Heavy Post-translational modifications - Multiple bands on gel (Jun/28/2011 )

Dear Bioforums,

I have YFP tagged proteins comprising 4 members of the same membrane-protein family. I am expressing them in HEK293 cells, solubilizing in triton, pulling them down with a rabbit anti-GFP antibody and blotting with a mouse anti-GFP. For two of the four proteins, as well as having a couple of bands around the expected point for un-modified and N-glycosylated forms (around 50kDa) there is a host of heavier, reproducible bands at around 70, 150 and heavier ~300kDa?, (beyond my ladder). The two members without bands act as good controls showing no cross reaction of antibodies.

For a few reasons I suspected it might be ubiquitination, but mutation of lysines and over-expression of mutant ubiquitin has no effect. I presume this also rules out things like neddylation and sumoylation. The samples are prepared by boiling for 3minutes in Laemmli buffer which presumably will remove disulphide bridges and prevent aggregation of the proteins (results are the same with 10 minutes boiling).

The only other thing I could think of was some kind of proteoglycan attachment (chondroitin, heparan), but as far as I can tell these only attach to protein regions with a big clump of acidic bases which my proteins don't have.

Any other suggestions about what modifications I might have (or what errors I've made!!) would be much appreciated,
Thanks,
CR

I don't know if this is a valid concern or not since I have never done IP before. If I am totally in left field then at least I will learn something new about IP

Is it possible that some portion of IgG or IgG+Protein A/G are still complexed with your protein when you run the gel? I know that IgG will degrade into its light/heavy chain components when it is subjected to reducing conditions (βME or DTT and boiling). If for some reason those complexes aren't degraded, fragments of Ab and/or Protein A/G still attached to your protein could maybe explain your mw increases.

proteaMatt on Tue Jun 28 14:17:54 2011 said:

Thanks for posting Matt, I hope that my sample preparation conditions would preclude this possibility though. The link between the antibody and target protein or the link between Ab and Protein G are non-covalent and normally also dependent on tertiary protein structure. Both complex structure (boiling) and non-covalent interactions (strong detergents in the buffer) should be removed.

I am definitely not ruling any artefacts 100% out at this stage though. I am currently trying with PNGase F to totally remove N-linked glycosylations. Should know by this time tomorrow if this helps the diagnosis.
CR