Hey Pascal,

One important reason for these corrections is that for inter-molecular FRET pairs there is no guarantee that the stochiometry of the donor-to-acceptor remains the same for the two different timepoints in your region-of-interest. For instance, let's say due to the dynamic nature of cytoplasm if a fraction of acceptor molecules move from other part of the cell to your ROI. Then you will see an increased acceptor emission even without a FRET, just due to cross-excitation. In such a scenario, the way to get an idea whether the increased emission in acceptor channel is due to FRET or stochiometric change is to do an acceptor-excitation/acceptor-emission correction. Therefore, corrections could be critical for inter-molecular FRET measurements.

Certainly, you may not face such issues if you are using an intra-molecular FRET construct, such a kinase biosensors. Therefore, if you are using an intra-molecular FRET construct then you are fine without many of the corrections.

Best regards,

AN

Hey Anu,

thanks for your advice. I'm going to detect cAMP changes with a regular CFP-EPAC-YFP biosensor. I did FRET-FLIM before and moved somewhere else where I can't use FLIM detectors, but to bridge the time for now I looking for a method to still can bring up a statement of the distribution pattern of cAMP, so I'm kind of new to FRET experiments. So basically, the shift will happen in any case. About the acceptor-exicitation/acceptor-emission correction, how do I perform it? This correction is including the acceptor and donor-only ones, isn't it?

Pascal

Hi Pascal,

All the correction for the cross-excitation and bleedthrough would actually require the coefficient measured from donor-only and accpetor-only cases. However, if you are planning to use the EPAC-based biosensor then I would say that corrections are not that critical. Since, you will use EPAC biosensor you do not have to much worry about the stochiometry changes because even if additional EPAC molecules move into your ROI the relative ratio (stochiometry) of donor and acceptor will remain the same, as every molecule of sensor contains both the fluorophores.

I have used EPAC-based sensors but always without any corrections. As I understand, in our case these corrections never made any big difference in the relative signal changes. These corrections will change the absolute FRET-ratio values but the normalized values relative to appropriate Rmin and Rmax remain more or less similar as the uncorrected normalized numbers. However, if the signal is too week in your preparation then consider background correction.

AN

For your EPAC-biosensor experimental setup: How did it look like then? You calibrated each single channel (YFP/CFP) with the same laser gain/pinhole settings, or did you calibrated somehow to compensate brightness differences that occur in e.g. the YFP/CFP pair of 1:4? Did you do post modifications like the correction in case of channel shift etc. and analyzation (including normalization) with ImageJ? Or did you only imaged each single channel and did the after-math afterwards?

I really appreciate your help,

Pascal

In our lab, we usually kept the parameters for illumination such that the brightness is compensated for the two channels. It seems that the normalized ratio values have a good tolerance for several experimental parameter variations. In our case, we were also doing a rough analysis along with the experiments real-time to be informed about any drift and focus changes etc. This particularly helps when one does long live-imaging to know whether the drugs, perfusion etc are working and the image field is stable. But if you have only two time points to measure then you could very well do the analysis afterwards.

AN

That sounds reasonable. Another thought would be, when I’m doing shift correction and only going for the absolute FRET values (that are not comparable between different samples but also didn’t need to), I can say about only one time point that we have different molecule concentration at different places in the cell, by showing the different FRET-ratios at different points. Or did I can’t do this statement?

Pascal