We re-cloned some of our cytoplasmic His-tagged genes into pTT-based vectors for expression in 293-6E cells but when we try to purify them via Ni2+ IMAC at small scale from whole cell lysates we always see major contamination with 4 or 5 proteins, SPQF (11 histidines in one region AND a 3C protease sensitive site!) and its partner NONO being the major bands. We have seen the same protein profiles also in 293/HEK grown in roller bottles. This level of contamination makes screening almost imposible if we have a protein oner 50kDa.
Does anyone else expression screen/purify cytoplasmic proteins from HEK cells via IMAC or was this just a stupid thing to do?
Any tips, tricks or perfectly blunt opinions would be appreciated-different lysis/IMAC binding conditions, different ion etc., otherwise it will be back to the cloning into OneStrep tag-encoding vectors for us.
I'm afraid I don't have a quick solution for your problem. Just a suggestion:
Is it possible to do a quick screen for conditions where those proteins bind less tightly? E.g. You could use a lysate in aliquots at 4-6 different buffers with lower pH than usual (e.g. pH 6.3, 6.5, 6.7, 6.9, ..) and maybe also different salt conc. (later). Binding and elution behaviour should be different under these conditions. You might lose some of your target protein binding but not all, since I would assume that the His-cluster in the His-tag behaves slightly different (i.e. binds more tightly) than the histidines in those clusters. Probably His-10 would be even better than His-6.
I hope that helps.
We had the exact same problem with protein expressed in HEK-cells (not 6E) with an N-terminal his-tag (6xhis). In our case the target protein was expressed in low amounts, so it was hard to do additional purification without loosing too much. We tried several different things, purification on Co-beads, ion exchange, SEC, tag-cleavage and reverse IMAC etc, but none of the approaches worked really well. For a while our solution was to do an initial ion exchange, where we got rid of NONO and then continue with IMAC. But the main problem for us was the low yield, which made it difficult to get any pure protein in the end, so our final solution was to change to the OneStrep tag as you suggest yourself.
Let me know if you are interested in any specifics (gels, column running conditions etc).