protein precipitates:HELP !

Clemens Suter-Crazzolara un691cs at
Thu Oct 13 05:48:24 EST 1994

I wrote:
> >I use over expression in E. coli to produce a HIS-tagged protein,
> >which is then denatured in urea and purified over a Ni-column.
> >The urea can be removed without problems, that is the protein does
> >not precipitate. This probably because I added a spot of Tween-20.
> >However, if I then concentrate the protein (I put the dialysis
> >tube on PEG15000) the protein precipitates like crazy.
> >I need this protein at very high concentrations, like 10
> >mg/ml (quite a lot).C
> >What can I do ? most literature says that top circumvent
> >precipitation, the best is to keep the protein in dilute form.
> >
Laura Lea answered:
>  Hi there, I also am having problems with my histidine tagged protein
> precipitating at concentrations greater than 2 mg/ml. I have Triton X
> 100 in my buffer, so I wonder if detergent is involved in the pptn. We
> are planning to try different PH conditions, we are currently at 7.3 and
> plan to go down to about 6.5, to see if this will help. Otherwise maybe
> adding a metal ion to coordinate the histidines?  Good luck with this.
> If you find a general solution please let me know. Thank you, Laura Lea

Inprincipal, I don't think it is the histidine tag that is causing the
problems. I am also testing the pH, currently at 8.9, will test the
more acid buffers also. Triton, or tween is apparently not so good:
CHAPS seems to be the detergent of choice. I added 10 mM to my buffer,
at different pHs it gets quite turbid: and my protein still precipitates.
On the other hand, I heard from a colleague that 1-2 M urea will keep
most proteins in solution, and they remain usually biologically active
(that is probably what you want too, isn't it ?). So this I will test
also: to first concentrate the protein, add urea to 6 M and slowly dialyse
down to 0 urea and check when it percipitates.
I have found no references in the literature yet about this problem. Perhaps
some other readers out there could help us ?

lots of luck,

clemens, heidelberg

More information about the Methods mailing list