• Primary structure: the genetically determined sequence of amino acid residues

• Secondary structure: regular, local conformations, chiefly a-helices and b-sheets, stabilized by hydrogen bonding between backbone NH and C=O groups

• Tertiary structure: folding of the protein chain into a compact shape stabilized largely by lipophilic interactions between amino acid side chains

  • Electrostatic interactions ("salt bridges") between charged side chains also play a role

  • The interior of the structure is largely close-packed

• Quaternary structure: assembly of multiple protein chains into a functional protein, again stabilized largely by lipophilic interactions

Although classification systems vary, proteins appear to be constructed of a limited number of domain types, built from a limited number of simple motifs. That is, evolution has not made use of the full range of conformation space available to proteins.

S.-H. Kim's group at Berkeley et al. [Proc. Nat. Acad. Sci., 2003, 100, 2386; 2005, 102, 618; 2005, 102, 3651; Protein Sci., 2006, 15, 1723] attempted to define the conformation space used.

• They calculated measures of pairwise structural similarity among the 1898 proteins in the PDB Select Database

• The calculations used 25,000 cpu hours on an IBM RS 6000

• The results were mapped onto a three-dimensional paramaeter space, roughly described as a, b, and a/b:

The Use of Protein Conformation Space

Here's the same idea, mapped onto Ramachandran diagrams:

Ramachandran Diagrams of Conformation Space

• For these diagrams, the proteins were broken into 6-9 residue segments

• The second diagram transforms the first from 0-180 degrees to 0-360 degrees.

Soding and Lupas [BioEssays, 2003, 25, 837] note that structural homology is substantially more conserved through evolution than is sequence homology. They argue that:

• Modern proteins evolved by fusion of simpler "building block" proteins

• Each of these simpler proteins represents a conserved fold

• Here are some of their suggestions:
  

Some issues that are not completely resolved:

• To what extent do crystal structures reflect structures in solution ?

  • Substantial evidence now exists that many proteins are conformationally heterogeneous in solution

  • However, many small organic molecules that are conformationally heterogeneous in solution crystallize in a single conformation

• To what extent do structures in solution or in crystals reflect structures within the cell ?
  • Many proteins exist within cells in complexes with other proteins or bound to structural elements of the cell