Team:Groningen/Brainstorm/Glucose Sensing

Originally the Rainbow bacteria, we now no longer plan to use the Cre/Lox system, but rather focus on glucose sensing. This choice was made because we have no good way to control the cre/lox recombination the way we would need for glucose sensing.

The idea now is to create a bacterium that senses the outside glucose concentration and produces an inversely proportional response.

Requirements
We use the MoSCoW prioritization method:


 * Must have this:
 * Response should be inversely proportional to the outside glucose concentration.
 * It should be measurable on-line (visually for example).
 * Should have this if at all possible:
 * Response should be a gradient (not just on/off).
 * Delay in the response should not exceed roughly 5-10 minutes.
 * (Model) parameters should be known or obtainable.
 * Could have this if it does not affect anything else:
 * Response should reasonably accurately follow the concentration (after a delay).
 * Won't have this now but would like in the future:
 * Expand to other metabolisms.
 * Produce something more useful than a colour.

Design
Our bacterium would have the following components:


 * The PTS system, which is already present in ...
 * A promoter that hooks into the PTS system. Can be either:
 * Constitutively active, but repressed by CcpA.
 * Activated by CcpA.
 * Product behind promoter as response.
 * Degradation-tag attached to the product to increase turnover rate.

Modelling
First of all we are trying to model the PTS system as is. We're looking at the following existing models:


 * A limited model consisting of only 9 substances and 7 reactions, as described in Neves1999.
 * A more complete model by M.H.N. Hoefnagel, J. Hugenholtz and J.L. Snoep (hoefnagel2), discussed in Hoefnagel2002.

Our attempts at using these models are available in our SVN repository.

Once we have a model of the PTS system we will try adding one or more "reactions" for the HPr phosphorylation/dephosphorylation under the influence of FBP and ATP, as this leads to HPr-ser46, which lets Ccpa bind to a cre site more easily. For the moment it is assumed that HPr and Ccpa are available in abundance.