Immunofluorescence and Western Blot don't match - (Apr/02/2009 )
Hi,
I have read a couple of topics where people (Lab rat and Telomerase) mentioned that they have detected their protein in the nucleus by immunofluorescence, but when they tried to confirmed by Western Blot, they only detected it in the cytoplasmic fraction. I HA tagged my protein of interest and I can detect it in the cytoplasm and in the nucleus by IF (z-stack confirmed). However, when I performed nuclear fractionation on the rest of the cells (i.e. I put coverslips on my plates for the IF, the rest is used for extracts), my protein was barely detected in the nucleus (and all fractionation controls were good; tubulin only detected in the cytoplasm and Histone H4 only in the nucleus). Has anyone heard about proteins leaking out of the nucleus during fractionation? I don't understand how come with such a strong nuclear signal by IF (with HA antibody), I cannot detect it by western. I have done nuclear extracts for 8 years and it is the first time I see that, although it seams to have happened before! 
Thank you very much.
yes - it depends on the type of antibody.
some antibodies are not suitable for WB but they work very fine for IF, or opposite. you would need to check the company info for any antibody you buy.
Curtis on Apr 3 2009, 11:09 AM said:
some antibodies are not suitable for WB but they work very fine for IF, or opposite. you would need to check the company info for any antibody you buy.
People routinely misuse and misunderstand cytoplasmic/nuclear fractionation. Many proteins, even those that are strongly nuclear by IF, can be found in cytoplasmic fractions. The reason is that this type of cytoplasmic/nuclear extract isolation is very leaky. Only proteins associated tightly with nuclear structures (such as DNA, export channels, etc.) will stay in the nuclear fraction. Proteins that are nuclear but not tightly associated with nuclear structures will end up in the "cytoplasmic" fraction. A better way to think of it is cytosolic fraction (which includes the nuclear compartment cytosol vs cytoplasmic) and nuclear extract. A good example is the aryl hydrocarbon receptor. In response to ligand, the AHR rapidly moves to the nucleus. If its dimerization partner, ARNT, is present, the AHR binds to ARNT and binds to DNA and can be found predominantly in nuclear extracts. In the absence of its partner, the AHR still moves to the nucleus (as it still contains an NLS and is transported by the importins), but as it does not heterodimerize or bind DNA, this nuclear AHR cannot be found in nuclear extracts though it is clearly nuclear by microscopy. Don't fall prey to the fractionation myth--look at all data!
The antibodies are good and work for both IF and WB (I have a good signal in the cytosolic fraction). I saw it with both a Flag and HA antibodies.
I agree with you Dr. Teeth, it makes sense.. I do believe my IF data, the problem is that I want to study the role of my protein in the nucleus... so I have to be able to isolate it from there. I guess I could crosslink before doing my extracts.. or do my experiments with total extracts, but then nuclear binding partners get diluted with the cytoplasmic ones.
I will read the AHR literature... thanks a lot!
Dr Teeth on Apr 3 2009, 07:48 AM said:
Curtis on Apr 3 2009, 11:09 AM said:
some antibodies are not suitable for WB but they work very fine for IF, or opposite. you would need to check the company info for any antibody you buy.
People routinely misuse and misunderstand cytoplasmic/nuclear fractionation. Many proteins, even those that are strongly nuclear by IF, can be found in cytoplasmic fractions. The reason is that this type of cytoplasmic/nuclear extract isolation is very leaky. Only proteins associated tightly with nuclear structures (such as DNA, export channels, etc.) will stay in the nuclear fraction. Proteins that are nuclear but not tightly associated with nuclear structures will end up in the "cytoplasmic" fraction. A better way to think of it is cytosolic fraction (which includes the nuclear compartment cytosol vs cytoplasmic) and nuclear extract. A good example is the aryl hydrocarbon receptor. In response to ligand, the AHR rapidly moves to the nucleus. If its dimerization partner, ARNT, is present, the AHR binds to ARNT and binds to DNA and can be found predominantly in nuclear extracts. In the absence of its partner, the AHR still moves to the nucleus (as it still contains an NLS and is transported by the importins), but as it does not heterodimerize or bind DNA, this nuclear AHR cannot be found in nuclear extracts though it is clearly nuclear by microscopy. Don't fall prey to the fractionation myth--look at all data!
OliM on Apr 2 2009, 06:37 PM said:
I have read a couple of topics where people (Lab rat and Telomerase) mentioned that they have detected their protein in the nucleus by immunofluorescence, but when they tried to confirmed by Western Blot, they only detected it in the cytoplasmic fraction. I HA tagged my protein of interest and I can detect it in the cytoplasm and in the nucleus by IF (z-stack confirmed). However, when I performed nuclear fractionation on the rest of the cells (i.e. I put coverslips on my plates for the IF, the rest is used for extracts), my protein was barely detected in the nucleus (and all fractionation controls were good; tubulin only detected in the cytoplasm and Histone H4 only in the nucleus). Has anyone heard about proteins leaking out of the nucleus during fractionation? I don't understand how come with such a strong nuclear signal by IF (with HA antibody), I cannot detect it by western. I have done nuclear extracts for 8 years and it is the first time I see that, although it seams to have happened before!
Thank you very much.
try subcellular fractionation
that is what I did.. I am just looking for a protocol that will prevent nuclear protein leakage out of the nucleus during fractionation. I think I will try to crosslink before, hoping that my protein binds to a partner that will keep it in the nuclear fraction... If not, I have to find a condition where it binds tightly to DNA or the matrix (like for AHR in presence of its ligand).
MaggieRoara on Apr 3 2009, 07:49 PM said:
OliM on Apr 2 2009, 06:37 PM said:
I have read a couple of topics where people (Lab rat and Telomerase) mentioned that they have detected their protein in the nucleus by immunofluorescence, but when they tried to confirmed by Western Blot, they only detected it in the cytoplasmic fraction. I HA tagged my protein of interest and I can detect it in the cytoplasm and in the nucleus by IF (z-stack confirmed). However, when I performed nuclear fractionation on the rest of the cells (i.e. I put coverslips on my plates for the IF, the rest is used for extracts), my protein was barely detected in the nucleus (and all fractionation controls were good; tubulin only detected in the cytoplasm and Histone H4 only in the nucleus). Has anyone heard about proteins leaking out of the nucleus during fractionation? I don't understand how come with such a strong nuclear signal by IF (with HA antibody), I cannot detect it by western. I have done nuclear extracts for 8 years and it is the first time I see that, although it seams to have happened before!
Thank you very much.
try subcellular fractionation