|
| |
I would certainly agree that several new model proteins are available to
replace lysozyme. Thaumatin and con-A are suggested, but are not
particularly desirable in our experience. The reasons are relatively
straightforward, and give some indication of what one should look for in
a model protein. In the case of thaumatin it is a) expensive and b)
requires an "unusual" agent - tartrate - for crystallization. Con-A a)
already crystallizes to very high resolution (a complaint often leveled
against lysozyme), b) requires two different metal ions (extra
dimensions in the solubility diagram), and c) - worst of all the
crystals are difficult to redissolve.
There are a number of factors we consider when looking at potential
model proteins, not the least of which is cost. We also prefer single
chain or monomeric proteins - you don't want to be worrying about the
proteins "normal" association-dissociation kinetics on top of even more
association-dissociation kinetics during growth or nucleation. Not
having cofactors or prosthetic groups is also desirable - they're
additional complications in the solubility diagram among other things.
Ditto for environmentally sensitive groups like a heme - just how
certain are you that ALL your heme's are in the same oxidation state?
With those that are not giving different protein configurations, etc.
We also tend to rule out proteases due to autolysis.
Also, the protein's preparation and crystallization should be as
"idiot-proof" as possible. There are a lot of people in the crystal
growth end of this field that are not biochemically competent (I still
vividly remember one call for help with lysozyme which eventually
contained the exclamation of "so that's why they dialyze their
protein!!").
Being an historically well studied protein is an asset from the crystal
growth side. It probably does not make the protein very interesting to
the crystallographic community (witness lysozyme), but having a lot of
ones background research already in the literature is very useful.
Proteins that we're currently looking at for models - GFP, alpha
amylase, serum albumins, PYP, con-A (despite its negatives hope springs
eternal and we give it a try on an almost yearly basis), RNase. We're
planning on going to recombinant proteins in much of our work.
Marc Pusey
--
Biophysics SD48
NASA/MSFC
Huntsville, AL 35812
|
