Abstract
Heavy and light chain cDNAs that encode for a bovine anti-testosterone monoclonal antibody have been cloned and sequenced. Testosterone binding by the antibody encoded by the cDNAs has been verified by expression in COS-1 cells and subsequently identifying bovine anti-testosterone IgG in transfected cell culture media. The cDNAs are the first to be described for functionally rearranged bovine Ig genes. The heavy and light chains were identified as C-gamma-1 and lambda respectively. The coding region of the V-domains include J-regions which suggests that the germline organisation and the mechanisms of V-exon assembly in cattle are likely to be similar to those of other mammalian species. The derived protein sequence of the heavy chain has been compared with those of bovine IgG2 and human and rabbit IgGs, in order to identify potential binding site(s) for Fc-gamma-1 receptors and complement on bovine IgG1. A three-dimensional molecular model of the antigen binding site of the bovine anti-testosterone antibody has been constructed. The major structural feature of the model is a cavity formed by residues from five of the CDRs. Based on the model, testosterone is predicted to bind into the cavity and establish a hydrogen bond through the D-ring hydroxyl with CDR-H3 Tyr-97. Site-directed mutagenesis at CDR-H3 Tyr -97 confirmed the importance of this residue for the high affinity of the antibody. Substitution of phenylalanine at this site resulted in an antibody with an affinity reduced by nearly three orders of magnitude, whereas substitution of glutamate resulted in an antibody that still binds testosterone, but with a greatly reduced affinity. The reduction in binding energy associated with the replacement of phenylalanine for CDR-H3 Tyr-97 was calculated to be 3.7 kcal mol-1., which is consistent with the loss of a single hydrogen bond. The model has been used to propose specific amino acid substitutions design to enhance the antibodies affinity for testosterone.