To truly apply engineering principles to synthetic biology, it is of utmost importance that the DNA parts used perform as expected for all who choose to use them. The goal of the 2017 Interlab study is to determine how reliably known parts can conform to expectations and previous results. We were delighted to be one of the hundreds of teams participating in this years study, where ourselves and teams around the globe quantified GFP expression for various devices, with each device theoretically conferring precise gene expression that should be easily replicated.
Materials and Methods
We performed the Interlab study with no deviation whatsoever from the 2017 Plate Reader protocol. We transformed DH5alpha competent E. coli cells with each device (and controls) from the 2017 distribution kit, took samples every two hours, and then measured fluorescence and absorbance using our BMG FluoStar Plate reader.iGEM Plate Reader Protocol
As stipulated by the protocol, first we used the LUDOX and FITC provided to get standard measurements for both OD600 and Fluorescence. Results were as expected, with LUDOX absorbance values similar to example data, and fluorescein dilution creating a linear curve (when it is taken to account that the maximum reading of our plate reader for fluorescence was 2600, meaning the un-diluted sample produced too much fluorescence to be accurately measured. Data is summarised below.
Colonies of each device had their fluorescence and OD600 Measured every two hours for six hours total. Results are summarised below. Error bars are standard error of the mean.
Some devices produced very noisy measurements, for example device 1, but most tended to stabilise in terms of data output after 6 hours. Once again, device 1 was incredibly noisy at the first two time points measured, but final measurements had less deviation between measurements (as evidenced by smaller error bars) and produced fluorescence output of a similar magnitude to the positive control. Device 3 and Device 6 produced relatively little fluorescence output. A possible reason for this could be that the different RBS sites used in each construct make translation less efficient than the RBS sites used in other devices. These data do seem to imply that the type of RBS used does have an impact on signal to noise ratio. For example, there is minimal noise in Device 2 compared to other devices, and this device also produced strong fluorescent output. It is also possible that expression time has an influence on precision of gene expression, as, for the most part, GFP fluorescence stabilised over time in each device.