Principles of Protein Structure - G. Schultz 1982

Amino Acids
Empirical Analogies Among Amino Acid Residues

Biological diversity simplifies analysis. The vast multitude of biological species will probably never be fully explained. The unique features of each species depend on such A large number of random events (events that remain unknown to us and are therefore considered random) that a purely historical approach has gained great importance in biology. Although at first glance this approach may seem ill-suited for explaining fundamental principles, in practice it has proven extremely useful.

Substitution frequencies reveal similarities. Regarding side chains, for instance, an analogy can be found by analyzing the frequency of Amino Acid Substitutions in the sequences of corresponding Proteins from homologous organisms. To obtain a more precise measure of similarity, the observed frequencies are normalized—that is, compared to the frequency of substitutions that might occur by chance. Due to the uneven distribution of Amino Acids in nature, the latter characteristics are specific to each substitution: an Ala-Leu substitution is 13.0 ∙ 7.8/3.3 ∙ 2.2 = 14 times more likely than a Phe-Tyr substitution (see the Column "Relative Abundance in E. coli proteins" in Table 1.1).

Class="center">Table 1.2 Matrix of mutual substitution frequencies along the evolutionary path of 256 accepted point Mutations per 100 residues (PAM)^a

^a Data taken from [20]; the PAM unit is defined in Section 9.1. Each value represents The ratio of the probability that a given pair of amino acids occurs at a specific position in proteins of a given type to the probability that this would happen by chance. All values are multiplied by 10. For example, in two sequences at an evolutionary distance of 256 PAMs, Leu and Met are found 2.1 times more frequently than would occur by chance. The Asp-Tyr pair occurs only 0.1 times as frequently as by chance, meaning that for a given folded chain, the Asp → Tyr or Tyr → Asp substitution is highly unfavorable and therefore forbidden. The Amino acids are arranged such that the value ∑mij (t — j)² is minimized. Thus, amino acids that frequently substitute for each other at a given position in the folded chain are clustered together.

Groups of interchangeable residues. The observed substitution probabilities are given in Table 1.2. The amino acids are arranged such that the second moment about the diagonal, ∑mij(j—i)², is minimized. This arrangement reveals interchangeable groups. The substitution of one amino acid for another occurs predominantly within such groups. Therefore, these amino acids are mostly similar to one another regarding their overall effect on protein structures.

The empirical CHARACTERISTICS OF THE residues in Table 1.2 show that the aromatic amino acids Phe, Tyr, and Trp form a group of interchangeable residues and that their roles are closely related. Additionally, the positively charged residues Lys, Arg, and His form a similar group. Notably, the negatively charged residues Glu and Asp are clearly segregated from the positively charged ones. Thus, substitutions between oppositely charged residues are rare. A high substitution frequency would be expected if charged residues were always located on the molecular surface and served no other function than to increase Protein solubility.

Side chain size is important. Clearly, the large aliphatic nonpolar residues Val, Leu, and Ile (but not Pro!) form a group of interchangeable amino acids that also includes the mildly polar Met and Cys. All small residues (Ser, Thr, Asp, Asn, Gly, Ala, as well as Glu, Gln, and Pro) form a separate group. Consequently, dissimilar residues such as the nonpolar Pro and the strongly polar Asp are grouped together. This suggests that for protein structures, side chain size is almost as important as its chemical nature.