Protein Stability

The term protein stability refers to the energy difference between the folded and unfolded state of the protein in solution. Remarkably, the free energy difference between these states is usually between 20 and 80 kJ/mol, which is of the magnitude of one to four hydrogen bonds. Although this suggests that proteins are only marginally stable, the stability is sufficient to prevent spontaneous unfolding at normal temperatures.

TABLE II Conformational Angles of the Major Secondary Structural Elements

Secondary structure   ø ψ    
Helical conformations          
  α-Helix −57 −47    
  310 Helix −49 −26    
  Collagen Helix −78 +149    
β-Strands          
  Antiparallel −139 +135    
  Parallel −119 +113    
β-Turns   ø (i + 1) ψ (i + 1) ø (i + 2) ψ (i + 2)
  Type I −60 −30 −90 0
  Type I´ +60 +30 +90 0
  Type II −60 +120 +80 0
  Type II´ −60 −120 −80 0
  Type III −60 −30 −60 −30

Protein stability is determined by an enormous number of weak interactions observed in the folded state which have to be balanced against an almost equivalent set of interactions with water in the unfolded state. This is a complex problem since every amino acid residue has potential for polar interactions via the peptide bond and a variety of ionic, polar, and nonpolar interactions through its side chains. This accounts for the difficulty in predicting structure directly from its amino acid sequence since the errors in any energy computation are far larger than the net stability of the protein.

A. Hydrophobic Effect
The major driving force in protein folding is the hydrophobic effect. This is the tendency for hydrophobic molecules to isolate themselves from contact with water. As a consequence during protein folding the hydrophobic side chains become buried in the interior of the protein. The exact physical explanation of the behavior of hydrophobic molecules in water is complex and can best be described in terms of their thermodynamic properties. Much of what is known about the hydrophobic effect has been derived from studying the transfer of hydrocarbons from the liquid phase into water; indeed the thermodynamics of protein folding closely followthe behavior of simple hydrophobic molecules in water.

Studies of hydrocarbon models demonstrate that at room temperature the insolubility of hydrophobic compounds is dominated by entropic rather than enthalpic considerations. This is often explained as awater-ordering effect where the insertion of a hydrophobic molecule into an aqueous environment induces a diffusely ordered water shell surrounding the molecule, akin to the formation of clathrates around noble gases and simple hydrocarbons. This shell forms because the hydrophobic compound cannot form hydrogen bonds to the water that surrounds it. Consequently those water molecules have a more restricted set of neighbors withwhomto fulfil their hydrogen bonding capacities. This reduces their degrees of rotational freedom and thus leads to a reduction in entropy. This simple explanation based on ordering of water is inadequate to fully explain the behavior of hydrophobic molecules in solution since both the standard changes in the entropy and enthalpy for the transfer of a hydrophobic molecule to water are strongly temperature dependent. Surprisingly the value for the free energy change for the transfer of a hydrocarbon to water is rather temperature insensitive as a consequence of compensatory changes in the entropy and enthalpy. These temperature dependencies are a consequence of the large temperature-insensitive heat capacity of hydrophobic molecules in solution. Thewaterordering effect appears to be the source of the anomalously high heat capacity of hydrophobic compounds in water.

Although there is no general agreement on the molecular basis of the hydrophobic effect, there is a good correlation between free energy of transfer of an organic molecule to water and its hydrophobic surface area, whereas there are no general correlations between the molecular features of the solute and entropic and enthalpic changes. Many studies have shown that the transfer of 1 Å2 of hydrophobic area to water is accompanied by an unfavorable increase in the free energy of 80–100 J/mol. The fact that this correlation is found to be fairly independent of the nature of the hydrophobic solute clearly indicates that the hydrophobic effect is a fundamental property of water.

Studies on the heat capacities changes observed at the protein folding transition showthat protein denaturation is analogous to the transfer of hydrophobic molecules to water. Furthermore it is well established that the stability of a protein is directly proportional to the difference between the exposed hydrophobic surface area in the unfolded and folded state. In recent years, site-directed mutagenesis has demonstrated the same thermodynamic relationship between the hydrophobic buried surface area and stability as observed for the transfer of organic hydrocarbons into water. That is, each buried 1 Å2 of hydrophobic surface area contributes ∼80 J/mol to the stability of the protein when the only difference is the change in surface area. Clearly the behavior of a protein in solution is more complex than that of a simple hydrocarbon. As noted earlier, proteins are only marginally stable. It would appear that the change in exposed hydrophobic surface area that accompanies protein folding slightly more than compensates for the decrease in entropy of the polypeptide chain as it adopts a well-defined conformation. This explains why it is so difficult to predict protein structure.

B. Hydrogen Bonds
Hydrogen bonds (D—H· · · ·A) are primarily electrostatic in nature and involve an interaction between a hydrogen attached to an electronegative atom (D—H) and another electronegative acceptor atom (A) that carries a lone pair of electrons. In biological systems the electronegative atoms in both cases are usually nitrogen or oxygen. The distance between the donor and acceptor atoms is usually in the range 2.8–3.1 Å where the D—H bond tends to be collinear with the lone pair of electrons. There is some variability in the geometry of the hydrogen bond, which is consistent with the predominantly electrostatic nature of this interaction. For example, in the α-helix and antiparallel β-sheet the N—H is approximately colinear with the C=O bond rather than be aligned with the lone pairs of the carbonyl oxygen. Many of the hydrogen bonds in proteins occur in networks where each donor participates in multiple interactions with acceptors and each acceptor interacts with multiple donors. This is consistent with the ionic nature of hydrogen bonds in proteins.

Originally it was believed that hydrogen bonds made an important contribution to protein stability on account of the extensive hydrogen bonding observed in α-helices and β-sheets. Indeed, virtually every hydrogen donor and acceptor in a protein are observed to form an interaction within the folded structure or to the external solvent. However, protein stability is the difference in free energy between the unfolded state and the folded state. In the unfolded state the polar components are able to form perfectly satisfactory hydrogen bonds to water that are equivalent to those found in the tertiary structure of the protein. Thus hydrogen bonding is energetically neutral with respect to protein stability, with the caveat that any absences of hydrogen bonding in a folded protein are thermodynamically highly unfavorable.

Although hydrogen bonds do not contribute to stability they are a major determinant of protein conformation. The necessity to form hydrogen bonds accounts for the α-helices and β-strands that abound in protein structures.

C. Disulfide Bonds
Many extracellular proteins contain disulfide bonds. In these proteins the presence of disulfide bonds adds considerable stability to the folded state where in many cases reduction of the cystine linkages is sufficient to induce unfolding. The source of the stability appears to be entropic rather than enthalpic. The introduction of a disulfide bond reduces the entropy of the unfolded state by reducing the degrees of freedom available to the disordered polypeptide chain. This stabilizes the folded state by decreasing the entropy difference between the folded and unfolded state. This suggests an obvious strategy for increasing the stability of a protein through the introduction of disulfide bonds. Although this might seem a simple task, the geometry of the disulfide bond is rather restricted. As a consequence the number of locations that can accommodate the replacement of two residues by cysteines in a protein without reducing the thermal stability of the folded protein are quite limited. Studies on T4 lyszoyme have shown that if suitable locations can be found, the degree of stability introduced into a protein is proportional to the size of the closed loop generated by forming a disulfide bond.

D. Ionic Interactions
The association of two oppositely charged ionic groups in a protein is known as a salt bridge or ion pair and is a common feature of most proteins. Typically these interactions contribute very little to protein stability since the isolated ionic groups are so effectively solvated by water. As a consequence very few unsolvated salt bridges are found in the interior of proteins. Furthermore, salt bridges are rarely conserved in orthologous proteins.

E. Dipole–Dipole Interactions
Dipole–dipole interactions are weak interactions that arise from the close association of permanent or induced dipoles. Collectively these forces are known as Van der Waals interactions. Proteins contain a large number of these interactions, which vary considerably in strength.

The strongest interactions are observed between permanent dipoles and are an important feature of the peptide bond. In the peptide bond the dipoles associated with the peptide carbonyl and amide group are aligned and give rise to a significant dipole moment (3.5 Debye units for a peptide bond versus 1.85 for a water molecule). These interactions fall off with the inverse of the second to third power when the dipoles are fixed and to the sixth power when they are free to rotate. So, for example, there is a substantial positive dipole at the amino-terminal end of an α-helix where the dipoles are constrained and aligned. As a consequence the N-terminal end of an α-helix is often utilized to bind negatively charged ligands in enzyme active sites.

Permanent dipoles may also induce a dipole moment in a neighboring atom or group. This is a stabilizing interaction, but is much weaker than that observed between permanent dipoles. These type of interactions are important since they change the charge distribution of neighboring atoms which in turn can profoundly influence activation barriers in enzyme catalyzed reactions.

London or dispersion forces are the weakest of all of the dipole–dipole. These are best described in quantum mechanical terms, but may be viewed qualitatively as the consequence of the transient asymmetry in the charge distribution in a neutral atom that induces a favorable dipole in a neighboring neutral atom thus leading to a weak attraction. These forces are inversely proportional to the sixth power of the separation. Although these forces are very weak there are an enormous number present within a folded protein such that they to contribute significantly stability of the folded state. As a group, the Van der Waals forces are important for stabilizing interactions between proteins and their complementary ligands whether the ligands are proteins or small molecules.