A quantitative framework reveals traditional laboratory growth is a highly accurate model of human oral infection

Abstract

Significance Most microbial knowledge comes from experiments in laboratory models, despite the assumption that these artificial systems alter microbial physiology relative the native environment. We tested this assumption for an oral pathogen, Porphyromonas gingivalis , using 93 metatranscriptomes from periodontally healthy and diseased patients and 122 transcriptomes from experimental models. We discovered that a simple in vitro model, midlogarithmic growth in rich media, highly recapitulates P. gingivalis gene expression in the human oral cavity, outperforming other models, including a murine infection model. These results support the biological relevance of decades of laboratory experiments with this pathogen and validate an accessible experimental model for studying P. gingivalis biology. Furthermore, these data provide a conceptual framework for understanding in situ gene expression across microbes.