The Microbiome in Populations with a Low and High Prevalence of Caries

Abstract

The oral microbiota was compared between Romanian adolescents with a high prevalence of caries and no dental care and Swedish caries-active and caries-free adolescents in caries prevention programs and with a low prevalence of caries. Biofilm samples were analyzed by FLX+ pyrosequencing of the V1 to V4 hypervariable regions of the 16S rRNA gene and polymerase chain reaction (PCR)/quantitative PCR (qPCR) for Streptococcus mutans and Streptococcus sobrinus. Sequences obtained blasted to 9 phyla, 66 genera, and 401 human oral taxa (HOT) in the 16S rRNA Human Oral Microbiome Database, of which 295 were represented by ≥20 sequences. The Romanian adolescents had more sequences in Firmicutes and fewer in Actinobacteria phyla and more sequences in the genera Bacteroidetes [G-3], Porphyromonas, Abiotrophia, Filifactor, Peptostreptococcaceae [11][G-4], Pseudoramibacter, Streptococcus, and Neisseria and fewer in Actinomyces, Selenomonas, Veillonella, Campylobacter, and TM7 [G-1] than the Swedish groups. Multivariate modeling employing HOT, S. sobrinus and S. mutans (PCR/qPCR), and sugar snacks separated Romanian from Swedish adolescents. The Romanian adolescents' microbiota was characterized by a panel of streptococci, including S. mutans, S. sobrinus, and Streptococcus australis, and Alloprevotella, Leptotrichia, Neisseria, Porphyromonas, and Prevotella. The Swedish adolescents were characterized by sweet snacks, and those with caries activity were also characterized by Prevotella, Actinomyces, and Capnocytophaga species and those free of caries by Actinomyces, Prevotella, Selenomonas, Streptococcus, and Mycoplasma. Eight species including Streptococcus mitis and Streptococcus species HOT070 were prevalent in Romanian and Swedish caries-active subjects but not caries-free subjects. In conclusion, S. mutans and S. sobrinus correlated with Romanian adolescents with caries and with limited access to dental care, whereas S. mutans and S. sobrinus were detected infrequently in Swedish adolescents in dental care programs. Swedish caries-active adolescents were typically colonized by Actinomyces, Selenomonas, Prevotella, and Capnocytophaga. Hence, the role of mutans streptococci as a primary caries pathogen appears less pronounced in populations with prevention programs compared to populations lacking caries treatment and prevention strategies.

Keywords: Romania; Sweden; adolescents; mutans streptococci; oral microbiota; pyrosequencing.

Figure 1.

Rarefaction curves. (A) Species richness, the number of types of sequences in a sample, in individual Romanian adolescents (blue lines) with a high prevalence of caries (n = 14), Swedish adolescents with caries activity (yellow lines, n = 12), and caries-free Swedish adolescents (red lines, n = 11). (B) Mean number of types of sequences in the 3 groups.

Figure 2.

Caries group separation by principal coordinates analysis (PCoA) and principal component analysis (PCA). (A) PCoA plot displaying the distribution among the 37 samples based on USEARCH clustering of the 942,774 sequences at 98.5% similarity against the Human Oral Microbiome Database (β diversity). (B) PCA modeling displaying subject clustering using the 295 human oral taxa (HOT) with ≥20 cluster sequences, Streptococcus sobrinus and Streptococcus mutans by polymerase chain reaction (PCR), and intake of sweet snacks. Blue dots for Romanian adolescents with a high prevalence of caries (n = 14), yellow dots for Swedish adolescents with caries activity (n = 12), and red dots for caries-free Swedish adolescents (n = 11). The scores t1 and t2 are the new partial least squares (PLS) regression–created variables summarizing the x-variables. The oval circle illustrates the tolerance ellipse based on the Hotelling T² distribution; any observation located outside of the ellipse would be an outlier.

Figure 3.

Partial least squares (PLS) regression of the microbiota associated with the caries groups. The PLS model used caries grouping as the y-variable and the 295 human oral taxa (HOT) with ≥20 cluster sequences, Streptococcus sobrinus and Streptococcus mutans by polymerase chain reaction (PCR)/quantitative PCR (qPCR), and intake of sweet snacks (3 variables) as the independent variable block. (A) The PLS scatter loading plot illustrating the clustering of Romanian adolescents with a high prevalence of caries (blue dots, n = 14) versus Swedish caries-active (yellow dots, n = 12) and caries-free (red dots, n = 11) adolescents. The scores t1 and t2 are the new PLS-created variables summarizing the x-variables. The oval circle illustrates the tolerance ellipse based on the Hotelling T² distribution; any observation located outside of the ellipse would be an outlier. (B) The PLS loading plot illustrating variables associated with each of the 3 caries groups. w*c[1] and w*c[2] represent loading for the 2 first components. The HOT with Variable Importance in Projection (VIP) values are shown in Appendix Table 4.