Dental caries is one of the most prevalent childhood diseases in the United States and worldwide (Dye et al., 2015; Kassebaum et al., 2017). Although considered a preventable disease, the persistence of high caries necessitates the improvement in diagnostics and therapeutic interventions for young and vulnerable populations. A multifactorial etiology of dental caries comprises of a complex interaction of acid-producing bacteria, fermentable carbohydrates, and host factors (Selwitz et al., 2007).

From the microbiological perspective, Streptococcus mutans has been considered a key pathogen in the initiation and progression of caries (Tanzer et al., 2001). Hence, S. mutans levels have been used as part of caries risk assessment (Guo and Shi, 2013; Edelstein et al., 2016). However, contradicting reports exist in the literature regarding the correlation between S. mutans counts and caries experience. Previous studies established a causal relationship between S. mutans levels in plaque or saliva and caries experience in children (Alaluusua and Renkonen, 1983; Loesche, 1986; Damle et al., 2016; Edelstein et al., 2016), while others reported distinct populations whose caries status is not correlated with S. mutans levels have been observed (Carlsson et al., 1985; Matee et al., 1992, 1993). Furthermore, studies in subjects with rampant caries did not reveal detectable S. mutans levels (Aas et al., 2008). These seemingly contradicting results raised awareness of previously untapped interesting caries discordant clinical population: caries-free children with high S. mutans levels, as well as those with low S. mutans levels who are severely affected by dental caries.

One possible explanation for the existence of discordant populations could be the ecological plaque hypothesis (Marsh, 1994), which echoes the fact that S. mutans may not be the sole etiological factor. In this scenario, caries would result from a disruption of the homeostasis of the resident microflora driven by changes in local environmental conditions (Wade, 2013). Emerging evidence associated additional bacterial species within the salivary microbiome with caries experience in children. Besides S. mutans, other acid-producing microorganisms such as certain species of the genera Veillonella, Scardovia, and Lactobacillus have also been identified in caries-affected children populations (Guo and Shi, 2013; Jiang et al., 2016; Xu et al., 2018). Additionally, there are studies linking Actinomyces and Bifidobacterium genera with cavitated caries lesions (Brailsford et al., 1999; Aas et al., 2008; Mantzourani et al., 2009; Guo and Shi, 2013). Despite a considerable body of evidence, the microbiological ecology of the caries etiology is not fully understood, limiting our ability to explain the existence of the caries discordant population. Moreover, a comprehensive investigation of the oral microbiome in this population is scarce.

In this study, we investigated the caries concordant-discordant clinical populations for their salivary S. mutans levels and the respective microbiome. These unique clinical populations may hold the key to a better understanding of the ecological relationship between S. mutans and other microbial species and the impact on dental caries in children.

Dental caries is a multifactorial oral disease affecting children worldwide, for which the complete pathogenesis is not yet fully understood. The existence of caries discordant populations, in which S. mutans levels do not correlate with the observed caries experience, poses a compelling challenge for caries management. In this study, we use saliva as a source of S. mutans and oral microbiome isolation because it reflects the composition of the oral microbiome, mirrors the complex microbial ecosystems of the oral cavity (Belstrøm, 2020), and is an excellent non-invasive diagnostic biofluid. During the initial phase of random participant recruitment for this study, we observed that only 60% of the clinical population could be categorized into caries concordant groups (HSHC and LSLC). In contrast, a substantial portion of the population (40%) displayed caries discordant phenotypes (HSLC and LSHC), implying that the presence and quantity of S. mutans may not be the sole determinant of disease risk. In this study, the possible role of S. mutans levels and/or other microorganisms on caries-associated ecology was explored in concordant and discordant populations.

Our results indicate that S. mutans levels may play a role in shaping the community as significant differences in the groups with high and low counts of this cariogenic pathogen were observed. In the HS groups, apart from S. mutans several other genera and species were significantly higher than in the LS groups (Table 1). The genera with significant differences included the acid producers, Bifidobacterium (Henne et al., 2015) and Atopobium (Kolenbrander et al., 2009), the lactic acid consumer Veillonella (Mashima and Nakazawa, 2014), and Megasphaera, which is closely related to Veillonella (Nallabelli et al., 2016). Previously, these genera have been identified in close association with high S. mutans levels (Aas et al., 2008) that they are thought to either assist in the production or metabolism of lactic acid. Of note Bifidobacterium dentium was found to be present in 30.8% of caries active individual but was not present at all in caries-free individual (Henne et al., 2015), suggesting that this species should be considered alongside S. mutans as another aciduric taxa contributing to dental caries. Although several studies have recognized the contribution of the genus Veillonella in dental caries (Becker et al., 2002; Aas et al., 2008; Gross et al., 2012; Jiang et al., 2016; Agnello et al., 2017; Xu et al., 2018; Hurley et al., 2019), differential roles may exist at species level (Do et al., 2015). In this study, the relative abundance of two different Veillonella species were significantly different between groups. While high levels of V. dispar were detected in the HS groups, V. rogosae was significantly higher in LS groups. Similar to our findings, previous studies have reported an increased abundance of V. dispar in caries-active children (Arif et al., 2008; Hurley et al., 2019) and V. rogosae in caries-free children (Arif et al., 2008). There is little doubt that the ability of Veillonella to consume lactate can contribute to caries development. Interestingly, M. micronuciformis, a low abundance species which was significantly increased in the HS groups were previously reported in association with caries (Kalpana et al., 2020).

Other pertinent bacterial species were observed in high abundance in the HS groups. For Streptococcus species, besides S. mutans, S. parasanguinis, and S. intermedius were significantly higher in the HS groups. However, their association with health or disease is controversial as elevated levels of these species were reported in both caries-active and caries-free children (Becker et al., 2002; Corby et al., 2005; Aas et al., 2008). In addition, significantly higher levels of caries-associated species such as Actinomyces sp. and Prevotella spp. were also observed in the HS groups, both of which have been reported in both the saliva and plaque of children with severe caries lesions (Becker et al., 2002; Corby et al., 2005; Jiang et al., 2013, 2016; Ma et al., 2015; Zhang et al., 2015; Zhu et al., 2018; Kalpana et al., 2020). Particularly, a higher abundance of Prevotella spp. was correlated with the group characterized by elevated S. mutans and caries experience levels. A previous study investigating the microbiome of dentin carious lesions reported similar results (Hurley et al., 2019).

Focusing on the LS groups, no significant trends were observed at the genus level. Besides the health-related V. rogosae mentioned above, at the species level, Alloprevotella spp. and Haemophilus sp. were also significantly increased in the LSLC group. While our findings are consistent with previous reports (Hurley et al., 2019), other studies suggested an association of these species with caries-affected children (Xu et al., 2018; Zhu et al., 2018), yet another group published no distinct association with either one (Jiang et al., 2016).

The microbial community diversity in both the caries concordant and discordant populations were compared based on clinical groups (Figures 4A,D), S. mutans levels (Figures 4B,E), and caries experience (Figures 4C,F). Overall, when comparing between clinical groups or caries experience, there were no significant differences in species abundance, richness, diversity, and no particular clustering pattern was observed. Previous studies that applied the same alpha-diversity index only found a correlation with caries experience when the disease severity progressed into dentin (Bong-Soo et al., 2018; Hurley et al., 2019). Findings from recent reports were consistent with our results that revealed no clustering of microbial communities based on caries status (Jiang et al., 2016; Zhu et al., 2018; Hurley et al., 2019). Further community diversity analysis based on S. mutans levels showed no significant differences in alpha-diversity (Figure 4C). However, the most striking result to emerge from our data was the microbial communities significantly cluster based on S. mutans (HS vs. LS) (Figure 4E). The saliva from children with high S. mutans counts exhibited similar microbial community structure independent of their carious status, while in individuals with low S. mutans levels, the communities were more diverse. This result further supports the influence of S. mutans in shaping the composition of the microbial communities.

Intriguingly, S. mutans accounts for less than 1% of the community composition yet plays a crucial role in shaping the entire community by influencing the abundance of other species and shifting the oral microbiome toward disease. However, our findings highlight other disease-associated species such as V. dispar, S. parasanguinis, S. intermedius, and Prevotella spp., which were significantly increased in HS groups and may contribute, in combination with S. mutans, to the caries progression. In contrast, other health-associated species exhibited higher abundance in the LS groups, such as V. rogosae, Haemophilus sp., and Alloprevotella spp., but their contribution to the balance of health and disease is unknown.

This study is without limitations. The sample number per group could be increased to improve statistical power of the study. Although 16S rRNA gene sequencing can provide adequate phylogenetic information to identify the bacteria at the species level, for some taxa such as different streptococcal species, the identification is limited by the discriminatory power of the hypervariable regions and the lack of a consensus protocol. Furthermore, diet, especially the frequent intake of high carbohydrate beverages and food, is known to contribute to a sharp decrease in pH that drives a proliferation of cariogenic bacteria and microbiome dysbiosis leading to dental caries. Additional mechanistic studies should be undertaken to fully grasp the interactions between these cariogenic species and their respective contribution to caries development. The presence of different S. mutans strains may also account for the discordant population’s phenotypes. Previous studies demonstrated that different strains of the same species might have different contributions to caries in children (Al-Hebshi et al., 2019). Similarly, clinical isolates of S. mutans showed distinct phenotypic traits and thus differences in caries severity (Valdez et al., 2017). Therefore, future work should examine other potential factors contributing to the existence of the caries discordant population such as dentition stage, specific S. mutans strains, diet, and host susceptibility.

This research has raised an important consideration of the additional species as health/disease microbial markers in conjunction with S. mutans. Our present unique study substantiates the understanding of the caries discordant populations microbiome and provides the first insight into how different microbes may contribute to the balance between health and disease in these populations. We believe that our research will serve as a base for future studies to unravel additional microbial targets that may help improve diagnosis and caries management of the caries discordant population.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: ENA; PRJEB48455.

The studies involving human participants were reviewed and approved by the University of California, Los Angeles Institutional Review Board. Written informed consent to participate in this study was provided by the participants’ legal guardian/next of kin.

MD contributed to conception, data acquisition, analysis and interpretation, and drafted and critically revised the manuscript. MA and LC contributed to data acquisition, analysis and interpretation, and drafted and critically revised the manuscript. BS drafted and critically revised the manuscript. XH, WS, DW, and RL contributed to conception, data interpretation, and critically revised the manuscript. NT contributed to conception, design, data acquisition, analysis, and interpretation, and drafted and critically revised the manuscript. All authors gave final approval and agreed to be accountable for all aspects of the work.

WS was a former Chief Scientific Officer of C3J Therapeutics, a biotech start-up company, where Delta Dental of Michigan is an investor. DW is a consultant to Wrigley/Mars, Colgate-Palmolive, and has equity in Liquid Diagnostics LLC. This does not alter our adherence to Frontiers Microbiology policies on sharing data and materials. This study received funding from Delta Dental. The funder was not involved in the study design, collection, analysis, interpretation of data, the writing of this article or the decision to submit it for publication. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.