Salting Out. Many proteins are fairly hydrophobic, several seed
storage proteins, for example, such as those from soya beans. Most of these
are insoluble in water, especially at their isoelectric pH. Their solubility is
primarily determined bysolvent quality. The chemical potential of a protein
in solution is to a large extent determined by the solvation free energy of the
groups in contact with solvent, which is positive for apolar groups, and
about zero or negative for most other groups. In first approximation its
value per unit apolar surface areaðAapÞmay be equated to the surface
tension of the solvent, since air also is ‘‘apolar.’’ By and large, solutes that
lower the solubility of proteins also increase the surface tension of the
aqueous phase. The surface tension mostly increases linearly with solute
concentration by an amountsper mole of solute. The increase in surface
free energy would then be proportional toAaptimess. If the solute is a salt,
its concentration can be expressed as ionic strength. As discussed above [Eq.
(7.8)], ionic strength also affects solubility in another way. Combining the
two relations, and further simplifying (7.8), would roughly lead to
ln
csat
c*&C 1 z^2 HIC 2 sI ð 7 : 9 Þwherecsatis the molar solubility,cthe hypothetical solubility forI¼0,
andC 1 andC 2 are constants. Likec, they depend on the type of protein,C 2
being proportional to the molar hydrophobic surface areaAap. s is a
characteristic of the salt present, and all constants depend on temperature.
It follows that for smallIthe first term is overriding (salting in), and
for largeIthe second one (salting out). This is illustrated in Figure 7.13a.
The trends predicted are observed, but the agreement between theory and
observation is not perfect. One reason is that a solution of zero ionic
strength cannot be obtained (except at the isoelectric pH), because the
protein is charged and contains counterions; especially the protein itself
contributes toI(see Section 6.3.1, item 3). This implies thatccannot be
determined with accuracy. Moreover, Eq. (7.9) overestimates the salting-in
effect for highz. Another complication is that at high ionic strength, say
above 0.3 molar, ionized groups at the protein start to bind, i.e., form ion
pairs with, counterions, thereby lowering the electric charge. Nevertheless,
the salting-in effect is about the same for different salts (if the ionic strength
rather than the molar concentration is used), whereas salting out greatly
depends on type of salt (Figure 7.13b), as predicted. The linear relation
between logðcsat=cÞpredicted by Eq. (7.8) at highIis also observed (Figure
7.13c).
The effectiveness of salts to reduce solubility greatly varies and
generally follows theHofmeister series(see Section 3.2). The salts that