In the Fifth International InterLab Measurement Study, the Measurement Committee tries to remove another source of variability in our measurements: the number of cells in a sample. Usually, in order to determine the mean expression level of GFP per cell (to divide the total fluorescence by the number of cells), we compute the “optical density (OD)” of the sample as a plausible approximation of the number of cells. However, OD measurements are highly variable between labs and it is unclear how good of an approximation it actually is. Therefore, a more direct method is needed, more specifically:
1) fluorescence measurements by the normalization to absolute cell count
2) fluorescence measurements by the normalization to colony-forming units (CFUs)
Before these cell measurements, three sets of unit calibration measurements are required: an OD600 reference point, a particle standard curve, and a fluorescein standard curve. For all of these calibration measurements, we used POLARstar Omega Microplate Readers from BMG LABTECH, with the same plates (Corning 96-Well Clear Bottom Black), the same volumes and settings (e.g., filters or excitation and emission wavelengths) as in the later cell-based assays. This includes all settings that affect the amplitude of the signal collected: filters settings; slit widths (30 nm); gain settings (62), measurement from the top.
Calibration 1: OD600 reference point
As is required in this year’s protocol, we used LUDOX CL-X (45% colloidal silica suspension) as the single point reference. This is to say, using the value for 100μL of LUDOX CL-X in a well of a standard 96-well flat-bottom black with clear bottom plate (0.063) as the reference, any Abs600 measurement from a plate reader can be transformed into the corresponding comparable OD600 measurement.
Table 1 is from the OD600 Reference Point tab of our InterLab Excel sheet. All cell density readings hereafter using this instrument with the same settings and volume can be converted to OD600 by multiplying by (in our case) 1.775.
Calibration 2: Particle standard curve
Measurements here allow us to construct a standard curve where an Abs600 measurement can be converted into an estimated number of cells. This is possible by the preparation of a Microsphere Stock Solution, which contains known amount of particles per volume. After the serial dilution and then Abs600 measurements of the Microspheres, we have Figure 1. According to our measurements, the resulting relationship (log scale) is not exactly a straight line with a slope of 1:1. The problem could be from two: on the one hand, imprecise pipetting during the serial dilution leads to a consistent difference between what we thought and what the volumes actually were. On the other hand, when concentration of the Microsphere gets too low, the Abs600 value also gets too small, exceeding the effective measuring range, which renders the last several sets of data senseless.
Figure 1. Standard curves generated by measuring the Abs600 of serial dilutions of monodisperse silica microspheres (μL) to demonstrate a linear relationship.
Calibration 3: Fluorescein standard curve
In order to compare fluorescence output of test devices between teams, it is necessary for us all teams to each create a standard fluorescence curve. Much like during Calibration 2, we prepare a dilution series of fluorescein. After measuring these fluorescence in our plate reader, we generate a standard curve of fluorescence for fluorescein concentration (Figure 2). This curve allows us to convert our later cell based readings to an equivalent fluorescein concentration.
Figure 2. Standard curves generated by measuring the flourescence of serial dilutions of fluorescein stock (μM) to demonstrate a linear relationship.
For measuring GFP, we have the filter that has the bandpass width at 30 nm, excitation 485 nm and emission 520 nm.
Fluorescence measurements by normalizing to absolute cell count
Now we understand the measurement process and therefore are able to proceed the cell measurements.
Figure 3. Workflow as demonstrated in Plate Reader and CFU Protocol
As shown in Figure 3, cell measurements are more complicated than Calibrations. First we transform a certain strain of Escherichia coli with respectively 8 different plasmids (including Test Devices 1-6, positive and negative control). Test Devices 1-6 and the positive control are all GFP expressing constitutive devices but with different promoters. From each of the transformation, we pick 2 colonies and inoculate in new liquid culture media. After a certain time of growth, we dilute the cultures to a target Abs600 of 0.02, 500µL of which each being a sample. At both time point 0 hours and 6 hours, we take such a sample and measure its Abs600 and fluorescence. This is a total of 8×2 eppendorf tubes per time point. Thus we have the Raw Readings needed for submission.
Fluorescence measurements by normalizing to CFUs
This procedure is quite easy. We grow two Positive Control (BBa_I20270) cultures and two Negative Control (BBa_R0040) cultures, prepare samples with certain OD600 and measure their CFUs.
This certain OD600 is through dilution of our overnight cultures. Within the linear range, first we make sure of a “Starting Sample” with a 0.1 OD600 measurement. Then we gradiently dilute this Starting Sample and inoculate with the last three dilutions. After such dilution we can visibly count the CFU and subsequently calculate out the supposed CFU/mL in the Starting Sample (OD600= 0.1), hence the normalization of fluorescence measurements to E.coli CFUs.
Figure 4. Gradient Dilution and The Spread Plate method as demonstrated in Plate Reader and CFU Protocol
We thank the iGEM Measurement Committee for providing us with this wonderful opportunity to be part of such an international effort. We look forward to seeing the results after data from all participating teams are systematically analyzed.