Dear all, I am measuring changes in fluorescence lifetime of my chromophore in cells expressing also GFP and I would need to filter off the GFP signal. After playing with the parameters in confocal microscope, collecting the emission from 620 nm further allowed me to see what I need, almost no GFP and still enough strong signal from my compound. I would like to do the same in FLIM so I bought this filter: 625nm 25mm Dia., High Performance Longpass Filter (https://www.edmundoptics.com/p/625nm-25mm-dia-high-performance-longpass-filter/27815) but it still does not seem to work: the GFP-cells themselves give the same strong signal of GFP whether I use this new filter or the old one (600+-50 nm). As far as I know, the arrow on the filter should point towards the specimen and away from the detector eye but just to be sure I tried both direction - without success. What could be the problem? Is there a technical parameter which I overlooked in the presented filter? Does one typically need to do some changes also in the software? I’m a FLIM beginner so I will really appreciate any suggestion. Thank you very much!
(Our device: Nikon Eclipse Ti A1R microscope upgraded with a FLIM kit from PicoQuant, equipped with a laser 485, PicoQuant, LDH-D-C-485 at 20 MHz)
What is your chromophore? From the 625 detection I guess it is further red than GFP. If yes adding a lightsource with a longer wavelength would be the best option to not even excite the GFP.
Otherwise getting completely rid of the GFP signal will not be feasible, you can just optimize the ratio from your signal and the GFP signal. For this calculate the ratio of the emissions spectras: Your chromophore /GFP and select a filter which has its transmission fitting the maxima of this curve.
Greatings to switzerland,
Ralf
Dear Ralf, thank you very much for your suggestions!
My chromophore could be possibly excited also a bit more in the red, but we only have the 485 nm pulsed laser so that option is not feasible for us.
You’re right with the 625 nm filter I’m not that close to the maximum ratio of the two emissions so I will give it another try! But since the staining in general is not great and I’m afraid I might still see too much of GFP I’m considering to increase the number of parameters while fitting the fluorescence decay and force one of them to be the decay of GFP itself (measured in a control experiment). The problem here is that the decay of GFP itself is not monoexponential (although the monoexponential fit is not that awful, X2 around 1,5-1,8) and also my chromophores typically decay biexponentially. Then about accuracy of a 4-exponential fitting I’m not so sure. However, since the number of events coming from GFP will be anyways a bit smaller than from my chromophore and the monoexponential fitting of GFP is not that bad, maybe approximating the decay of GFP as monoexponential wouldn’t bring too much error in the end. What do you think? Can this be done?
Thank you very much and greetings to Germany!
Karolina
the components you want to fit the more photons you need... If possible I would try to find an experimantal solution.
I guess you are using the GFP only to localize you cells of interest, and you do not need its lifetime information, correct?
In this case you could substitute GFP against a fluorescent protein not excited by the 485 nm laser, prefrentialy blue shifted. CFP, TFP, mTurq2 etc might be an option. Just check if any of theses are available in your lab and are compatible with your other light-sources & filter:
https://www.fpbase.org/search/?default_state__ex_max__lte=470
best
Ralf
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Physical Phenomena
- Luminescence
- Fluorescence