Bacteriome analysis of Aggregatibacter actinomycetemcomitans-JP2 genotype-associated Grade C periodontitis in Moroccan adolescents

Abstract

Background: Grade C (previously aggressive) periodontitis (GCP) in adolescents is prevalent in certain parts of Africa where it is associated with JP2 genotype, a highly virulent strain of Aggregatibacter actinomycetemcomitans. The aim of this study was to characterize the subgingival bacteriome in Moroccan subjects with GCP positive to A. actinomycetemcomitans JP2 genotype.

Methods: Subgingival plaque samples were collected from shallow and deep pockets of 8 subjects with GCP (17.2 ± 1.5 years) and from gingival sulci of 13 controls with no periodontitis (14.6 ± 1.1 years). Identification and genotyping of A. actinomycetemcomitans was performed using PCR analysis of the ltx operon, while bacteriome profiling was done by 16S rRNA gene sequencing (V1-V3 region). Groups were compared in terms of microbial diversity, abundances, and dysbiosis.

Results: The shallow and deep pocket sites from GCP cases had a significantly altered microbial composition compared to controls. Species associated with health included Haemophilus parainfluenzae, Lautropia mirabilis, Streptococcus spp., Gemella spp., and Rothia spp. While known periodontal pathogens, including Porphyromonas gingivalis, Tannerella forsythia, Treponema spp. and Fretibacterium spp., were significantly enriched in GCP, non-conventional taxa, including Pseudomonas oral taxon C61 and Enterobacter cloacae were more abundant and showed stronger association with the disease. Less significant differences in abundances of individual taxa were observed between shallow and deep pockets. Overall dysbiosis measured in terms of Subgingival Microbial Dysbiosis Index (SMDI) differentiated between GCP and no-periodontitis with 95% accuracy.

Conclusions: The results suggest that several periodontal pathogens involved in the adult-type periodontitis also play a role in JP2 genotype-associated GCP. The potential role of non-conventional taxa in the pathogenesis of GCP warrants further investigation.

Keywords: biofilm; dysbiosis; high-throughput nucleotide sequencing; microbiota; periodontitis.

Figure 1

Average microbial profiles. DNA extracted from subgingival plaque samples was sequenced for the V1–V3 region of the 16S rRNA gene using 2 × 300 bp chemistry. The resultant sequences were merged, quality-filtered and classified to the species level using a BLASTn-based algorithm. The stacked bars show the average relative abundances of all phyla, top genera and top species identified in the study groups.

Figure 2

Species richness and diversity. Species count table were rarified and used to calculate observed species richness (A), and alpha diversity [Shannon index, (B)]. The data were then centered log-ratio transformed and used for beta-diversity analysis based on Principle Component Analysis [PCA plot, (C)]. Significance of differences between the groups for (A) and (B) were sought using pairwise Mann-Whitney or Wilcoxon signed rank test as appropriate; *P ≤ 0.05. Significance of differences between the clusters in (C) were assessed with PERMANOVA.

Figure 3

Top differentially abundant species between GCP and health. Centered log ratio (CLR) transformed data were analyzed with MaAsLin2 to identify differentially abundant species between the controls and shallow pockets from the GCP cases (A) and between the controls and deep pockets from the GCP cases (B) Differences with FDR ≤ 0.1 were considered significant. The top 40 species based on magnitude of association (coefficient) are shown in (A) and (B).

Figure 4

Relative abundances of selected species. Combined box and strip plots depicting the relative abundances of selected known periodontal pathogens as well as non-conventional species that were identified as significantly different between GCP cases and control subjects. *Statistically significant difference based on MaAsLin2 analysis; FDR ≤ 0.1.

Figure 5

Subgingival microbial dysbiosis in GCP compared to health. Centered log ratio (CLR) transformed data were used to calculate SMDI for each sample as previously described (36) and presented as combined box and strip plot for each group. Significance of differences between the groups were sought using pairwise Mann-Whitney or Wilcoxon signed rank test as appropriate; *P ≤ 0.05; **P ≤ 0.01.