Fourth iGEM InterLab Study - 2017
Standardisation is one of the key principles in engineering and one of the major aims of synthetic biology. However, reliable comparison of fluorescence data from laboratories across the globe remains a problem to be solved. The reason for this problem is not only to be found in the way fluorescence is measured but also how the data is processed. Therefore, researchers often try to approach an interlab comparison by looking at the relative expression.
In order to improve scientific reproducibility and comparability of data the “Fourth iGEM Interlab Study” aims to tackle this issue by yielding standardised and comparable absolute values for green fluorescent protein (GFP) measurements in a plate reader. Therefore, researchers have been provided with the same detailed protocol and explicit data analysis form.
To achieve the aims of the InterLab Study, the experimental setup consisted of three consecutive experiments.
First, we determined the reference point of optical density at 600nm (OD600) by calculating the ratiometric conversion factor based on the absorbance of LUDOX-S40 (provided by iGEM) at 600nm. Secondly, we created and plotted a fluorescein fluorescence standard curve to obtain a fluorescence reference value. This was done by measuring the fluorescence intensity of a 1:2 serial dilution of fluorescein. Finally, we measured the test devices (see figure 1) containing GFP under the control of different promoter combinations. For this, we transformed each individual test device into E.coli DH5α. Thereafter, two colonies per test device were picked for further cultivation. After overnight incubation, the cultures were diluted to a target OD600 of 0.02. Then, OD600 and fluorescence intensity were measured after 0, 2, 4 and 6 hours of the OD600 normalisation step. By integrating all generated results, we could calculate a standardised fluorescence to OD600 ratio.
Materials and Methods
Constructs and Strains
Six devices were tested (BBa_J364000, BBa_J364001, BBa_J364002, BBa_J364003, BBa_J364004, BBa_J364005) with constitutive expression of GFP, only differing in their promoter (J23101, J34106, J23117) and bicistronic design element (B0034, J364100) variants. Whereas GFP is also expressed constitutively in the positive control (BBa_I20270), the negative control (BBa_R0040) is a sequence for pTet inverting regulator under the control of a constitutively active promoter. All devices were delivered in high copy number pSB1C plasmid backbone and transformed into E.coli DH5α.
OD600 and fluorescence measurement
OD600 and fluorescence were measured with FLUOstar OMEGA plate reader (BMG Labtech) in black 96-well plates with flat and clear bottom (Costar96, Corning). To achieve consistent plate reader measurements, the determined plate reader settings from the OD600 calibration and the fluorescein fluorescence measurement were not changed for the subsequent cell measurement. The exact instrument settings can be found in the appendix.
Calibration - OD600 Reference point
Absorbance of LUDOX-S40 (provided by iGEM) and H2O was measured in four technical replicates with plate reader at 600nm. LUDOX-S40 absorbance 600 values (Abs600) were corrected against the H2O blank. Then, the ratiometric conversion factor was calculated by dividing the corrected Abs600 by the reference OD600 value (provided by iGEM).
Fluorescein fluorescence standard curve
A standard curve measurement was performed through 1:2 serial dilution in four technical replicates starting from 50uM fluorescein (provided by iGEM).
The positive and negative control as well as the six test devices (provided by iGEM) were transformed into E.coli DH5α (New England Biolabs) and grown overnight at 37°C. From each plate, two colonies were picked and inoculated in Luria Bertani (LB) media (5mL) + Chloramphenicol (25ug/ml) at 37°C for 18 hours at 220 rpm. After the incubation, OD600 was measured in four replicates with plate reader to dilute the cultures in Luria Bertani (LB) media (12mL) + Chloramphenicol (25ug/ml) to a target OD600 value of 0.02. Cultures were then again incubated at 37°C at 220 rpm. Thereafter, OD600 and fluorescence were measured for each colony (n=2) in four technical replicates at time point 0, 2, 4 and 6 hours.
Measurement data was collected following the instructions from iGEM HQ which prepared and provided an excel sheet containing the calculations for the data evaluation and statistical analysis (file can be found in appendix). Graphs were made in GraphPad Prism 7.
Results and Discussion
The first step of the InterLab Study was to determine the OD600 reference point. As shown in figure 1, we calculated the corrected Abs600 by subtracting the H2O from the LUDOX-S40 absorbance measurement data (=0.015). As a result, the ratiometric conversion factor (refOD600/Abs600) was 2.881. In a later step, this factor was used to standardise the OD600 values of the test devices transformed into E.coli DH5α.
The second step of the InterLab Study was to determine a fluorescein fluorescence standard curve. To plot the graph, the mean values of the replicates at each concentration were taken. We expected a straight line on linear scale which is confirmed by linear regression R² = 0.984. From the linear regression, the mean of medium-high fluorescence levels (between 1.5625 - 25 µM fluorescein) = 0.000243 µM fluorescein / AU was calculated. These values of medium-high level points were chosen (instructions by iGEM) for final scaling, since they are likely to be less impacted by saturation or pipetting errors . This value will be used to further standardise the GFP fluorescence measurements of the test devices transformed into E.coli DH5α.
The last step of the InterLab Study is to measure OD600 and fluorescence of each test device (inclusive positive and negative control) transformed into E.coli DH5α. After transformation, two colonies of each device were cultured overnight. Lastly, cultures were normalised to an OD600 = 0.02 and then OD600 and fluorescence were measured at time points 0, 2, 4 and 6 hours after normalisation.
By measuring OD600, we gain information about the bacterial growth. Although figure 3 shows an “exponential-like” bacterial growth for almost devices, both colonies of device 1 grew very little. Furthermore, large differences in bacterial growth were not only remarkable between colonies carrying the same device but also between the different devices. Since measurement evaluation of only two colonies is very vulnerable to high variability and outliers, we suggest triplicates for future InterLab Studies to improve significance and robustness.
We gain information about the promoter strength represented in the expression of GFP by measuring its fluorescence (figure 4). The negative control (a pTet inverting regulator) had the lowest GFP expression in both colonies. Test device 2 and 4 displayed the strongest GFP expression. Unlike the other devices, test device 1 has not only a linear increase in GFP expression, but also has a much higher fluorescence value at time point 0h. Interestingly, when comparing OD600 with fluorescence, we observed a tendency of inverse proportionality. The more the bacteria grew, the lower was the GFP expression. This finding could argue that GFP expression has an impact on the cell viability. With this in mind, the high fluorescence levels of test device 1 could explain its little growth. Furthermore, the positive control expressed GFP at different levels for each colony. Although, both colonies grew equally strong, only colony 1 showed an increased GFP expression with time, but not colony 2. The observed variation may be explained by excessive transformation of DNA plasmids into recipient bacteria leading to a heterogenous plasmid numbers in the same batch of transformed bacteria.
Gathering the results from the performed experiments, we calculated a standardised version of the measured data represented as µM Fluorescein/OD600 ratio (figure 5). The ratiometric conversion factor (see figure 1) means that our measured Abs600 is 2.88 times higher than the reference OD600 provided by iGEM.
In figure 2, we determined the fluorescein fluorescence standard curve, revealing a mean value for fluorescein levels which serves as reference (= 2.43E-04 µM fluorescein / AU). This value means that we need 2.43E-04 µM fluorescein per measured fluorescence intensity unit.
Equation 1. Calculation of fluorescence to OD600 standardisation.
Dividing the GFP fluorescence measurement multiplied with the fluorescein reference value (2.43E-04 µM fluorescein / AU) by the OD600 measurement multiplied with the ratiometric conversion factor (2.88) we get the µM Fluorescein/OD600 ratio for each device at each time point. The high ratio of test device 1 can be explained by its linearly increasing fluorescence at high starting level combined with its low growth. Also, the stated discrepancy between the positive control colonies are represented in the fluorescence/OD600 ratio. Furthermore, we saw that device 1 has the highest ratio, followed by device 2, 4 and 5. The ratio for device 3 and device 6 is hardly distinguishable and close to the negative control. Notably, we could only observe a time-dependant increase in the the ratio for test device 4. All the other devices did not have a consistent ratio increase over time.
Taken together, we observed not only a high variability between the colonies but also between the devices. This finding supports the importance of multiple colony testing to obtain more robust data when studying the activity of a biobrick. Thus, we would suggest performing the experiments with three colonies to assure meaningful data. But, since the scope of the InterLab Study was to standardise fluorescence measurements from labs across the globe, we cannot conclude on the study’s success without comparing our data with other participating teams.
- Mutalik, V. K. et al. "Precise and reliable gene expression via standard transcription and translation initiation elements." Nature Methods 10, 354–360 (2013).
- Beal et al. “Reproducibility of Fluorescent Expression from Engineered Biological Constructs in E.coli.” PLoS ONE 11(6): e0157255, (2015)