Construction of a Dynamic Flux Balance Analysis Model for Streptococcus Gordonii

Periodontitis is a chronic and inflammatory disease of the tooth supporting tissues. Environmental changes can bring oral biofilms from a healthy state to dysbiosis, which potentially evolves to periodontitis. Overall, we aim to construct a biofilm community model that brings more insight in dysbiosis of the oral biofilm and the influence of environmental factors. We develop a dynamic flux balance analysis (DFBA) model for Streptococcus gordonii which forms the first step in the development of an oral biofilm community model linked to periodontitis.

S. gordonii produces H₂O₂ and lactate, inhibiting pathogen outgrowth and enabling cross-feeding in the biofilm, respectively. First, the automated reconstruction taken from the AGORA database^[1] was curated using literature information and databases (e.g., KEGG, RHEA, BIGG). FBA was used to evaluate growth rates, and to analyze carbon routing through the central carbon metabolism and amino acid biosynthesis pathways, resulting in a GSMN with 1278 reactions and 1078 metabolites, reproducing realistic FBA predictions. Secondly, the GSMN was transformed into a planktonic model in DFBAlab^[2]. Simulations with glucose as main C-source predict the expected coupled production of lactate and H₂O₂, and show that protoheme depletion diverts carbon resources to lactate, thus triggering lactate production. Protoheme is an essential compound that cannot be synthesized by S. gordonii. H₂O₂ production is predicted with and without protoheme limitation, although concentrations are low. Analyses of growth on amino acids showed enhanced production of H₂O₂ for growth on glucose and arginine, without influencing lactate production. Finally, DFBA simulations in a S. gordonii biofilm are established. Simulations still show the coupled production of lactate and H₂O₂. In a following stage, additional community members are added to the model.

^[1]S. Magnúsdóttir et al., Nat. Biotechnol., vol. 35, no. 1, 2017.

^[2]J. A. Gomez, K. Höffner, and P. I. Barton, BMC Bioinformatics, vol. 15, no. 1, 2014.