Summary:
Introduction: Endodontic infections have been linked to the commensal oral
microbiota, which colonize and proliferate in the root canal system as a
consequence of pulp necrosis secondary to caries, tooth trauma, defective
restorations or due to a failed endodontic treatment. A thorough understanding
of the microbial etiology and characteristics of endodontic infections is a
necessary step in developing effective intra-canal antimicrobial protocols.
Nevertheless, the exploration and identification of endodontic pathogens
remains one of the most challenging aspects in endodontic microbiology, with
the majority of bacteria still unknown or uncultivated.
The development and application of molecular biology methods has
facilitated the identification and linkage of specific bacterial species with
periradicular disease and thus have led to the discovery of novel endodontic
pathogens. Next-generation sequencing is now part of the toolbox available for
16S rRNA-based bacterial diversity analyses. The technology enables a large
number of reads in a single run, providing increased sampling depth compared to
other techniques and has the major advantage of enabling the detection of
low-abundant genera. So far, only eight studies have used this approach to
investigate different types of endodontic infections. From those, only two
investigations have examined the endodontic microbiome in teeth with failed
endodontic treatment.
Aim: The aim of the present study was to investigate the microbial composition
of symptomatic and asymptomatic primary and persistent infections in a Greek
population, by using 16S amplicon pyrosequencing.
Materials and Methods: 16S amplicon pyrosequencing of 48 bacterial samples was
conducted and sequencing data were analyzed using an oral microbiome-specific
(HOMD) and a generic (Greengenes; GG) database. Differences in bacterial
diversity [Phylogenetic diversity (PD), observed species, Chao1, and Shannon
indices] between the 4 EI groups were tested using a mixed-effects linear
regression model, accounting for clustering of observations within samples and
individuals, and applied a Bonferroni multiple-testing correction to account
for multiple pairwise comparisons. To formally test between-group differences
in microbial communities Analyses of Similarity (ANOSIM) were employed to
calculate R and P values using the phylogeny-based unweighted UniFrac distance
metric.
Results: A total of 406,070 sequences were obtained from the 48 samples after
quality filtering and de-noising corresponding to 8,460 reads per sample
(range:2,500-19,024). A total of 339 Operational Taxonomic Units (OTUs) were
assigned to 11 phyla, 60 families, and 109 genera. The most abundant phyla were
Bacteroidetes (36.2%), Firmicutes (32.9%), Actinobacteria (8.1%), Synergistetes
(7.4%), Fusobacteria (7.4%), Proteobacteria (5.2%), Spirochaetes (1.9%), and
Tenericutes (0.5%). Persistent infections were significantly enriched for
Proteobacteria (6.4% vs. 4.0%; P=0.02) and Tenericutes (0.1% vs. <0.05%;
P=0.03) compared to primary ones. Tenericutes were detected in 42% of
persistent infections versus 12% of primary infections (P<0.05). Using the GG
database, we identified 18 additional less-abundant phyla, all at less than
0.2% abundance. Among those, Cyanobacteria (0.018%) and Acidobacteria (0.007%)
were the most abundant, and were significantly enriched among persistent
infections: Acidobacteria were detected in 42% of samples with an abundance of
0.1%; Cyanobacteria were detected in 67% of samples with an abundance of 0.3%.
At the genus level, Bacteroidaceae_unclassified, Pyramidobacter, and Parvimonas
were the most abundant in primary infections whereas Fusobacterium,
Bacteroidaceae_unclassified, and Prevotella were the most abundant in teeth
with persistent infections. Significant differences were observed for 14 taxa
including increased enrichment of persistent infections for Lactobacillus,
Streptococcus, Sphingomonas and Ralstonia. In primary infections, symptomatic
ones were more diverse that the asymptomatic ones; in persistent infections the
opposite was found. Persistent infections showed higher Phylogenetic Diversity
compared to primary infections. ANOSIM indicated statistically significant,
albeit weak associations of infection type (R=0.087, P=0.005), symptoms
(R=0.055, P=0.047), and combined strata (R=0.093, P=0.007) with
UniFrac-assessed bacterial community composition. Using the GG database, in
primary infections, we identified a total of 24 phyla and 280 genera, whereas
these numbers were 28 and 347, respectively in persistent infections. In
primary infections, we identified on average 10 phyla, 50 genera, and 112
species-level phylotypes per sample, whereas these numbers were higher (12
phyla, 80 genera, and 162 species-level phylotypes) in teeth with persistent
infections. ANOSIM indicated statistically significant but small-in-magnitude
association of bacterial composition and infection type according to GG
database. Finally, bacteria classified as Elusimicrobia, OP3, OP8,
Planctomycetes and WS5 were not detected in primarily infected canals whereas
Gemmatimonadetes was the only phylum that was not found in
endodontically-treated teeth with persistent infections. Several additional
genera were detected using the GG database. Candidatus solibacter, Sharpea,
Methylobacterium, Novosphingobium, and Jathinobacterium were the most abundant
in this group.
Conclusions: In conclusion, the present pyrosequencing study offers a novel,
detailed characterization of the endodontic microbiome both in primary and
persistent infections. The present results suggest a high bacterial diversity
of endodontic infections and a more diverse bacterial community profile in
persistent versus primary infections. In addition, primary symptomatic
infections tended to be more diverse than primary asymptomatic infections; in
contrast persistent symptomatic infections were less diverse than persistent
asymptomatic ones. Using the GG database, a substantial number of
microorganisms was not possible to be taxonomically classified and may be
associated with the development of apical periodontitis. Further endodontic
microbiome studies are warranted to identify and characterize these
microorganisms.
