A reproducible microcosm biofilm model of subgingival microbial communities

Objective: To develop a reproducible subgingival microcosm biofilm model. Material and Methods: Subgingival plaque samples were collected from four deep pockets (probing pocket depth >= 6 mm) in each of seven patients with periodontitis and from shallow pockets (probing pocket depth <= 3 mm) in two periodontally healthy donors. An active attachment model and a peptone medium (Thompson et. al., Appl Environ Microbiol 2015;81:8307-8314) supplemented with 30% serum was used. Biofilms were harvested at 2 and 4weeks. DNA of dead cells was blocked for amplification by propidium monoazide treatment. Composition was analyzed using 16S rRNA gene amplicon pyrosequencing. Similarities between the biofilm samples were assessed by non-metric multidimensional scaling using the Bray-Curtis similarity index and similarity percentage analysis. Data from duplicate experiments, different biofilm sources and different biofilm age were compared. Results: The non-metric multidimensional scaling revealed a strong clustering by the inoculum source, the donor and their periodontal status. Statistically significant differences were found between the sources of inoculum (P=.0001) and biofilm age (P=.0016). Furthermore, periodontitis biofilms (P) were distinct in composition from health-derived biofilms (H) by genera: Porphyromonas (P=19%; H=0%), Filifactor (P=10%; H=0%), Anaeroglobus (P=3%; H=0%), Phocaeicola (P=1.5%; H=0%), Parvimonas (P=19%; H=14%), Fusobacterium (P=2%; H=26%), Peptostreptococcus (P=20%; H=30%), Veillonella (P=7%; H=8%) and 57 other genera. Similarity distances (Bray-Curtis) (mean 0.73, SD 0.15) and the Shannon diversity index (mean 2, SD 0.2) revealed no differences between duplicate experiments (P=.121). Conclusion: This biofilm model allows reproducible production of complex subgingival microbial communities.