If done carefully, confocal microscopy tends to be quite reliable provided antibodies are specific enough.  

In subcellular fractionation, especially if homogenization or cell lysis is performed in small volume, cross-contamination of components from various compartments is a common issue. If you are seeing almost an equal amount in nuclear and cytosolic portion after subcellular fractionation, and given that you have other house keeping controls that track clean, cross contamination would seem unlikely. 

The logical troubleshooting step will be to test how specific is your antibody that is used in western blots and if washing conditions could be improved to eliminate non-specific binding, assuming that is the case. 

Best...

Hi Rüstem,

Could you just tell us how you permeabilized your cells for the IF staining? Are you sure your antibody had access to the nucleus? Because this might be the most obvious explanation I'd say...

Microscopic data have being annoying every biochemist and molecular biologist for a long while, as microscopists are cinical enough to tell (me, for example) that they can provide with whatever results you need, because they'are observing zillion of different images. Moreover, submitting "whatever"pictures to real researcher -biochemist and molecular biologist, they demanded to be not in a supporting cast technicians list, but only among main authors of every research, thus becoming "famous scientists" without even having any idea what scientific topics they are "famous" for.

Ilya Tsyrlov,

What you are stating mostly applies to incompetent people. Most competent people are more than familiar with limitations of each technique they employ in their work.

All techniques have their limitations, including subcellular fractionation, gel electrophoresis, transfer to membrane and detection by immunological methods.

Dear Dr Tsyrlov,

I'm sorry to hear that you appear to have had some bad experiences collaborating with "microscopists" before, but please don't throw away the baby with the bathwater. Not everyone using immunofluorescent stainings is fabricating results...

Immunofluorescence can be a powerful and revealing tool, and as long as it is backed up with biochemical or cytometric methods (depending on what you're looking at), I don't see any reason to be so pessimistic about it.

Even better: this topic started with a colleague who appears to have some troubles confirming IF with biochemistry or vice versa, and who is worried by this problem. A reflex that we should all encourage, no? And wouldn't you agree that providing evidence through two different and independent methods should be preferred over focusing on just one method? Therefore, I would suggest to use this forum to help fellow researchers with their practical problems rather than talk about personal frustrations. 

Hi, 

Thank your for your replies. 

I fixed the cells in 4% PFA and permeabilized with 0.2% Triton-X in PBS. I think there is no problem in access to nucleus because I had a nuclear staining of another protein- a transcription factor although I used a different antibody. Besides, using only GFP tagged protein (without any antibody usage) also shows similar cytoplasmic localization. 

Then, about western blot, I ve been using the antibody already quite a time now and I am positive about the specificity, so I doubt that it is a product of nonspecific binding exactly at the same size with my protein.  

Hi Rüstem,

Of course I could advise you to try some other fix/perm protocols (e.g. saponin, methanol or acetone) as their efficiency differs between antibodies and target proteins. But if indeed the GFP tagged protein is only visible in the cytoplasm, there should probably be a different explanation.

Would there be a possibility that your protein of interest is cytoplasmic but associated with (complexes at) the nuclear membrane?

Tausif Alam,

What the coincidence! The microscopist who said those immoral (and “immortal”) words about “whatever pictures”, is now working ,,,on the same University of Wisconsin-Madison campus, maybe just hundreds meters from you. It’s a laughing point. Two other points are serious, both making your “suggestions” look pretty wrong. First, the person is more than competent in microscopic field, as being  originated from Russian branch of Anitchkov school of morphologists. Secondly, my many talks with him (and some other competent microscopists) revealed their narrow thinking, all around limitations of each technique they employ in their work. Therefore, I’m sure that because microscopy is still the technical field of using microscopes to view objects and areas of objects, so the right place in scientific papers for you guys must be in the Acknowledgments section, with “thanks for technical support”.  

Ilya Tsyrlov,

In (relatively) objective scientific investigation if one can gather more evidence by using multiple tools/techniques, the higher the possibility that the conclusion reached is valid. If two techniques yield results that are not consistent and this is not for technical reasons, one needs to find or devise another way to resolve the issue.

You take a particularly dim view of microscopy, without even knowing whether or not Rustem's technique was flawed, which I find troubling. As for your general thrust about microscopy, similarly unenlightening comments can be made just about anything.

An honest and competent scientist doesn't look through fields to find a rare event that supports his preconceived notion. One tries to show a representative picture of the overall observation. 

Dear Pieter,

I dont know if it is associated with (complexes at) the nuclear membrane as it is a pretty  uncharacterized protein. Do you have an idea how to test it? Do you think if that is the case the protein would go into the nuclear fraction? At the end of the fractionation, I extract the nuclear proteins from the pelleted nuclei with a nuclear extraction buffer, then centrifuge to pellet nucleic acids etc. and use the supernatant as nuclear fraction.

I am very curious about figuring out whats going on, so thank you, I appreciate your ideas. 

Rustem, before extracting nuclear proteins, one should always wash the nuclear fraction to reduce cross-contamination with other cellular components. This can be done by gently resuspending the nuclear pellet in the same buffer you now use to homogenize the cells (usually an isotonic buffer without detergents) and pelleting the nuclei at the same low speed as you used the first time. This process can be repeated as many as 3 times. Your yield of nuclear pellet will decrease every cycle but at the end what you will retain will be substantially free of soluble cytosolic proteins that are not interacting and associated with nuclear membrane.

Best... 

A possible reason for the discrepancy is that some proteins will localize differently when the cell is in suspension (during the fractionation) or well spread (during microscopy). The 10 minutes it takes to trypsinize and wash your cells before putting them on ice and starting the fractionation could change the localization of your protein. We've seen this with a few different proteins.

Hi Rustem and everybody! have you solved your problem? because I have a very similar problem right now and I´m becoming desperated...

 Dear Kyle, could you tell me your articles about this topic? Thanks

Dear Marcelino, I havent "solved" the problem or done any additional tests, but based on my controls and considerations, I hold on to Kyle's response as the explanation. In case you figure out some other points, please let me know too. And good luck ;)

Hi Rüstem, you must check once by synchronizing your cell to S-phase before preparing slide their may be possibilty of cell cycle specific localization of your protein. in general after transfection it takes at least 24 hrs for slide preperation and over confluent cells may be arrested to G0/G1.