1.0 Background and History

Soon it was demonstrated that all nearly all naturally occurring protein fibers could be grouped into one of three classes according to their X-ray diffraction patterns: the alpha-type containing (among others) the proteins of unstretched hair, fingernail, horn and bacterial flagella; the beta-type characterized by stretched hair and silk fibroin; and the third type called gamma containing the protein collagen. Unfortunately, these fibrous proteins yielded only relatively ill-defined diffraction patterns compared to crystals of very much smaller molecules which would hamper further efforts of structural characterization of these molecules. It is now clear that the success in characterization of these fibrous proteins was due to the fact that the proteins comprising them were of a single secondary structure type. Some soluble, globular proteins were known to crystallize already by the late 1920's and the inability to interpret the X-ray diffraction patterns of these crystals were taken as evidence of little long-range order in the polypeptide chain.

Meanwhile, in America, researchers at the California Institute of Technology set out on another, albeit indirect path. Recognizing the limitations of the X-ray technique as applied to fibers, they turned their attention to crystals of amino acids and simple polypeptides with the hopes of learning enough about the covalent geometry of the polypeptide chain to permit a guess as to how the folded polypeptide chain might look. By 1950, they had produced the structures of a few amino acids, simple polypeptides and related molecules at atomic resolution. From these data, the general characteristics of the polypeptide chain were revealed. Bond distances and bond angles were measured with an accuracy of 0.02Å and from these data it was determined that the atoms comprising the peptide bond were in a trans configuration, within a few hundredths of an Ångstrom of lying in a common plane. The geometry of the polypeptide hydrogen bonds involving the peptide backbone atoms were also analyzed and shown to be fairly independent of sidechain influences. Based on the assumptions that:

1. the geometry of the polypeptide backbone was the same as that found in the X-ray crystal structures of amino acids and related compounds

2. all residues were in equivalent positions

3. each residue participates in at least one hydrogen bond

Closely following these developments was a Danish researcher K. Linderstrøm-Lang who, with great insight, reasoned that there should be at least four levels of structural organization present in protein structure. In Linderstrøm-Lang's hierarchy of protein structure (see Linderstrøm-Lang, 1952; Linderstrøm-Lang & Schellman, 1959) each level was characterized by a particular type of organizing force and higher levels of organization were composed of elements described by the previous level. We now know that this organization is an oversimplification, but the organization of structure into levels is still useful from a pedagogical viewpoint. The definitions of the structural hierarchy as proposed by Linderstrøm-Lang are as follows:

primary structure: The chemical structure of the polypeptide chain or chains in a given protein - i.e., the number and sequence of amino acid residues linked together by peptide bonds.

secondary structure: Any such folding which is brought about by linking the carbonyl and imide groups of the backbone together by means of hydrogen bonds.

tertiary structure: A organization of secondary structures linked by "looser segments" of the polypeptide chain stabilized (primarily) by sidechain interactions. Disulfide bonds are included in this level.

quaternary structure: The aggregation of separate polypeptide chains into the functional protein.

Fred Richards has written a very interesting account of his personal interactions with Linderstrøm-Lang as a post-doctoral fellow in 1954 (Richards, 1992). Contains a brief historical perspective of work on protein chemistry. Highly recommended!