Research Project
9347936
Abstract Therapeutic proteins are the fastest growing class of pharmaceuticals ? they provide specificity that is difficult to achieve with small molecule drugs. However, protein therapeutics present their own difficulties in development and manufacturing. They require physical stability at high concentrations (150 mg/ml for IgG) for efficacy and low viscosities (<50 cP) conducive to delivery by injection. Additives, identified as inactive ingredients by the Food and Drug Administration (FDA), can modify both the solubility and viscosity of protein formulations. The goal of this project is to identify screening techniques to reduce the time and protein required to identify formulations with low viscosity and high solubility. Three primary techniques will be evaluated: viscosity measurements at low concentration, osmotic second virial coefficient (B value) and the diffusion virial coefficient (Kd). Protein formulations generally follow an exponential relationship between viscosity and protein concentration. First we determine the minimal range of protein concentrations required to be measured to establish an exponential relationship sufficient that viscosities of formulations at high protein concentration can be ranked. Next we evaluate the ability of B value (sum of protein interaction forces) and Kd (diffusion tendancy) to also rank viscosity at high concentration. Each parameter represents a potential contribution protein formulation viscosity which can be rapidly measured using relatively little protein. In addition to direct ranking of formulation viscosities they will be evaluated in the context of the ? ( ?? ) Mooney viscosity equation = ? 1??? where ? / ?0 is the ratio of formulation viscosity to solvent ?0 viscosity and A and B are hydrodynamic and self-crowding factors. The measurement techniques will be evaluated using five different proteins (including IgGs,a pharmaceutical biologic and a protein known to form high order aggregates). After comparing the ability of each technique to predict viscosity a general methodology will be designed based on the minimal use of protein to successfully rank order viscosity at high concentration. The screen will method will favor high-throughput B value and Kd measurements as long as they can reliably discriminate between formulations with respect to lower viscosity. Otherwise all protein and time reductions will come from low concentration viscosity measurements as a substitute for high concentration viscosity measurements. The screen to increase solubility and decrease viscosity will be evaluated on three additional pharmaceutical protein. The resulting screen requiring less time and protein for the identification of low viscosity and high solubility formulations will help to alleviate a significant problem in the development of protein therapeutics.
