Hi John,
This is really an amazing wild west story: the man
who crystallizes faster than his protease! I really must compliment you with how
you successfully performed these experiments!
Unfortunately, proteins usually do not crystallize that
fast (at least not in my hands), so in these cases other methods have to be
used. As has mentioned before, protease inhibitors are the way to go. Especially
with autolysis, as one protease cuts another one, the speed of the reaction goes
with the square of the protease concentration. Whereas in dilute solutions not
much happens, as soon as you start to concentrate towards crystallization
conditions, say 10 mg/ml, degradation suddenly goes very fast.
There
are 2 cases to consider:
1) the
protein you want to crystallize is a protease and is destroying itself. In
this case you need to cocrystallize with a potent and specific inhibitor. With
serine proteases, Wolfram Bode was very successful by using
chloromethylketone-containing peptides (e.g. PPACK). These compounds would make
covalent links with both the active site serine and histidine, effectively
killing any protease activity.
2) the
protein you want to crystallize is not a protease and it is a contaminant which
is causing the problems. In this case I would add a protease inhibitor coctail
in an earlier step of the purification to block the protease before the final
purification steps. I would also add some small broad protease inhibitor e.g.
PMSF to the protein solution used for crystallization.
Herman
From: CCP4 bulletin board
[mailto:[log in to unmask]] On Behalf Of John Domsic
Sent:
Wednesday, January 16, 2013 2:22 PM
To:
[log in to unmask]
Subject: Re: [ccp4bb] protein degradation in
crystal
Hi Lisa,
Speed is definitely a big factor here. With a
protein I work with I can get large crystals in myriad conditions that only
diffract to about 4-5 Ang. What I ended up doing was taking these
crystals and seeding entire screens. I found that not only would
crystals appear sooner but it revealed novel crystallization conditions.
These seeded crystals would appear within minutes as Preben described and
diffracted to better than 2 Ang. Another thought would be to try limited
proteolysis to see if you can identify a more stable
construct.
-John
--
John Domsic
Postdoctoral
Fellow
Gene Expression and Regulation Program
The Wistar
Institute
Philadelphia, PA 19104
On Wed, Jan 16, 2013 at 7:17 AM, jens Preben Morth
<[log in to unmask]> wrote:
Dear Lisa
It is not uncommon to see breakdown products
when you run crystals on gel. Espesially if they are older crystals,
sometimes you even see higher molecular bands, these are probably due to
intra molecular cross links formed over time.
If you are worried about
stability, try to increase the crystallization speed, we have one example
where we see a clear difference in both crystal quality and even space group
depending on when we fish the crystals. The crystals appear within 5 min,
the best quality data sets come from crystals we fish after only
30-60 min.
You may also have a little protease contamination of course,
to prevent this add protease inhibitor, or DTT, or EDTA to you protein
before you set it up.
cheers Preben
On 1/16/13 12:14 PM, LISA wrote:
Hi All,
I have an 36KD protein which can be
crystallize in two days. Most of the crystals are very big. But all
cystals have poor resolution,lower than 3.8 A. I picked some crystals,
washed them in the mother solution and then run SDS-PAGE. It is surprised
to find that different cystals have different components. Some crystals
have several samll bands below the band of the protein. And in some
crysals the bigger size band (as the construct should be) almost
disappared and have smear. Does the protein was degradated in the
crystals? Did someone met the similar problem as I? Thanks
All the
best
lisa
--
J. Preben Morth,
Ph.D
Group Leader
Membrane Transport Group
Nordic EMBL
Partnership
Centre for Molecular Medicine Norway (NCMM)
University of
Oslo
P.O.Box 1137 Blindern
0318 Oslo, Norway
Email: [log in to unmask]
Tel: +47 2284
0794
http://www.jpmorth.dk