Date of Award

Spring 1-1-2011

Document Type


Degree Name

Doctor of Philosophy (PhD)


Chemical & Biochemical Engineering

First Advisor

Theodore W. Randolph

Second Advisor

John F. Carpenter

Third Advisor

Richard D. Noble


The development of protein based biopharmaceuticals has resulted in the ability to treat many serious conditions, including endogenous protein deficiencies, cancer and autoimmune disorders. However, in order to be successful drug candidates, these more complex molecules require stable formulations that can last from manufacturing through transportation to patient administration. Fusion proteins, which are constructed from unrelated proteins or protein domains, present additional formulation challenges. In these proteins, domains did not co-evolve for stability and thus conditions favoring the stability of one domain may destabilize the other.

Our current hypotheses are that the least stable protein will also be the least stable domain in the fusion and that selective stabilization of this least stable domain can be used to reduce overall aggregation of the fusion protein, with preferential binding of cosolutes to the native state being one method of achieving selective domain stabilization.

Abatacept, an Fc-CTLA4 fusion protein, is used as a model to investigate the stability behavior of Fc fusion proteins. Fc-CTLA4 exhibits markedly different aggregation rates with only a small shift in pH. Changing from pH 7.5 to pH 6 causes a two order of magnitude increase in rate of monomer loss during incubation at 40 °C. The aggregation behavior during accelerated stability studies at elevated temperature was found to be controlled by conformational instability of the protein. Conditions which destabilized the CTLA4 domain and the CH2 portion of the Fc domain lead to a decreased activation energy for aggregation and an increased aggregation rate.

Fc-CTLA 4 was also studied to examine how conditions which increase conformational stability impact the progression of aggregation due to other physical stresses (e.g., exposure to air/water and water/ice interfaces and freeze concentration during freezing). No changes were detected in the amount of monomer lost or number of particles formed between solution conditions at pH 6.0 and pH 7.5, despite large differences in conformational stability under these solution conditions.

Two HSA fusion proteins (HSA-hGH and HSA-GCSF) were studied to investigate small molecule ligand binding as a selective stabilization strategy, with octanoic acid used as the ligand. Addition of octanoic acid resulted in increased conformational stability for HSA, HSA-hGH and HSA-GCSF. Repulsive protein-protein interactions were only increased for HSA and HSA-hGH; HSA-GCSF protein-protein interactions remained unchanged. Reductions in aggregation were seen in the case of HSA-hGH. Thus it appears that specific binding to a domain of a fusion protein can reduce aggregation in more than one way (i.e. increasing both colloidal and conformational stability). No changes in HSA-GCSF aggregation were observed with the addition of octanoic acid, despite the increases in conformational stability, indicating the ligand binding stabilization approach does not appear to be applicable to all HSA fusions.