*** For details on how to be removed from this list visit the ***
*** CCP4 home page http://www.ccp4.ac.uk ***
Dear All,
When trying to model different oxidised cysteines I encountered
varying library descriptions for CSO (single bond; Sg-Od), CSX
(double bond; Sg=Od), CSW (2 double bonds; od1=Sg=od2) in the ccp4
monomer library and for libraries generated by PRODRG.
Initially, I just noticed that the ccp4 library description for CSX
(where there should be a double-bond between the Sg and the Od) seems
to be incomplete. It essentially concerns the double-bond between the
sulfur Sg and the Od; whereas PRODRG indicates the corresponding bond
length at 1.500 A (which I presume is more realistic), creating
library descriptions with the ccp4i sketcher gives 1.800 A for this
bond, which I suppose corresponds to a single bond (strangely enough
the ccp4 library description for entry CSO, which should have a
single bond between the Sg and the Og, contains 1.780 A (i.e. shorter
than for the assumed double bond in CSX)). I e-mailed Garib about
this and I think this will be taken care of.
However, I then tried to create all oxidised cysteine descriptions
from PRODRG (either using the link from the HIC-Up database or by
pasting coordinates for an oxidised cysteine). It appears to be me
that PRODRG always creates a double-bond between the Sg and the Od
when I try to create descriptions for S-hydroxy cysteine (i.e. single
bond between Sg-Od). This applies to both the CSO and CEA entries in
the HIC-Up database which, as far as I understand, should both
describe a single bond between the Sg and Od. I also tried to draw
an s-oxycysteine residue using the JME editor, and there, though a
single bond is indicated between the Sg and the Od, it has a distance
of 1.500 A. Again this would seem to me to be closer to a double bond.
Furthermore, for the CSW (two oxygens with double bond; Od1=Sg=Od2)
entry I noticed that with the libraries created from PRODRG when
refining in Refmac it seems to move one of the two oxygens out of the
density. To illustrate that point: looking down from the Cb towards
the Sg (in a particular orientation), significant electron density
can be seen towards the left and right of the Sg; i.e. Od1=Sg=Od2),
and with the library descriptions created from the CSW entry with the
ccp4i sketcher Refmac also refines it well enough. However, with the
CSW library descriptions from PRODRG, the refined coordinates looks
more like that:
Od1
||
Sg=Od2
The Od1 ends up in a more perpendiular orientation to the Od2 (and
compared with the Od1 with the ccp4 libraries) where there is hardly
any electron density. I suppose this will have to do with angle
restraints and/or chiral volumes but so far have not been digging
deeper there.
As I have several of the above mentioned oxidised cysteine species in
my current model, I so far I ended up combining libraries from ccp4
(CSO and CSW) and PRODRG (CSX), but this I guess is really just a
workaround.
I would be rather curious as to why these differences in the library
descriptions exist, and which of the libraries are the more
appropriate to use. I was also wondering whether PRODRG creates
adequate descriptions for s-oxycysteine. I doesn't seem so to me like
that, at first sight.
Finally, I have a question concerning the potential determination of
the different species from X-ray data (the oxidation state may be
particularly relevant biochemically). My current dataset has minimal
Bragg spacing of 2.5 A, and thus does not suffice to unequivocally
determine the oxidation state of the cysteines. However, (in addition
to biochemical and mass-spectrometry data) I thought I could go about
to model the different species (led by difference Fourier maps; and
introduce only this change in my model and then calculate maps,
density fit, and Rfree for the different modeled oxidised cysteine
species). I would eventually model the species that would give me the
lowest Rfree. The Rfree differences are rather small (in the range of
0.06) and may thus not be significant. I was thus wondering whether
this seems like a reasonable procedure?
I apologize for this lengthy e-mail and thank you already for bearing
with me that long.
Any comments would be greatly appreciated!
Many thanks in advance.
Best regards,
Florian
--------------------------------------------
Florian Schmitzberger
Medical Biochemistry and Biophysics
Karolinska Institute
Scheeles vaeg 2
SE-171 77 Stockholm, Sweden
Tel: +46-8-524-86875
