1i1p: Difference between revisions

[[Image:1i1p.gif|left|200px]]<br /><applet load="1i1p" size="350" color="white" frame="true" align="right" spinBox="true"  

Accurate experimental phases derived from a MAD experiment may be useful to enable the identification of solvent molecules during the course of an atomic parameter refinement. The structure of a double-stranded DNA decanucleotide bearing a cisplatin interstrand cross-link at 1.6A resolution, whose phases were first determined experimentally using the L(III) edge of the Pt atom, was refined by various methods. The previously published structure resulted from a least-squares refinement using the structure-factor magnitudes and stereochemical restraints (program SHELX). In this paper, these previous results are compared with a model obtained by the likelihood-maximization method (program REFMAC) which allows the combination of the observed magnitudes with experimental MAD phases. This solution corresponded to a lower R(free) (18.8 compared with 20.3%), a lower R factor and accounted for 135 water molecules and one spermine molecule collected by the program wARP during refinement. The previously published SHELX solution exhibited no spermine molecule and accounted for 92 water molecules, only 74 of which are also present in the model obtained with the MAD phases. In order to verify that these improvements were actually related to the use of the MAD phases, the same type of procedure without the MAD phases was applied starting from the initial model. The resulting solution had a higher R(free) (20.3%), which could be related to the loss of 22 water molecules and the addition of 20 new ones. MAD phases therefore seem especially helpful in preventing the model bias which may affect the solvent molecules. All models have in common a hydration cage of nine water molecules which surround the platinum residue. In addition to the spermine molecule, the model obtained with the MAD phases allows description of the water-molecule organization, with reproducible motifs related to the base pairs and to the phosphodiester backbone.

1I1P is a [http://en.wikipedia.org/wiki/Protein_complex Protein complex] structure of sequences from [http://en.wikipedia.org/wiki/ ] with <scene name='pdbligand=CPT:'>CPT</scene> and <scene name='pdbligand=SPM:'>SPM</scene> as [http://en.wikipedia.org/wiki/ligands ligands]. Full crystallographic information is available from [http://oca.weizmann.ac.il/oca-bin/ocashort?id=1I1P OCA].  

''Page seeded by [http://oca.weizmann.ac.il/oca OCA ] on Thu Feb 21 13:07:04 2008''