Endy:Translation demand/Current status

8/11/06
After 3 runs of the primary experiment (GFP in MG1655) and 2 runs of the "Voltmeter", I have successfully been able to place a range of demands on the cell as evidenced by the clear and consistent levels of fluorescent expression from each RBS, but I have not been able to establish a clear pattern of growth rates. I think this could be due to the existence subpopulations with lesser or no plasmids. Regarding the Voltmeter, it seems that the strain I'm using, C86 (derived from X90), cannot repress expression of mCherry. Our friends at the Sauer Lab have suggested that this may be due to the loss of an F Plasmid expressing the repressor. I'll be looking into these two problems over the next few weeks.

Cells expressing Ampicillin resistance do so by excreting Beta-Lactamase into the extracellular environment to destroy the antibiotic. This allows one cell to "protect" another by creating a pocket free of antibiotic. When one combines the growth advantage given to plasmid free cells with this abilitiy to survive without providing resistance for itself, the possibilty of losing the plasimd generationally becomes stronger. In order to test for plasmid instability, I've grown cultures and plated them in media both with and without antibiotics. The idea is that if subpopulations of cells within the culture have lost the plasmid, they will grow only on the plate without antibiotics. In my first run of this plating experiment, I plated high copy mCherry constructs in C86, a strain of bacteria with a GFP chromosomal insert. The result was as expected; the constructs with stronger RBS's grew more colonies on the LB plate (no antibiotic) than the Ampicillin plate meaning that conditions were present to select for plasmid free populations. A disparity in the number of colonies became indistinguishable in the mid and low strength RBS's meaning perhaps that the demand placed on the cell was too weak to cause this negative selection. When this experiment was repeated with high copy GFP constructs in MG1655, there seemed to be little or no difference in the number of colonies suggesting a more stable plasmid situation.

Though plasmid loss is perhaps the most striking manifestation of instability, there are other, more subtle forms. The mechanism of plasmid segregration in bacteria isn't clearly undestood. Kristala Jones Prather has suggested that especially when using high copy vectors (100 - 300 copies/cell), it is possible that there is an asymmeteric segregation of plasmids during cell division. If daughter cells are receiving disproportionate amounts of the plasmid, the variation in copy number will produce too many unique subpopulations to pin down a pattern in growth rates for a particular construct. I'll be reading some literature on the mechanism of plasmid segregation to see to what degree this phenomenon may be affecting my data. If it proves to be significant, I'll look into reducing the variability of copy number in a particular cell while somehow maintaining the range of demands allowed by using the high copy vector.

In the mean time, I'm going to make two major changes to my experiment to see if I can get more consistent growth rate data. I'm going to move my GFP and mCherry scaffolds to a mid copy vector with Kanamycin resistance, pSB3K3. It's my hope that by moving to a mid-copy vector, the variability in plasmid segregation will be much reduced, but protein yield will be enough to produce a range of growth hits on the cell. As far as I understand, Kanamycin works intracellularly (not excreted into the media), thereby eliminating the potential for plasmid free colonies to survive by clustering around KanR+ subpopulations. I hope to run these constructs in the plate reader within the next week or so.

As for the Voltmeter, the Sauer Lab has given us the parent strain of C86 (X90) lacking the F Plasmid. If the Voltmeter's inability to repress my constructs is due to the loss of the plasmid, I would expect that this X90 F- strain would demonstrate the same phenotype. After transforming the X90 F- with my mCherry constructs, I grew cultures and induced half with IPTG and not the other. I found that both inuduced and uninduced cultures expressed mCherry at the same level (qualitative observation), but that this level isn't nearly as high as the mCherry expression observed in the Voltmeter. I'd like to run the plate reader in the near future with X90, X90 F-, C86, and perhaps MG1655 expressing high copy mCherry constructs to see how the expression levels quantitatively compare.

To Dos

 * 1) Test for plasmid instability
 * 2) *Grow cultures used for plate reader expts. 1-3 and plate them on Amp and on LB plates.
 * 3) *Use the same growth protocol as used for the full experiments.
 * 4) *Do this for the 5 RBSs.
 * 5) Select non-pink colonies from the voltmeter plates and see if some of those have the F-plasmid.
 * 6) Compare mCherry expression levels in X90, X90 F-, and C86 in a plate reader run.
 * 7) *Modify the plate reader protocol to have lesser time between repeats.
 * 8) *See if the Voltmeter was unable to repress the mCherry constructs on low copy.
 * 9) Construct mCherry/GFP-pSB3K3 constructs.