Molecular Biology: Protein Structure and Function - Stepanov V.M. 2005

Secondary Structure
On the Relationship Between Primary and Secondary Structures

As already mentioned, when describing Secondary Structure, amino acid side chains and long-range interactions are typically neglected. However, as our understanding becomes more refined, these factors must be taken into account. For instance, moving from Glycine to Alanine—an amino acid with the simplest side chain—drastically narrows the range of possible Conformations. This restriction becomes even more severe for peptide bonds involving the imine nitrogen of Proline. A more nuanced consideration of individual amino acid properties reveals specific preferences for The formation of particular secondary structures.

This is driven not only by the stereochemical influence of side chains on conformational stability, but also by how they pack On the surface of the emerging structure and the resulting pattern they form. For example, if hydrophobic residues are distributed in the Primary Structure such that they are separated by two to three polar amino acid residues, they form a distinctive ridge of hydrophobic radicals along one side of the a-helix. The interaction of such ridges can drive the Formation of the Spatial Structure, which in turn stabilizes the secondary structure. Thus, the preferential formation of a specific type of regular secondary structure can be determined not only by The Nature of The amino acid residues, but also by their spatial arrangement within a given segment of the peptide chain.

Statistical analysis of established protein spatial structures has proven extremely useful, allowing researchers to assign individual amino acid residues parameters that reflect their propensity to incorporate into an a-helix, a ß-Structure, or a ß-turn. By scanning a newly determined primary structure using a "sliding window" corresponding to several consecutive residues, one can identify characteristic clusters of residues prone to a-helix formation, suggesting that this region of the protein is folded into an a-helix. This hypothesis is then verified by checking whether the residues at the ends of the helix match the Amino Acids typically found in such positions. Further refinements are provided by stereochemical considerations.

Several approaches have been developed to predict Protein secondary structure from primary structure. Among these, the Methods proposed by P. Chou and D. Fasman, O. Ptitsyn and A. Finkelstein, and V. Lim have gained the widest acceptance. As tests on a range of Proteins have shown, the probability of successfully predicting a-helical regions is quite high, whereas predictions for ß-structural elements and turns are less reliable.

The Challenge of making unambiguous secondary structure predictions is understandable, because once a secondary structural element forms, it must integrate into the tertiary structure—that is, become part of a network of long-range interactions. These interactions can either stabilize the secondary structure or disrupt it. Consequently, there is always a certain probability that a prediction will fail. Indeed, analysis of numerous protein spatial structures has shown that identical short Amino acid sequences can sometimes form the same secondary structures in different proteins, as established, for example, for eight hepta- and hexapeptide segments. However, cases where identical sequences yield different secondary structures in proteins of varying architecture are also common; this has been observed for seven hexapeptide fragments. In particular, the exact same sequence

Class="center">Asn—Ala—Ala— Не— Arg— Ser

adopts an a-helical conformation in Phosphofructokinase, nestled between two other a-helices, whereas in Thermolysin it forms part of a ß-pleated sheet alongside four other peptide chain segments.

Thus, long-range interactions within the tertiary structure can significantly modulate the primary-to-secondary structure relationship, making it far from strictly deterministic. Nevertheless, the fact that many secondary structures—especially a-helical regions—can be successfully predicted clearly demonstrates that this relationship is real. Secondary structure prediction is particularly valuable when comparing proteins, especially in cases where primary structure alignment fails to reveal a significant percentage of identical amino acid residues (e.g., falling below 15%), yet there is reason to suspect that the proteins are evolutionarily related and share a similar three-dimensional fold. In the absence of direct data on the polypeptide chain folding of such proteins, similarities in predicted secondary structural elements can serve as a compelling argument. Frequently, secondary structure prediction also helps uncover distinctive features of specific protein regions; for instance, hydrophobic a-helices may turn out to be transmembrane segments.