Prenucleation clusters and crystallization control by additives
Helmut Cölfen · Acta Crystallographica Section A Foundations of Crystallography · 2009
The study of human protein complexes has proven to be particularly difficult, in part due to the fact that many individual eukaryotic proteins, when examined in isolation, are seen to have large continuous regions that are intrinsically disordered.An interesting example of this can be seen in a family of neural cell adhesion proteins, which are single-pass transmembrane proteins, and bear substantial sequence similarity to cholinesterases (ChEs).The regions of sequence similarity correspond to only parts of their complete sequences, thus establishing the ChE domain as a modular domain incorporated into a group of proteins that we have called 'cholinesteraselike adhesion molecules' (CLAMs).CLAMs are devoid of catalytic activity, since they lack residues crucial for catalysis.They play, however, a key role in the earliest stages of the development of the central nervous system (CNS) and mutations in the ChE domain of one of them, neuroligin, are associated with autism and mental retardation.The cytoplasmic domains of CLAMs bear no sequence homology to any known protein.In silico studies on the analysis of the structure of CLAMS, via FoldIndex © [1] (http://bioportal.weizmann.ac.il/fldbin/findex), has predicted which regions are likely to be unfolded.These results are compared to physicochemical studies, which demonstrate experimentally that the cytoplasmic domains of the CLAMS are, in fact, Intrinsically Disordered Proteins (IDP) [2],[3],[4].FoldIndex © is also being used routinely, at the Israel Structural Proteomics Center (http://www.weizmann.ac.il/ISPC), to aid in crystallization of proteins by first predicting which regions of a protein sequence is likely to be intrinsically disordered, and subsequently eliminating these stretches from the construct that is cloned.