Principles of Protein Structure - G. Schultz 1982

Protein–ligand interactions
Ligand-binding sites of immunoglobulins

Proteins exhibit selectivity in their interactions with cellular components. Unlike naturally occurring proteins, chemically synthesized Polypeptides behave like unruly children: they indiscriminately "poke," bind, and disrupt many low-molecular-weight metabolites [614]. Natural polypeptides—proteins—have been "trained" by evolution to interact with only a limited set of molecules [586]. This could only happen because, unlike synthetic polypeptides, proteins acquired The ability to form specific, compact structures. Specific binding is an intrinsic property of proteins, and the most general outcome of protein structural Organization has turned out to be a tendency toward non-binding rather than binding.

The only proteins that have acquired the ability to form binding sites for virtually all types of molecules are IMMUNOGLOBULINS (Ig). The best-studied immunoglobulins include plasma Antibodies as well as myeloma proteins, which are produced by Cancer Cells but otherwise behave like normal immunoglobulins. It has been demonstrated that the reactions by which many human and mouse myeloma proteins bind ligands* closely parallel those characteristic of the Ligand–antibody system (see works [540] and [617]).

* In biochemistry, a "ligand" refers to a small molecule attached to a larger one. In coordination chemistry, ligands are molecules or ions clustered around a central entity, such as an iron ion. In this chapter, the term "ligand" is used in both Senses.

The three-dimensional Structure OF THE binding sites has been determined for two different myeloma proteins: human immunoglobulin New, which binds vitamin K derivatives [618], and mouse immunoglobulin McPC 603 [607], which binds phosphorylcholine. Based on these studies, the following Conclusions were drawn.

Random pairing of n₁ H-chains and n₂ L-chains* results in The formation of n₁ ∙ n₂ different antigen-binding sites.

In both proteins, the ligand-binding sites are located between the V_L domain of the light chain and the V_H domain of the heavy chain (Fig. 4.2, c). Thus, the binding site is formed by two different polypeptide chains. Utilizing two chains may be a solution to The problem of generating a virtually unlimited number of distinct binding sites from a limited amount of genetic material. If the number of different L- and H-chains is n₁ and n₂, respectively, there are n₁ ∙ n₂ combinations of V_L and V_H domains, i.e., n₁ ∙ n₂ different binding sites [540]. The value of n₁ ∙ n₂ has been estimated at 10⁷ [542], which means that n₁ and n₂ must be on the order of 10³ — 10⁴.

There are six hypervariable segments, each containing 5 to 10 residues per binding site. The chain folding in all variable domains of immunoglobulin New and immunoglobulin McPC 603, as well as in other immunoglobulins, is very similar [540, 543, 544]. Structural differences are concentrated primarily in the so-called hypervariable loops [619]—three segments containing 5 to 10 residues in both the V_L and V_H domains. The amino acid residues that form the binding pocket complementary to a given ligand belong to these six hypervariable segments.

The pattern of insertions and deletions characteristic of the hypervariable regions determines the dimensions of the binding site. In the case of Ig New, the binding site is a shallow (~6 Å) groove measuring 15 x 6 Å. The phosphorylcholine-binding site in Ig McPC 603 is located at a depth of 12 Å.