Getting standard fluorescent curves is trickier than getting standard absorbance curves. This is because fluorescence is measured in relative units (relative fluorescence units = RFU). RFU values depend on the concentration of the sample but also the gain on the machine. Higher gains yield higher RFU values for the same sample concentration, and vice versa. The detector in the Tecan must have an optimal linear response ending at ~ 40000 RFU. Therefore, you always want the highest sample concentration to register ~ 40000 RFU. Ideally, sample concentrations less than this concentration will fall linearly to 0. This is a protocol that I developed to find the linear region of fluorescence samples by adjusting the gain to lower sample dilutions. It is based on this experiment.
1. Take your fluorescent compound of interest and find its excitation and emission wavelengths, if not known
1a. Do an absorbance scan
1b. Set excitation wavelength ~10 nm lower than max ABS
1c. Scan emission starting at ~30 nm greater than the excitation wavelength
1d. Note the wavelength of maximum emission
1e. Scan excitation starting ~50 nm lower than the max ABS and ending ~10 nm lower than the emission wavelength
1f. Emission should be set to wavelength of maximum emission + 20nm.
1g. Hopefully excitation and emission wavelengths match
1h. ALWAYS keep excitation and emission wavelengths at least 30nm from each other! This accounts for bandwidth overlap.
2. Make a serial dilution (e.g., 1:10) of the fluorescent compound starting at a very high concentration (say, a few orders of magnitude greater than what you need)
3. Do a fluorescence intensity measurement with the gain set to the maximum sample concentration (this is 1 of 3 options for setting the gain)
4. After the scan is done, ensure that the gain is >60. The PMT is not accurate if the gain is less than 60! Also, note that the sample concentration on which the gain was calculated was set to ~ 40000 RFU.
5. Plot the standard curve, it likely has a linear region and a plateau because the gain is too low
6. Redo the fluorescence intensity measurement, but this time, set the gain to the next highest sample concentration
7. Repeat steps 4-6. The plateau of the standard curve should be disappearing and the concentrations should be getting more linear. For example, in this graph, the plateau region is disappearing and the curve is becoming more linear (from maroon to green to purple to sky blue). As you adjust the gain to less concentrated samples (thus increasing the gain), some of the more concentrated samples may be too fluorescent for the detector and read as overflow. That’s OK.
8. At some point, the well from which the gain was calculated will read ~ 40000 RFU and all less concentrated wells will fall linearly from that well. This is the ideal gain. In the above graph, this happened when I set the gain to a well containing 0.000125% sodium fluorescein.
9. If you want to detect lower fluorescent signals, set the gain to an even less concentrated well.
10. Remember this gain. In subsequent experiments, you can manually set the gain to this value and eliminate the need for a standard curve with each experiment.
11. Beware of irregularities in clear multiwell plates. If you weren’t wearing gloves, some of the wells might be partially covered with fingerprints. I’ve also noticed irregularities in the polystyrene well bottoms. For example, the curve belowwas obtained with the same solutions as the curve above. Of course, this could be due to pipetting errors, but also check your plates.
