Virtual Molecule: P0 Myelin Glycoprotein. I. Homology Modeling and Prediction of the Secondary and Tertiary Structure - Molecules: Nucleation, Aggregation and Crystallization - page 238

Introduction
P0 glycoprotein is the major structural protein of peripheral nerve myelin;
it is thought to modulate inter-membrane adhesion by providing the stable
link between apposed cytoplasmic membrane surfaces in peripheral
myelin (Inouye et al ., 1985). Most of the studies on P0 protein — an abun-
dant structural protein of peripheral myelin — have been focused on
structure-function correlation in higher vertebrates: bovines and human
(Luo et al ., 2006). It has even been observed that conformational changes
of P0 protein are due to protonation-deprotonation of His 52 at P0's puta-
tive adhesive interface (Luo et al ., 2006). Removal of the fatty acids that
are attached to the single Cys 153 residue in the cytoplasmic domain of P0
has not changed the PNS myelin structure of frog and mouse, suggesting
that the P0-attached fatty acyl chain does not play a significant role in
PNS myelin compaction and stability (Luo et al ., 2007; Luo et al ., 2008).
The P0 protein of peripheral myelin is very hydrophobic, and has one
transmembrane segment. The soluble part, representing two-thirds of the
total mass of the protein, has been cloned, expressed, and crystallized for
three-dimensional structure determination (Shapiro et al ., 1996).
Methods and Tools for Homology Prediction
A virtual model of the complete P0 protein has been built using a homol-
ogy modeling method (HMM), secondary structure prediction and
threading. BLAST on NCBI, SWISS-MODEL (An Automated
Comparative Protein Modelling Server: First approach mode) and
Project (optimize) mode on the ExPASy Proteomics Server, Verify3D,
WhatIf and the DeepView Swiss-PdbViewer off-line tool have all been
used to perform homology modeling. ClustalW on EMBL-EBI on-line
service, BioEdit and JalView have been used to analyze the protein
sequences. SWISS-MODEL, JPred and SOSUI have been used to predict
secondary structure and the transmembrane region. RCSB PDB, Pfam,
ChemSketch, RasMol, Swiss-PdbViewer, PyMOL, VMD and NAMD
and Cn3D have been used to visualize, verify, optimize and rearrange
3D models of the protein.