Saliva Versus All-Site Microbiome and Proteome Mapping in Periodontitis

Abstract

Aim: Saliva is a diagnostic surrogate for microbial and host biomarkers in periodontitis, but whether it reflects subgingival plaque, gingival crevicular fluid (GCF) or serum remains debated. This study profiled GCF and subgingival plaque from all sites of periodontitis patients, comparing them to saliva and serum.

Materials and methods: Saliva, serum, subgingival plaque and GCF were obtained from three patients with stage III, grade C periodontitis, having 23, 25 and 27 teeth, respectively, with six sites sampled per tooth. All plaque and GCF samples were pooled per patient. Shotgun sequencing and mass spectrometry proteomics were used for microbiome and proteome analysis, respectively.

Results: Totally 277 microbial taxa were collectively identified in saliva and plaque, 93 of which were differentially abundant between the two. Saliva exhibited higher overall species diversity, but lower periodontal pathogen abundance. A total of 1153 host proteins were identified (saliva: 803; GCF: 932; serum: 195) across the three biological fluids, with 685 shared among saliva and GCF and 109 among all three. Saliva contained slightly fewer proteins than GCF, but shared several common immune, metabolic and enzyme regulation pathways.

Conclusion: Saliva is effective for broad microbiome and proteome screening; whereas plaque delivers greater precision in identifying specific periodontal pathogens directly associated with a periodontal pocket.

Keywords: gingival crevicular fluid; liquid chromatography tandem mass spectrometry‐based proteomics; plaque; saliva; serum; whole genome shotgun sequencing.

FIGURE 1

Illustration of sample collections and their corresponding high‐throughput analyses. Samples collected from paper points strips, unstimulated saliva, GCF and serum were first processed by centrifugation. The pellets from saliva and paper points were subjected to whole genome sequencing to examine the oral microbiota. The supernatants of saliva, GCF and whole serum were subjected to LC/MS2 to determine host protein profiles.

FIGURE 2

Microbiome analysis between plaque and saliva. (A) Venn diagram based on identified species. (B) Chao1, (C) Simpson and (D) Shannon index between plaque and saliva (E) Mean bacterial relative abundance in plaque and saliva. Species with a relative abundance of less than 3% are classified as ‘other’. (F) Bacterial relative abundance for key species in both plaque and saliva. The relative abundance of the top 10 species in plaque or saliva samples plus T. denticola (the only red complex species not among the top 10), while the rest species are summed into ‘other’. Abundance showing a significant shift (p < 0.05) between the two sites, as determined by LEfSe analysis, is highlighted with an ‘*’. Bacteria with a greater discriminatory presence in plaque are shown in red, while those with a greater discriminatory presence in saliva are shown in blue. Three complex species are highlighted in red rectangles.

FIGURE 3

Host protein profiles and functional analyses. (A) Venn diagram based on the identified proteins between saliva, GCF and serum. (B) PCA analysis based on the serum counts of saliva, GCF and serum. (C) The relative protein abundance based on the normalised total spectra. Proteins with a relative abundance of less than 3% are classified as ‘other’. (D) The top 10 significantly enriched GO terms based on proteins overlapping between all three tissues. (E) The top 10 significantly enriched GO terms based on proteins overlapping between all three tissues, along with the corresponding proteins contributing to the enrichment of these terms. (F) All significantly enriched GO terms based on proteins solely identified by saliva and GCF. (G) Five death‐related GO terms based on proteins solely identified by saliva and GCF, along with the corresponding proteins contributing to the enrichment of these terms. (H) Three cellular response‐related GO terms based on proteins solely identified by saliva and GCF, along with the corresponding proteins contributing to the enrichment of these terms. (I) All significantly enriched GO terms based on proteins identified solely in serum. (J) All significantly enriched GO terms based on proteins identified solely in serum, along with the corresponding proteins contributing to the enrichment of these terms.