Separating GST from GST-fusion protein

Allison via (by allison from
Mon Apr 20 09:44:29 EST 2009

DK wrote:
> In article <mailman.899.1239820815.13724.methods from>, "Paul J. Phelan" <Paul.Phelan from> wrote:
>> I think I asked this question before on this forum, but I thought I'd 
>> try it again just in case anyone has any new "tricks" or ideas, because 
>> I'm having the same old problem again.  I am trying to purify a 
>> GST-fusion protein, and it's contaminated with GST.  The GST-fusion has 
>> a mol. wt. of 40,000 and GST is 26,000, but on a Sephacryl S100 gel 
>> filtration column, both proteins are co-eluting in a single, beautiful 
>> peak.  Anyone would think that such a textbook peak would only contain 
>> pure protein, but no - half of it is GST!  I even re-ran the gel 
>> filtration column in a buffer containing 0.4 M NaCl to try to separate 
>> the two proteins, but I got the same single peak, so the GST must be 
>> still there.
>> Now I am not talking about digesting a GST-fusion and then removing 
>> GST; I have no problem doing that, but I am trying to purify some 
>> intact GST-fusion, without taking GST with it: does anyone have a magic 
>> trick to get rid of GST?
> Because GST forms dimers (which is surprisingly little known and 
> responsible for probably countless incorrect conclusions in published 
> papers), you can only do it two ways: 1) cleave off GST if your 
> expression construct allows it, 2) run chromatography that 
> separates GSG from your GST fusion in the presence of high urea 
> concentration (at least 1M and as high as 2M may be necessary). 
> This is only an option of your fusion partner survives such 
> conditions. 
> DK

Getting rid of trace GST after cleavage is not trivial IMO.   After 
expressing a GST fusion protein we cleaved away the GST using a thrombin 
cleavage site and ran an affinity column to get rid of uncleaved fusion 
protein and free GST.  The resulting flowthrough has a single band on 
Coomassie blue stained gel but still has GST present as measured by ELISA.
We've switched to plan B - a different construct.


More information about the Methods mailing list