No Subject

[Palmira daSilva <dasilva@scripps.edu>]

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

Hi dear all:

I have three complexes of a 45 KD Fab ( two domains with a flexible hinge 
region). Two of them are solved, the third has been a headache for some 
time now.

complex I:  space group C2; 1 mol/asyu; solvent content 59%
complex II: space group I422; 1 mol/asyu; solvent content 59%; cell 
dimensions 122.7x122.7x146.7

The third crystallizes in C orthorhombic, unit cell 122.7x256.7x146.8, 
which processes and scales quite decently to 2.5 A. I presume similar 
solvent content to I and II which leads to 4 mol/asyu. I tried finding the 
MR solutions assuming both C222 and C2221 with EPMR, MOLREP and AMORE using 
the complex I as a model. MOLREP is the one that seems to work better. It 
appears that I have two sets of molecules related by translation. I have 
two strong rotational peaks so two solutions are always quite OK but I 
cannot find the other two correctly. In C222 after refinement the Rs are 
above 50%; in C2221 R/Rfree is 40/47. The log file of minimization and 
annealing states that parts of two NCS molecules superimpose with symmetry 
mates. The maps (sa_omit and composit_omit) for C2221 look awful! I used 
domain-domain NCS restraints in all refinements.

I refined and rebuilt the model with only two mol/asyu assuming C222 ( 
C2221 leads to R's quite above 50% after the first round of refinement). 
The sa_omit_maps are great, I see perfect density for the ligand in the 
active site, I almost thought I only had two molecules (80% solvent, ouch) 
in the asymmetric unit but... R/Rfree got stuck at 31/33! I used this dimer 
as search model (again in epmr, molrep and amore) with no better luck.

I processed and scaled and looked for molecular replacement solutions in 
all other possible space groups (the details would be too lengthy but I am 
sure is C orthorhombic); I collected data on crystals frozen in different 
conditions and cryoprotectants and as for trying other crystallization 
conditions...I've got this crystals after 150 trays, I am stuck with this 
crystal form.

I have been thinking that my problem might be the hinge angle in the two 
elusive molecules being so different from the search model that baffles the 
search programs. I am a neophyte in protein crystallography so I would 
appreciate any suggestions on:

1) How do I perform a search for my 4 molecules allowing for flexibility in 
the hinge region?
2) If the above fails... any suggestion is more than welcome.

Thanks
Palmira