consists in determining the dependence of protein solubility in the
given aqueous solvent on the concentration of PEG in the solvent (1)
–protein (2) –PEG (3) system. Extrapolation of this dependence to
CPEG= 0 gives the value for the effective activity of the protein in its
saturated solution (log Cbiopolymer). Evidence for the validity of this
extrapolation includes (a) the experimentally observed linearity of
log solubility versus PEG concentration plots, (b) the extrapolation
of such plots to correct activities in the situation where protein
activities can be experimentally determined, and (c) the independence of the extrapolated activities on protein concentration over
a wide range. A more detailed analysis [35] makes it possible to
relate the activity to the value of the second virial protein coefficient
characterizing the protein–solvent interaction. The data obtained
represents the activity of BSA in its saturated solution and gives—in
the case of water as a solvent—an indication of the hydrophily of the
biopolymer. In thermodynamic terms, Rosenberg’s methods relate
the activity of the protein to the second virial coefficient A12that
characterizes the water (1)-biopolymer (2) interaction [34].
Experimentally, the method consists of preparing binary solutions of PEG and BSA at the required temperature, and pH. After
mixing for 1 h, a separation of phases is established by means of
centrifugation at 50.000 × g for 30 min. The weight concentration of
BSA in the supernatant is determined by determination of absorption at 279 nm.