Lambert Instruments

Using FLIM as a technique to detect e.g. protein interactions inside cells, or dynamics of certain proteins is a very sensitive way of performing experiments. Lifetimes are quite sensitive to many factors, which should be considered carefully.

Concentration
Generally, the decrease in CFP fluorescence intensity and lifetime at intracellular concentrations approaching the mM regime is reminiscent of the famous, although poorly understood, “self-quenching” or “concentration quenching” phenomenon. For diluted, non-interacting chromophores, fluorescence lifetimes should be independent of the chromophore concentration, and this is actually one of the expected major advantages of FLIM techniques for the quantification of FRET. [1]
However, significant concentration-dependent quenching of the CFP fluorescence lifetime was observed in cells expressing simultaneously the YFP acceptor. This probably reflects primarily intermolecular energy transfer between unlinked CFPs and YFPs. But also in the absence of YFP acceptors this quenching was observed, which might be attributed to pseudo-homo-FRET. [1]
Homo-FRET may lead to significant perturbations of the measured fluorescence decay, if different sub-populations of the chromophore are associated to slightly shifted emission spectra, which we know is actually the case for purified CFP. Therefore, it is called pseudo-homo-FRET. In this situation, energy transfer will preferably take place from the blue emitting sub-populations, and will lead to a specific decrease in the apparent fluorescent lifetimes determined on the blue edge of the emission spectrum (<495nm emission). This decrease in fluorescence lifetime will be maximum if the collected fraction of the emission spectrum does not include any contribution from the “red” acceptors emission. [1]

Photobleaching
Photobleaching the acceptor is a method widely used for quantifying energy transfer in intensity-based FRET imaging. However, in FLIM-based FRET imaging it is very hard to obtain a full recovery of the donor lifetime only, after photobleaching the acceptor. In next example, the CFP-YFP FRET pair was used; the lifetime of CFP only was 2,5ns and the lifetime of CFP in presence of YFP was1,68ns. After 98% photodestruction of YFP, the lifetime of CFP only increased up to 1,83ns. However, irradiating YFP in the absence of CFP lead to a significant increase in the fluorescence detected in the “CFP” detection channel (up to 28% above background after an 80% photobleaching of YFP). Therefore, it seemed that some photoproducts were created during YFP irradiation, which contributed to fluorescence in the same wavelength ranges than CFP, leading to a contamination of the CFP signal. [1]

Autofluorescence
Media and cells can contain substances that are autofluorescent. Examples are NADH, riboflavins, collagen, lipofuscins [5]. This means that those substances can be excited at the same wavelength as used to excite the fluorophore of interest. Every substance that is excited and emits photons (thus fluorescence) has a lifetime. The emission of the autofluorescence material mixes with the emission of the fluorophore of interest. If it is not filtered out, it will influence the measured lifetime. Therefore, FLIM users generally use phosphate buffered saline (PBS) as medium for the lifetime measurements. One can also use phenol red-free medium.

Fixation
It would not be surprising if fixation does affect fluorescence lifetimes. The important thing is that you need to use the appropriate controls for your experiments. If you are looking for FRET in fixed cells then you need to have donor only and donor + acceptor cells treated under the same conditions. [3]
Various labsuse paraformaldehyde (PFA) fixation protocols, methanol fixation and combined methanol/PFA fixation protocols; which protocol to use is mostly determined upon the protein of interest. You should check that the protein localisation is not perturbed by the fixation method and also that the fluorescence is not completely diminished. [3]

More
There are more factors that should be taken into account. The work of Klaus Suhling [4] shows that the refractive index of the medium can influence the fluorescence lifetime of GFP. Also GFP expressed alone can have a different lifetime to when it is part of a protein construct.
Even the temperature has severe influence on lifetime measurements. Therefore, the temperature of the microscope stage should be kept as stable as possible, e.g. by use of a climate control chamber.

Reference list