavity had distinct bacterial diversities that changed with disease status, as reported previously (Paster et al., 2006).

By comparing the alpha diversity indexes among groups using Student's t-test, we found that the ACE richness index in the NC group was significantly higher than in the HC group (p = 0.01), indicating that the bacterial diversity of healthy dental plaques was higher than that in caries. This finding was in agreement with results from previous studies (Li et al., 2005; Preza et al., 2008). Considered from a bacterial point of view, a more diverse community represents a more healthy and stable ecosystem. The carbohydrate-driven lowering of the pH from lactate produced by acid-producing species could lead to suppression of acid-sensitive species and overgrowth of acid-tolerant species, resulting in decreased bacterial diversity in supragingival plaques as caries progressed, as well as a decreasing number of species capable of surviving harsh conditions (Gross et al., 2010). Li et al. also reported that the species richness and Shannon diversity index were greater in caries-free children than in those with severe early childhood caries (Li et al., 2007). Our study provided additional support for the “ecological catastrophe” of dental caries occurring between commensal microbiota and the host, confirming a decline of bacterial diversity as the severity of caries increased (Marsh, 2003). Some bacteria in dental caries might become lost, inhibited, or replaced, although these possibilities require further study to establish their validity.

The existence of a “core microbiome” was first proposed by Turnbaugh et al. (2007) and referred to the organisms, genes, or functions shared by all or most individuals in a given human habitat, such as the oral cavity, nasal cavity, skin, and intestinal tract. One goal of the Human Microbiome Project was to determine whether an identifiable core microbiome exists among individuals. Most studies conducted to date have focused on the human intestinal tract and investigated the relationships between the core gut microbiome and diseases (e.g., obesity, diabetes, and metabolic syndrome) (Tschöp et al., 2009; Turnbaugh and Gordon, 2009; Neu et al., 2010; Tilg, 2010). Zaura et al. provided the first insight into the oral core microbiome using 454 pyrosequencing technology and found 387 overlapping OTUs in 3 healthy individuals (Zaura et al., 2009). In our study, 7522 overlapping OTUs were shared by the four groups in a Venn diagram (Figure 6A), defined as plaque core microbiome, regardless of whether it had high or low abundance. These widespread core microbiomes might play important roles in the stability and functions of the plaque microbiota. By PICRUSt analysis, we predicted bacterial functions mainly enrichment in membrane transport, carbohydrate metabolism, replication and repair, amino acid metabolism, translation, and energy metabolism. As suggested by Shade (Shade and Handelsman, 2012), beginning with a Venn diagram representing membership, layers of complexity are added by incorporating the composition, phylogeny, persistence, and connectivity to define core microbiomes. The appropriate definition of a core microbiome depends on the ecological question addressed. Our study mainly aimed to determine the presence of a core microbiome, so we typically defined a core as the group of members shared among the microbial community and represented it by the overlapping areas of circles in Venn diagrams. In a future study, we will apply multiple definitions to the core microbiome to enhance the ecological meaning.

Our results indicated that similar community structures of supragingival plaques were present in the four groups according to PCoA and hierarchical clustering analysis (Figure 4 and Figure S4). Samples from the healthy group tended to cluster together, whereas the caries microbiota appeared to be more variable, which was consistent with the results of Yang's study (Yang et al., 2012). The bacteria composing each group were relatively constant. It can be inferred that bacterial components of dental plaques were not markedly affected by the disease status or pathological factors. In spite of the presence of similar bacterial members, the abundances of some bacteria differed significantly among the groups. Twelve bacterial phyla were represented in all supragingival plaque samples. Six predominant phyla were Bacteroidetes, Actinobacteria, Proteobacteria, Firmicutes, Fusobacteria, and TM7, constituting 99% of the total microbiota (Figure 1). These dominant phyla were largely the same as those previously described for childhood dental caries using pyrosequencing (Jiang et al., 2013).

At the genus level, 122 different genera were identified in the plaque samples of dental health and caries. Seventeen genera had abundances higher than 1% (Figure 1), of which the genera Actinomyces, Streptococcus, Rothia, Selenomonas, and Veillonella have been detected in early caries studies (Munson et al., 2004; Tanner et al., 2011). The dental plaque biofilm was composed of complex microbial communities, instead of a simple superposition of bacteria. After the formation of acquired pellicles on the teeth, the free bacteria in the oral cavity adhered to pellicles in a definite sequence, thus governing the formation of mature dental plaques (Li et al., 2004). The initial colonizing bacteria was Streptococcus sanguis, followed by members of the Actinomyces, Neisseria, Veillonella, Capnocytophaga, Leptotrichia, and Haemophilus genera, collectively known as early colonizing bacteria. These bacteria can quickly adapt to new environments by regulating specific gene expression and providing new adhesion receptors for later colonizing bacteria, such as members of the Fusobacterium, Treponema, Tannerella, Prevotella, and Porphyromonas genera (Haffajee et al., 2008). With an increasing number of bacteria in dental plaques, the concentrations of signaling molecules increased, which can activate the expression of related genes, such as virulence factors and mucopolysaccharides. Certain bacteria gradually decrease in abundance, due to the antagonism of predominant bacteria. Consequently, the ecological balance of plaque bacterial microbiota may be broken, eventually leading to the occurrence and development of dental caries (Reading and Sperandio, 2006). We clearly observed that some bacteria in the supragingival plaques of dental caries showed a significantly reduced abundance, such as those in the Aggregatibacter, Cardiobacterium, Corynebacterium, Lachnoanaerobaculum, Porphyromonas, and Tannerella genera. Aggregatibacter, Porphyromonas, and Tannerella genera have been associated with dental health (Alcaraz et al., 2012; Jiang et al., 2014). However, the abundance of other bacteria significantly increased, such as Atopobium, Cryptobacterium, Lactobacillus, Mogibacterium, Ochrobactrum, Pseudomonas, Rhizobium, Alloprevotella, Bacteroides, Centipeda, Campylobacter, Megasphaera, and Mycoplasma; thus, they were regarded as potentially cariogenic bacteria. Some are well-recognized cariogenic bacteria (Lactobacillus) or have been identified at significantly higher levels in dental caries (Cryptobacterium, Atopobium) (Jiang et al., 2014; Obata et al., 2014). While others may be recognized as new cariogenic bacteria, such as Catonella, Centipeda, Fretibacterium, Rhizobium, Ochrobactrum, and Mogibacterium. Furthermore, we screened potential biomarkers using LEfSe analysis, including Desulfomicrobium, Mycoplasma, Corynebacterium, Clostridales, Gammaproteobacteria, Cardiobacterium, and Pasteurellaceae. Therefore, dental caries is a polymicrobial disease caused by various bacterial consortia that are present commensally in healthy individuals at lower levels.

According to our PLS-DA analysis, 27 key genera were identified that were responsible for differential distribution with VIP > 1 (Table 2). Two bacteria with a VIP > 2 (Cardiobacterium and Corynebacterium) were also identified as health-related genera. Cardiobacterium is a member of the group of HACEK bacteria (Haemophilus, Actinobacillus, Cardiobacterium, Eikenella, and Kingella), which are part of the normal oropharyngeal flora and have common characteristics that easily lead to endocardial infections (Hagiya et al., 2014). Although Cardiobacterium bacteria have been reported to relate with aortic valve infective endocarditis and ventricular septal defects (Vanerková et al., 2010; Choudhury et al., 2013), its relationship with oral health and disease has rarely been studied. The Cardiobacterium genus has only 2 species, namely C. hominis and Cardiobacterium valvarum. In our study, C. hominis was identified as health-related species that might play an important role in maintaining dental health. Moreover, Corynebacterium bacteria were considered as part of a candidate signature for dental health (Peterson et al., 2013). We found that Corynebacterium matruchotii was significantly more abundant in dental health than in caries and was identified as a health-related species, which was consistent with the results of Gross's study (Gross et al., 2010).

S. mutans was recognized as the main cariogenic bacteria because of its strong capacity for acid production and acid tolerance, as well as its capacity for exopolysaccharide production (Loesche, 1986). Thus, it was surprising in our study that S. mutans showed a low abundance in supragingival plaque (only 0.046% in all samples), and showed no significant difference among the groups. In Aas's study, S. mutans was not detected in 10% of dental caries using checkerboard hybridization (Aas et al., 2008). Furthermore, Aas et al. found that the proportion of S. mutans detected in cavitated dentin lesions was high, whereas the proportion in surfaces of intact enamel and white-spot lesions was very low, which agreed with our results. We speculate that the distribution of S. mutans in different oral ecological sites might be heterogeneous and regional. Determining the importance of S. mutans in caries development requires a comprehensive consideration of its virulence factors. Streptococcus sobrinus was also recognized as cariogenic bacterium. Nevertheless, we observed a low proportion of S. sobrinus in the plaque samples. Hirose et al. found that S. sobrinus in saliva of children could not only cause pit-fissure caries, but also smooth-surface caries, and suggested that S. sobrinus has stronger cariogenic ability than does S. mutans (Hirose et al., 1993). It has been demonstrated that S. sobrinus is significantly more acidogenic than S. mutans and is capable of sustained acid production at pH levels below 6.0 (de Soet et al., 1989; Madison et al., 1991). In our study, S. sobrinus was statistically more abundant in caries. Interestingly, we found that S. sobrinus was absent in the NC group and was thus regarded as a caries-specific species. Thus, S. sobrinus perhaps deserves more attention as cariogenic bacterium.

In summary, 454 pyrosequencing technology has greatly expanded our knowledge regarding the bacterial diversity and community structure of supragingival plaques in adult dental health and caries. We observed a vast bacterial diversity of supragingival plaques with 2,261,700 high-quality sequences and 10,365 OTUs at 97% identity, which belonged to 453 species, 122 genera, 66 families, 34 orders, 21 classes, and 12 phyla. Bacterial diversity in dental health was higher than in caries and gradually decreased with the severity of dental caries. Our results indicated the occurrence of similar community structures of supragingival plaques among groups. An identifiable core plaque microbiome was determined among the four groups with 7522 overlapping OTUs. Furthermore, we identified a list of caries-related and health-related bacteria, whose specific functions need further testing and verification. Given the polymicrobial and dysbiotic nature of dental caries, diagnostic and preventive strategies directed against specific bacteria would not be universally effective. Regarding preventive and therapeutic approaches, passive immunization strategies are predicted to be inefficient, and the anticipated effectiveness of active immunization strategies, such as caries vaccination, is also uncertain. We believe that the treatment of dental caries should involve balance restoration and that strategies for modulating interactions in the oral microbiota (i.e., molecular interactions, biological microbe-microbe and microbe-host interactions, and microbial ecological interactions), should be further developed.

Conceived and designed the experiments: JL and ZH. Performed the experiments: CX and SR. Analyzed the data: CX and ZH. Contributed reagents/materials/analysis tools: CX and SR. Manuscript preparation: CX. Manuscript revisions: JL, ZH, CX, and SR.