Measurement
In our project, for the measurement of our quorum sensing response, we used a system based on a reporter fluorescent protein (EYFP) and a constitutively expressed control one (ECFP). As it would be subject to the same extrinsic variation as the reporter gene, but would not respond to our tested stimuli, it would allow us to use the CFP measurement to normalize YFP. This gives us the possibility of a much more reproducible result, less dependant on growth media and cell density (a rather unreliable parameter to measure) at the time of each measurement. Another great benefit of using this is that dividing two fluorescence measurements gives us an adimensional parameter, that has greater capacity for comparison with other experiments and constructs.
In our measurements of quorum-sensing-responsive promoters activity, we compared the variance of our controls, which should have a constant value for fluorescence due to a constant amount of leakiness, when normalizing by the OD600 value and the CFP fluorescence measurement. We found, in this and other experiments, that the normalization using CFP presented significatively less variance. In this example, we found a variance that represented, when reaching equilibrium value, a percentage of the mean twice as small as that of the normalization by OD.
Another way to raise the precision of our measurements was to not use LB for measuring in the plate reader, as it had high autofluorescence and commonly exibited fluorescence higher than our tests at low CFP values. Also, LB may change its color depending on how old it is and other environmental interactions, further tempering with the measurements. By centrifuging the cells for each sample and resuspending them in molecular grade H2O, we had a blank with much less autofluorescence and less variance between measurements. For kinetic experiments, in which we could not wash the medium in each measurement, we had to use M9 medium supplemented with glucose to obtain a precise fluorescence measurement, although M9 still presented a bit of autofluorescence compared to water, and it did make our time-dinamics go slower, taking up to twice as much as the original amount of time to reach the same response levels.
References
- T.J.Rudge, J.R.Brown, F.Federici, N.Dalchau, A.Phillips, J.W.Ajioka, J.Haseloff. “Characterization of Intrinsic Properties of Promoters” ACS Synthetic Biology 5 (1), 89-98, (2016) doi:10.1021/acssynbio.5b00116