The combining site of an antibody is entirely contained in the papain digest fragments known as Fab (see Cohen, figure 1, this discussion). These are formed from the light chains and the amino terminal half of the heavy chains known as the Fd fragments. Evidence from many laboratories suggests that the Fd fragment contains the combining site, but it is still uncertain whether the light chain is directly involved in the site, or whether it plays a secondary role in stabilizing the structure of the Fd fragment (see Fleischman 1966). We are attempting, therefore, to determine the chemical structure of the Fd fragment. The evidence of Haber (1964) and Whitney & Tanford (1965) showed that the secondary and tertiary structure of the Fab fragment of several rabbit antibodies could be completely disrupted by reduction of all the disulphide bonds in 6 M guanidine, and that, if such fully denatured molecules were allowed to refold under appropriate conditions, there was a small but significant recovery of affinity for the antigen. These experiments suggest that the specificity of the antibody-combining site is determined by the amino acid sequence of the peptide chain and hence that each antibody specificity would be expected to be reflected in a unique sequence in certain sections of the Fd fragment. It was expected that elucidation of the sequence of this part of the heavy chain would be difficult, as it was evident from earlier work on the amino terminal amino acids of immunoglobulin G (IgG) from several species that mixed sequences were to be expected, and that these may be unrelated to antibody specificity. Hence, considerable variability was probable, only part of which was likely to be related to antibody specificity. On the other hand, some sections of constant sequence should be present, as the Fd fragment contains the interchain disulphide bond between the heavy and light chains and possibly a heavy-heavy interchain bond, as well as the allotypic antigenic sites of the rabbit IgG heavy chain (see Oudin, this discussion). All three features are common to all molecules in a preparation of IgG from molecules homozygous at the allotypic locus, and hence it would be expected that they would be located in stable sections of amino acid sequence in Fd fragment. The problem has been tackled in two ways. (1) Work has been commenced with a pathological human IgG which is believed to have a single amino acid sequence and hence can be studied by conventional techniques. Homology between the sequences of several proteins from different mammalian species, suggests that the results will be a valuable guide to what may be expected in IgG from other species such as rabbit, from which purified antibodies may be much more easily prepared. In particular we were fortunate in obtaining a pathological human protein the heavy chain of which had a blocked amino terminal amino group, as is found in the rabbit IgG heavy chain and hence the sequence of this protein is being studied. (2) There has never been any convincing demonstration, however, that pathological immunoglobulins have antibody activity, and it is possible that this property may be dependent on some special feature not present, or difficult to recognize, in the pathological protein. Work is also being carried out therefore on the sequence of the Fd fragment of rabbit IgG which has been prepared from non-immunized rabbits, from rabbits homozygous at the allotypic locus and from purified rabbit antibodies. Mixed amino acid sequences are found in all these preparations. This precludes a straightforward solution of the sequence and hence attention has been directed to the sequences adjacent to known points such as the amino and carboxyl terminals and the interchain disulphide bonds where it is possible to estimate the relative content of different sequences at these fixed points, and to determine if there is any relation to the known variables of antibody specificity and allotypy.
Differential reduction of interchain disulphide bonds of mouse immunoglobulin G