:-) guillaume your protein probably precipitates on the column...300mM imidazole is already a lot...

if you put 5mM EDTA over the column, the column will turn white (the c02+ has been eluted and the likely bound prot!) and the eluate will be pink...if the prot is not in this eluate...it means the protein is truly precipitates..so if the protein is soluble in the extraction buffer (when i mean soluble...it must stay in the SN after 45min a 45000rpm) , keep this buffer along your purification including the elution...and think about how to get rid of the triton (i bet this is what you want to get rid ) after the elution...on peut en discuter

Dear Guillaume,

two things that spontaneously came to my mind are ...

1) I agree with Didier about the possibility of precipitation.

2) If it is not precipitation but too strong binding, you could try to shorten the His-tag perhaps from six to four His residues? I would be strongly interested to know if this approach works.

Good luck!

 Dear Daniel,

I doubt that strong binding could be involved. Indeed, 6His-GFP is perfectly eluted at 150 mM imidazole (obtained 2 mg from 250 ml infected Sf9). I'm clarifying my lysates at 60 000 g right now to see if my protein precipitates within those conditions. If not I will use same buffer all over the purification process and if it doesn't change anything then will try to strip the column with EDTA.

Thanks anyway and if any of you get other ideas you're welcome to share them.

Cheers.

is sf9 expression absolutly required (any ptm involved!)..if not, parallel cloning and ecoli may give some results.see

Yes Sf9 are required because it's highly degraded in bugs and we think not properly folded. After having done some tests here are the results (I'll send you an image later). After spinning my lysates at around 68 000 g my protein seems to still be there. Using the same buffer during all the procedure allowed me to recover 50% of my protein by elution with 150 mM imidazole. An additional elution with 300 mM imidazole doesn't improve anything. How did I estimate those 50%? Well after all those elution I striped cobalt with 5 mM EDTA analysed what I removed and what was left on beads. Treatment with EDTA doesn't elute anything and I retrieved the remaining 50% on the beads... In parallel I purified 6His-GFP and did not any of those problems, everything is eluted as it should.

Very strange then. How my protein can interact non specifically with my Cobalt beads while it's perfectly soluble?