Epitope tags are often considered fool-proof methods to avoid having to raise your own antibodies, and we have tried to use epitope tags to compare the abundance of receptors and ligands in vivo. Whilst doing so, we noticed that the popular HA tag changes its affinity to anti-HA antibodies depending on the context. For instance, a C-terminally HA tagged reporter protein was detected with a certain affinity, but adding just a small tetrapeptide between the HA tag and the stop codon reduced the sensitivity of the western blot by a factor 10. Neither HA tagging nor the addition of the tetrapeptide affectewd the specific enzyme activity of the reporter protein, but in the westernblot (SDS-denatured proteins) the signal was 10-fold reduced when the HA tag was not completely at the C-terminus. I understand that epitope tags can be context dependent when doing immmunoprecipitations of native proteins, but after boiling in SDS and running an SDS PAGE, the proteins should be unfolded. Has anybody else seen similar context dependencies, and if yes, which epitope tag was used? Are there any epitope tags available that are more robust than the HA tag? Any suggestions or free discussions are most appreciated..
Hi Jürgen,
so far if have only been sometimes facing significant differences in expresssion levels due to N- or C-terminal localisation of epitope-tags like HA-, FLAG- and myc-tags. Since you added a tetrapeptide N-terminal to your HA-tag of your protein of interest, this might be causing the problems of detecting the HA-tag. You might consider to add the tetrapeptide ether c-terminally with your HA-tag at the N-terminus, or the other way around. But as long as you don't have an antibody detecting you POI directly you gonna have pretty much difficulties to proof antibody-based evidence when only relying on epitope tags and the overexpression of epitope-tagged proteins.
Best,
Kai
Thanks Kai, the problem is that I want to detect both the ligand and the receptor with the same antibody, originally with the idea that the two totally different proteins could be detected with the same affinity, so that we can work out receptor-ligand ratios in vivo. Making an antibody to each of them is comparing apples and pears, because every antibody is different, and every purified protein is different, particularly when comparing a soluble ligand with a 7 transmembrane domain receptor. So we thought, let's tag both receptor and ligand with the same epitope tag, but it turns out that even with the same protein (the soluble ligand), the tagging works different depending on whether the receptor-binding site (the tetrapeptide) is present or absent. It means that although we can detect both ligand and receptor with the HA antibodies, we can't trust that we are seeing equal numbers of molecules when the signals on westerns are the same.
Dear Jürgen,
The context of the tag you described is not surprising me. With one protein I had an experience was possible to detect by western only after stripping of the previous antibody. The harsh stripping procedure (SDS, BME, 70 degrees) suggested me than some epitopes of my protein were "visualised" for the antibody during this procedure (it could be the trick for you). I agree you boil the protein, than SDS-PAGE, transfer with methanol, are the conditions does not allow to renature the protein. But when it reaches the membrane you immediately switch into the native conditions. It is not excluded at this step the proteins are bound partially to the membrane can form the structures at their tails if they are not bound. This are my speculations only, hope it will help you to think about the differences you observed with your protein tagged with/without the linker. I was searching for the picture from SEM, to find if the proteins bind the membrane in the linear or rather globular form... It would judge my speculations if found :(
Regards - Jan
Hi Jan, yes that's definitely an avenue I'll explore, excellent suggestion. If the protein partially re-folds after transferring to nitrocellulose, even 4 amino acids could partially mask the tag. I have noticed in the past that staining a nitrocellulose with ponceau red can improve subsequent western blotting signals. So may be a stripping protocol as pre-treatment to primary antibody could help. We are lucky that our ligand construct is a very robust enzyme reporter and so we know how much we have loaded in each lane. It still worries me that tagging a soluble protein and a membrane spanning protein may not give equal affinity to the anti-tag antibody... We may have to repeat the experiment with different tags to have a better chance to not fool ourselves...
Dear Jürgen,
Completely agree with improvement after Ponceau, since TCA or acetic acid is there. Anyway two samples loaded in multiplication, the strips after slicing, you can check at least ten conditions in one experiment. Only I’m afraid is refolding of the small HA… On other hand big tag with polyclonal antibodies (GST etc) will destroy your receptors and ligands…
Keep fingers cross - jan
- Badges
- Science method