(507h) A Second-Generation Genome Scale Model for Clostridium Acetobutylicum

The development of genome scale models (GSM) for a variety of microbial, eykaryotic and multi-cellelar organisms [1] is proceeding with an accelerating pace. Automated reconstruction tools such as ModelSEED [2], Pathway Tools [3], and others are increasingly being utilized to quickly arrive at draft compilations of metabolic models followed by detailed curation to improve upon modeling components crucial for answering the queries relevant to the specific organism and study. Typically, simulations are carried out for an idealized growth substrate typically containing a single carbon substrate (e.g., glucose). Regulation is customarily handled by shutting off specific reactions in the presence of a specific nutrient environment. For example, in the iAF1260 E. coli model, there are specific reaction restrictions for anaerobic, aerobic, and aerobic with glucose growth conditions [13].

In this talk, we present a second-generation genome scale metabolic model for Clostridium acetobutylicum. The model contains over 750 genes and 1200 reactions. The reactions are all charge and elementally balanced. We made extensive use of MetRxn (http://metrxn.che.psu.edu), a web-based knowledgebase that includes standardized metabolite and reaction descriptions and integrates information from 8 databases and 44 metabolic models into a single unified data set. In addition to metabolic biotransformations, the model also includes regulatory information. Establishment of detailed gene-protein-reaction associations enables us to properly localize regulatory interactions at the transcriptional, protein or metabolite level and rely on the gene-protein-reaction (GPR) connections to propagate their action through metabolism. We categorize regulatory interactions into either up or down and of moderate or high strength. We make use of transcriptional data obtained under various stress conditions. In analogy to GPR associations, we define and populate stressor-regulator-operon-gene (SROG) associations that pinpoint which genes are affected by a specific stressor. We discuss the use of gene expression data throughout model development such as during the GapFill process. The model allows for phenotype predictions for growth under various media and conditions and enables the generation of testable hypotheses on the response to substrate changes and genetic alterations on biofuel production.