PURPOSE

The aim of the study is to determine the degree of coupling between protein unfolding rate and system viscosity at low temperatures in systems relevant to freeze-drying.

METHODS

The cold denaturation of both phosphoglycerate kinase (PGK) and beta-lactoglobulin were chosen as models for the protein unfolding kinetics study. The system viscosity was enhanced by adding stabilizers (such as sucrose), and denaturant (guanidine hydrochloride or urea) was added to balance the stabilizing effect of sucrose to maintain the cold denaturation temperature roughly constant. The protein unfolding kinetics were studied by both temperature-controlled tryptophan emission fluorescence spectroscopy and isothermal high-sensitivity modulated differential scanning calorimetry (MDSC) (Tzero). Viscometers were used to determine the system viscosity. To verify the predictions of structure based on protein unfolding dynamics, protein formulations were freeze-dried above the glass transition temperatures, and the protein structures in dry products were determined by fluorescence spectroscopy of reconstituted solids by extrapolation of the solution data to the time of reconstitution.

RESULTS

Empirical equations describing the effect of sucrose and denaturant (urea and guanidine hydrochloride) on protein cold denaturation were developed based on DSC observations [X. C. Tang and M. J. Pikal. The Effects of Stabilizers and Denaturants on the Cold Denaturation Temperature of Proteins and Implications for Freeze-Drying. Pharm. Res. Submitted (2004)]. It was found that protein cold denaturation temperature can be maintained constant in system of increasing sucrose concentration by simultaneous addition of denaturants (urea and guanidine hydrochloride) using the empirical equations as a guide. System viscosities were found to increase dramatically with increasing sucrose concentration and decreasing temperature. The rate constants of protein unfolding (or the half-life of unfolding) below the cold denaturation temperature were determined by fitting the time dependence of either fluorescence spectroscopy peak position shift or DSC heat capacity increase to a first-order reversible kinetic model. The half-life of unfolding did slow considerably as system viscosity increased. The half-life of PGK unfolding, which was only 3.5 min in dilute buffer solution at -10 degrees C, was found to be about 200 min in 37% sucrose at the same temperature. Kinetics of protein unfolding are identical as measured by tryptophan fluorescence emission spectroscopy and by high-sensitivity modulated DSC. The coupling between protein unfolding kinetics and system viscosity for both proteins was significant with a stronger coupling with PGK than with beta-lactoglobulin. The half-lives of PGK and beta-lactoglobulin unfolding are estimated to be 5.5 x 10(11) and 2.2 years, respectively, even when they are freeze-dried in sucrose formulations 20 degrees C above Tg'. Thus, freeze-drying below Tg' should not be necessary to preserve the native conformation. In support of this conclusion, native PGK was obtained after the freeze-drying of PGK at a temperature more than 60 degrees C above the system Tg' in a thermodynamically unstable system during freeze-drying.

CONCLUSIONS

Protein unfolding kinetics is highly coupled with system viscosity in high viscosity systems, and the coupling coefficients are protein dependent. Protein unfolding is very slow on the time scale of freeze-drying, even when the system is freeze-dried well above Tg'. Thus, it is not always necessary to freeze-dry protein formulations at temperature below Tg' to avoid protein unfolding. That is, protein formulations could be freeze-dried at product temperature far above the Tg', thereby allowing much shorter freeze-drying cycle times, with dry cake structure being maintained by the simultaneous use of a bulking agent and a disaccharide stabilizer.