*** For details on how to be removed from this list visit the ***
*** CCP4 home page http://www.ccp4.ac.uk ***
I'm not sure what "failed to diffract" means. You will always see
something on the detector. What did it look like? Ice? Glass? 50A? No
spots?
Dr. Thomson is right. It is important to know if the crystal diffracts
in the cryobuffer at room temperature. This will tell you if the
crystal is damaged by the transfer into cryoprotectant or by the cooling
process.
(However, I would like to point out that "bending magnet" does NOT imply
"weak", but I digress...)
References:
- In general, it is always a good idea to look at Elspeth's papers. ;)
Rob Thorne [JACr v39 244-251 (2006)] and Constance Schall [JACr v38
412-419 (2005)] have also done recent systematic studies of cryocooling
crystals. Also, Kim and Gruner [Acta Crys. D61, 881-890 (2005)] have
revived the ancient art of cryocooling without the need for any
cryoprotectants, if you have the right high-pressure equipment.
- There is a very good book by Pierre Douzou called "Cryobiochemistry:
an introduction" Academic Press (1977), which reviews much of the
founding work and "lost techniques" for cooling biological samples. A
great deal of what I know about cooling crystals I learned from this book.
Suggestions:
- There is always oil. Just make sure you get all the liquid aqueous
phase off of the crystal before you plunge it. I recommend uncut
Paratone-N and some practice working in a viscous medium.
- cross-linking with gluteraldehyde makes the crystal more stable and
will generally extend the range of solutions you can expose it to.
- You can try smaller concentrations of a wider variety of cosolvents.
For example, 50% glycerol will still have a sharp ice ring, but 15%
glycerol with 5% PEG will cool as a glass. Quarternary mixtures can
work even better (and a few have already been suggested).
- pH* is an often overlooked problem. Not many people realize that tris
and glycine buffers will go to pH* ~ 14 when cooled to 100K. Yes, that
strong temperature dependence of the pH of a tris buffer continues all
the way down.
- The change in dielectric constant of water with temperature was once
shown to be the most important factor in crystal destruction upon
cooling. Douzou's book describes one of Greg Petsko's techniques where
he slowly cooled crystals and slowly added methanol to correct the
dielectric constant back to 80. This apparently worked very well, but
you need a way of slowly adjusting the temperature to do this.
- Thermal shock will be a problem if the warm_volume/cold_volume ratio
of your cryobuffer is different than the warm_volume/cold_volume ratio
of your crystal lattice. The mismatch will cause your crystal lattice
to strain, and strained lattices don't diffract very well. I think a
good way of adjusting the cold_volume of your cryo is to work near a
phase transition in the cryobuffer. That is, use a series of cosolvent
concentrations in the neighborhood of the threshold between when the
diffraction pattern of your cryobuffer goes from sharp ice rings to a
broad, glassy "water ring". If you adjust your cosolvent concentration
very carefully, you will see "ice rings" get broader and broader as you
increase the cosolvent concentration (or decrease the drop size). The
density of this "nanocrystalline ice" will be somewhere between ice Ih
and low density amorphous ice. Your "sweet" density might be in there
somewhere. The transition is very sharp, but managable.
- Osmotic shock is a common problem when moving a crystal into
cryoprotectants. Something that worked for me once in a situation where
osmotic shock was such a big problem that the crystals couldn't be
transferred into ANYTHING without cracking, was to just expose the drop
to air and allow it to slowly dry down until the PEG concentration was
high enough for cryoprotection. The crystals were "happy" over this
gradual transition. You know when the PEG concentration is high enough
by looping up some of the drop without a crystal and checking to see if
if "freezes clear".
In general, when checking cryos: It helps to move your bowl of lN2 onto
the microscope stage. Shine a light into the bowl, and focus the
microscope on the surface of the lN2. After plunging your sample, raise
the loop to just above the liquid surface (where the N2 gas is still
colder than a cryostream) and you can get a clear look at your loop
(without monopolizing an x-ray diffractometer).
-James Holton
MAD Scientist
Vincenzo Carbone wrote:
Hi guys
Sorry for the non ccp4 based question but i was wondering if anyone
has any tips on producing a suitable cryo buffer for crystals grown in
peg8000. My crystals come out quite quickly in 15 to 10% peg and can
be quite large. Ive tried to increase the peg concentration,
replacing the peg with glycerol or using a combination of peg8000 and
glycerol but each time the crystals fail to diffract when flash
frozen. Previous experience with these crystals have shown diffraction
of up to 2A when dry mounted.
thanks,
vince
