Michael Myers myersm at
Tue Jul 6 19:28:42 EST 1993

In article <munro.741763893 at sbsu1> munro at (Gordon Munro) writes:
>sbhattac at (Dr. S. Bhattacharya) writes:
>>I'm affinity purifying a GST-fusion protein expressed in E.coli using a 
>>Glutathione-sepharose 4B matrix from Pharmacia.  I get a nice 52 kd band
>>which I've blotted and is a fusion protein as it is detected by an anti-GST
>>antibody.  I also get a 69 kd band co-purifying; this is not a fusion protein.
>>It does not appear when I express the GST alone. Changing the cell type
>>(DH5alpha and TOPP-1) does not affectit.  Washing the matrix (before elution)
>>with increasing stringency of NaCl (upto 1M) also does not help. Pharmacia
>>tell me that this hasn't been reported to them before. 
>>Has anyone else had a similar experience, I would be glad to hear.  Also any 
>>suggestions on how I can get rid of the extra band would be much appreciated.

I used the GST system quite a bit to purify various deletion derivatives of 
human Hsp70 expressed in E coli. I often found a ~70 kd protein co-purifying
with the fusion. Western blotting confirmed my suspicions that this protein 
was DnaK, the E coli homologue of hsp70. We later found that this was probably
not specific to 70 heat shock proteins, as another lab using the GST system 
to purify a non-heat shock related protein routinely saw co-purification of
DnaK. My ultimate conclusion was that some of the GST fusions do not fold well
and therefore tend to be bound by DnaK. For example: I tested the ability of DnaK to bind directly to glutathione-agarose (not trivial since it's a tripeptide
and 70k hsps can bind to some peptides). This indicated that the DnaK had be 
binding to the fusion protein or one of its respective "halves". After cleavage
with thrombin, I passed the mixture of proteins over a gel filtration column and found that DnaK eluted by itself. This suggested that DnaK only recognized the
intact fusion protein, not GST nor the released fragment of human Hsp70.Of course its possible that the lifetime of a DnaK-GST or DnaK-HumHsp70 complex was shorter than the duration of the treatments and chromatography, but that's what I got.

Now for the practical part: If your 70 kD band is really DnaK, you can quite
easily adsorb it out by passing the material over ATP-agarose (make sure you useresin with linkage to C-8 of the ATP molecule).I hope your protein doesn't bind

Michael Myers
The Laboratory of Genetics
The Rockefeller University
myersm at

More information about the Methods mailing list