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

Abstract

Bacterial behavior and virulence during human infection is difficult to study and largely unknown, as our vast knowledge of infection microbiology is primarily derived from studies using in vitro and animal models. Here, we characterize the physiology of Porphyromonas gingivalis, a periodontal pathogen, in its native environment using 93 published metatranscriptomic datasets from periodontally healthy and diseased individuals. P. gingivalis transcripts were more abundant in samples from periodontally diseased patients but only above 0.1% relative abundance in one-third of diseased samples. During human infection, P. gingivalis highly expressed genes encoding virulence factors such as fimbriae and gingipains (proteases) and genes involved in growth and metabolism, indicating that P. gingivalis is actively growing during disease. A quantitative framework for assessing the accuracy of model systems showed that 96% of P. gingivalis genes were expressed similarly in periodontitis and in vitro midlogarithmic growth, while significantly fewer genes were expressed similarly in periodontitis and in vitro stationary phase cultures (72%) or in a murine abscess infection model (85%). This high conservation in gene expression between periodontitis and logarithmic laboratory growth is driven by overall low variance in P. gingivalis gene expression, relative to other pathogens including Pseudomonas aeruginosa and Staphylococcus aureus. Together, this study presents strong evidence for the use of simple test tube growth as the gold standard model for studying P. gingivalis biology, providing biological relevance for the thousands of laboratory experiments performed with logarithmic phase P. gingivalis. Furthermore, this work highlights the need to quantitatively assess the accuracy of model systems.

Precise spatial structure impacts antimicrobial susceptibility of S. aureus in polymicrobial wound infections

“…While it is difficult to perform direct comparisons due to the lack of quantitative spatial structure data in human chronic wounds, collectively these studies suggest that P. aeruginosa and S. aureus spatially segregate in wound infections. Future studies will focus on using the quantitative spatial structure data generated in this study as a benchmark for comparison to in vitro models and human wounds using a similar framework to that recently developed in our laboratory using transcriptomic data ( 39 , 58 , 59 ).…”

Iron deficiency modulates metabolic landscape of Bacteroidetes promoting its resilience during inflammation

“…As inflammatory conditions lead to reduced iron levels as well as elevated oxidative stress mediators released by inflammatory cells, adaptation through elevation of FldA and replacement of Frd seems like an intuitive strategy benefiting persistence of the bacteria in such conditions. Our data may also have translational significance as FldA is highly upregulated in B. thetaiotaomicron in the colitis model as well as it’s transcript is very abundant in P. gingivalis identified in patients with periodontal disease 24 . Such data point to a novel metabolic mechanism employed by P. gingivalis that relies more on iron independent enzymes during active disease status and thus contrasts with thus far favored metabolic model that is employed by the bacterium when grown in laboratory in iron rich conditions 41 .…”

“…First, the locus coding for a flavodoxin (FldA – BT0517) and DUF2023 protein (BT0516) was drastically upregulated in both, our model and the colitis model (Tables 5, 7) 2 . The P. gingivalis flavodoxin encoding gene, fldA, was found to be one of the most highly expressed genes in microbiomes derived from human subjects with periodontitis indicating that in the extracellular inflamed environment genes upregulated by iron deficiency are also highly expressed 24 .…”

Polymicrobial Interactions of Oral Microbiota: a Historical Review and Current Perspective

“…For example, the 10-species “Zurich model” has been used to study topics such as biofilm development, virulence factors, and microbial interactions ( 54 , 55 ). The advancement of sequencing technologies and microscopy techniques offers further opportunities to design and customize in vitro models to match the oral community composition, physiology, and spatial patterning ( 39 , 42 , 56 ), and there are significant opportunities to use these model systems to probe diverse questions ( Table 1 ). In contrast, top-down approaches that directly analyze human patient samples will continue to be essential for understanding the functional and spatial dynamics of oral microbial communities, as discussed in “Current and future perspectives” above.…”

“…Additional support for “the community as pathogen” comes from community-wide metatranscriptomic studies that found organisms traditionally considered commensals transcribe the majority of virulence factors during periodontal diseases ( 37 , 38 ). Metatranscriptomics can also provide insight into the in situ functionality of individual species ( 39 ). One striking finding is that a species can have similar relative abundance but different functional activities across different conditions.…”