persistent 70kDa contaminant (dnaK?)

[R.N. Leach]

I'm expressing a soluble 38 kDa 6His fusion protein in E.coli. I've found that 
an ammonium sulfate cut of Ni2+-NTA column fractions followed by HIC on 
butyl-sepharose resolves my protein from all contaminants EXCEPT for a 
persistent ~70kDa protein. I'm assuming this is the chaperone-like dnaK. Prior 
washing of the Ni2+-NTA column (at 4degreesC) with a buffer containing Mg2+ 
and ATP appears to show some success in removing some of the 70kDa protein 
(another indication it may be dnaK) but a lot of it is obviously still 
remaining associated with my protein. How can I break this association? Does 
it require more than just Mg2+ATP? If anybody has experience of this problem 
I'd be glad of any advice.

Rob Leach 
School of Biochemistry & Molecular Biology
University of Leeds