Research Projects

Proteins are the macromolecules responsible for most enzymatic, structural, and signal transduction functions in biology. Because of this, the medical and biotechnology fields are heavily investing in trying to manipulate protein function. I am interested in the phase behavior of protein solutions. Simply put: what makes proteins stick together, how can we control how much they stick, and what structures they form when they do stick? Some specific applications I am interested in are:

Biological phase transitions
There is a growing list of cellular structures are formed by the spontaneous phase separation of biomolecules. These structures have many common features, such as the presence of multivalent scaffold molecules. However, they also perform many different functions within the cell and have evolved to optimize these specific functions. We are studying how structure/function properties propagate from the nanometer-scale characteristics of the individual molecules to the characteristics of a macroscopic phase.




Protein crystallization
X-ray crystallography is the workhorse technique for determining the three-dimensional structures of folded proteins. Typically the slowest step in this process is the growing of the crystals, which is currently accomplished by high throughput screening (i.e. trial-and-error). Crystallization is a "Goldilocks" problem: the proteins must stick together strongly enough to stabilize the crystal, but not so strongly that the aggregation is uncontrolled. To understand this we need two things. First, we need to understand the forces that make proteins stick together, and how these forces are affected by experimental variables like temperature, pH, and salts. Secondly, we need to understand the kinetic pathways that lead to crystallization and the pathways that lead to non-crystalline aggregates and gelation.


Self-assembly
Most biological machines are built by "bottom-up" processes, where building blocks spontaneously assemble into a final product. The complexity of machines observed in nature speaks volumes of the potential of bottom-up assembly in nano-manufacturing and medicine. Experimental groups have made tremendous progress recently in designing nanoscale building blocks with novel shapes and binding affinities. However, in order to mimic the complexity of assembled objects found in nature, we need a better understanding of the kinetic factors that lead to either successful assembly or undesirable aggregates.


Amyloid fibrils
Many diseases, such as Alzheimer's, Type II diabetes, and Mad Cow, are associated with the deposition of fibrillar protein aggregates called "amyloids". These aggregates have a striking degree of morphological similarity despite little sequence homology in the precursor proteins. While a few cases of these diseases are genetic, the large majority are sporadic. What is is that determines whether aggregation begins in middle age, late in life, or not at all? Another complication is the growing evidence that disease progression is driven, not by the fibrils, but by soluble oligomer states. Are these oligomers precursors to fibril formation, or off-pathway aggregates? The goal of this project is to understand the physics of amyloid aggregation in order to develop strategies to remove protein from the most toxic states.


Drug formulations
Protein-based pharmaceuticals are the fastest growing segment of the drug discovery industry. These products are very well tolerated and can offer remarkable specificity. However, proteins are fragile molecules and there are numerous stresses involved with manufacturing, transportation, storage, and delivery. The final formulation must be stable against denaturation and/or aggregation. At the same time the formulation must be compatible with the delivery mechanism. Concentrated solutions are cheaper and easier to administer, but at high concentrations proteins are extremely viscous. This project is a collaboration with Amgen to understand the aggregation and viscosity of protein solutions in order to develop formulations that are both stable and easily administered.