Engineering a dysbiotic biofilm model for testing root caries interventions through microbial modulation

Abstract Background This study aimed to engineer and optimise a dysbiotic biofilm model to develop in vitro root caries for investigating microbial modulation strategies. The model involved growing complex biofilms from a saliva inoculum collected from four volunteers using two strategies. In the first strategy (“pre-treatment strategy”), bovine root slabs were used, and two natural compounds were incorporated at time 0 of the 10-day biofilm experiment, which included sucrose cycles mimicking the cariogenic environment. In the second strategy (“post-treatment strategy”), mature biofilms were grown in a modified Calgary biofilm device coated with collagen and hydroxyapatite for 7 days and then were exposed to the same natural compounds. The metatranscriptome of each biofilm was then determined and analysed. Collagenase activity was examined, and the biofilms and dentine were imaged using confocal and scanning electron microscopy (SEM). Mineral loss and lesion formation were confirmed through micro-computed tomography (μ-CT). Results The pH confirmed the cariogenic condition. In the metatranscriptome, we achieved a biofilm compositional complexity, showing a great diversity of the metabolically active microbiome in both pre- and post-treatment strategies, including reads mapped to microorganisms other than bacteria, such as archaea and viruses. Carbohydrate esterases had increased expression in the post-treated biofilms and in samples without sugar cycles, while glucosyltransferases were highly expressed in the presence of sucrose cycles. Enrichment for functions related to nitrogen compound metabolism and organic cyclic component metabolism in groups without sucrose compared to the sucrose-treated group. Pre-treatment of the roots with cranberry reduced microbial viability and gelatinase (but not collagenase) activity (p