Western Blot and Inmunofluorescence contradictions - (Apr/28/2014 )
Hi there,
I've been analyzing for a few months a protein both by western-blot and by inmunofluorescence and I'm having what it seems to be contradictory results. Let me explain:
-When I generate protein extracts by cell-fractionation (cytoplasm-nucleoplasm-chromatin) and analyze them by western blot i detect my protein at all the 3 fractions. As a control for the fractionation I use GADPH for cytolplasm (mostly cytoplasm with a minor staining at nucleoplasm but I've heard that there's always minor contamination) and ORC2 for nuclei (mostly chromatin but also in nucleoplasm).
-When performing an IF protocol (formaldehid fixation) I only see staining at the cytoplasm.
-Also the antibody I use for WB and IF it also works for inmunoprecipitation so we can rule out that the antibody is not recognizing the native protein.
I dont know how to explain it. Maybe i have to modify the IF protocol or the cell-fractionation?
Thank you in advance for your time and answers,
Gonzalo
No need to double post - other one deleted. Bob.
It could be that the cells aren't properly permeabilized by your protocol. What conditions do you use. Have you tried other fixation/permeablization methods? Is your antibody capable of IP and/or detecting native protein? Are you using the same concentration for IF as you are for the western blotting? How do your controls look?
For permeabilization I use Triton 0.5% in 5% BSA 10 minutes a RT and gently mixing after fixation with 4% formaldehid in PBS during 15 minutes. I haven't tried other protocols but this is some options I have in mind. Yes, the antiboy recognizes the native protein since it works for IP. The concentration I use for IF is higher than the one used for WB.
PS: Sorry for the double post.
Higher concentrations of antibody can lead to artifacts where there is staining but it is non-specific/off-target, similar to what you would see when you do a western with too high a concentration of antibody. Ideally you would use the same concentration as you are using for your western, that you know gives you specific banding.
But the problem is that I'm seeing by WB a localization of my protein that doesn't match de IF results. If the concentration used for the WB (I use 1/2000) was higher than for the IF (I use 1/500) I could understand that the localization in the nucleus by WB could be an artifact due to excess of antibody, but that's not the case. I don't know if equalling concentrations for both experiments could solve this problem.
PS: Thanks for your answers.
The artifact could be that the staining of the cytoplasm is so intense that you don't easily see the nuclear staining...
Your other conditions look fine. You could try up to 1/2 an hour for the permeabilization, but that shouldn't have a major effect, I use a 10 min one myself.
If you haven't validated your antibody there is the possibility that what you think is IP is actually non-specific binding - the simplest way to test is to tag (e.g. FLAG, HA) your GOI in a plasmid and transfect, then pull down with your AB and detect for the presence of the tag.
About the antibody: i have the protein with HA and His-tag. What I've done is to pull down with HA and detect with my AB and I saw an specific band, so I think I've validated the AB, true?
What I could try is to perform an IF with a pre-extraction step before fixing. When i do that i use this protocol: Triton 0.5% in 5% BSA 5 minutes at RT and gently mixing. But i don't know if it will work, I'll let you know. Do you think changing how I fix he cells could also work?
ghernandez on Tue Apr 29 09:10:24 2014 said:
the concentration used for the WB (I use 1/2000) was higher than for the IF (I use 1/500)
1/500 is 4x 1/2000. you're confusing dilution factor with concentration.
mdfenko on Tue Apr 29 12:26:07 2014 said:
ghernandez on Tue Apr 29 09:10:24 2014 said:
the concentration used for the WB (I use 1/2000) was higher than for the IF (I use 1/500)
1/500 is 4x 1/2000. you're confusing dilution factor with concentration.
True, but in this case are equivalent.
It does sound like your antibody is specific.
I don't really know about your pre-extraction idea. It may be a viable alternative, but it sounds tricky - you may well rupture the cells too much for them to be usable. I would try using a different fix, and you could also try some different incubation buffers. If you are using PBS and the protein is a phosphoprotein (could be that the nuclear stuff is phosphorylated?), then you should switch to TBS (phosphates in PBS inhibit binding to the phosphoproteins), but there are a few others that may work better as well.