A computational systems approach has been used to investigate the biogenesis of the
outer envelope of Escherichia coli. Specifically, a deterministic model incorporating the
lipopolysaccharide, peptidoglycan and phospholipid components. The expanded model
now contains 195 reacting species, making up 145 biochemical reactions, making it the
largest dynamic model published. The model was used to understand key regulatory
elements underpinning pathway regulation and its findings extrapolated to account for
phenotypic effects such as growth rates and cell envelope health. The model was capable
of producing the known amounts of endpoint metabolites for each of the relevant
pathways- Indicating that the parametrisation is reasonable for this network with steady
state achieved for most reactions. However, there are numerous nodes which are fitted
which can be a source of error. In order to test model predictions a strain library of
tuneable CRISPR recombinant E. coli, which can downregulate specific genes (between
5-30%) within this network was generated. During strain validation it was found that
genetic perturbation resulted in an associated accumulation at the metabolic level. This
strongly suggests that the effect observed in these recombinant strains is likely a result of
the perturbation of the specific gene of interest, making it an appropriate method to
constrain the metabolic model. Intriguingly, one of the recombinant strains (MsbA- LPS
flippase), when repressed caused a build up of DSMP which is the substrate for LpxK.
This association has not been published previously and potentially represents a novel
regulatory mechanism. However, more work is required in order to constrain and test the
model accuracy.