Protein-protein interactions on the surface of immunoglobulin molecules

Design, synthesis, and application of a protein A mimetic

Low-molecular-weight synthetic molecules that mimic the activity of native biological macromolecules have therapeutic potential, utility in large-scale production of biopharmaceuticals, and the capacity to act as probes to study molecular recognition events. We have developed a nonpeptidyl mimic for Staphylococcus aureus Protein A (SpA). The specific recognition and complexation elements between the B domain (Fb) of SpA and the Fc fragment of IgG were identified from the x-ray crystallographic structure. Computer-aided molecular modeling was used to design a series of biomimetic molecules around the Phe132-Tyr133 dipeptide involved in its binding to IgG. One of the ligands binds IgG competitively with SpA in solution and when immobilized on agarose beads, with an affinity constant of 10(5)-10(6) M-1. The immobilized artificial Protein A was used to purify IgG from human plasma and murine IgG from ascites fluid, and to remove bovine IgG from fetal calf serum.

Electrostatic complementarity between proteins and ligands. 1. Charge disposition, dielectric and interface effects

Electrostatic interactions have always been considered an important factor governing ligand-receptor interactions. Previous work in this field has established the existence of electrostatic complementarity between the ligand and its receptor site. However, this property has not been treated rigorously, and the description remains largely qualitative. In this work, 34 data sets of high quality were chosen from the Brookhaven Protein Databank. The electrostatic complementary has been calculated between the surface potentials; complementarity is absent between adjacent or neighbouring atoms of the ligand and the receptor. There is little difference between complementarities on the total ligand surface and the interfacial region. Altering the homogeneous dielectric to distance-dependent dielectrics reduces the complementarity slightly, but does not affect the pattern of complementarity.

Immune response to a hapten of fluorescein isothiocyanate in a single mouse analyzed by two-dimensional affinity electrophoresis

Immune response to a hapten of fluorescein isothiocyanate (FITC) in a single BALB/c strain mouse was analyzed by two-dimensional affinity electrophoresis (2D-AEP). Anti-FITC antibodies were induced by immunization with FITC-conjugated bovine serum albumin. The antibodies were separated into a large number of spots of IgG due to differences in their isoelectric points(pI) and binding affinities to the FITC ligand. These spots consisted of IgG families which were composed of several spots having an identical affinity to the ligand but a different pI. The spots were not clearly detected in the antiserum taken on day 7 after the primary immunization, but on day 21 the spots of IgG were clearly detected, with a high diversity and specificity for the ligand. The size and number of IgG spots were markedly increased by the secondary immunization; however, the third immunization did not increase the size and number of IgG spots. The IgG spots of each family were specifically stained with an antimouse IgG subclass antibody. Furthermore, a monoclonal antibody (FL-D6) was separated by 2D-AEP into a single family which consisted of seven IgG1 spots having an identical affinity to FITC but different pIs. Therefore, each of the IgG families of anti-FITC antibodies in the antiserum can be generated by a single clone of anti-FITC antibody-producing cells. The substitution of dextran T2000 or lipopolysaccharide for bovine serum albumin as a carrier for FITC induced much smaller amounts of anti-FITC antibodies with a low diversity but high specificity to FITC.

Protein-water interaction energies as predictor for antigenic determinants

Goodford's GRIN/GRID method is used for the prediction of antigenic determinants of lysozyme by calculation of protein-water interaction energies. The comparison of the regions of high interaction energy with experimentally determined contact surfaces in antigen-antibody complexes and epitopes obtained by cross-reactivity measurements shows a noteworthy agreement. The model is proposed to enlarge the basis of theoretical models for epitope prediction. It may contribute to the increase of the prediction value when applied together with other methods.