Principles of Protein Structure - G. Schultz 1982

Protein–ligand interactions
Ligand-binding sites of immunoglobulins
Binding model in the immunoglobulin–ligand system

IMMUNOGLOBULINS with a known Structure are used to study immunoglobulins of different Specificity. The V_L /V_H immunoglobulin domains can be viewed as consisting of a rigid framework to which hypervariable loops are attached. This provides a basis for the comparative analysis of immunoglobulin–Ligand systems by building a model using the framework coordinates (derived from regions of known V_L/V_H structure) and the Amino acid sequences of the hypervariable Regions of the immunoglobulin under investigation. The first protein studied using this approach [291, 620] was the mouse myeloma IgA MOPC 315, which is known to bind 2,4-dinitrophenyl derivatives. These results were later refined by Dwek et al. [621], who combined Model Building with nuclear magnetic Resonance, electron paramagnetic resonance, and chemical modification studies.

* The letters H and L denote the heavy and light chains, respectively.

In principle, a systematic analysis of any immunoglobulin–ligand system is feasible. Binding models can be investigated systematically because in mammals, such as rabbits or goats, antibody synthesis can be induced against virtually any carefully selected molecule (whether of low or high molecular weight). The resulting Antibodies are then purified using Affinity Chromatography. Typically, this Procedure results in a degenerate Immune Response, i.e., the synthesis of several types of antibodies by multiple clones of so-called plasma Cells [542]. These antibodies differ in their ligand affinity constants, which is also reflected in the chemical, spectral, and immunological properties of their binding sites. Subsequent Amino Acid Sequence analyses revealed that these differences are caused by Amino Acid Substitutions in the hypervariable regions of the V_L and V_H domains [622]. At first glance, a degenerate response might seem to complicate the Spatial Organization of the binding sites. In practice, however, this situation actually allows data from Multiple binding sites to be combined and mutually cross-checked, which significantly increases the reliability of identifying correct models.