Re: [ccp4bb]: Radiation damage problem

[James Holton]

***  For details on how to be removed from this list visit the  ***
***          CCP4 home page http://www.ccp4.ac.uk         ***


It is definitely true that some sensitive residues decay quite a bit 
faster than the high-resolution spots.  This has been repoted in several 
papers now.  What has not been repoted (to my knowledge) is a convincing 
connection between the decay of side chains involved in a crystal 
contact and the loss of high-resolution data.  In all of the studies I 
have seen that measure the rate of loss of resolution, it is always 
tightly related to dose.  If you know of a study where it has been shown 
to be related to something else, I would like to hear about it. 

 I, personally, do not think that there is a connection between 
specific damage near crystal contacts and loss of resolution.  My 
reasoning is that a cryo-cooled crystal is permeated by a matrix of 
glassy water.  Glassy water is a SOLID.  So, once cooled, the crystal 
lattice is supported by a lot more than just crystal contacts.  Kind of 
like an insect preserved in amber.  Breaking bonds near the contacts are 
not going to do much more distortion to the overall lattice than broken 
bonds in the core or even in the solvent.  Diffraction is a process that 
occurs over dozens of unit cells.  There must be long-range distortions 
to impact it.  Site-specific damage, formally, only impacts the B-factor 
of that site.

-James Holton
MAD Scientist


Udupi Ramagopal wrote:


I think some times the sensitivity of residues involved in critical 
cystal contacts
can disturb these valuable suggestions and calculations.

Ramu

On Mon, 24 Apr 2006, James Holton wrote:

***  For details on how to be removed from this list visit the  ***
***          CCP4 home page http://www.ccp4.ac.uk         ***



Disclaimer:

Radiation damage is a complex phenomenon, and many aspects of radiation
damage at low temperature are still hard to explain, let alone predict.
You DEFINITELY need to know things like the beamline's brightness and
the concentrations of any atoms in your sample heavier than oxygen to
even have a ghost of a chance of understanding the origin of your 
problem.

However...

If you are used to this beamline and you "feel" that these crystals are
decaying faster than "normal", then the most likely problem is that you
have a heavy atom in your sample. This will cause your crystal to absorb
a lot more x-rays than normal (doing more damage). Anything heavier than
oxygen is a potential problem, and stuff in the solvent counts! The
"rule of thumb" is that the "absorptivity" of an atom is roughly
proportional to the atomic number. So, try doing things like replacing
potassium with lithium or iodide with fluoride (each a factor of 6).
Obviously, the more concentrated a heavy atom is, the bigger the problem
will be.

Since beamline brigthnesses vary by a factor of 30,000 from place to
place, and crystal absorptivity can vary by a factor of 10 or more,
radiation damage questions can be hard to answer without knowing flux,
beam size and sample composition.

But, if you do know these things...

The decay in protien crystals is generally proportional to absorbed dose
(Joules/kg or "Gray"). Dose, in turn is proportional to fluence
(photons/mm^2) and the "extinction coefficient" of your crystal in the
x-ray range. You get fluence (photons/mm^2) from flux (photons/s) by
dividing by the beam size to get brightness (photons/mm^2/s) and
multiplying by the total exposure time (in seconds). Given the beamline
flux, wavelength, beam size, and crystal composition the expected
lifetime of your crystal can be computed with a program called RADDOSE
(Murray et. al. JSR (2005), 12, 268-75). Or you can "google" for 
RADDOSE.

For example, you can expect the lifetime of your crystal will be
(roughly) cut in half by the addition of 4M NaCl, 5M NH4SO4, 1 Se per 30
residues (at 0.97A), or 150 mM CsI.

-James Holton
MAD Scientist


Uhnsoo Cho wrote:

Dear bulletin members,
Sorry for non-CCP4 related questions.
Recently, I've got crystals which have a radiation damage problem.
It also has a weak diffraction because of the big unit cell dimension
(about 300A in one direction), so I have to overexpose crystals even
with the synchrotron beam.
The problem is that during the data collection, the resolution decaded
gradually from 3 to 6A after certain number of frames (about 30 or 40
frames).
Do you have any experiences or any suggestions that I can evade this
radiation damage?
Any suggestions will be grateful.
Thank you.
Best regards,
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Uhn-Soo,Cho
Graduate student in Biological Structure Department
University of Washington.
HSB G514
1959 NE Pacific St.
Seattle, WA 98195-7420
Box. 357420
Fax #206-543-1524
Tel # 206-221-2435
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~