Multi-Omics Data Integration Reveals Key Variables Contributing to Subgingival Microbiome Dysbiosis-Induced Inflammatory Response in a Hyperglycemic Microenvironment

Abstract

Subgingival microbiome dysbiosis promotes the development of periodontitis, an irreversible chronic inflammatory disease associated with metabolic diseases. However, studies regarding the effects of a hyperglycemic microenvironment on host-microbiome interactions and host inflammatory response during periodontitis are still scarce. Here, we investigated the impacts of a hyperglycemic microenvironment on the inflammatory response and transcriptome of a gingival coculture model stimulated with dysbiotic subgingival microbiomes. HGF-1 cells overlaid with U937 macrophage-like cells were stimulated with subgingival microbiomes collected from four healthy donors and four patients with periodontitis. Pro-inflammatory cytokines and matrix metalloproteinases were measured while the coculture RNA was submitted to a microarray analysis. Subgingival microbiomes were submitted to 16s rRNA gene sequencing. Data were analyzed using an advanced multi-omics bioinformatic data integration model. Our results show that the genes krt76, krt27, pnma5, mansc4, rab41, thoc6, tm6sf2, and znf506 as well as the pro-inflammatory cytokines IL-1β, GM-CSF, FGF2, IL-10, the metalloproteinases MMP3 and MMP8, and bacteria from the ASV 105, ASV 211, ASV 299, Prevotella, Campylobacter and Fretibacterium genera are key intercorrelated variables contributing to periodontitis-induced inflammatory response in a hyperglycemic microenvironment. In conclusion, our multi-omics integration analysis unveiled the complex interrelationships involved in the regulation of periodontal inflammation in response to a hyperglycemic microenvironment.

Keywords: computational biology; cytokines; dysbiosis; hyperglycemia; inflammation; metagenomic; metalloproteases; microbiome; omics integration; periodontitis.

Figure 1

The hyperglycemic microenvironment stimulates pro-inflammatory cytokine and matrix metalloproteinase secretion in a gingival coculture model. Pro-inflammatory cytokine and matrix metalloproteinase secretion of (A) IL-1β, (B) TNF-α, (C) IL-6, (D) MMP3, (E) MMP8 and (F) MMP9 by the coculture model following 24-h stimulation with subgingival microbiomes from healthy participants (healthy) and patients with periodontitis (periodontitis) in 5 mM and 25 mM glucose. The dashed line represents the baseline secretion by the untreated coculture model in both 5 mM and 25 mM glucose. Results are the mean ± SEM. One independent replicate and two technical replicates were undertaken for each condition. Two-way ANOVA with the Tukey post hoc test. * p < 0.05, ** p < 0.01, **** p < 0.0001.

Figure 2

The subgingival microbiome highlights the enrichment of particular amplicon sequence variants (ASVs). (A) Relative abundance of most abundant genera following 16s rRNA gene sequencing of subgingival microbiome from healthy participants and patients with periodontitis. (B) Principal component analysis (PCA) of the ASV analysis of microbiomes from healthy participants and patients with periodontitis. (C) Enrichment of ASV of microbiomes from healthy participants and patients with periodontitis. Samples from healthy donors are identified h1 to h4 while the samples from patients with periodontitis are identified p1 to p4.

Figure 3

Multi-omics correlations by computational biology. (A) Arrow plot. This plot is the superposition of 3 PCAs corresponding to the 3 layers of omics data. For each sample, one arrow represents a set of omics data and the mean of the 3 layers is the centroid. If the arrows are close, the 3 layers of data are similar. On the contrary, if the arrows are apart, then the 3 layers are different. (B) Varplot. Positively correlated variables are grouped together. Negatively correlated or anti-correlated variables are positioned on opposite sides of the plot origin (opposed quadrants). Samples from healthy donors are identified h1 to h4 while the samples from patients with periodontitis are identified p1 to p4.

Figure 4

Multi-omics correlations by computational biology analyses reveal key variables contributing to periodontitis-induced inflammatory response in a hyperglycemic microenvironment. (A) Projection of the most important variables contributing to the varplot’s component 1. (B) Projection of the most important variables contributing to the varplot’s component 2.