InterLab Measurement Study

Reproducibility is one of the most crucial parts of reliable research. However, labs are not equal and reproducing data obtained in one lab might not be achieved in the same way and extent in another lab around the world. The iGEM 2015 Measurement Interlab Study addresses the question of how consistent research results are in a global context. All participating teams in the Measurement Interlab Study have been given the same task to construct three promoter-GFP devices and measure their fluorescence.

We assembled all three constructs in the high-copy number plasmid pSB1C3 and transformed E. coli TOP10 cells to express GFP signal. The only difference between the constructs is the promoter used to drive the GFP expression. We measured the GFP signal at an OD of approximately 0.6 in a 96-well plate. The obtained results showed a broad variability of signals within one construct, most probably due to impurities on the colony plate. However, we can see that the device 1 consisting of the biobricks J23101 and I13504 are expressing by far the highest signal. We also see that device 2 (J23106 + I13504) show the second strongest signal, even though only twice as strong as device 3 (J23117 + I13504) which we would have expected to be 10-fold stronger when comparing with data given in the iGEM registry and the strength of each single promoter.

Experimental design

For this study, we were supposed to use four different biobricks and the three promoters (J23101, J23106 and J23117), and each one of them should be combined with a GFP sequence downstream of the promoter. For each single part we extracted purified biobricks from the distribution kit. Using the 3A assembly method (following the iGEM protocol), we constructed all three devices in the high-copy number plasmid pSB1C3 and transformed them into E. coli TOP10 cells.

Each colony selected has been prior analyzed by gel digestion with AvrII and PstI to confirm the presence of the promoter and by the fragment size the presence of the GFP in the target plasmid. After confirming the strains, we restreaked them on fresh LB-agar plates and cultured them until measurement.

After constructing all devices, we grew them in severated E-flasks containing LB medium and chloramphenicol. We grew them to an OD of approximately 0.6 and measured their fluorescence.

As a control we used pure LB-medium to subtract the fluorescence coming from the medium. In retrospective, it would have been also valuable to have a proper non-fluorescent bacteria as control.

Materials and Methods

Constructs and Strains

All transformations for this study have been performed in E. coli TOP10 cells. All biobricks were isolated from the DNA Distribution Kit and transformed in order to amplify the plasmids which contain the genetic elements necessary for our device assembly. The chemocompetent cells were consequently transformed by Heat-Shocking at 42 °C using a 5-x KCM buffer for competence and transformation efficiency.

The amplified plasmids were then recovered using the QIAprep Spin Miniprep Kit (QIAGEN). In a next step, we followed the 3A assembly protocol suggested from iGEM in order to create the devices. Therefore, we cut the promotor parts (J23101, J23106, J23117) with EcoRI and SpeI whereas the GFP sequence (I13504) has been cut with XbaI and PstI. After ligation, we used 2 µL of the ligation product for transformation. All constructs were plated on chloramphenicol plates.

Figure 1: Schematic overview of the devices in pSB1C3 backbones for the InterLab Study 2015.

Subsequently, colonies were picked at the next days and analysed by digestion with AvrII, a promoter specific restriction enzyme, and PstI. As the GFP sequence is rather long with approximately 900 bp, we can see by size of the fragments whether GFP has been inserted. Furthermore, the specific cleavage of AvrII within the promoter sequence will give us indication of the presence of the promoter. Sequencing could unfortunately not be conducted for any biological clone.

Inoculation and Cultivation

Strains were cultured in 250 mL E-flasks filled with 10 mL LB medium and the corresponding antibiotic (here: Chloramphenicol 25 µg / mL taken from a 1000-x stock solution). The cultures were kept at 37 °C and incubated at 150 rpm.

Pre-cultures were taken from LB-agar plates and incubated overnight at 37 °C under permanent agitation. Using a photospectrometer, the OD was permanently measured.

GFP measurement

All strains were picked from LB-agar plates and left for overnight in separate 250 mL E-flask in order to let them grow until they reach an OD > 1. We diluted the sample then to an OD of approximately 0.6. For the OD measurements, an Eppendorf BioPhotometer was used. When the sample has been diluted to the right OD, we harvest cells by centrifugation of 1 mL of culture for 5 min, 13000 rpm. We removed the supernatant and resuspended the pellet was subsequently resuspended in 1-x PBS. Each sample was applied in technical and biological triplicates (expect of device 3 where we could find only one positive clone after cloning). The fluorescence was measured using a Clariostar (BMG LABTECH) and analysed using MARS Data Analysis software.

Statistical Analysis

Signal was corrected against the medium blank and then normalized by the corresponding OD. For each technical triplicate an average has been calculated. Subsequently, we took the mean of the three means describing the three biological replicates (for device 1 and 2; one biological replicate for device 3) in order to get a value for each construct. The standard deviation has been calculated from all three means representing the biological triplicate. Statistical significance was calculated by Student's T-Test.

Results and Discussion

As we could not unfortunately sequence all our constructs, we verified our assembly by digestion using the promoter-specific AvrII and PstI. This digestion gave us a two fragment of a size of 920 bp and 2068 bp. To our own pity, one of our digestion results have not been saved, in place we show the fluorescence expressed on the plate as a proof of a functioning construct.

Figure 2: Verification of Promoter-GFP constructs (A, B, C) using PstI and AvrII, as well as verification of GFP expression in replicate 2 and 3 of Device 2 (B).

We have seen the proper sizes for all constructs tested. However due to time constraints, we were unable to find the last two replicates for the device 3. After 16-hours incubation, the GFP signal was measured and normalized by the sample OD.

As displayed in the graph below, we see quite some strong fluorescence coming from GFP which is however mostly due to one biological replicate explaining the huge standard error of this bar. Device 2 about two-times as strong as device 3 and therefore the order or devices is following their pre-described strengths. In comparison to the documented promoter strengths, the difference between device 2 and 3 is still too small as a ten-fold difference would have been expected. One of the reasons for this small difference might be that device 3 contains only one data point as only one biological replicate could be produced. Due to huge error bars, it is however impossible to conclude from this data.

Figure 3: GFP Fluorescence from TOP10 E. coli transformed with either device 1, device 2 or device 3.

In order to produce a more reliable data set, we would have needed to re-investigate our strains used and double check their purity. Furthermore, we would need to acquire the last two replicates for device 3 and redo our experiments.