Physiologically relevant coculture model for oral microbial-host interactions

Abstract

Understanding microbial-host interactions in the oral cavity is essential for elucidating oral disease pathogenesis and its systemic implications. In vitro bacteria-host cell coculture models have enabled fundamental studies to characterize bacterial infection and host responses in a reductionist yet reproducible manner. However, existing in vitro coculture models fail to establish conditions that are suitable for the growth of both mammalian cells and anaerobes, thereby hindering a comprehensive understanding of their interactions. Here, we present an asymmetric gas coculture system that simulates the oral microenvironment by maintaining distinct normoxic and anaerobic conditions for gingival epithelial cells and anaerobic bacteria, respectively. Using a key oral pathobiont, Fusobacterium nucleatum, as the primary test bed, we demonstrate that the system preserves bacterial viability and supports the integrity of telomerase-immortalized gingival keratinocytes. Compared to conventional models, this system enhanced bacterial invasion, elevated intracellular bacterial loads, and elicited more robust host pro-inflammatory responses, including increased secretion of CXCL10, IL-6, and IL-8. In addition, the model enabled precise evaluation of antibiotic efficacy against intracellular pathogens. Finally, we validate the ability of the asymmetric system to support the proliferation of a more oxygen-sensitive oral pathobiont, Porphyromonas gingivalis. These results underscore the utility of this coculture platform for studying oral microbial pathogenesis and screening therapeutics, offering a physiologically relevant approach to advance oral and systemic health research.

Fig. 1

Development of an asymmetric gas coculture system utilizing an immortalized gingival epithelial cell line as a barrier. a A schematic illustration depicting the gas composition within the apical and basolateral chambers of an asymmetric coculture setup housed in a vinyl anaerobic chamber, as well as the overall composition of the entire asymmetric coculture system. (Left) Input gas containing 10% oxygen enters the basolateral chamber to support cell monolayer respiration, progressively depleting the oxygen. Concurrently, the apical chamber maintains an anaerobic environment that closely replicates the surrounding anaerobic conditions, thereby ensuring the optimal growth of anaerobic bacteria. (Right) The plate cap ① covers the gas-permeable plate ④, preventing contamination and leakage. The apical chamber of the Transwell inserts ② is filled with coculture medium (CCM, see Table S1), which supports the stable growth of the cell monolayer ③ and oral anaerobes. The lower compartment (i.e., the basolateral compartment) of the Transwell insert is filled with cell culture medium. The gas-permeable membrane ⑤ at the bottom ensures unidirectional oxygen diffusion. Basolateral gas flow containing 10% oxygen enters through the gas inlet ⑨, spreading evenly through the asymmetric coculture chamber ⑧ with a magnetic stirrer ⑦. Exhaust gas is discharged through the gas outlet ⑥, completing the system’s airflow. The figure was created using BioRender. b A physical picture of the asymmetric coculture chamber. c Comparison of fluorescence intensities of FITC-dextran in the basolateral chamber at 24 h after adding FITC-dextran to the apical chamber of Transwells containing TIGK monolayers. Undifferentiated (negative control) and differentiated TIGKs under the normoxic culture conditions (referred to as “normoxic”) were compared to differentiated TIGKs under the asymmetric culture condition (referred to as “asymmetric”). The background fluorescence intensity of the blank culture medium was subtracted for each condition. Undifferentiated TIGK monolayer cultured under the normoxic condition before switching to a differentiation medium containing Ca²⁺ serves as the negative control (N.S., P > 0.05, *** P < 0.001, n = 2 technical replicates, N = 3 biological repeats). d The morphology of TIGK monolayers in Transwell inserts maintained in a cell culture incubator under normoxic conditions or cultured in an asymmetric coculture chamber for 24 h. The collagen coating is known to affect bright field imaging due to the presence of the collagen fibers, which can obscure fine details of the cells or structures being observed compared to uncoated surface. e Comparison of cell viability in TIGK monolayers cultured under normoxic and asymmetric culture conditions. Heat-treated cells are the negative control (N.S., P > 0.05, ** P < 0.01, n = 3, N = 3)

Fig. 2

The asymmetric gas setup maintains the viability of Fn. Representative flow cytometry results for a Fn 23726 and b Fn 25586 under asymmetric gas coculture (with TIGK monolayer), normoxic coculture (with TIGK monolayer), anaerobic condition (in the anaerobic workstation), and dead cell negative control after heat treatment. The MOI was 1. SSC: side scattering. c Viability quantification of Fn strains 23726/25586 assessed by flow cytometry following different treatment groups, including the asymmetric gas coculture, normoxic coculture, anaerobic culture (positive control), and dead cell negative control after heat treatment (N.S., P > 0.05, ** P < 0.01, *** P < 0.001, n = 2 technical replicates, N = 3 biological replicates). CFUs of d Fn 23726 and e Fn 25586 in the apical chamber under normoxic and asymmetric coculture conditions at 0, 12, and 24 h time points (n = 2, N = 3). The MOI was 1

Fig. 3

Differential responses of TIGKs challenged by three Fn model strains under normoxic and asymmetric gas coculture conditions. Morphological comparison of TIGK monolayers in a the medium control (CCM only) and after coculture with b Fn 23726, c Fn 25586, and d Fn 10953 under the asymmetric gas coculture system. The MOI of Fn was 1. Red arrows indicate the adherence of bacterial aggregates to the TIGK monolayer. Due to the presence of the collagen fibers, collagen coating can obscure fine details of the cells or structures being observed compared to uncoated surface. e Representative flow cytometry results for TIGKs after coculture with three different Fn strains, as assessed by the SYTOX green staining. f Quantification of TIGK viability after asymmetric coculture with Fn 23726, Fn 25586, and Fn 10953 based on the above flow cytometry results (0.01 < *P < 0.05, n = 2, N = 3). g Comparison of fluorescence intensities of FITC-dextran in the basolateral chamber at 24 h after adding FITC-dextran to the apical chamber containing TIGK monolayers receiving different treatment conditions. Undifferentiated TIGKs cultured under the normoxic condition served as the negative control due to the lack of an intact monolayer. Differentiated TIGKs with or without Fn 25586 challenge under normoxic or asymmetric culture conditions were compared. The background fluorescence of the culture medium was subtracted. The MOI was 1 (N.S., P > 0.05, *** P < 0.001, n = 2, N = 3)

Fig. 4

Analysis of Fn invasion and its effect on the secretion of proinflammatory factors by cell monolayers. a Illustration of Fn transitioning from a planktonic state to adherence and invasion to gingival epithelial cells. The figure was created using BioRender. Confocal images showing the interactions between Fn and TIGK cell monolayer under normoxic (b) and asymmetric coculture conditions, colocalized with Fn strains 23726 (c), 25586 (d), and 10953 (e). Fn was prelabeled with a red fluorescence dye DiD, and TIGK cells were post-labeled by Alexa Fluor™ 488 Phalloidin for cytoskeleton. MOI of Fn was 50. f Differential invasion of three Fn strains into the gingival epithelial cells (TIGKs) under normoxic and asymmetric coculture conditions for 24 h. The MOI of Fn was 1 (*0.01 < P < 0.05; n = 2, N = 3). g Enzyme-linked immunosorbent assay result of CXCL10 (IP-10) expression levels after infection with different Fn strains. The value was measured by assessing absorbance at 450 nm, with a reference at 570 nm (n = 3, N = 3). h–j Luminex outcome of the difference in expression levels of different pro-inflammatory factors after infection with different Fn strains (*0.01 < P < 0.05, ** P < 0.01; n = 1, N = 3). k Intracellular bacterial counts of Fn 25586 and Fn 10953 after treatment with amoxicillin, metronidazole, and their combination (n = 2, N = 3)

Fig. 5

The asymmetric coculture system further enables oxygen-sensitive anaerobe Pg to proliferate normally in coculture with TIGK cell monolayers. a CFUs of Pg 33277 in the apical chamber under normoxic and asymmetric coculture conditions at 0, 24, and 48 h time points (n = 2, N = 3). The MOI was 1. b Representative flow cytometry results for TIGKs after coculture with Pg 33277, as assessed by the SYTOX green staining. c Quantification of TIGK viability after asymmetric coculture with Pg 33277 based on the above flow cytometry results (*0.01 < P < 0.05, N.S., P > 0.05, n = 2, N = 3)