A common technique in a structural lab is to grow up a protein in bacteria and purify it. It's pretty much essential to all structural work (NMR of packed bacterial cell pellets is a different matter, and We Don't Do That). To make things easier for ourselves, some people like to grow 'their' protein fused to something else that makes purification easier.
For example, a protein that we'll call GST binds to a chemical known as glutathione. If you can immobilize glutathione on some sort of supporting matrix, then you can use this to pull GST out of a complex mixture of cellular protein (pretty specifically, in a lot of cases). So if you make a 'fusion protein', consisting of your protein linked to GST, you get a quick and simple way of purifying the thing you're after.
A complication arises if you then, as we often do, need to separate the GST from your protein. We do this by using specific proteases, that you can think of as a pair of molecular scissors which 'cut' the fusion protein at a certain place. In this example we have engineered (yes: "genetic engineering". It's What We Do) the fusion to contain this 'certain place' between the GST and your protein. So in theory you can bind the fusion to the supporting matrix/glutathione, treat with the relevant protease, and obtain just your protein.
There are, as ever, technical issues and nothing's ever that straightforward, but this is what goes on all the time in my lab.
Now, one of the proteases we use has the name human rhinovirus 3C protease. A certain company sells it to us under a tradename, let's call it "Scissors"™. Buying "Scissors"™ is expensive - it works out at around Aus$60 per prep. So the Black Queen finagled a plasmid from someone, worked out a method, and now makes our own "Scissors"™, for about a tenth of the price. Thing is of course that Certain Company gets really upset if we say "Scissors"™ — in a publication or whatever — and didn't actually buy it from them. BQ has been telling everyone that we can't call it "Scissors"™.
"Human rhinovirus 3C protease" is rather a mouthful, though: It's a lot easier to ask "Where's the "Scissors"™" than "Where's the human rhinovirus 3C protease?". To keep everyone happy, we've come up with a new name for it. We're now referring to human rhinovirus 3C protease as K-Zyme, which is easy to remember, say and type, and isn't a tradename. So anyone can use it.
And if you would like some K-Zyme, and instructions on how to make it, then apply to the usual place.
Comments
In our lab some folks were working on Chrystellase for a while before her paper came out and she was again free to call the enzyme by its proper name. Different reasons. Now I want to do GST pulldown just to request some reagents from y'all.
Posted by: Alethea | May 7, 2008 10:41 PM
ahh... protein purification. How I remember those sweet 'pouring' days. I think I spent more that half my time during one project trying to produce and purify those contructs we had. Purify over column and cleaving.... and no robots. Just a small PhDstudent ;)
Posted by: chall | May 8, 2008 02:31 AM
I'm doing something very similar with a protein called "BFF-MBP". Haven't worked out a protocol for purifying "Magic Xa" yet, as this whole "fusion-protein-thing" is very new to our lab. SWAM wouldn't approve it until I found one that we could cut off Every. Extra. Amino. Acid. Still in the "testing WT protocol and demonstrating perfect activity retention" phase. Previously, I had been doing purifications the "grow 250L of saturated culture and do a shit ton of columns" method.
Posted by: Maxwell's Demoness | May 8, 2008 06:52 AM
Y'know, MD, traditional methods are perfectly adequate if you get the expression right.
40 mg purified protein (two columns) from a litre of culture, and crystals from it? That should do you.
Contact me privately for help - with that or the fusion thing -, do.
Posted by: bk | May 8, 2008 10:12 AM
ahh... 40 mg from one litre of culture? sweet! One of my constructs wasn't really cooperating with me, if I remember correctly I had one liter of culture and 2 mgs in the end... lucky for me, that wasn't the important protein we needed for all the immunisations ;)
Posted by: chall | May 9, 2008 01:37 AM
I have no trouble with traditional methods, but 40mg per liter?!?! With my protein I am ***lucky*** to get 0.5mg per liter, often less. It's got a really low solubility and is highly toxic to the cell. Hence trying to work out the fusion, MBP is good at enhancing solubility, which seems to be working so far...
Posted by: Maxwell's Demoness | May 9, 2008 03:06 AM
Of course, every protein is different and 40 mg/l is unusual, if not the best I've ever had.
The trick is in the expression: get 30 - 50% of your total cell protein to be the protein you want and purification should be a doddle.
I strongly recommend people try auto-induction, and playing with the temperature. The thing about auto-induction is that the protein is not expressed until the culture is good and thick, and the cells are starting to die anyway. If expression at 20C does not give you soluble protein, then purify from the inclusion bodies and refold. With an enzyme, you've got a ready assay for the success of refolding (much more difficult to bring an enzyme back from the dead in the structural world).
And if that doesn't work - try different constructs, different species, if those options are available. Like I say, getting good expression is the key to a happy protein chemist.
Posted by: bk | May 9, 2008 09:34 AM