[irrumator11]

Hi all,

I'm doing a multi-step purification to purify a protein (or proteins!) responsible for a particular biochemical activity. I do not know the identity of the protein, so I am doing the purification by following the activity only. Here is my purification table.

Step	Total protein   (mg)	Specific   activity   (units/mg)	Yield (%)	Increase   in purification
Processed crude	57.6	35	100	=1
Ammonium sulfate	36.9	74	87	2.13
Phenyl   sepharose	6.64	1560	29	44.9
Hydroxyapatite	2.8	1463	7.9	42.1
Heparin   sepharose	0.46	2215	2.5	63.8
Gel   filtration	0.082	2356	1.9	67.8

As you can see, something funky is happening with the Hydroxyapatite column, because the specific activity does not increase at that step. I am stumped as to what that can mean. Is the hydroxyapatite column fractionating out a cofactor or another protein that is necessary for activity? This is my first time doing a purification like this. Is a yield of 1.9% considered colossally shitty? It seems I am losing the most yield at the Phenyl sepharose step -- does that mean I should eliminate that column from my purification?Any help is appreciated! Thanks in advance!

[Arkcon]

A fundamental fact is the more you manipulate a biological chemical, the more risk you have of degrading it.  Either by oxidation in the air, temperature effects, or reaction to the pH of the medium or the eluents needed.  You also could definitely be purifying out a co-factor, or concentrating an inhibitor, or encouraging the protein to bind to itself, blocking it's own active site.  You don't have to do any step that you think may be the cause of a problem.  There's certainly a value in trying to make a guess about the structure of your protein, and selecting which purification methods to include or exclude based on that.  If you guess wrong, you've at least excluded something.

[irrumator11]

Thanks for your reply, Arkcon.  I will modify my selection of columns for my next purification.

However, if the hydroxyapatite column is separating out a co-factor, then I would definitely be interested in pursuing that.  I guess I can run my sample through hydroxyappatite, find the active fractions, and then add the "inactive" fractions to it one by one in my assay to see if I can get the activity to go even higher.

Any other potential reasons as to what could be happening during the hydroxyapatite step, or why the overall yield is so low?

[Yggdrasil]

I agree that if your hydroxyapatite column is removing a cofactor, you can find that cofactor by adding inactive fractions back to your active fractions to see whether any of the inactive fractions contain such a cofactor.

It could also be the case that the hydroxyapatite column is not really doing much to separate your protein or complex of interest from the contaminating proteins.  Similarly, it looks as if the gel filtration step is not really helping improve the purity of your protein as well.  Have you analyzed the fractions after each step with SDS-PAGE?  Is your purified sample after all of the purification steps still a complex mixture or do you recover a single band or a few bands that seem to be present in stoichiometric amounts?  If multiple proteins are present in your sample, did they all elute from the gel filtration column together at a retention volume that would suggest they are forming a stable complex?

It seems like you could try just a three-step purification consisting of ammonium sulfate precipitation, a phenyl sepharose column, and a heparin column.  Interestingly, these steps that seem to help increase the specific activity the most also seem to affect the yield the most.

[irrumator11]

Yggdrasil, I agree with you that the final gel filtration step is also not very hopeful in terms of increasing specific activity.

I only checked my fractions by SDS-PAGE after the gel filtration step, and I see a single band by silver-staining.  The location of the band correlates well with the single peak I see on the gel filtration chromatogram.
I sent the protein out for mass-spec, and it turned out to be GAPDH (although the activity we're assaying for is novel). GAPDH uses NAD as a cofactor.  Now I am wondering whether gel filtration separated out free NAD.  In fact, wouldn't hydroxyapatite also bind free NAD?  Maybe that's what I am fractionating out.