Lecture 21 March 2005

Function can be explained on several levels: The function of a kinase enzyme is phosphotransferase activity, but on a higher level it might be part of signal transduction system. (Fig. 2-1)

Also, some proteins have several functions, in some cases, both well-known functions and newly-discovered functions.

Recognition, Complementarity and Active Sites

Binding is important for proteins involved in signaling, catalysis or transport.

1. The molecule that is bound, the ligand, can be small or large (another protein).

2. Binding is non-covalent, and uses the standard interactions

3. Specificity is usually high (the affinity relative to that of similar molecules)

4. Specificity is a consequence of complementarity of shape and polarity

(Fig. 2-2) (Jmol)

A protein can form a binding pocket for a ligand that is quite different from that found on the surface, or in water itself.

Such specialized environments are particularly important for enzymes. For example, enzymes often have groups with shifted pKa values to facilitate catalysis. We are familiar with groups of opposite charge forming an ion pair, which might contribute to the stability of a protein.

1. In an enzyme two groups that could ionize might be held too far apart, and in a nonaqueous environment, to transfer the proton, leaving both in the unionized state. That way each might participate in catalysis, one as a strong acid, and one as a strong base. This could not generally occur in solution.

2. The environment of a single group can cause a shift in its pKa. Two lysine residues that are close to each other, will make it more difficult to protonate the second one. (Fig. 2-3) (It becomes a stronger acid, its proton affinity down.) This can facilitate its role in proton transfer. Likewise for a lysine found in a nonpolar environment. This will favor the uncharged, unprotonnated state. This can facilitate some chemcial steps, such as a nucleophilic attack.

Flexibility and Protein Function

Proteins are commonly thought of a "molecular machines", but in fact they are generally much more flexible than macroscopic machines made of metal. The early metaphor of Emil Fischer of the "lock and key", for an enzyme and its substrate has some validity, but is not entirely accurate. Most enzymes are highly specifc for their substrates, while their are some exceptions. However, many enzymes change conformation after binding one of their substrates. This is known as induced-fit, as we have seen previously.

(Fig. 2-4) Tight fit between enzyme and substrate

Proteins are inherently flexible because their conformations are maintained by numerous weak interactions. At amy moment many of them might be broken, leading to local, transient rearrangements. This flexibility can allow an enzyme to bind a variety of slightly different substrates, but it also allows a variety of different inhibitors to bind to a single enzyme. See the example of the HIV-protease. (Fig. 2-5)

Binding of ligands can range from loose binding (K_D = 10^-3 M) to extremely tight binding (K_D = 10^-12 M) Tight binding is usually associated with subtle conformational changes that allow greater surface complementarity. Denaturants of proteins (urea or SDS) usually eliminate very tight binding.

One property of enzymes that function at very high temperatures in thermophilic organisms is greater rigidity. (Fig. 2-6) This can be seen as shorter N- and C-termini, and shorter loops, in some cases.

The flexibility of proteins varies greatly. Some proteins undergo large conformational changes, for example, the binding of AMP by adenylate kinase. (Fig. 2-7)

Location of Binding Sites

Binding sites vary in size and in other characteristics, depending also whether the partner is a small molecule, another protein, or a nucleic acid. Protein-protein interaction surfaces can be quite large, hundreds of square Å, and discontinuous surfaces, (Fig. 2-8). (Fig. 2-9)

Small ligands often bind in cavities. This allows special environments that might be necessary for tight binding. Protein flexibility is often necessary to allow the ligand to enter the cavity. (Fig. 2-10)

We have previously discussed the location of active sites in several standard folds. From another point of view, the binding of ligands or substrates often occurs at the interface of two domains, or of two subunits. (Fig. 2-11)

For example, the binding of ATP in the ATP synthase occurs at the switch region near the C-terminal ends of the parallel β-sheet (Rossmann fold), but it also occurs at the interface of two subunits.

Nature of Binding Sites

Prediction Server Q-site Finder

Laurie AT, Jackson RM.
Q-SiteFinder: an energy-based method for the prediction of protein-ligand binding sites.
Bioinformatics. 2005 Feb 8; [Epub ahead of print]

One common feature of a ligand-binding site is hydrophobic surface. Due to the hydrophobic effect, the binding of a ligand is more favorable if a larger amount of hydrophobic surface is buried upon binding.

Hydrophobic surfaces on proteins can lead to non-specific associations between proteins, otherwise known as aggregation, which is dangerous. Binding sites for small ligands are usually concave, with the hydrophobic surface less accessible. (Fig. 2-12)

Many protein-protein interactions are transient, which requires that they have only moderate stability. Relatively small hydropobic patches will usually lead to moderate affinity. Transient binding is important in signal transduction pathways. More stable complexes can arise when a complex of two proteins creates a new binding site for a third protein. The 3-subunit complex could be more stable than any combination of 2. Covalent modification, such as phosphorylation, can alter conformation, and weaken binding, allowing proteins to dissociate and find new partners. (Fig. 2-13) This is refered to as partner swapping.

Domain swapping allows more stable complexes to form. As illustrated in (Fig. 2-14), in a trimeric complex, part of one subunit invades another and displaces the corresponding part. The domain-swapped complexes can be transient or long-lived. (Jmol)

Proteins should generally be considered to have a layer of water molecules bound to their surfaces. This can be demonstrated by ultracentrifugation, and the water molecules will appear in the crystal structures, as we have seen. (Jmol). Therefore, when a ligand binds to a protein, it must not only be desolvated (stripped of water), but must also displace the waters residing at or near the binding site. The energetics of this process is in general complex.

Distinguishing between interactions that contribute to specificity and to affinity, is based on whether the interactions are directional or not. Hydrogen bonding is directional, so it usually contributes to specificity. Nonpolar interactions are not directional, and therefore they will contribute to affinity, but not specificity. In the case of a fatty acid binding protein, specificity is with respect to the carboxyl group of the fatty acid. (Fig. 2-15) (Jmol)

Structural Proteins

Cells are highly organized, largely due to structural proteins inside.

Well-known structural proteins include silk, collagen, keratin, elastin.

They can be held together by numerous weak, noncovalent interactions (the usual H-bonding, ion pairing, and hydrophobic), and also by covalent linkages (disulfides, or other specific, enzyme-catalyzed bonds). (Fig. 2-17)

Some structural proteins are also enzymes. Muscle is such an example in which ATPase enzymes are coupled to the filaments.

Some structural proteins function as scaffolds, for the binding of numerous other proteins. Ste5p forms a scaffold for the MAP kinase signalling pathway of response to mating type pheromones in yeast. It not only binds 3 kinases, which function sequentially, but also binds a β subunit of a heterotrimeric G protein. (Fig. 2-18)

Comments/questions: Email me