covalent-bound HIS complex can not be pull down: help needed - (Aug/20/2012 )

Hi again,
I am trying to purify some protein for mass spectrometry and I have a problem. My protein has FLAG and HIS tags. When I check the supernatant of my cells (it is a secreted protein) by western-blot (after SDS-PAGE under denaturating conditions) I can see the main band of but, under certain conditions, I also see a higher molecular weight band. Another proteins from this family are reported to make covalent complexes with other proteins so I imaging that is what is happening with our protein of interest. This is extremelly interesting so I decided to pull down the protein using the HIS tag (to avoid having problems with heavy and light chains from anti-flag antibodies). The problem is that although I am able to pull-down the main band (single protein) the complex does not bind to the colum (well... only a tiny fraction of it can be detected after pull-down). I tried different buffers, salt, detergent, temperatures, etc, etc... and the result does not change.
I wonder if highly packed complexes bound by covalent bonds could, somehow, make the HIS tag unaccesible for the resin.... Does anyone experience something similar?
If I use denaturing conditions for the pull down (e.g. 8M Urea), do you think I would be able of pull the complex down? I used urea before for pull down a membrane protein but I do not know if covalent complex are urea-resistant...
Thank you for your help

covalent complexes should be urea resistant. It might be that the His-tag is unaccesible to the resin if the protein terminus on which the tag is, is involved in the interaction. I had this problem before and I solved it by having the His either on the other end or linked to the protein with a long linker ((GS)n). But this needs recloning. However, in your case, urea might do the trick.

Andreea

Thanks a lot Andreea,
C-terminus is my only choice because a tag in the N-terminus influences localisation and function of the protein... it is a secrerted protein and there is a signal peptide, so N-terminus tag is not an option. I am making, just in case, a new construct having an internal tag between the signal peptide and the N-terminus of the secreted, processed protein. In the meantime, I will try with UREA and I hope it works.
Best