As part of iGEM’s five year project by the Measurement Committee to reduce interlaboratory variability and increase repeatability in the quantification of green fluorescent protein, Team St_Andrews participated in the 2018 Interlab Study to meet the criteria for the characterisation and contribution section of the Bronze Medal award.
Despite Green Fluorescent Protein (GFP) being one of the most common reporter proteins in synthetic biology and in most cellular research, the reproducibility of fluorescence quantification proves inefficient. This is because fluorescence values vary with equipment and are often reported in varying forms due to different methods of data processing. Plate Readers, which are commonly used as a high-throughput method of measuring absorbance or fluorescence, also prove problematic as they determine the fluorescence of a total population of cells yet exact cell counts are difficult to determine. Current methods involve using the Abs600 of a sample to calculate the optical density as an approximation of cell count, however this is an inadequate method as it is not a direct cell count, and it is unclear how accurate the approximation is.
To increase the accuracy and precision of fluorescent measurements, and thus their reproducibility, iGEM developed a 3-step calibration process for labs worldwide that could provide a method of standardising fluorescence data by removing variable factors such as equipment, indirect cell counts, or data processing methods. In order to achieve this, the procedure was tested in labs across the world as part of the 2018 Interlab Project to determine if the values would in fact be somewhat ubiquitous. The aim of the overall study was to reduce interlaboratory variability in fluorescence measurements through a standardised procedure involving normalising the data to absolute cell count or colony forming units (CFU) instead of optical density (OD).
Fundamentally, the first step was to create a sterile working environment. This involved ensuring the necessary equipment was autoclaved, the space was neat and disinfected with 70% ethanol, and Bunsen burners were present to circulate the air. Often a sterile fume-cupboard was also utilised when possible.
The CCMB80 buffer was produced according to the iGEM Competent Cells Production Protocol to produce 200ml. Table 1 shows the amounts used for each of the reagents. The initial pH measured 8.56, so 0.1M HCl was used to reduce the pH, despite individual drops the pH decreased rapidly to 5.6 so KOH was added to compensate and bring the pH to the target pH6.4. KOH was used to ensure there was no addition of novel salts. Distilled water was added to produce 200ml before the buffer was filtered through a sterile filter due to a yellow discolouration. An unidentified small amount of brown insoluble precipitate was filtered out.
Table 1: Composition of the CCMB80 Buffer – The table shows the desired concentration of each of the compounds, the molecular weight given from each of the compound containers and the calculated mass of each reactant that was used.
A working stock of DH5α was produced from a frozen seed stock by inoculating 2x10ml LB media with the thawed cells under aseptic conditions. These samples were left shaking overnight at 37oC. A 10µl aliquot of these samples were plated on individual SOB plates which were left overnight at room temperature according to the iGEM Seed Stock Protocol. When observed the following morning however, little had grown so the plates were transferred to their optimum temperature (37oC) for a few hours. From the plates, three further inoculations were produced; 1 colony per 20ml of SOB media, and again the samples were left shaking at 37oC.
From this working stock, 2.5ml was taken to inoculate 250ml SOB media in a 1L flask at 37oC. The OD600 of the sample was taken at regular intervals until it reached 0.3, after which the sample was divided equally into 5 falcon tubes before being centrifuged (4oC, 3000G) for 10 minutes. The supernatant was removed before 8ml of the ice-cold CCMB80 buffer was added to each tube to resuspend allowing for the five tubes to be combined into two equal tubes. 20ml of the buffer was then added to each tube to resuspend before the samples were incubated, centrifuged and 5ml of the buffer was added to each tube. The samples were then combined. A sample of 50µl cells and 200µl LB was measured using Nanodrop resulting in a total sample OD of 2.15. the sample was incubated on ice for 45 minutes before it was aliquoted into 100µl amounts which were immediately frozen using liquid nitrogen before being labelled and frozen at 80oC. These would be used in order to produce the competent cells. 15% Glycerol was added to the remaining working stock before it was aliquoted into labelled 1ml cryotubes, immersed in liquid nitrogen and frozen at 80oC, producing another seed stock.
1ml of Chloramphenicol was made using 0.025g of Chloramphenicol and the appropriate volume of ethanol, and this was used to create the agar plates. The first batch of plates made were somewhat lumpy due to a stubborn cloud of agar that refused to melt, however successive batches were much smoother.
The competency of the cell stock was measured using iGEM’s Competent Cell Test Kit and associated methodology. This involved transforming the cells using a heat shock method with different concentrations of plasmid DNA BBa_JO4450 and calculating the subsequent efficiency of the cells based on the number of colonies on the agar plates. The cells were determined to be in the higher end of the acceptable efficiency range.
Three calibrations were performed to the plate reader was set appropriately, and to allow a comparison for the data to be obtained in the cell measurement stage. Each calibration plate was measured using a SpectraMax M3 Plate Reader. The plate reader had path-length correction disabled and operated on filters of excitation 485nm, and 530nm emission, with a bandpass width of 30nm.
The first was performed with LUDOX CL-X – 45% Colloidal Silica Suspension. This offered a reference point from which a conversion factor could be determined to interconvert the Abs600 obtained by the plate reader into the OD600 Measurement that would be obtained from a spectrophotometer. These readings are not interchangeable as the absorbance of the plate reader is dependent on the fluid depth to determine the path length whilst the spectrophotometer has a fixed path length dictated by the cuvette (typically 1cm). The plate was structured as shown in Diagram 1.
Diagram 1: LUDOX Calibration Plate Structure- where L (red) indicates Ludox, and H2O (blue) indicates distilled H2O. Each well contained 100µl.
The second calibration was done using Silica Microspheres which mimic the size and optical characteristics of E.coli. Thus a particle standard curve could be developed using a known number of particles allowing for the conversion of absorbance results with estimated cell count. A serial dilution of the silica beads was performed in the plate as shown in Diagram 2.
Diagram 2: Silica Microsphere Calibration Plate Structure- where S (purple) indicates Ludox, with gradients of colour indicating the serial dilution of the silica beads. The dilution ratio is given on each well. H2O (blue) indicates distilled H2O. 100µl of water was added to wells 2-12, 200µl of the stock Silica Microsphere solution was added to well 1. 100µl of the stock was pipetted into well 2 and mixed by pipetting gently, before 100µl of the solution was transferred to well 3, and so forth.
The final calibration used Fluorescein to produce a standard fluorescence curve that could be used to normalise the machine-specific values obtained from the Plate Reader in the plasmid studies. GFP, although used within the plasmids, was not used as a calibration tool as GFP is expensive to purify. Fluorescein was used due to its similar optical behaviours. The Fluorescein sample came at a 10x concentration, so was diluted with PBS Buffer to produce the 1x concentration. The plate was structured as shown in Diagram 3.
Diagram 3: Fluorescein Calibration Plate Structure- where F (yellow) indicates Ludox, with gradients of colour indicating the serial dilution of the fluorescein solution. The dilution ratio is given on each well. PBS (grey) indicates PBS Buffer. 100µl of water was added to wells 2-12, 200µl of the 1xFluorescein solution was added to well 1. 100µl of the stock was pipetted into well 2 and mixed by pipetting gently, before 100µl of the solution was transferred to well 3, and so forth.
The Competent DH5α cells made in the Preparation Stage were transformed with the experimental plasmids using a Single Tube Transformation Method for each plasmid. The plates were left to incubate overnight at 37oC. The colonies were screened for the transformations using a QiaPrep Spin Miniprep Kit and colony PCR.
From the resulting transformation plates, two colonies were taken and used to inoculate 2x10ml LB media and Chloramphenicol. The cultures were left to incubate on a shaker overnight at 37oC. The cultures were then diluted ten fold using LB + chloramphenicol media. Using a plate reader, the absorbance of the cultures was determined at 600nm, before the cultures were diluted in order to achieve an Abs600 of 0.02 in a final volume of 12ml. The process was done as quick as possible to limit light exposure of the samples.
A 500µl aliquot of each diluted culture was taken and placed on ice before the remaining cultures were left to incubate for 6hours. Following the six hours, a 500µl aliquot of each culture was taken and again placed on ice. These were then pipetted onto respective plates (0hr and 6hr). The plate structure is shown below in Diagram 4.
Diagram 4: Cell Measurement Plate Structure. Four replicates for each of the two colonies were used. The contents of each well are labelled using the iGEM device name with ‘T.D.’ defining Test Device. Each well contained 100µl.
Using the settings established in the calibrations, the two plates (0hr and 6hr) were measured using a Plate Reader for fluorescence and absorbance.
Colonies from the positive and negative transformation plates were used to inoculate four more LB and Chloramphenicol mediums (two each), and the cultures were left overnight at 37oC. The samples were then diluted eightfold with LB/Cam for the absorbance of the samples to be measured at 600nm using the plate reader. Each well contained 200µl.
Using the values obtained, each solution was diluted to an OD600 of 0.1 in a total 1ml solution before being measured using the plate reader to ensure the values were 0.1 (or approximate). Triplicates of each dilution was used producing 6 measurements for each control (positive and negative).
Each of the OD600 0.1 samples were then used in a dilution series where 100µl of the starting sample was added to a 1900µl LB/Cam solution to produce a twenty-fold dilution. The solution was vortexed to mix before the process was repeated twice to produce two further twenty-fold dilutions. Two tenfold dilutions were then performed to produce a final dilution factor of 8x10-5. The latter three dilutions of dilution factors 8x10-3, 8x10-4, and 8x10-5 respectively were then each plated onto a LB+Cam plate and left to incubate at 37oC overnight. The resultant colonies were then counted allowing for the calculation of colony forming units per 1ml of OD600 = 0.1 Culture (CFU/ml/OD) under the assumption that one bacterial cell produced 1 colony.
After locating the single Flow Cytometer in the University (and some who knew how to use it), and eventually obtaining the SpheroTech Rainbow Calibration Beads, the Flow Cytometer Protocol was attempted.
A 50µl sample of the SpheroTech beads was added to the A10 well of the two plates prepared in the Cell Measurement Protocol. The Flow Cytometer was then used to measure each well at 10,000 events per well.
Results were submitted to the iGEM Measurement Team and were accepted on the 17th August 2018.
While most of the experiment progressed as expected, there were several instances in which issues arose, particularly with regards to the first part of the cell measurement protocol. In the first attempt, the fluorescence values for the positive control after six hours did not differ from the negative control suggesting an issue with the expression of GFP within the positive control cells. Based on the data obtained in the transformation screening and in the colony forming unit’s aspect of the Cell Measurement Protocol, it was known that there was in fact no issue with the expression of the plasmid or in the fluorescence of GFP, and thus the relevant protocol was repeated.
In the second repetition of the Cell Measurement Protocol, time constraints meant that the six hour period was reduced to five. Although this was not expected to alter the results too drastically, further issues were encountered when it came to reading the absorbance values using the plate reader. Every value appeared ubiquitously as 4.00, yet when measured individually using a spectrophotometer the values were all between 0.00 and 0.4. It was suspected that this was due to the use of a slightly different type of plate for the plate reader. Despite this, the flow cytometry results obtained appeared accurate to what was known about the experimental devices. It was however decided that the cell measurement protocol should be repeated again with the familiar, iGEM recommended plate type.
The third repeat was scheduled based on the information garnered in the second repeat, this ensured enough time for the 6hour incubation and time in case anything else went wrong. Both the fluorescence and absorbance results of this repeat appeared consistent with that was expected; the positive control was greater than the negative control, and there were no extreme or unexpected values. The results obtained in the flow cytometry protocol also appeared consistent with the data, however there was unexpected fluorescence in the LB/Cam media. The media should not have shown to fluoresce, suggesting possible contamination despite every effort to prevent contamination of the samples. The fluorescence however was deemed insignificant in comparison to the test devices and thus the data was submitted and accepted.
Through the Interlab Study we learned and developed many new skills and gained a working knowledge of lab equipment such as Flow Cytometers and Plate Readers that we would not have been exposed to in our lab classes. We also had the opportunity to understand the importance of keeping a lab-book and practise this fundamental lab skill. The iGEM Interlab Project was incredibly interesting and enjoyable to be a part of, and the skills we gained throughout our experience will inevitable help us in our university careers and in the future.
With thanks to:
Clarissa Czekster for patiently teaching and guiding us through the techniques.
Fiona Cooke for her time; operating the Flow Cytometer and her invaluable explanation of the output.