IGEM:Harvard/2007/Two Component Systems

=FecA Two Component System=

In the beginning, our team decided to use cells as biosensors, used to bind to targets, like breast cancer cells. We were later compelled to up the ante. So, we directed our thinking toward a 2 component system where E Coli would bind to a target and then produce a reporting signal. From this came the search for E Coli outer membrane receptors that are already part of a signally pathway, trying to do as little signal pathway re-engineering as possible. There are very few outer membrane receptors that suit our purpose. In fact, there is only one that we could find: FecA. Using the FecA receptor from the outer membrane of Escherichia Coli, we hope to bind to given targets and produce a reporting signal. We originally planned to insert a random peptide library into the FecA protein and see which n-mer binds to our target. And this receptor-ligand binding should set off the FecA signalling pathway. The concerns presented are:

1) Where should the random library be inserted? FecA has several loops which serve as potential sites for insertion.  Literature suggests that the conformational changes of loops 7 and 8 are most critical to binding and signal production.  So then, would it be best to insert a library into these loops?  Would it be best to use several locations and several libraries at once to get the correct response?

2) Random library insertion is a game of chance, given the number of possibilities that can be produced and biases in different methods of producing random libraries. And how does the insertion of new peptides effect specificity? THUS....

3) Computational methods of predicting which sequences produce the desired binding and signal are attractive, but are at this point unknown to us. These methods (IPRO, CHARMM, dead end elimination) have not been attempted with FecA, as far as we can tell.  They are also fairly new methods.  Our advantage is that FecA is well characterized liganded and unliganded, which is much easier than creating a protein from scratch that will bind to a particular target.  In this vein, we have contacted several researchers who have produced papers on computational design of proteins, receptors in particular.

Brainstorming
6/28/07

Given the complexity of reengineering a receptor binding site such that it binds with a target other than its wild type ligand, we could create a "negative gate." The target will block ferric citrate from binding, thus turning off the FecA signal. Loops 9 and 10 look promising because they are accessible and do not participate in the binding of ferric citrate (they do not contain binding residues). Loop 11 would be another choice except that it has binding residues.

7/6/07

From convo between George K. and Shaunak -Use IPRO for binding FecA to a small molecule or ion, can make a biosensor. Small molecule or metal ion should be simple, biologically relevant, and have no existing biosensors for it (something new). Shouldn't be too hard for IPRO to handle. - Novel: bind FecA to a small polypeptide. Need small polypeptide ligand that has been crystallized bound to another protein before for "active site transplantation" (different from IPRO). It should be biologically relevant, but need structure bound to another protein. -They would like us to test FecA against chemically similar molecules to give them a baseline etc.

Completed Work
See Two Component System Protocol for completed work.

Readings

 * 1) FecAPorins1 pmid=10931321
 * 2) FecAPorins2 pmid=9368371
 * 3) FecAPorins3 pmid=12948769
 * 4) FecAPorins4 pmid=16718597
 * 5) FecAPorins5 pmid=17197416
 * 6) FecAPorins6 pmid=11872840
 * 7) FecAPorins7 pmid=15292131
 * 8) FecAPorins8 pmid=12948487
 * 9) FecAPorins9 pmid=16313612
 * 10) FecAPorins10 pmid=16923915
 * 11) FecAPorins11 pmid=16177795
 * 12) FecAPorins12 pmid=12736688
 * 13) FecAPorins13 pmid=14500902
 * 14) FecAPorins14 pmid=11243829
 * 15) FecAPorins15 pmid=17208966
 * 16) FecAPorins16 pmid=16513775
 * 17) FecAPorins17 pmid=9826510