Thermodynamics of Partitioning 2421 Phase Transfer

Dearden [31] pointed out in his review in 1985 that a thermodynamic analysis of partitioning cannot explain the partitioning process fully, because each parameter (enthalpy, entropy) reflects the difference between behavior in each phase, and tells us nothing directly about the absolute contributions in each phase. It is only by the thermodynamic investigation of gas-water and gas-lipid (octanol) partition that the contribution of each phase can be properly assessed. In terms of the driving force for partitioning from water to octanol, Cramer's investigations discussed above [62], and a wealth of experimental solvation data collected since [67, 68], have indicated that the dominant contributions for nonpolar solutes come from interaction with the lipid phase. In partitioning, then, it is not "hydrophobicity" but lipophilicity which is the driving force.

Studies on the role of packing interactions in stabilizing the folded form of proteins have led to the same conclusion regarding the relative importance of interactions in the aqueous phase, and interactions (packing) in the folded state.

In 1990 Dill [69], in an excellent review of the forces responsible for protein folding, concluded that the dominant force is "hydrophobic" but only provided that the term is operationally defined in terms of the transfer of nonpolar side chains from water into a nonpolar environment.

In 1992, Sneddon and Tobias [70] carried out a molecular dynamics simulation of the thermodynamics of interconverting isoleucine and valine side chains in the core of the protein, ribonuclease Ti- They concluded that burial of nonpolar side chains in the in terior of the protein is favorable not so much because of the aversion of nonpolar groups towards water ("hydrophobicity") but rather because these groups can participate in favorable packing interactions (enthalpic, van der Waals interactions) within the core of the folded protein. So, with protein folding as with partitioning of solutes between aqueous and lipid phases, the emphasis is still on lipophilicity as the driving force.

Dearden's review [31] covered pertinent thermodynamic investigations of simple solute partitioning up to 1985. In 1990, Burgot and coworkers [71] investigated the thermodynamics of octanol/water partitioning of a series of ^-blockers (Fig. 3), ranging in logP from 0.16 (atenolol) to 3.37 (propranolol). They measured the enthalpy of transfer, and derived entropy from the independently measured partition coefficients. Apart from one compound, sotalol, which contains the highly polar sulfonamide group, transfer was dominated by the entropy term. This result was in accordance with the earlier thermodynamic investigations by Weiland et al. [72] of binding to the [i-receptor itself, which lead them to suggest that antagonist binding is entropy-driven, whereas agonist binding is enthalpy-driven. The quite reasonable expectation of opposite behavior (in terms of enthalpy-versus entropy-driven binding) of agonists and antagonists is not, however, consistently observed in all studies [73, 74].

An intriguing study of the thermodynamics of partitioning of an extensive series of substituted benzoic acids, between octanol and water, was made in 1992 by Da, Ito and Fujiwara [75]. From van't Hoff plots (note that Beezer, Hunter and Storey [76] have pointed out that errors may be considerable when thermodynamic quantities are measured by use of van't Hoff plots), enthalpies and entropies of transfer were obtained, atenolol atenolol sotalol sotalol propranolol

Figure 3. Selected |3-blockers. log P values increase from top to bottom.

and the original Hansch n constant was separated into jth (enthalpic) and tts (entropic) parts according toEq. (17):

For the great majority of substituents, theenthalpic component is dominant, that is | jth | > | jrs | . However, for alkyl and higher alkoxy substituents, entropy dominates, so | ,7rs | > | nH | . There is no correlation between jth and jts. The study shows clearly that substituent constants are not determined exclusively by enthalpy or by entropy, but that the two terms contribute cooperatively. Furthermore, use of separate enthalpy or entropy parameters rather than free energy parameters was shown in several instances to enhance the correlation coefficient in QSAR equations. The Aqueous Phase and the "Hydrophobic Bond"

The attraction of two nonpolar groups for one another in water has a strong parallel with the process of partitioning of a nonpolar solute between water and a nonpolar solvent, and it is widely believed that this force, the "hydrophobic effect" or "hydrophobic bonding" is due to some special properties of water. Controversy persists over both the physical origin of this attraction, over the properties and extent of "ordered" water that is widely believed to exist in the vicinity of nonpolar solutes, and over the use of the term "hydrophobic" or "hydrophobic bond" to describe the phenomenon. Kauzmann [77] in 1959 was the first to propose that the aversion of nonpolar groups for water is the most important factor in stabilizing the folded state of proteins, and he introduced the term "hydrophobic bond" to describe the apparent attraction between two nonpolar groups in water. Use of such a term was criticized by Hildebrand [78] in a letter to the Journal of Physical Chemistry in 1968. Hildebrand pointed out that the noun, "bond", was inappropriate because the apparent attraction between alkyl groups in water has none of the features which distinguish a chemical bond from van der Waals forces. Further, he regarded "hydrophobic" inappropriate, on the grounds that alkyl chains in micelles of soap are not bonded together because of phobia for surrounding water, for they stick together just as strongly in the absence of water. Hildebrand suggested one speaks simply of alkyl interaction free energy, or entropy. The net free energy of solution of hydrocarbons in water at room temperature is dominated by a large, negative entropy term, believed to arise from the increased ordering of water molecules in the immediate vicinity of the nonpolar solute.

Nemethy, Scheraga and Kauzmann [79] replied to this criticism by saying that they did not wish to argue on a point of nomenclature, for the term "hydrophobic bond" had proved useful as shown by its frequent occurrence in physical, chemical, and biochemical nomenclature. Moreover, because of the entropy factor, the water-hydrocarbon system does differ in a unique manner from other systems of low miscibil-ity - so why not use a unique term to describe it?

So what explanations have been offered, at the molecular level, for this unique behavior? The review by Dill [69] in 1990 can usefully be consulted, and also in 1990 there was published an excellent discussion by Taylor [80], who goes into the various interpretations that have been put forward to explain Cramer's [62] results.

The traditional explanation of NĂ©methy, Scheraga and others is that water molecules around the nonpolar solute arrange themselves like an "iceberg" or "flickering cluster", an ordered state of low energy, low entropy, with good water-water hydrogen bonds. When two nonpolar residues approach one another, two solvent shells merge into one, with release of some ordered water molecules to "bulk" water, consequent increase in entropy, and decrease in free energy. This view has frequently been challenged as over-simplistic, and attention has been drawn to the role of attractive (en-thalpic) forces in water which cannot be dismissed, and to the contribution to entropy changes made by flexible solutes (alkyl chains) which are more restricted in solution -any solution - than they are in the gas phase [80].

A new view of the hydrophobic effect was advanced by Muller in 1990 [81, 82] and in 1992 a review by Muller [83] summarized the way in which divergent opinions on the nature of the effect have arisen. According to Muller, all available data are consistent with the idea that some structural reorganization of water adjacent to nonpolar groups does occur, and that this is indeed responsible for the folding of proteins and drives the association of nonpolar groups. However, the organization of this "hydration shell water" is not iceberg-like. Muller's view, in essence, is that one may distinguish between "hydration shell" and "bulk" water molecules in the following way: hydration shell H-bonds are enthalpically stronger than bulk H-bonds, but a greater fraction of them is broken. This view leads to a neat explanation for the well-known effect of urea in increasing the solubility of hydrocarbons larger than ethane. It is postulated that urea reduces hydrophobic hydration by two mechanisms [82]. First, it occupies space in the "hydration shell" that could otherwise accommodate water molecules, and second, it alters the contribution of van der Waals interactions to the enthalpy of solvation.

0 0

Post a comment