Re: [ccp4bb]: Phasing with FeS cluster

[Paul Hubbard <phubbard@post.its.mcw.edu>]

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

Thanks for the respones. Sorry that I didn't mention the program I tried out
was with SHARP.

Yes, I am quite serious about getting help (hence this e-mail), so here is an
abridged list of what I've tried so far:

1) Molecular replacement of a protein with 26 % identidy (I cut out the
domains before making a search model) - tried Amore, MolRep, EPMR and CNS at
4A and 5A. No luck.

2) Rigid body refinement of an ideal 4Fe-4S cluster in PHASES (using GREF)
with MAD data, then refining in SHARP using f' and f'' values taken from an
EXAFS scan. I used only the Fe positions from GREF, and refined them in
SHARP. The atoms move such that the Fe-Fe distances are all wrong. I know 3A
should be good enough to separate out the Fe atoms - but I can't at the
moment (hence this e-mail)

3) I tried using SOLVE with a fine grid search for 4 Fe atoms (ntol_site 1,
ntol_soln 1). It found 2 of them, so I tried placing an ideal 4Fe-4S cluster
in its position so that 2 Fe atoms match the results in SOLVE. Couldn't get a
decent fit in the SOLVE electron density blob.

4) The current method described in my previous e-mail was done in SHARP,
which I have been told is very good at handling f' and f'' refinement.

5) Space group is P212121 - 66A x 108A x 111. 1 molecule per ASU, with 1
4Fe-4S cluster.

NOTE!!!! I have Os, Hg, and Pb derivatives - I am NOT trying to solve the
structure just from Fe MAD (I doubt its possible).

Sorry if I have forgotten any relevant information. Please let me know if
I missed anything.

Thanks

AGS

"David J. Schuller" wrote:

> On Wed, 9 Jan 2002, Paul Hubbard wrote:
>
> > I am having trouble doing initial phasing for a protein with an FeS
> > cluster. MAD data at the Fe abs. edge is to 3A, and in-house native
> > anomalous is to 2.65A.
> >
> > Due to the low resolution, I have been treating the 4Fe-4S cluster as a
> > single Uranium atom. Therefore I use f' and f'' values for U taken from
> > the Sasaki table. I fix these parameters for the in-house anomalous and
> > remote wavelength data sets, and allow both f' and f'' to be refined for
> > the inflection and peak wavelengths.
> >
> > The problem is that the "y" coordinate moves by ~2A, and I get stupid
> > final values for f' and f'' for the inflection and peak wavelengths
> > (i.e., f' at inf. is 35.1671 and f'' is 10.405)
> >
> > Any thoughts would be very much appreciated.
> >
> > P.S: I have read the fancy-looking manual, and asked a number of people.
> > No luck there.
>
> even if you wanted to model this as a single atom, i don't see why you
> would use the f' and f" values from U, these are dependent on the
> wavelength and the element and the edges for U are nothing like those for
> Fe, your dominant scatterer.  i would say use atom type Fe and let your
> occupancy go above 1.0.
>
> but with 3.0 - 2.65 A you should be able to resolve the single atom
> positions.  you might try using a 4Fe-4S cluster from a known structure
> (if you are willing to presume its similarity) and run it through
> molecular replacement.  you might even give a shot at positioning the Fe
> atoms with direct methods.
>
> if you are serious about wanting help, you might also disclose a few more
> details. for example, what software have you have been using?  that would
> enable experienced folk to toss in comments like "program X can't handle
> refining f' and f", try program Y instead".  what is your space group?  if
> your space group is monoclinic, any movement of Y could be laughed off.
> some idea of the size of your asymmetric unit and how many 4Fe-4S clusters
> you expect to be present would also be very helpful.
>
> cheers,
>
> =======================================================================
> "Now that I'm a cranky, constipated old man I can afford to say that
>  the younger generation of scientists makes me sick to my stomach."
>                                                   - C.M. Kornbluth
> =======================================================================
>                         David J. Schuller
>                         modern man in a post-modern world
>                         MacCHESS, Cornell University
>                         djs63@cornell.edu