Our previous studies of hydrogen exchange of monoclonal anti-yeast iso-i- cytochrome c (mAbs) have indicated that binding of the antigen to the mAb stabilizes the domains of the Fab fragment which are remote from the antigen-binding site. The Fab consists of the light chain (VL and CL, the variable and constant regions) and the VH and CH1 (the variable and the first constrain region) of the heavy chain. In the studies of protein folding (see another report) we have advanced a hypothesis that the protein structure is stabilized by the residue-residue interaction associated with the ordered hydrophobic core which is called the core group interaction. Based on these we speculate that the hypothetical core group interaction associated with the VL-VH interface and the VL and VH cores may influence the interaction at the antigen-antibody interface. To test this hypothesis, we wish to substitute the residues of the VL-VH interface and the VL and VH cores and determine if the substitution changes the affinity of the Fab to the antigen. To begin this study, we have sequenced cDNAs encoding VL and VH of three anti-yeast iso-i- cytochrome c mAbs 4-74-6, 2-96-12 and 4-128-6. The exception is that the sequence of 26 nucleotides at the 5' end was not determined for mAbs 4- 74-6 and 4-128-6. Then, the amino acid sequences were deduced. It is striking that the amino acid sequences of VL and VH of mAb 4-74-6 including complementarity determining regions (CDRs) are significantly more homologous to anti-lysozyme HyHEL5 (Sheriff, S. et al. (1987) PNAS 84, 8075) than mAb 2-96-12. This is true despite the fact that the epitopes of mAbs4-74-6 and 2-96-12 are closely related to each other as shown in the previous studies. Thus, the present results confirm the current idea that substitution and insertion of a limited number of residues of CDRs change the specificity of antigen recognition. Homology modeling of the VL-VH dimers has allowed us to assign the residues of the VL-VH interface and the VL and VH cores. Furthermore, it has suggested the following possibility. The overlapping epitopes of mAbs 4-74-6 and 2-96-12 may share the structural element consisting of Asp60-Glu61. This shared epitope element may interact with homologous structural elements of the antigen-binding sites of these two mAbs consisting of argininine and serine. These elements of the antigen-binding sites may be recognizable only in the three-dimensional structure but not the amino acid sequences of the CDRs.
