Hello, Sabine.
I believe the technique that Kendall is referring to is Surface Entropy
Reduction, which involves replacing large, highly entropic residues on the
surface of the protein with lower entropy residues to reduce the entropy
shield and facilitate crystallization. Traditionally, the replacement
residue was an alanine, but we have recently found that tyrosine and
threonine also mediate crystal contacts. Multiple alanine mutations can
reduce the solubility, which in your case might not be a bad thing. About
18 difficult to crystallize proteins that we know of have been solved with
this technique (6 of them were targets that fell out of the MCSG pipeline).
We have collaborated with the Eisenberg lab to develop a server to suggest
mutations to make. The server incorporates the three primary criteria that
need to be considered when selecting sites for mutagenesis (as well as some
minor ones). These are 1) the entropy of the residues within a continuous
stretch, which can be represented as an entropy profile graphed by residue
number, 2) the predicted secondary structure, and 3) the sequence
conservation.
The server is http://nihserver.mbi.ucla.edu/SER/
I should mention that I am in the Derewenda lab, so I am not an unbiased
source, but I am a firm believer in the technique. The paper on using
different target residues should be in Acta Cryst before too long (our
"minor corrections" have been submitted).
Good luck,
David Cooper
On Thu, 22 Feb 2007 10:03:33 -0500
Kendall Nettles <[EMAIL PROTECTED]> wrote:
Sabine,
There are protocols to modify surface residues that can help with
crystallization, and make the protein less soluble. Unfortunately, I¹m
drawing a blank on the details. I remember someone in Andrzej Joachimiak¹s
group was working on this as a rescue approach for the structural genomics
pipeline, and it had been previously published by others.
Have you looked at the protein with dynamic light scattering?
Are their cysteines? Are you using reducing agents? Sometimes mutating
cysteines to serines can help.
How much purification have you done? Try ion exchange and gel filtration.
Do you have more than one ligand? I think there is quite a lot of
variability in how different ligands promote crystallization. Are you
adding
the ligand in excess? Try a few different molar ratios. If if has high
affinity, you might want to try removing excess unbound ligand at the end.
Good luck!
Kendall
On 2/22/07 9:08 AM, "Schneider Sabine" <[EMAIL PROTECTED]> wrote:
Hi everyone,
I am trying to crystallise an extremely soluble and charged protein. It is
~30kDa and has an estimated PI of 5.2 and theoretical charge over pH range
4-10 from + 24 to -29. It is still happy at a concentration of 190mg/ml
and
fully reconstituted with its ligand.
I have tried high throughput crystallisation with 10 different screens
from
Nextal with concentrations of 60, 100 and 150mg/ml with no NaCl and NaCl
concentrations of 100mM, 300mM and 1M in either Hepes pH 8 or Tris-HCl pH
7.5.
The distribution of heavy precipitation, light crystalline precipitation
and
clear drops through out the screens locks like I am in the right
concentration
range around the 100mg/ml, but I am not getting any real hit. There are
some
drops with extreme phase separation. I also tried changing the temperature
from 20C to 4C.
I chased up a few conditions with this strong phase separation (or where I
imagined little objects...) by manual screening and also adding additives
like
3% Succrose, 50-200mM LiCl, 100mM EDTA, varying the PEGs (1500, 3350,
4000,
6000, 8000) as well as adding NaCl to the reservoir solution in sitting
as
well as hanging drop screens. But I am just getting nowhere - either just
precipitation or the drop stays clear with the strong phase separation.
I also re-cloned it with chopping of a few more residues on the N-term
where
according to a secondary structure prediction a helix starts and it is
still
very happy at high concentrations, but again nothing in the
high-throughput
screens.
Has anyone any suggestions what else I could try?
Thanks!
Sabine
This message has been checked for viruses but the contents of an
attachment
may still contain software viruses, which could damage your computer
system:
you are advised to perform your own checks. Email communications with the
University of Nottingham may be monitored as permitted by UK legislation.
.
-------------------------------------------
David Cooper, Ph.D.
University of Virginia
Molecular Physiology and Biological Physics
Jordan Hall, Box 800736
1300 Jefferson Park Ave
Charlottesville, VA 22908-0736
Phone: (434) 982-3151
Fax: (434) 982-1616