you mean non-reducing but still denaturing conditions? (SDS in loading buffer and in the gel). The molecular weight markers if these are the condition are still reliable. They are generally denatured proteins, without any -S-S- that must be reduced. Why you see a band shift then...it is another question :) (interesting by the way)

Thank you for your answer Fulvio... you are right, they are in denaturing conditions. I'll look at other possibilities. Thank you again.

If you are using non denaturing conditions is possible that the tri dimensional structure of the protein is interfering with the migration or that your protein is forming complexes... That way the molecular weigh read in the gel is bigger than expected. there are markers for non-denaturing conditions

http://www.sigmaaldrich.com/etc/medialib/docs/Sigma/Bulletin/mwnd500bul.Par.0001.File.tmp/mwnd500bul.pdf

I need to know first how you calculated that the protein is 180 kD?

Try to stain your gel after you are done so that you are sure that this is the only band and there is no other band interefering. It might be the 3D structure is more resistible to electrophoresis.

Do you use a purified (recombinant) protein or a lysate ? Do you have any reducing agent to prevent intermolecular disulfide bonding? If you work with a lysate or crude fraction and do not use a reducing agent for storage you could have a smaller protein bound to your protein of interest via a intermolecular disulfide bond.

What is your basis for the correctness of the westernblot result? Do you know that your antibody really binds to your protein of interest? A lot of commercial antibodies are made against peptides and create problems when used for the detection of the corresponding protein. Why do you have to use a non-reducing blot in the first place? Or ist a human/animal serum of which you "think" that it binds to the protein of interest? Does the antibody supposedly not bind to the reduced form?

What is the basis for the calculation of 180 kDa? If it is based on the primary structure and a theoretical calculation: have you considered the mass of post-translational modifications adding to the apparent mass?

Last but not least: the apparent mass as estimated by SDS-PAGE is exactly that - an apparent mass.

It is hard to say. Maybe your protein is heavily glycosilated.

I would like to stress on the question of Jay Boniface , since some time proteins do not migrate accordig to the expected molecular weight it might be due to glycolysation and also the molecular weight we calculate depending upon the amino acid number do not give the exact molecular wiegt. Any way molecuar weight markers are free from disulfide bonds ( which are actually targeted by reducing agent) so the migration should not be affected. For your protein you have to see full litearture.

Happy working

Thank you all of you for your help. I have tried to do the same experiment I have found in some papers. I have purchased a positive control and it will tell me if I'm detecting the protein I expect. I made this question because I see (in papers) that the protein is detected at 180 kDa. Since I detect only one band at 220 I wonder if weight markers are suitable only for reduced samples o there are specific markers for these conditions. When the positve control arrive and I do the experiments I will tell you what I found. Indeed, it appears that the protein is glycosylated 180 kDa. The unglycosylated have a weight of 140. I have thought dimers, but it seems that the protein does not dimerize. I' tell you. Thank you all again!!!

Another thing: I use non-reduced conditions because it is an animal serum that only recognizes the non-reduced protein. This serum has auto-antibodies because an illness and it is published that only detects the non-reduced form. In my case it is also true, because I do not see the band at 220kDa when protein is reduced. The only difference is the weight. And I know the sera is positive for the auto-antibody because I have seen it by immunofluoresce. Thank you again!!

If it is a specific band, I would not worry about it too much. Marker is not always accurate anyway.