28 February 2005
Biol 6312
Quaternary Structure
Many proteins are composed of multiple polypeptide chains. They might be identical chains, or nonidentical chains, or a mixture: a2, ab, a2b2, ab2c10
Such assemblies are called oligomers. The examples a, b, and c above are called monomers or subunits. Dimers (2), Trimers (3), Tetramers (4), Pentamers (5), Hexamers (6), Heptamers (7), Octamers (8), Nonamers (9), Decamers (10), Undecamers (11), Dodecamers (12)
We can distinguish between Homodimers (a2) and Heterodimers (ab). In the case of hemoglobin the 2 different subunits resemble each other. In many heterodimers the 2 subunits are unrelated.
How do oligomers form? In general, they form spontaneously. This requires specificity and strength.
Specificity requires complentary interfaces. Correct interfaces must be favored over incorrect ones. Complementarity requires more than matching shapes.
Polar groups must be matched with polar groups, so that H-bonding is maximized.
Nonpolar groups must be matched with nonpolar groups, with van der Waals interactions maximized.
Positive charges must be matched with negative charges, or possibly with dipolar groups.
Since these tend to be weak interactions, a strong binding oligomer must have many such interactions.
Example of complementarity (Fig. 1-66) Interleukin-4 and its receptor. (Jmol)
We have already considered coiled-coils, which have repeating patterns of interfacial residues.
Coiled-coil (Fig. 1-67)
Leucine zipper (Fig. 1-68)
The strength of the interactions determines whether the oligomer is a transient complex or a stable one.
The secondary structures might be stabilized by the formation of the complex.
How do subunit interfaces differ from the interiors of proteins?
The same interactions exist:
H-bonding, ion pairs, hydrophobic interactions, disulfide bonds, and metal ligation
The first 3 are the common noncovalent interactions. Hydrophobic interactions are important, just as they are in folding proteins. A greater hydrophobic surface buried, generally indicates a stronger interaction. This is due to the hydrophobic effect. Ion pairs are also common at subunit interfaces. Hemoglobin is a well-known example. Binding of oxygen leads to the breaking of ion-pairs, some at the subunit interfaces. This is part of the mechanism of cooperativity.
So, ion pairs can be regulated by pH. At low or high pH they can be broken, causing subunits to dissociate, or to more subtly change their positions.
| High pH | |
| Ion pair | |
| Low pH | |
What is the contribution of H-bonding at the interface of subunits? Specificity
H-bonds are short range and rather directional. That means that making the H-bonds at a subunit interface will tend to align the 2 subunits precisely, and keep them that way.
H-bonding between subunits can also involve joined secondary structure, usually in the form of a beta sheet. (Fig. 1-70)
Water molecules are also often found at subunit interfaces. For example, they can bridge two polar groups by H-bonding. (Fig. 1-69) shows the dimeric pre-albumin (PDB 1bm7).
Both water molecules and secondary structure joining are illustrated in the co-crystal of 2 subunits from the ATP synthase. (Jmol)
Inappropriate quaternary interactions due to mutation may lead to disease states
Sickle-cell anemia is the most well-known example. This is the first characterized "molecular disease" and was investigaed by Linus Pauling and colleagues. See the historical article:
Strasser BJ.
Linus Pauling's "molecular diseases": between history and memory.
Am J Med Genet. 2002 Aug 30;115(2):83-93.
Another article, not available to us.
In sicle-cell anemia, one surface residue of the beta chain is changed from Glu to Val. Ordinarily this might not be a severe problem. Although Val is hydrophobic, a single such change is usually not enough to disrupt function. In this case, it is not the disruption of function that is the problem. Rather it is the ability of the mutated beta chains to mediate fiber formation. Normally Hb consists of 4 chains, 2 alpha and 2 beta. Hb is found at extremely high concentration in the red blood cell. The sickle cell mutation allows a complementary surface to form, leading to beta-beta interactions between tetramers, and fiber formation. This surface forms only in the deoxy-state of Hb. Therefore fibers tend to form in the capillaries, distant from the heart and lungs, where oxygen concentration is low. Capillaries are most susceptible to damage because they are so narrow. (Fig. 1-71)
So, in sickle-cell anemia, an oligomeric protein becomes a fibrous protein, due to an increase in complementary surfaces.
Oligomeric proteins are also sensitive to the dominant-negative effect. If a protein can no longer function due to a mutation, for example a domain is missing, it might be lethal in a homozygous situation, but not so in a heterozygous situation. That is, the normal gene can produce enough normal protein for the cell to survive. But if the protein functions in an oligomer, one non-functional monomer might be enough to inactivate the entire oligomer. In that case, the mutation is called a dominant-negative. Illustrated in Fig. 1-72.
Reference:
Herskowitz I.
Functional inactivation of genes by dominant negative mutations.
Nature. 1987 Sep 17-23;329(6136):219-22. Review.
In this example, the mutated protein dimerizes, and each subunit independently binds another protein (blue). Both of the blue proteins must be present for function. If one of the dimeric proteins cannot bind a blue protein, but it can still dimerize, then it will be dominant-negative. So, if the functional and non-functional proeins are produced at equal levels, only 25% of the dimers will be active.
Geometries of Quaternary Structure
Protein complexes that are built from identical subunits tend to have symmetry, while those built from several non-identical subunits will lack symmetry.
In a complex built from identical subunits, each subunit will have a binding surface A and a complementary surface A'. This could result in dimers or complexes of greater numbers, depending upon the arrangement of the binding surfaces. Some complexes have subunits with additional binding surfaces, B and B', allowing higher orders of symmetry. (Fig. 1-75)
Some complexes are mixed, such as the human growth hormone bound to its receptor, which is found in two copies. (Fig. 1-73) (Jmol)
A symmetric complex is built from identical protomers, where a protomer is one or more different subunits. In the case of Hb, the protomer contains 2 subunits, one alpha and one beta. The protomer must be asymmetric.
We will not discuss symmetry rigorously. In fact, most symmetric complexes are formed from a relatively small number of protomers. And most often they have simple rotational symmetry. Of course, proteins cannot have mirror or inversion symmetry since the amino acids are exclusively L-form. Let's look at some examples. (Fig. 1-74)
Jmol: Dimer, Trimer, Planar tetramer, Tetramer, Pentamer, Planar-hexamer, Hexamer-trimer of dimers, Heptamer, Octamer-tetramer of dimers, Dodecamer-trimer of tetramers
The simple planar symmetries have the A-A' interfaces only, while the others have B-B' (and more) in addition.
Protein Flexibility
The protein structures we have seen have been presented as static images, but in fact, proteins have some flexibility. This refers to a variety of motions that tend to be very fast in comparison to the time it takes to collect data for a protein structure determination. Therefore, structures in the protein data bank, are in general, average structures.
Three kinds of motion can be described:
The larger motions tend to be slower (Fig. 1-76)
Ligand binding often causes local conformational changes. For example, a loop might fold down over the substrate of an enzyme. This can be a hinge type movement, although many other small motions might be involved. Overall, such changes tend to be local, i.e., they involve parts of the protein near the substrate binding site. (Fig. 1-77)
An excellent web site exists that catalogs motions in protein.
Link to the Molecular Motions Database for Triosephosphate Isomerase
Protein motions occur in the interior among non-covalently bonded atoms. These are usually small motions on the order of 1Å. For motions to occur in the interior, numerous fluctuations must occur at the same time. For example, an aromatic ring can flip in about 1 picosecond (10-12 sec), but this occurs only about once in 109 picoseconds (1 millisecond).
So, although protein interiors are densely-packed, they are more dynamic than typical crystalline salts. Amino acids on the surface are not subject to such limitations on motion.
Some proteins have several well-defined conformations. Energetically there must be a small barrier to pass from one to the other. Such a transition would be promoted at higher temperature. (Fig. 1-79)
Link to the Molecular Motions Database for T4 Lysozyme
Ligand-induced Conformational Changes are important movements. These movements often occur between an inactive form and an active form.
One example is called Induced-Fit. This usually occurs with enzymes that have 2 substrates. The binding of one substrate cause the enzyme to change conformation, now allowing it to bind the second substrate. This mechanism prevents hexokinase from binding ATP, and hydrolyzing it before glucose binds. (Molecular Motions Database)
Link to the Molecular Motions Database for Aspartate Aminotransferase (Fig. 1-80)
Ligand-induced conformational changes can affect the quaternary structure of proteins.
One example is Hb, which changes in subtle ways when oxygen binds.
More drastically, the oligomeric state of actin can change upon ATP binding, forming long polymers.
Comments/questions: Email me
Copyright 2005, Steven B. Vik, Southern Methodist University