Question
Asked 9 September 2012

GFP fluorescence after fixation

Does anybody have experience with GFP fluorescence preservation following aldehyde or other kind of fixation? In our hands formaldehyde fixation (from freshly hydrolyzed 4% paraformaldehyde), rapidly quenches fluorescence.

Most recent answer

I think the first question that should be answered definitively is if GFP fluorescence quenching should be EXPECTED after formaldehyde fixation. 
In all my years working with GFP, I've only had this happen when the tissue was exposed to organics like methanol or over-fixed using glutaraldehyde (which I was told also increases auto-fluorescence but not sure if I buy that). 
I mean, I've stored GFP+ embryos for over a year in 4deg/PBS without quenching after fixing in properly pH'ed, fresh 4% paraformaldehyde in PBS (sic). 
The only time I've ever had to use methanol was for cryo-sectioning, which I hate doing unless it must be done.
So, why is it quenching?  This is surprising to me.

Popular answers (1)

Brian Lin
Tufts University
There are a ton of different things that could solve your problems, quite a few listed above. I would also recommend two other things.
If your application does not require 4% PFA, please try using 1%. We do a ton of IHC, to DAB as well as fluorescence, and 1% PFA is much better for a large group of proteins and fluors. 1% is as simple to make as 4%, so it is usually an easy switch if you do not specifically need 4%. More generally, however, we have found 1% PLP to be an extremely good fixative for almost every single antibody we use in our repertoire as it combines the ample but not overdone fixing power of 1% PFA, with a background/fluorescence decreasing properties of lysine and periodate.
-----
Unfortunately, it's more involved than simply 1% PFA so unless you do a ton of this and rely on your IHC, this may not be worth it.
There is literature out there for 1% PLP, but I don't currently have access.
Briefly, to make 100 mls, dissolve 1 g PFA in 60 ml of 60C dH2O. Solubilize with a few drops of 1 N NaOH on low heat and stirring.
Weigh out solid .1M Phosphates: 1.014 g Na2HPO4 (dibasic) and .392g NaH2PO4 (monobasic) and add it to the 1% PFA with stirring.
Simple paper filter it, and add to it 1.37 g of lysine mono-HCL, predissolved in 10 ml H20.
Finally, add in .21g Na Periodate and QC to 100 ml.
(like I said, this is very laborious, but it is an excellent fixative for people who are doing careful IHC, DAB or fluoresence in tissue. Also, the lysine and periodate appear to do some kind of complexing reaction, so the fixative needs to be made fresh, at most 6 or 7 hours before use.
----
If you are still with me, the other thing you should consider is using anti-fade. There are a slew of them out there, our favorite is pretty cheap and homemade. It is 90% glycerol with n-propyl gallate. Our recipe for this is:
- Prepare a 10X PBS stock solution.
- Prepare a stock solution of 20%(w/v) n-propyl gallate (Sigma P3130) in dimethyl formamide or dimethyl sulfoxide. (Note: n-propyl gallate does not dissolve well in water-based solutions.)
- Thoroughly mix 1 part of 10X PBS with 9 parts of glycerol (ACS grade 99-100% purity) and slowly add 0.1 part 20% n-propyl gallate dropwise with rapid stirring.
Hopefully these and the suggestions by other people above me help fix your issue! I know well the pain staining issues :(
54 Recommendations

All Answers (112)

Aurore Dubuffet
Université Clermont Auvergne
Have you ever tried using anti-quenching mounting medium?
1 Recommendation
Is your vector stably or transiently transfected? It should express GFP unless there is a problem with the vector. Did you check the live cells for a signal before you fixed them? You should look at them if possible to make sure they are expressing the protein as expected. It is important to determine whether the problem is with the fixative or the initial transfection or expression. If you don't see fluorescence in the unfixed cells, you should sequence the vector to look for potential problems.
if you see fluorescence in Live cells check the following:
If the pH of Paraformaldehyde is too basic, it quenches the GFP signal.
The recipe i follow is
2.7 ml 37% formaldehyde
0.4 ml 25% glutaraldehyde
46.9 ml PBS
Incubate the cells with fixative for 5 min at 4 degrees, rinse with PBS, and store the cells in PBS. This should enable you to see native GFP fluorescence. You use any fixative you want if you are going to do immunocytochemistry with an anti-GFP antibody.
1 Recommendation
Helene S Robert
Central European Institute of Technology-Czech Republic
4% PFA is done in PBS pH 7.
If the GFP signal is sufficiently strong in live imaging, it will resist the fixation.
I fix 1hr under vacuum, wash twice with PBS and mount in an anti-fading solution (DABCO 25mg/ml in 90% glycerol in PBS, pH 8).
Unless you will have to do an immunolocalization of the GFP with an anti-GFP antibody.
Brandon Henderson
Marshall University
For this GFP is this in cells, neurons, mice? I use mice that express a GFP-tagged receptor and I have found that after fixing with 4% PFA, perfusing with 5% sucrose improves my fluorescent yield. I got this from the following paper:
Renda, A., Nashmi, R. Spectral Confocal Imaging of Fluorescently tagged Nicotinic Receptors in Knock-in Mice with Chronic Nicotine Administration. J. Vis. Exp. (60), e3516, DOI: 10.3791/3516 (2012).
2 Recommendations
Brandon Henderson
Marshall University
Another thing to consider, are you protecting your samples from light? If you are imaging your samples without immunostaining, this is even more important. Whether your samples are cells, slices, etc., make sure you minimize the time that they are exposed to light.
I also agree with everyone who has mentioned mounting medium. I have had good results with vectashield and mowiol.
1 Recommendation
Jerilyn A Timlin
Sandia National Laboratories
I have experienced this before as well and had success with the glutaraldehyde+ formaldehyde recipe described above using vectashield mounting medium. My understanding from talking with others is that the success is somewhat dependent on the tissue. good luck.
Thomas E Finger
University of Colorado
GFP survives fixation with 4% buffered PFA for several hours, but the fluorescence will decrease with increasing times in fixative. The GFP will not fluoresce effectively as long as any fixative is present or if the pH is not neutral, hence it is necessary to wash the specimen or slides thoroughly in buffer after fixation in order to optimize the signal. The most reliable way to detect the presence of GFP after fixation is to use immunocytochemistry. Commercial fluorescein-labeled anti-GFP antibodies are available.
2 Recommendations
Ahmed Abu siniyeh
University of Jordan
why don't you try to wash with 0.1M Glycine for 15 min at RT, after fixation.
Because Glycine is quenching the formaldehyde and terminates the cross-linking reaction.
2 Recommendations
Cathleen Teh
National University of Singapore
You can also try decreasing the length of fixation, this will improve the retention of GFP fluorescence.
Brian Lin
Tufts University
There are a ton of different things that could solve your problems, quite a few listed above. I would also recommend two other things.
If your application does not require 4% PFA, please try using 1%. We do a ton of IHC, to DAB as well as fluorescence, and 1% PFA is much better for a large group of proteins and fluors. 1% is as simple to make as 4%, so it is usually an easy switch if you do not specifically need 4%. More generally, however, we have found 1% PLP to be an extremely good fixative for almost every single antibody we use in our repertoire as it combines the ample but not overdone fixing power of 1% PFA, with a background/fluorescence decreasing properties of lysine and periodate.
-----
Unfortunately, it's more involved than simply 1% PFA so unless you do a ton of this and rely on your IHC, this may not be worth it.
There is literature out there for 1% PLP, but I don't currently have access.
Briefly, to make 100 mls, dissolve 1 g PFA in 60 ml of 60C dH2O. Solubilize with a few drops of 1 N NaOH on low heat and stirring.
Weigh out solid .1M Phosphates: 1.014 g Na2HPO4 (dibasic) and .392g NaH2PO4 (monobasic) and add it to the 1% PFA with stirring.
Simple paper filter it, and add to it 1.37 g of lysine mono-HCL, predissolved in 10 ml H20.
Finally, add in .21g Na Periodate and QC to 100 ml.
(like I said, this is very laborious, but it is an excellent fixative for people who are doing careful IHC, DAB or fluoresence in tissue. Also, the lysine and periodate appear to do some kind of complexing reaction, so the fixative needs to be made fresh, at most 6 or 7 hours before use.
----
If you are still with me, the other thing you should consider is using anti-fade. There are a slew of them out there, our favorite is pretty cheap and homemade. It is 90% glycerol with n-propyl gallate. Our recipe for this is:
- Prepare a 10X PBS stock solution.
- Prepare a stock solution of 20%(w/v) n-propyl gallate (Sigma P3130) in dimethyl formamide or dimethyl sulfoxide. (Note: n-propyl gallate does not dissolve well in water-based solutions.)
- Thoroughly mix 1 part of 10X PBS with 9 parts of glycerol (ACS grade 99-100% purity) and slowly add 0.1 part 20% n-propyl gallate dropwise with rapid stirring.
Hopefully these and the suggestions by other people above me help fix your issue! I know well the pain staining issues :(
54 Recommendations
Jacob Goodwin
Karolinska Institutet
Depending on what you are looking at you might also want to try cold 3:1 methanol/acetone for 15 min to fix your cells. A lot of my previous work was looking at transiently transfected cells expressing a GFP fusion construct. As long as you wash the cells well before mounting there I had no problem imaging Z-stacks through the cells. And have also performed ICC of these cells and gotten some quite beautiful images.
1 Recommendation
Josef Gotzmann
Max Perutz Labs Vienna
GFP and paraformaldehyde should work properly - never use MetOH for fixation. Any scavenger in the mounting medium will help (glutathione; vitaminE) or commercial anti-fading formulation. the problem is that GFP fluorescence comes along with oxygen radical production. sometimes reduction of PFA down to 1% also might be helpful.
josef
1 Recommendation
Yaroslav Tsytsyura
University of Münster
I will agree with Josef Gotzmann, concerning "never use MetOH for fixation" - it destroys the ultrastructure completely turning everything in some ugly mess (checked with TEM). For GFP fixation you may check the papers from Markus Grabbenbauer - he made the photooxidation of GFP in the presence of DAB to visualize it for EM: http://www.nature.com/nmeth/journal/v2/n11/full/nmeth806.html
Or, as it was advised before, use the anti-GFP AB. Works well.
2 Recommendations
Brian Lin
Tufts University
that is actually super interesting. I've tried MeOH fixation for a few antibodies which were super picky, and wouldn't work in anything else, and it turns out the the epitope they were raised on was a synthetic peptide fragment that was pretty small compared to the actual target protein.
Yaroslav Tsytsyura, do you mind sharing the TEM picture if you don't plan on using it elsewhere?
1 Recommendation
Johannes A. Schmid
Medical University of Vienna
From my experience the fixation with paraformaldehyde does not destroy the fluorescence of GFP and its variants (but methanol or ethanol reduce it by about 90%). As mentioned by others: the mounting might be crucial. We observed a very drastically accelerated bleaching of ECFP after mounting in DAKO flourescent mounting fluid, while other GFP-variants were not destabilized. For more info see our paper at: http://www.plosone.org/article/info:doi/10.1371/journal.pone.0018586
4 Recommendations
Alexander E Kalyuzhny
University of Minnesota Twin Cities
The easiest way would be using anti-GFP antibodies conjugated to Cy2 or Alexa 488 (same colro as GFP). The advantage of such approach is that you can "pull out" very low abundant and almost invisible GFP signal. In addition, you can combine your anti-GFP-Cy2 with TSA (tyramide signal amplification) protocol that will give you a very strong amplification.
Yaroslav Tsytsyura
University of Münster
Brian Lin, I'll try to find it. It was tested once about 4-5 years ago, so I have to dig through the archives. Once I find it, I'll attach it to the answer.
Brian Lin
Tufts University
Oh man, that was so long ago, don't search too hard then, I was mainly just curious :D
Yaroslav Tsytsyura
University of Münster
Brian Lin, here it is. I agree, MethOH is OK or "acceptable" for fluorescent immunostainings, but if you hope to use this method of fixation for TEM, you can forget it. It would be a way better (though, much more harmful to the experimentator) to use Acrolein instead. This would give you a possibility to make correlative light and electron microscopy, if you need. And as far as I remember despite being so nasty in terms of hazard acrolein is a very potent fixative and does not kill the epitope. We worked with it around the same time but did not find any benefit over regular PFA fixation for our particular case and quit it because of hazard.
The slide is attached. Original images are buried somewhere, but one PowerPoint slide I found.
1 Recommendation
The following paper (Schnell et al., (2012) Nature Methods 9, 152–158)
http://www.nature.com/nmeth/journal/v9/n2/full/nmeth.1855.html?WT.ec_id=NMETH-201202 contains some data on EGFP fluorescence intensity in MDCK cells after fixation with PFA (and subsequent permeabilization). See Figure 3.
The also analyzed ultrastructural changes after fixation (PFA vs. Methanol vs. Glutaraldehyde) and permeabilization by electron microscopy.
1 Recommendation
Yaroslav Tsytsyura
University of Münster
Andreas Nerlich, Thanks. Seems like they were not lazy to make a paper out of it. We never have done pre-fix with PFA and permeabilisation with MethOH, just fixing in -80°C methanol for the fluorescence immunostainings. But it was not for long, once we checked our samples with TEM...
Elena S Nadezhdina
Russian Academy of Sciences
Preparation of TEM samples usually includes post-fixation with OsO4 and long procedure of resin embedding. If sample is not fixed with aldehydes, these procedures destroy its fine structure. For example, even formaldehyde fixation is not strong enough to prevent microtubule destruction with OsO4, and glutaraldehyde fixation is required.
To see EGFP fluorescence, I recommend using 1% formaldehyde directly solved in warm PBS. Try either warm (apr. 37 deg. C) or RT solution for fixation. Never exceed 1 h in fixative, wash sample extensively with PBS. We also use fixation with absolute methanol at -20 deg.C for 5 min, immediately followed with 3% formaldehyde at 4 deg.C for 15 min and extensive washing with PBS.
Brian Lin
Tufts University
Thanks Yaroslav, our lab doesn't plan on doing EM at all as we are mainly looking for co-localization and identifying cell types through IHC, but the picture does make a point which makes sense too, since MeOH fixes through precipitation which would really screw up fine structure, but sometimes helps IHC by revealing hidden epitopes.
Cheers!
Niccola Funel
Università di Pisa
Dear George,
I recently performed the analyses on FFPE tissues containing Fluorescent signals. The routine procedure used to remove Paraffin, does not limited the fluorescence signal in the cells. So far I would like to suggest you to use ethanol to fix snap frozen tissue and cell cultures transfected previously with GFP signal. This procedure allows you to perform also a combination with different antibodies. Please use the glycerol mounting to have a look at the microscope.
See you, Niccola
Andrew Hamilton
University of California, Davis, School of Medicine
Hi George,
We've always had pretty good preservation of GFP fluorescence post 4% PFA fixation, but then we usually do use an anti-GFP + Alexa-488 secondary to amplify the signal as Alexander suggested. We have found that getting the samples to the microscope as quickly after mounting as possible helps with the brightness immensely, so you could try that as well. Good luck!
2 Recommendations
Stjepan Uldrijan
Masaryk University
Working with human cancer cell lines transiently transfected with plasmids coding for GFP- or YFP-tagged proteins we had good experience with less concentrated PF (3%) and shorter fixation time (max. 10 min). I hope this helps.
Xuejun Chen
National Institute of Allergy and Infectious Diseases, National Institutes of Health
I had used 1% PFA to fix tissue and preserved GFP. In contrast, methanol "killed" GFP.
3 Recommendations
Abbes Belkhiri
Baptist Health Sciences University
Fixation with 1% or 2% paraformaldehyde should do the trick.
Good luck!
Anne-Catherine Schmit
University of Strasbourg
I checked GFP-tagged proteins in fixed plant cells and I got a good preservation of the signal with 1.5% PFA + 0.5% glutaraldehyde for 30min in teh following buffer 50mM PIPES, 2mMEGTA, 2mMMGSO4 pH 7.2. After extensive washing the preparation is incubated 3x10min in NaBH4 1mg/mL and than if neede permeabilized using 0.05% tritonX-100 and immunostained or just DAPI stained.
The sigal was less intense than in vivo but still visible when the level of expression was high. In other cases I immunostained with alexa488 labelled anti-GFP antibodies, which helped to enhance the signal.
I hope you will also succeed
Kristine Atkinson
A/A Patent Investigations
Formaldehyde is notorious for quenching fluorescence, paraformaldehyde a close relative: I wouldn’t use either if a way around it can be found. You could try buffered glutaraldehyde alone (I think cacodylate is a no-no, but glycine is good—you might also be sure your buffers aren’t quenching), and if you’re using 4%, try 0.5% on tissue that is not floating in liquid media (you can put it on a filter or fine mesh and suction away the excess fluid). There are many references to autofluorescence appearing with glutaraldehyde fixation, so if your purpose in fixation is to denature proteins that would cause enzymatic activity, a flash of steam can sometimes do the trick (works on cell smears). Acrolein is hardly used these days, but worth a try. In all cases, fix as briefly as possible.
If you’re dealing with plant cells, the cell wall is an extra barrier that animal people don’t have to deal with. Can you use protoplasts? The idea of permeabilizing, as suggested by Anne-Catherine Schmit, sounds like a good idea, although turning to a second antibody is a headache. Damaging the emission intensity is a hazard both with detergents (a danger is vesicle solubilization) or with something mechanical like gentle heating, freeze-thaw or low pressure.
Verveer et al. offer solutions in “Imaging Protein Interactions by FRET Microscopy,” Chapter 32, in Protein-Protein Interactions (eds. Golemis and Adams). Cold Spring Harbor Laboratory Press, 2005. Sakhalkar et al. (Phys. Med. Biol. 52 (2007) 2035–2054; http://radonc.duke.edu/wysiwyg/downloads/Fluorescence_Preservation_in_optical-ECT.pdf ) seemed to have more of my own experience, that simple ethanol (not methanol) fixation was best—there may be shrinkage and not the best morphology, but different fixatives are best for different purposes-- unless you’re after a lipid that will disappear with ethanol!
This reference by Straight from Stanford http://straightlab.stanford.edu/PDF/fluorescent.pdf surely offers in-depth experience-- with the caveat that your own system may have its quirks. Their section warning about the problems of fixation include migration of the GFP as a result of fixation, in addition to quenching.
In all: do you have to fix? Can you use fresh or frozen sections? Or try turning the protocol upside down, by putting the tissue in a nanocage, photograph the GFP in vivo and then fix for morphology.
1 Recommendation
Brian L Brockway
Duke University Medical Center
The problem may actually be more than quenching alone. I read once that (cytoplasmic) GFP is very water soluble. So, when you fix with PFA, which introduces holes in membranes as a side product of its cross-linking, the GFP tends to leak out. More washes (PBS) make it even worse. In agreement with this, nuclear GFP is often more robust, and the newer membrane-targeted GFP's (farnesyl, palmitoyl, etc) are just awesome. I have seen non-quenched GFP after overnight fixation with cell membrane tagged GFP.
Glutaraldehyde is a faster fix than paraformaldehyde, so maybe that's why some people see better results. But glut is "heavy handed"; might easily overfix and quench. Rather than a cross-linking (aldehyde) fix, use precipitating fix (ethanol) as others suggest. Methanol is much worse, as it dissolves lipid/membrane more than ethanol. Be careful of commercially prepared 10% buffered formalin; it has methanol in it!
1 Recommendation
Matthew Lakins
University of Cambridge
I would definitely recommend using a lower amount of PFA in your fixative (if you use PFA at all), and preferably 1%. However, unless the GFP is tagged to something that is easily fixable, you are likely to get seepage..
I would also recommend using an anti-GFP antibody to amplify the signal as GFP can, as you say in your question, diminish after fixation.
Anne Cantereau
French National Centre for Scientific Research
we use anti-GFP antibodies for fixed samples, some work very well (invitrogen, abcam)
2 Recommendations
Paul A Lucas
New York Medical College
Formaldehyde, glutaraldehyde, alcohol, etc. does not "quench" fluorescence. Instead, it denatures the GFP such that it no longer functions and fluoresces. Any remaining fluorescence is either non-specific (many proteins auto-fluoresce after fixation) or it is a few GFP molecules that retain their original conformation. If you fix the tissue, then as Ann Cantereau states, you MUST use an anti-GFP antibody to detect the GFP. At that point, you are "seeing" the GFP by whatever tag you have on the secondary antibody.
54 Recommendations
Jose A. Garcia-Sanz
Spanish National Research Council
I fully agree with Paul Lucas
GFP does not like PFA especially the higher % of it, for example, for long period of fixation (eg 1-2 days), 4% gradually denatures the protein and you cannot get actual proteins that expected. But for long period (say 1-2 days), 1% PFA would be better. All of these depend on duration after fixation and amounts of GFP expressed.
I have brain sections of rat which were trnsduced with EGFP virus and were perfused, post fixed with 4% PFA overnight. I never had problem with quenching. I even dont use anti-GFP antobody. In general I can directly see GFP under fluorescent microscopy without any problems.
Kristine Atkinson
A/A Patent Investigations
We understand that the majority of researchers who haven’t encountered fluorescence problems from a kit used on their standard (animal) tissue wish to report that it works for them—of course it does, the kits are designed to work under most circumstances. But Dr. Komis has a specific situation and asks others for their experience in working around it—as we’ve seen, many suggest less paraformaldehyde or shorter exposure; but if he is using plant tissue, fixation times often are longer from the get-go, in order to surmount the cell wall and other problems. Fixation of plant material is challenging, as they have vacuoles, a lower concentration of precipitable (or linkable) protein in the cytoplasm, and may be impermeable to many standard buffers (see M.A. Hyatt, “Problems in Processing Plant Tissues” chapter in Principles and Techniques of Electron Microscopy: Biological Applications; Cambridge Univ Press 2000; pp 444-445). A few decades ago, fixatives such as permanganate and mercury were common.
Vitha et al. (Histochem J. 32:457-66) suggest it isn’t aldehyde fixation that destroys plant F-actin immunofluorescence, but other steps in the visualization process are critical to preserve it. Can we hear from plant histologists how they fix and prep for fluorescence microscopy? Do results vary by species? What buffers work best?
2 Recommendations
Juliana Lott Carvalho
University of Brasília
I suggest you to take a look at the article entitled: Preservation of the green fluorescent protein on decalcified bone tissue (Cienc. Rural vol.40 no.10 Santa Maria Oct. 2010 Epub Oct 29, 2010). Dr. Boeloni describes several attemps to process and preserve GFP fluorescence in rat bones. EDTA decalcification and cryomicrotomy was the only combination protocol which preserved GFP fluorescence.
2 Recommendations
Sereno Denis
Institute of Research for Development
We have the same problem with fluorescent parasites. We lost a large amoutn of fluorescence after 2% PFA. We are therefore looking for some protovol that will allowed us to recovert fluroscence after fixation. Does somebody have tried methanol fixation for GFP?
Christophe Huret
Université Paris Cité
I agree with Brian. We use 1% PFA for fixing infected cells expressing GFP and it's work well.
Laura Gonzalez Bosc
University of New Mexico
Fixation does not affect GFP expressing cells in our hands. However, if level of expression is low then it might reduced and you do not longer see it.
Franck Anicet Ditengou
University of Freiburg
The protocol described below is for 5-7 day-old Arabidopsis seedlings. It gives excellent results for most of GFP-fused proteins used in our lab. I am using freshly prepared 4% FA in PBST (PBS+ 0.1% Triton X-100). After 2 min vacuum, plants are fixed for 30 min at room temperature. After having thoroughly rinsed the samples (first in PBST and then in water, each step 10 min), seedlings can be mounted in any anti fading solution. I am using anti fading solution from Invitrogen. Note that the fixative should be kept at room temperature prior usage. I hope this works for you as well.
4 Recommendations
Emmanuel Dupont
Imperial College London
PFA as low as possible, time as short as possible, temperature as low as possible. There will be some loss of fluorescence no matter what you do.
1 Recommendation
Irfan Khan
Aga Khan University, Pakistan
I have problem in GFP protein expression in Mesenchymal stem cells. I m using adenoviral vectore for the epression of my desire gene in MSCs. GFP as the tag gene, for confirmation of my desire gene expression. Very good expression of desire gene but GFP express in few cells.?
Yiannis Drosos
St. Jude Children's Research Hospital
From my personal experience, it is not the fixation but the dehydration process that quench the GFP. That is why when you fix cells expressing GFP and proceed to fluorescence microscopy, you don't loose the signal.
So, George, if you could fix your tissue with 4% PFA but then proceed for cryosections, your GFP will be active.
I hope this helps.
2 Recommendations
Daniel Possin
University of Washington Seattle
In our experience, the amount of GFP fluorescence lost to PFA fixation is highly variable, depending on the construct created. If you wish to retain morphology and have a GFP fixation problem, the best solution is an anti-GFP antibody label.
1 Recommendation
Atsuyasu Sato
Kyoto University
I agree with Daniel. I used anti-GFP to detect GFP in 4%PFA-fixed lung tissues. You need antigen-retrieval for it. For cultured cells, 1%PFA might be better.
Libero Vitiello
University of Padova
In our hands, PFA (2%) does not appear to have any negative effect on GFP fluorescence in cultured cells (in the past I compared fresh cells with fixed cells from the same transfection experiment). However, in skeletal muscle we noticed that if we simply prepare cryo-sections from fresh samples we loose just about all fluorescence; in order to preserve it we need to fix the whole muscle in PFA 2% (4 degrees for one hour will suffice for ~50 mg of mouse tissue), followed by overnight saturation in sucrose 30% at 4degrees, followed by snap freezing and cryosectioning. Please note, however, that in our experience the use of an anti-GFP ab will always provide a higher signal than GFP's own fluorescence, both in sections and in cultured cells.
Fatemeh Shamsabadi
Universiti Putra Malaysia
I agree with Laura Gonzalez Bosc and Emmanuel Dupont. Based on my personal experiment, I used 4% PFA for 10 min at RT, washed two times with PBS and visualized the signals in transduced cells under fluorescent microscope. I was wondering if anybody can inform me about the positive GFP signals in live cells without fixation or any process like changing the media. Could be detected?
Thanks
Sascha Beneke
Universität Konstanz
Hi!
At least in our hands a 15 to 20 min fixation in 3.7% formaldehyde/PBS is not reducing so much eGFP fluorescence intensity. However, if you use precipitating agents such as methanol, eGFP is rapidly destroyed. We see a 50% drop compared to above formaldehyde (FoAl) fixation if we fix about 7 min in 100% methanol in a fridge - but for some experiments you have to take this. I would be careful about using short incubation times with low concentration PFA or FoAl - both rapidly induces DNA strand break signalling, as we published recently (Beneke et al., 2012 PLoS One). So, dependent on your target, this could change a lot.
Rihab Nasr
American University of Beirut
PFA Fixation did not affect GFP expressing cells in our hands.
Laura Le Bouil
L'Institut du Cerveau et de la Moelle Épinière, French National Centre for Scientific Research
Hi!
I have perfused animals for a year with PFA 4% and couldn't see GFP in the brain slides with and without Immunohistochemistry. After reading all the answers here we tried to perfuse with PFA 1% buy the same protocol and now we see it very well. We use emGFP which may be modified a lot compare to the original one and became sensible.
Good luck
2 Recommendations
Dennis Brown
Massachusetts General Hospital and Harvard Medical School
We have used both 4% para and PLP on kidney and cells expressing GFP and can easily detect it in cryostat sections. I agree that it is not detectable in unfixed cryostat sections. How are you sectioning the tissue Laura?
Laura Le Bouil
L'Institut du Cerveau et de la Moelle Épinière, French National Centre for Scientific Research
I'm sectioning the tissue in the cryostat, 10µm thicke for fresh frozen tissue and 15µm for perfused brains. In our hands, GFP is detectable in unfixed cryostat sections but it looks like a steam cloud with low intensity.
Moreover we have never seen GFP in PFA 4% perfused brain sections (cryostat, cryomicrotome). But with PFA 1%, it appears with high intensity before immunohistochemistry.
Dennis, do you use the original GFP or a modified one?
After imaging many EGFP-expressing cells fixed with PFA (1-4%) at confocal and epifluorescence microscopes without problems, what I've found is that the mounting medium that you use is critical to maintain EGFP (and other fluorochromes) fluorescence.
We only use "Prolong Gold Antifade" from Molecular Probes. Protects fluorescence during the observation (no fading effect), and keeps it for months provided that you maintain your slides at 4ºC and protected from light
1 Recommendation
Vanessa Nicolin
University of Trieste
Hi, I agree with Daniel
Gyorgy Lur
University of California, Irvine
As stated above, PFA should barely affect GFP fluorescence. I`ve been using 2-4% PFA to fix stuff ranging from monolayers of primary or cultured cells to brain tissue in perfused animals, no issues.
On the other hand I think I know what your problem is: organic solvents do destroy flourescent proteins. PFA is (almost) always contaminated with large amounts of methanol. That will do your GFP in... To avoid problems you should buy EM grade PFA (e.g. from Electron Microscopy Sciences, cat. # 15713). It`s more expensive but it should solve the problems associated with low quality PFA. Of course immunostaning for GFP works as well.
Good luck!
2 Recommendations
Phoebe Stavride
Foundation for Research and Technology - Hellas
I don't know what your downstream application is, but we used to stain GFP cells for observation in the microscope and they only lost a small fraction of the fluorescence. We usually fixed cells with PBS:acetone 1:2, but a 1:1 Acetone:Ethanol mix is supposed to retain GFP fluorescence better. That said, it also depends on how much fluorescence you have to begin with (our cells were transfected with GFP-constructs). But it's a simple and cheap thing to try, and you could know if it works for you in a day! Good luck!
1 Recommendation
Modesto Redrejo-Rodríguez
Universidad Autónoma de Madrid
I had the same problem and solved reducing the fixation time (from 60 to 15 min @ RT) and performing a second step of fixation with pre-chilled methanol (5 min @ -20ºC). I couldn't use only methanol fixation because it gave rise to artifacts in localization of nuclear proteins, but the double fixation works fine.
I have tried fixing with 3.7% PFA in 1X PBS for 3 mins and it works fine. Try limiting the fixation time to avoid quenching of fluorescence
How about localizing GFt the EM level ? I have no luck so far ..
We always detect the EGFP positive cells by confocal microscopy,our protocols is fix the cells in 3%PFA for 30 mins on ice,and neutralize with 50mM NH4Cl for three times,and we never face this "quenching" issue.
Thanks.
Wolfgang H. Muss
Paracelsus Medical University Salzburg
Good day, dear Arrhenius Lin,
just to comment on your acknowledged hint to use NH4Cl for "quenching".... what you are proposing and practically doing is what I have done all way since 1981 for my TEM-specimen preparations, when Roth J, Bendayan M, Carlemalm E, Villiger W, Garavito M. published their paper: <Enhancement of structural preservation and immunocytochemical staining in low temperature embedded pancreatic tissue.> in J Histochem Cytochem. 1981 May;29(5):663-71.
(find a free access pdf @ http://jhc.sagepub.com/content/29/5/663.long ).
I would like to call that "quenching" = <<deactivation of free aldehydes left from fixation>> rather than "neutralization" of fixative.
The process itself (quenching by 50mM hydrous NH4Cl =ammonium chloride - solution) in other publications also has been described by use of Na-borohydride (NaBH4) and other work has been done by using glycine buffer (0.3M).
Best wisehs and good luck.
1 Recommendation
@Wolfgang
Thanks a lot!I intended to use “neutralize” to express the meaning"deactivation of free aldehydes left from fixation",but that seems to be misleading.I do it just follow the extant SOP in our lab,I am not so sure where did it come from,thank you very much for pointing out the exact reference.
Wolfgang H. Muss
Paracelsus Medical University Salzburg
Dear Arrhenius, you're welcome!
My posting not intended to indoctrinate anybody! (:-))
best regards
Brian Lin
Tufts University
@Wolfgang
I had not previously heard of any deactivation of fixation treatment. Correct me if I am wrong, but I see that for the most part you do EM. Have you noticed any difference in immunostaining of tissue with or without the NH4Cl treatment?
We work with mostly 1% PLP fixed mouse tissue, sometimes 4%, and I am wondering if treating with NH4Cl might enhance staining or reduce background from unreacted aldehyde groups.
Thanks!
FYI: ...Never had a problem imaging (confocal) EGFP-construct transfected cells after fixing with 4% PF. We routinely fix cells using fresh 4% PF in PBS (RT/rocker/15 min.), followed by a 5% Glycine/PBS "quenching" step (RT/rocker/5 min.), and then a final series (3x) of PBS washes (5 min./rocker/RT) before blocking.
cheers,
Jeff
Wolfgang H. Muss
Paracelsus Medical University Salzburg
(Apologize for lengthiness)
Dear Brian, dear Jeffrey, I came back to this after having finished some other work, so apologize for perhaps somewhat late reply. I honestly must say that I am not familiar with IHC / LM, Fluor Micr. etc. as it is done nowadays. I am only a "dinosaur" dealing with TEM for the last year of my profession prior to retiring (and naturally with interest in methodological / technical questions in Histology, Ultrastructure, Spec.Prep.techniques etc.).
So it might be that my answer here is not quite the answer you want to find or get confirmed what you have been wondering about (cf. above). As far as it is my knowledge, all aldehyde fixatives used
[(severity perhaps depending also on concentration used and time of fixation, as perhaps also temperature. For FA (formaldehyde/paraformaldehyde etc): it has been reported that small (10 x 10 x 3 mm) tissue pieces fixed promptly in neutral buffered formalin for 6–24 hours will generally show good cytological preservation and immunolocalization, with a minimum of antigen masking. It is the great variation in time and conditions for fixation that cause the majority of problems in immunochemistry - so it might depend also on strictly adhering to exptl. protocols)]
may have a displeasing property: free aldehyde groups left in the tissue would e. g. bind the primary and secondary antibodies, leading to high background. Background due to free aldehyde groups has been reported more likely to occur when the fixative is glutaraldehyde or paraformaldehyde. Therefore older methods describe the <blocking of free aldehyde groups> by use of e.g. NH4Cl (= hydrous ammonium chloride 50 mM), glycine added to a buffer (0.3M). Additionally (from a websource): citation: <<Tissues that have been fixed with a glutaraldehyde-based fixative must be treated with inert amine-containing molecules prior to the immunoassay because free, unsaturated aldehyde groups are available to covalently link amine-containing moieties such as antibodies. The most efficient aldehyde blockers are ethanolamine and lysine.>>end of citation. in former times - if I remember correctly, also NaBH4 (Na-Borohydride) was used to block free aldehydes left in the tissue. Most of these aldehyde blockers (esp. glycine and NaBH4) have been reported to be quite elutive, so for EM-work these blocking agents haven't been in use for a lot of TEM-IHC/ICC applications. I have done the blocking with NH4Cl for preventive reasons (if tissue fixed for TEM should undergo further IHC-techniques) but also just to be on the safe side with a semithin staining on resin embedded sections (not to get a false positive metachromatic staining due to free aldehyde groups, cf. Schiff-reaction).
The latter mentioned blocking/quenching method most obviously is not the same as e.g. to block nonspecific antigenic sites found on tissue (where one uses other agents, cf. IHC-handbooks etc.).
Not to increase the length of my post here I would like to add one perhaps interesting article: "monthly" (June 2012)-question from the <Molecular Biology Forums> for all colleagues "out in the wild" dealing with similar problems described in this ResearchGate thread:
GFP Imaging in Fixed Cells
Kristie Nybo, Ph.D.
BioTechniques, Vol. 52, No. 6, June 2012, pp. 359–360
Full pdf: What fixative will allow visualization of GFP fluorescence? (Thread 13765)
Best regards and good luck, Wolfgang
3 Recommendations
Brian Lin
Tufts University
@Wolfgang
Thank you very much for the response, it is most enlightening!
I will try pretreating tissue with a lysine containing buffer before adding primary antibody and see if it improves anything. However, the thought now arrives that as we use a serum containing block, that any unreacted aldehyde groups would most likely be cleared off before addition of primary.
Again, thanks for the help.
Brian
Wolfgang H. Muss
Paracelsus Medical University Salzburg
Dear Brian, you're welcome!
Thanks for your kind reply, very best regards and good luck !
Wolfgang
EGFP fluorescence fades away very quickly due to to a physical phenomenon called "photobleaching" that is caused by the high energy of the confocal lasers (or mercury lamps) on the sample. To reduce photobleaching I use (and recommend) Pro-Long Gold as an antifade mounting medium. Just a few ul on the coverslip (on the sample side) will do it. In my hands this is the best anti-fade reagent.
Anne Laurencon
French National Centre for Scientific Research
EGFP is definitively denatured by several fixative, the worst being Methanol which erase most of the fluorescence, PFA also dim it in my hands (4%, never tried 1% on my samples). You might not be affected by this this if your signal intensity is strong anough. Two solutions worked for us: doing live imaging is one of them, the second is doing IHC, but there again, not all antibodies will recognize the protein when it has been denatured. Only two available on the market worked for me at the time: The rabbit antibody sold by Molecular Probes, and the Chicken antibody sold by Aves.
Of course antifades help during imaging but do not solve the problem of protein denaturation during fixation.
4 Recommendations
Attila Lehotzky
Research Centre of Natural Sciences
how is this start? old thing. brochure of company described 2% FA or PA as proper fixative to keep GFP fluo after processing and water based embedding. well i use chilled methanol to particle or membrane bounded fluo of GFP. it was correct. any other case, use GFP-antibody to detect the construct.
Enrico Marani
University of Twente
Contact Richard Horobin, specialist in stainings and fluorescence.
Troy A Hornberger
University of Wisconsin–Madison
Another thing to bear in mind is that GFP is extremely water soluble, so if you are working with tissue sections (no membrane surrounding the tissue to retain the GFP), the GFP will float away as soon as it hits an aqueous solution. To overcome this issue we have fixed cross sections in acetone containing 2% PFA. Here is the recipe for making this solution: Bear in mind that we work with skeletal muscle cross-sections (2-10uM thick) and it only takes about 10sec at -20C to fully fix the GFP in place. Anything longer than this will raise issues with antigen masking / over-fixation.
2% Paraformaldehyde in Acetone
Adapted from: Practical Methods in High-pressure Freezing, Freeze-Substitution, Embedding and Immunocytochemistry for Electron Microscopy (Mary K Morphew; Dept. of MCD Biology University of Colorado)
Paraformaldehyde is not soluble in acetone so:
1.) Make 20% solution of PFA in anhydrous MeOH (1 gram of PFA in 5mL of MeOH)
2.) In 100mL beaker, heat water to ~65 degrees and place tube of 20% PFA in MeOH into water; intermittently vortex
3.) Crush NaOH pellet and add approx. 1/8th of pellet to 20% PFA in MeOH
4.) Continue warming and vortexing until PFA goes into solution
a. May need to add more NaOH
b. Takes approx. 15 minutes
5.) Dilute to 2% with acetone
2 Recommendations
Atoshi Banerjee
Translational Health Science and Technology Institute
hi
i have personal experience working with GFP denaturation on paraformaldehyde fixation. We tried and we standardized we fix in 3% paraformaldehyde for 3-5 min (not more in any case) then we wash with excess PBS followed by permeabilize with methanol for 3 min at -20C (dont keep for more time) then following washing we block and proceed with antibody incubation. i work with cell lines for IF and we still see very strong GFP signal and works very good with us.
Enrico Marani
University of Twente
Look into intermittent microwave heating, which can bring fluorescence back or enhance fluorescence. see Kitayama et al 2000, en 1999. Look also in the Microwave Cookbook for microscopists by Kok and Boon several editions.
Samantha Stehbens
The University of Queensland
Are you looking in cells or tissue? In cells- I find it makes a difference when you dilute your PFA(I use the high purity EM grade which comes in 10ml snap vials, 16%-rt15700 from EMS) in a buffer other than PBS.
I have compared BRB80, CSK and PBS, side by side- and get the best results (native gfp- no anitibody recovery required), with BRB80. It's easy to make a 5x stock, filter it and keep it in the fridge.
(Tip- use KOH tablets to pH).
I also routinely permeabilise my cells and see no loss of fluorescence- however I image on a spinning disc which minimizes photo damage of my specimens. I haven't imaged native fixed EGFP on an LSM.
I mount my samples in a homemade mounting media which contains anti fade- like DABCO or NPG.
PROTOCOL TEXT:
Cold Spring Harbor Protocols
Mowiol-DABCO stock solution
1,4-diazabicyclo-[2,2,2]-octane (DABCO; Merck 803456)
Glycerol
Mowiol
Tris-Cl (0.2 M, pH 8.5) (see below)
Combine 2.4 g of Mowiol, 6 g of glycerol, and 6 mL of H2O. Mix for 3 h. Add 12 mL of 0.2 M Tris-Cl (pH 8.5). Incubate with mixing at 50°C for 10 min. Centrifuge at 5000g for 15 min to pellet insoluble material. Add DABCO to final concentration of 2.5% to the solution as antibleaching agent. Store in 500-μL aliquots at −20°C.
thaw to use, store at 4C or -20C
add 5ul per coverslip onto glass slide, invert coverslip onto drop
remove extra mowiol carefully around edges with a little aspiration or by capillary action with kim wipe
place in drawer overnight to harden
(if you're in a hurry.... place at 37C for an hour to harden)
store slides at 4C in slide box
--------------------
Tris-Cl
Tris base
HCl
To prepare a 1 M solution, dissolve 121.1 g of Tris base in 800 mL of H2O. Adjust the pH to the desired value by adding concentrated HCl.
pH HCl
7.4 70 mL
7.6 60 mL
8.0 42 mL
Allow the solution to cool to room temperature before making final adjustments to the pH. Adjust the volume of the solution to 1 L with H2O. Dispense into aliquots and sterilize by autoclaving.
If the 1 M solution has a yellow color, discard it and obtain Tris of better quality. The pH of Tris solutions is temperature-dependent and decreases ~0.03 pH units for each 1°C increase in temperature. For example, a 0.05 M solution has pH values of 9.5, 8.9, and 8.6 at 5°C, 25°C, and 37°C, respectively.
1 Recommendation
Amy Schutz Geschwender
LI-COR Biosciences
In my experience, glutaraldehyde is the fixative that causes the strongest autofluorescence. It's a good one to avoid, if autofluorescent background is masking the signal you want to see.
1 Recommendation
Amy Schutz Geschwender
LI-COR Biosciences
Are you staining for any targets other than GFP? Choice of fixative can drastically affect immunofluorescent staining. With one fixative, you may get strong and beautiful signals - but a different fixative gives you no signal at all. So, if other protein targets are also important, you'll want to consider that as well when choosing your fix. If the primary antibody is recommended for IF staining, the data sheet will usually recommend a fixative. This is a good place to start, if you're not sure what fixation conditions are best for that target. You may have to tinker with it a bit, to find conditions that make all of your antibodies happy.
Samantha Stehbens
The University of Queensland
you can quench gluteraldehyde autofluorescence with freshly prepared 0.2% sodium borohydride in PBS. When glutaraldehyde reacts with amino groups it forms a fluorescent component which needs to be reduced directly by hydrogen generated in situ by hydrolysis of sodium borohydride.
make sure it is fresh --> If it does not bubble it is no good. Rinse it several times over a 20-30 minute period.
I only use gluteraldehyde if my antibody works with it- very few antibodies like it. It destroys native GFP in my hands.
Hong-Xiang Liu
University of Georgia
1) I tried light fixation (4% parafamaldehyde) on cryostat sections from freshly frozen tissues several times. I could not see GFP labels even after immunoreactions, so timely fixation is important.
2) 4% PFA in PBS works very well for my tissues. Immunostaining can dramatically enhance the GFP signals. And it works for other antibodies.
3 Recommendations
Hi,
I agree with Atoshi Banerjee fot the fixation . we fix in 3% paraformaldehyde for 3-5 min (not more in any case) then we wash with excess PBS followed by permeabilize with Triton 0.1% for 10 min .........
4 Recommendations
Guannan Wang
University of Michigan
Does anyone try to see GFP signal on cryostat sections directly without fixation of any kind? From the comments here, it seems like fixation can interfere with the detection of GFP, so I'm wondering if I can skip the fixation completely. 
3 Recommendations
Timothy A. Norman
Boston University
Loss of GFP signal is due to dehydration or when fixatives such as NBF (10% formalin with various amounts of methanol) are used for FFPE tissues which leads to denaturation of the chomophore.
 I've used 4% PFA to fix mouse brains without loss of eGFP fluorescence when cryopreserved with either 10%/2% - 20%/2% glycerol:DMSO gradient or 10% & 20% sucrose followed by OCT embedding.
2 Recommendations
Emmanuel Dupont
Imperial College London
Freshly prepared PFA in PBS for 10 min at 4oC preserved some fluorescence, the longer the fixation the more fluorescence is lost.
1 Recommendation
Enrico Marani
University of Twente
Look into Van Valkenburg and Marani on fixation dopamine. Be aware that fluorescence also diminishes if Ab is made by coupling with another fixative than used for fixation.
oh my goodness...10 pages of suggestions but how many actual solutions?...
what do you do with so many possibilities?
Enrico Marani
University of Twente
Dear Jerry, if you had looked into handbooks from 1970-1990 you should have noticed that fluorescence is a difficult subject.
I think the first question that should be answered definitively is if GFP fluorescence quenching should be EXPECTED after formaldehyde fixation. 
In all my years working with GFP, I've only had this happen when the tissue was exposed to organics like methanol or over-fixed using glutaraldehyde (which I was told also increases auto-fluorescence but not sure if I buy that). 
I mean, I've stored GFP+ embryos for over a year in 4deg/PBS without quenching after fixing in properly pH'ed, fresh 4% paraformaldehyde in PBS (sic). 
The only time I've ever had to use methanol was for cryo-sectioning, which I hate doing unless it must be done.
So, why is it quenching?  This is surprising to me.

Similar questions and discussions

How do you detect GFP reporter fluorescence after fixation?
Question
8 answers
  • Lorena HidalgoLorena Hidalgo
Hi! I´d like to ask for a protocol to detect GFP fluorescence after fixation and permeabilization.
This is the situation:
I´m using a MSCV-MLL-AF9-IRES-GFP plasmid to induce acute myeloid leukemia into mice in vivo. To do so, I infect mouse hematopoietic stem cells (HSCs) with a retrovirus containing the plasmid and then I transplant them into mice. After 10 weeks, the mice get sick (as I can see GFP fluorescence in blood).
When I analyze leukemia cells in vivo, I can see a nice GFP signal. However, after fixation and permeabilization, the GFP fluorescence is lost. I tried to detect it with an anti-GFP antibody, but it doesn´t work. I´ve done this many times changing some parameters (fixation and permeabilization reagents, incubation times, etc.) but nothing seems to work. This is the detailed protocol of what I do, in case you identify a potential mistake:
- Isolation of bone marrow cells from mice and resuspension in staining media (HBSS + BSA)
- Fixation with 1% PFA for 15´ on ice in the dark
- Permeabilization with permeabilization buffer (containing saponin) for 5´ at room temp in the dark
- Incubation with the GFP antibody for 1 h
- Wash with permeabilization buffer x2
- Resuspension of stained cells in PBS+DAPI
- Analysis by flow cytometry
When I run the samples in the flow, I can see cells have been properly permeabilized since most of them are DAPI+. However, I don´t see any GFP signal (I use AF488 anti-GFP antibody).
Please some recommendations on how to solve this!!

Related Publications

Article
The green fluorescent protein (GFP) of Aequorea victoria is a unique in vivo reporter for monitoring dynamic processes in cells or organisms. As a fusion tag GFP can be used to localize proteins, to follow their movement or to study the dynamics of the subcellular compartments to which these proteins are targeted. Recent studies where GFP technolog...
Article
Intracellular molecules often localize specific environments or undergo specific movement in living cells. The visualization of molecular dynamics is very important to understand their function in cell biology. Green Fluorescent Protein (GFP) is often employed in protein dynamics quantification using Fluorescence Recovery After Photobleaching (FRAP...
Got a technical question?
Get high-quality answers from experts.