Stable reconstructed human gingiva–microbe interaction model: Differential response to commensals and pathogens

Background: To investigate human oral health and disease, models are required which represent the interactions between the oral mucosa and microbiome. Our aim was to develop an organotypic model which maintains viability of both host and microbes for an extended period of time. Methods: Reconstructed Human Gingiva (RHG) were cultured air-lifted with or without penicillin-streptomycin (PS) and topically exposed to Streptococcus gordonii (commensal) or Aggregatibacter actinomycetemcomitans (pathogen) for 72 hours in agar. RHG histology, viability and cytokines (ELISA), and bacterial viability (colony forming units) and location (FISH) were assessed. Results: The low concentration of topically applied agar did not influence RHG viability. Topically applied bacteria in agar remained localized and viable for 72 hours and did not spill over to infect RHG culture medium. PS in RHG culture medium killed topically applied bacteria. Co-culture with living bacteria did not influence RHG viability (Ki67 expression, MTT assay) or histology (epithelium differentiation, Keratin10 expression). RHG exposed to S. gordonii (with or without PS) did not influence low level of IL-6, IL-8, CCL2, CCL5, CCL20 or CXCL1 secretion. However, all cytokines increased (except CCL2) when RHG were co-cultured with A. actinomycetemcomitans. The effect was significantly more in the presence of living, rather than dead, A. actinomycetemcomitans. Both bacteria resulted in increased expression of RHG antimicrobial peptides (AMPs) Elafin and HBD-2, with S. gordonii exposure resulting in the most Elafin secretion. Conclusion: This technical advance enables living human oral host–microbe interactions to be investigated during a 72-hour period and shows differences in innate immunology triggered by S. gordonii and A. actinomycetemcomitans.