If you are obtaining dirty protien after an IMAC step this is the first step that needs looking at, any affinity step should provide you with clean protein, through adjustment of the tag (position and length) salt conc, imidazole washes and IMAC resin type. The next step is to track the purification with an activity assay and protein determination step, where do you lose protein? on the column or during concentration you should see as you purify the protein the specific activity increase. You can apply other affinity steps such as a Strep tag this can help or I have found a quick cleanup on a ion exchange before IMAC can help or an ammonium sulfate cut. I have found that when dirty protien is applied to a SEC then this can drag your protien out of solution as the impurities precipitate out of solution so clean the column and replace the filter before use. Alternatively your protien is just very sticky and bind heat shock protein etc if this is the case then its tough and you need it identify a method to dissociate the other protein....

Three purification steps are a lot for you? Poor kid, I guess you never really tried a real purification of endogenous protein.

But seriously, you cannot expect high purification fold from SEC. How does your chromatogram look? Do you get many peaks? How does it look on SDS-PAGE? It just may not make much sense to use SEC, because all the proteins you're left with are of similar size. That happened to me as well, so I couldn't resolve them.

The other option is, if you're purifying heterologous protein from E.coli that it's some chaperone bound and it won't your protein go just like that. We also had big troubles with that.

Anyway, the ionex should be of much bigger help than SEC.

@James Errey

you'll never get 100% pure enzyme after just one step. Even for such tags as Strep-tag they report about 95% purity. And with His-tag you will always get some endogenous proteins binding because of their amino acid content.

I see I hurt someone's feelings.

Is your protein dirty, or is it forming oligomers? This will also give multiple peaks from SEC. How many bands on SDS-PAGE versus peaks from your SEC?

protein is dirty. No multiple bands on Western (anti-His)