Discipline: Biological Sciences
Subcategory: Microbiology/Immunology/Virology
Shu Pan - University of Wisconsin-Madison
Co-Author(s): Kiel Nikolakakis, Min Pan, and Jennifer Reed, University of Wisconsin-Madison Edward Ruby, University of Hawaii
Vibrio fischeri ES114 is a gram-negative bacterium that can be found either free living or associated with the bioluminescent organ of a squid, Euprymna scolopes. The squid-vibrio system is used to study the metabolic maintenance of symbiotic relationships, which are found everywhere. For example, the biochemical activities of microbes that live inside or on the surface of our bodies are critical to our health. Using modern systems biology tools, we aimed to better understand the metabolism of this bacterium and, in particular, its symbiotic relationship with the squid, as a way to provide new insights into the general symbiotic mechanisms that underlie bacterium-host interactions.
We first constructed a genome-scale metabolic model (GEM) for V. fischeri. Some content of the V. fischeri model was extracted from one of the best-curated models for Escherichia coli (iJO1366). E. coli reactions and metabolites were directly transferred into the V. fischeri model and associated with 783 V. fischeri metabolic genes that were orthologous to E. coli genes. A total of 39 genes, 25 metabolites, and 66 reactions unique to V. fischeri were further added based on genome annotations, data in the literature, and our own experimental results. We measured the biochemical composition of V. fischeri, its growth rates, and metabolite consumption and production rates to parameterize the model. To validate the model, we performed high-throughput experiments to measure the growth phenotypes of V. fischeri under 181 different sole carbon-source conditions.
Our model is a systems-level representation of most known biochemical reactions and genes of V. fischeri ES114. It consists of 1725 reactions, 1019 unique metabolites, and 822 metabolic genes. The model correctly predicts 86.2% of the experimentally validated growth phenotypes. It also successfully categorizes 83.4% of the 822 metabolic genes into either essential or non-essential genes in a complex-nutrient medium. In the future, we will use this model, together with transcriptomics data of V. fischeri cells immediately after release from the symbiotic organ, to infer interactions between the bacterium and its host. Furthermore, various high-throughput in silico experiments, such as growth competitions between mutants, can be simulated using this model.
V fischeri model.docxFunder Acknowledgement(s): We thank the Gordon and Betty Moore Foundation (award #3396) for supporting this work.
Faculty Advisor: Jennifer Reed, jennifer.reed@wisc.edu