Content area
Full Text
About the Authors:
Casper Søgaard
Affiliations Department of Plant Biology and Biotechnology, Faculty of Life Sciences, University of Copenhagen, Frederiksberg, Denmark, Villum Kann Rasmussen Centre for ProActive Plants, Frederiksberg, Denmark
Anne Stenbæk
Affiliations Department of Plant Biology and Biotechnology, Faculty of Life Sciences, University of Copenhagen, Frederiksberg, Denmark, Villum Kann Rasmussen Centre for ProActive Plants, Frederiksberg, Denmark
Sophie Bernard
Affiliation: Laboratoire de Glycobiologie et Matrice Extracellulaire-EA 4358, University of Rouen, Mont Saint Aignan, France
Masood Hadi
Affiliation: Technologies Division, Joint BioEnergy Institute, Sandia National Laboratory, Emeryville, California, United States of America
Azeddine Driouich
Affiliation: Laboratoire de Glycobiologie et Matrice Extracellulaire-EA 4358, University of Rouen, Mont Saint Aignan, France
Henrik Vibe Scheller
Affiliation: Feedstocks Division, Joint BioEnergy Institute, Lawrence Berkeley National Laboratory, Emeryville, California, United States of America
Yumiko Sakuragi
* E-mail: [email protected]
Affiliations Department of Plant Biology and Biotechnology, Faculty of Life Sciences, University of Copenhagen, Frederiksberg, Denmark, Villum Kann Rasmussen Centre for ProActive Plants, Frederiksberg, Denmark
Introduction
The Golgi apparatus is an organelle that plays a central role in the assembly of glycans associated with various macromolecules (i.e. matrix polysaccharides, proteins, lipids) in eukaryotic cells [1], [2]. Biosynthesis of glycans requires concerted actions of enzymes and proteins including glycosyltransferases, modifying enzymes (e.g. methyltransferases, acetyltransferases, sulfatetransferases), nucleotide sugar transporters, and nucleotide sugar conversion enzymes, many of which are localized in the secretory pathway including endoplasmic reticulum (ER) and Golgi stacks. These enzymes and proteins must be oriented in the membrane so that the catalytic domains face the relevant sides of the membrane where the substrates are available and the products can be channeled to the enzymes and proteins in the proceeding steps during biosynthesis.
Because an experimental determination of protein membrane topology is often laborious, efforts have been directed towards bioinformatically predicting the topology of membrane proteins based on the structural and statistical evaluation of the amino acid sequences. The transmembrane domains of membrane proteins all have two common features: a hydrophobic middle section composed of mostly aliphatic amino acids [3], [4] and a flanking sequence composed of aromatic amino acids, mostly tryptophan and tyrosine [5]. With the inclusion of the positive-inside rules and machine-learning techniques, a dozen of algorithms for predicting topology has been established and is widely used, including poly-Phobius [6],...