Using Genome-Scale Metabolic Modeling to Compare Serotypes of the Foodborne Pathogen Listeria Monocytogenes

Listeria is a genus of rod-shaped, Gram positive, non-spore forming facultative anaerobic bacteria.  The disease caused by Listeria monocytogenes, a problematic foodborne pathogen, is called listeriosis.  There are an estimated 1,600 cases of listeriosis each year in the U.S. resulting in 260 deaths.  The mortality rate of listeriosis is extremely high, ranging between 20-30%, and listeriosis costs an estimated $2.8 billion in healthcare/year.  Certain sectors of the population, including pregnant women, the elderly, and the immunocompromised, are at high risk of contracting listeriosis.  95% of human listeriosis cases originate from three L. monocytogenes serotypes—1/2a, 1/2b, and 4b.  Over the past 20 years, serotyping of L. monocytogenes strains was conducted by pulsed field gel electrophoresis, a method that only looks at the banding patterns of genomic DNA fragments in gels following restriction enzyme digestion. The main goal of this project is to expand the use of genome-scale metabolic models (GEMs) to better understand the diversity of outbreak strains of the foodborne pathogen L. monocytogenes.  Separate GEMs were created for six different strains of L. monocytogenes, two from each of the three major serotypes.  These models were used in conjunction with Biolog assays to analyze nutrient utilization and to resolve model inconsistencies.  These analyses were also used to determine the metabolic differences between strains and to discover possible reasons for varying degrees of pathogenesis between strains.

GEMs were initially constructed using the semi-automated resource K-Base. Total reaction content shared by all six Listeria monocytogenes GEMs was >90%, and from the remaining reaction pool we have identified subsets and unique reactions for each serotype class and for each strain.  Initial GEMs had 19 carbon metabolites comparable to those on Biolog PM1 carbon plates, and manual curation increased computational metabolic capabilities for each GEM to contain 58 exchange reactions for metabolites contained on Biolog carbon source plates. This led to carbon source utilization comparisons between computational and experimental data and agreement was determined >82% for all of the L. monocytogenes GEMs. Essential reaction analysis revealed 351 predicted essential reactions shared in common by all six strain-specific GEMs, and the remaining predicted essential reactions are mostly not serotype specific.  We have now identified some metabolic differences and similarities between strains of the same serotype and of different serotypes (1/2a, 1/2b, and 4b). 

This project greatly advances the understanding of Listeria, including its evolutionary process and differential metabolic capabilities, and increases the understanding of inter-strain differences of L. monocytogenes strains—specifically, nutrient capabilities.  Finally, this project will lay the groundwork for future GEMs for relatives of Listeria, and lead to the development of new food safety techniques to possible control L. monocytogenes in foods, and to treat human disease.