| True solubility data for proteins in given solvent conditions are difficult to obtain. First, crystals have to be grown and the process of crystal growth has to proceed to completion. At this point, a limiting value for solubility at thermodynamic equilibrium is obtained. However, very often cessation of crystal growth can happen, which impedes the solubility measurements. There are a variety of methods in the literature describing measurement of protein solubility. Most of these methods are not practical and require sophisticated equipment, and even then reported values are larger than limiting values or true solubility. During the course of this work, a method was developed, which is of general use, does not require expensive equipment, no specialized reagents are needed, and small quantities of protein are sufficient. The method employs a hanging drop for crystal growth, a simple bench top laboratory centrifuge and a UV-spectrophotometer. For this research two highly purified proteins were used, equine serum albumin with MW of 65.7 kDa and sweet protein thaumatin I with MW of 21.5 kDa.; Experimental data for the solubilities of both proteins were compared with a theoretical model developed by Haas-Drenth-Wilson for the correlation of second virial coefficient values with the thermodynamic solubility. Excellent agreement between experimental data and the model was obtained for equine serum albumin. With sweet protein thaumatin I, quite satisfactory correlation between solubility data and second virial coefficient values were obtained, but not as good as in the case of equine serum albumin.; Solubility data in the literature are almost exclusively obtained with lysozyme and very few other proteins are characterized in terms of even limited phase diagrams. Therefore, data generated in this work may prove to be valuable for other crystal growers, and certainly represent an expansion of the existing data. The method used to generate the solubility data for these two proteins is promising with potential to be easily applied to any protein of interest. Even in a modestly equipped laboratory this method could be used. |
