University of Virginia Health System

Structural biology of membrane proteins:  Membrane proteins represent about ~25% of all proteins in any given genome carrying out many vital functions (transporters, receptors, channels, etc…).  Membrane proteins also represent ~50% of all drug targets, highlighting their functional importance.  However, membrane proteins are lagging behind water soluble proteins in terms of the number of the structures determined.  Currently there are about ~60,000 structures in the protein data bank (pdb) with only few hundred membrane protein structures. This is mainly due to the difficulty faced when working with membrane proteins.  Now with a variety of technical advances we are able to tackle this important class of proteins and produce high resolution structures using X-ray crystallography.  This allows us to examine their function and better understand how they work.  We are currently pursuing a variety of membrane proteins that have shown good levels of expression, some of which have produced initial crystals making for good projects.

Development of new methods:  Advancement of current methods and development of new ones is likely to play an important role in the determination of many membrane protein structures.  One of the challenges in membrane proteins crystallography is the hurdle of growing high quality crystals.  Membrane proteins are typically solubilized in detergents which cover their hydrophobic regions making them more water soluble.  However by shielding this hydrophobic region, these detergents make a large surface area unavailable for crystal contacts.  Thus it is common for membrane protein crystals grown from detergent media to be of poor quality having limited crystal contacts.  I developed the bicelles method for the crystallization of membrane proteins.  Bicelles are a mixture of a detergent and a lipid that form bilayer discs, and exhibit a variety of phase transitions that can influence the process of crystal growth.  Crystals grown from lipidic media such as the bicelles typically pack differently from detergent grown crystals, producing type I crystals where proteins can stack side by side as if in a bilayer.  In type I crystals the hydrophobic region of a membrane protein can make extensive contributions to crystal contacts producing high quality crystals.  I am interested in further development of the bicelle method for the crystallization of membrane proteins and utilizing it for the determination of high resolution structures.

Design of crystal lattices:  One aspect of protein design is the design of macromolecular assemblies.  It is possible to take advantage of the inherent symmetries of naturally occurring protein oligomers to design macromolecular assemblies.  Crystal growth is a form of self-assembly, and higher order oligomeric assembly can be a driving force for crystal growth.  With proper protein domain fusion it should be possible to organize an ordered crystal lattice.  We are interested in developing tools that would allow us to design specific macromolecular assemblies and pre-arranged crystal lattices.

Other Websites for this mentor:
http://people.virginia.edu/~sf3bb/