Many thanks Jacob and Mark for your questions/suggestions. In response: > So what happened with the non-reducing gel? (If the DTT was fresh, there > should be no problem, but if not...) The gel is very clear, it shows the same exact pattern as the reducing gels. Both high and low MW fractions run at the same MW. I used -80C stocks of the native protein for which the DTT was buffer exchanged out (3, ~15-fold "exchanges" via concentration/dilution). Technically, there is a tiny amount in there. But I don't think it's disulfides based on this experiment. Concerning the freshness of added DTT, it's pretty much as fresh as our solid DTT at -20C is fresh. I make all buffers containing DTT just before use and I do it by first adding all the components to ddH2O except the DTT; then I put the buffer at -20C until it gets cold ~4C. Then I pH it and add the DTT and move it to the column.
>you're right, I read too quickly over the 2 orders, and understood "2-fold". >Is it possible the oligomer peak contains higher aggregates that inhibit >crystallisation and do not >separate well on your gel filtration column? In >that case, perhaps very high-g centrifugation could remove them, or a gel >filtration column that separates better in that MW-range. >Or perhaps the tail >fractions of the oligomer peak on your current column might be already >conformationally pure enough. Concerning this point made by Mark, there might be something to this. On gels these fractions do tend to appear slightly "dirty" in that there are a couple very faint bands at high MW. Also, When I rerun individual peak fractions back through gel filtration, the high MW samples sometime result in additional "light-absorbing matter" at even higher MW closer to the void fractions. Crystallization drops for the high MW fractions appear less soluble even at lower concentrations than the lower MW fractions. However, the unclipped, lower MW samples also are less soluble than their clipped counterparts (which raises the issue of the importance of the tag). I'm still playing with optimization of these samples as well because who knows really.
