Author:

Abstract:

INTRODUCTION Electrospraying (ES) or electrohydrodynamic atomisation, i.e. the formation of tiny droplets from a jet of conductive liquid under the influence of an electric field, has been extensively studied following its discovery in the 17th century. Since then, ES been applied in various fields, of which its use for ionisation of large biomolecules in mass spectrometry is probably the best known. More recently, ES has gained attention for pharmaceutical and food-related applications for its particle engineering and encapsulation potential. The high encapsulation efficiency of ES could prove to be especially interesting for proteins, as techniques like coaxial ES, could allow for the protein to remain in aqueous solution during the encapsulation process, with a minimised contact between the protein and the potentially protein denaturing solvents of the polymer solution. Even though the use of an aqueous solution for ES would be beneficial for protein stability, physicochemical factors like the relatively high conductivity and surface tensions of these aqueous solutions have made attaining a stable Taylor cone-jet mode difficult compared to the organic solvent-based systems used for polymers and small molecules. This is reflected in the fact that the majority of proteins to be successfully electrosprayed like enzymes or bovine serum albumin (BSA), had to be either emulsified or dissolved in a mixture of water and an organic solvent. Because adding organic solvents eliminates a major benefit of the ES technique for proteins, we therefore investigated the feasibility of using fully aqueous media for ES with BSA selected as a model protein for particle engineering. In addition, we investigated the influence on the ES process of adding protein stabilisers commonly used in spray and freeze drying processes and the influence of adding high protein concentrations to the feed solution. This in order to evaluate the feasibility of the technique for future use with monoclonal antibodies, as, to the best of our knowledge, little or no research has been performed regarding the ES process and subsequent storage stability of these therapeutically important proteins. MATERIALS & METHODS Aqueous formulations containing one or multiple known protein stabilisers, were electrosprayed as such or with the addition of BSA, which served as a model protein (Sigma-Aldrich, St. Louis, Missouri, U.S.). The used ES setup (IME Technologies, Geldrop, the Netherlands), included a climate chamber, which allowed for the temperature to be kept at 25 ± 2 °C and the relative humidity at 35 ± 5 % during the ES experiments. Passivated stainless steel dispensing tips (Nordson EFD, East Providence, Rhode Island, U.S.) of 20, 25 and 27G were used as nozzles. Density, conductivity, viscosity (capillary viscometery) and surface tension (pendant drop method) were determined for a selection of the liquid feed solutions. Scanning electron microscopy (SEM) was used to evaluate powders resulting from the ES process. RESULTS & DISCUSSION Because the used ES setup did not allow for directly assessing the cone jet stability, the conducted experiments focused on obtaining solid particles on the collector. Results for the initial excipient screening suggested a concentration and conductivity dependence for the atomisation efficiency of the different formulations. Arginine HCl, sulfobutylether-β-cyclodextrin (SBE β-CD) and phosphate buffer solutions had conductivities in the 10-1 S m-1 range, far exceeding the reported conductivity limit for obtaining a cone-jet mode of 10-4 S m-1 limit 5. Since the addition of BSA or another protein would inevitably further increase conductivity, increasing concentrations of polysorbate 20 were added to the formulations since the atomisation mechanism of ES is based on competition between electrostatic repulsion and surface tension stress at the air-liquid interface. Following the positive correlation between polysorbate 20 concentration and ES process feasibility, further optimisation steps were taken by combining different excipients and subsequently adding BSA in different concentrations. These results suggested that the total solid content might be another critical factor for the ES feasibility for aqueous solutions of BSA, which will be further investigated using a mixture design of experiments incorporating ordinal responses. CONCLUSION An extensive excipient screening and physicochemical characterisation study was conducted which focused on defining relevant parameters for ES of aqueous protein formulations. The data suggested that lowering surface tension and increasing overall solid content of the feed solutions could potentially compensate for the high conductivity and relatively low evaporation rate of the aqueous protein formulations. This will be further investigated using a mixture design of experiments incorporating ordinal responses.