The interaction of protein A and Fc fragment of rabbit immunoglobulin G as probed by complement-fixation and nuclear-magnetic-resonance studies.

Protein-A-Fc-fragment complexes were observed in sedimentation-velocity experiments by ultracentrifugation. The interaction was studied by protein-fluorescence-quenching titrations of the Fc fragment with protein A, allowing the dissociation constant to be determined under a variety of conditions. The first component of the complement pathway, C1, is activated by complexes of protein A with rabbit IgG (immunoglobulin G), and the structural basis for this interaction was studied by using n.m.r. (nuclear magnetic resonance). The four Fc-fragment binding sites on protein A were shown to contain aromatic amino acids, and to be connected by mobile hydrophilic regions. Neither n.m.r. nor proton-relaxation-enhancement studies show evidence of a large conformational change of the Fc fragment on binding protein A, and this suggests that the cross-linking of the Fc fragments may be primarily responsible for the activation of component C1. This is supported by the inability of a univalent tryptic fragment of protein A to activate complement fixation by rabbit IgG.

The gross architecture of an antibody-combining site as determined by spin-label mapping.

1. A series of Dnp (dinitrophenyl) nitroxide spin labels was used to map the dimensions of the combining site of the Dnp-binding immunoglobulin A myeloma protein MOPC 315. The method compares the observed e.s.r. (electron-spin-resonance) hyperfine splittings with those calculated on the basis of different postulated motions for the spin label. The analysis is complicated by the sensitivity of the e.s.r. hyperfine splitting to the overall ;tumbling' time of the antibody-hapten complex and the polarity of the spin-label's environment. When these effects are considered quantitatively, it is then possible to determine the degree of mobility of each hapten which is allowed by the shape of the combining site. 2. The dinitrophenyl ring is rigidly held, and the depth of the site is 1.1-1.2nm and has lateral dimensions at the entrance to the site >/=0.6nmx0.9nm. The analysis of the results for spin-labelled haptens with chiral centres allows these lateral dimensions to be refined to 0.8nm and 1.1nm, and it is shown that the site is asymmetric with respect to the plane of the dinitrophenyl ring. 3. A polarity profile of the combining site was also obtained and a positively charged amino acid residue, possibly arginine-95(L) (light chain), was located at the entrance to the site. 4. The binding of Gd(III) to the antibody-hapten complexes results in quenching of the e.s.r. signal of the nitroxide. By using La(III) as a control, the paramagnetic contribution to the quenching is measured. 5. Analysis of the differential quenchings of the enantiomers of two five-membered nitroxide ring spin labels gives two possible locations of the metal-binding site. One of these is equidistant (0.7nm) from each of the three dinitrophenyl aromatic protons, and nuclear-magnetic-resonance relaxation studies, at 270MHz, on solutions of dinitrobenzene, Gd(III) and the Fv fragment (variable region of heavy and light chain) from protein MOPC 315 support this location for the metal site. 6. The e.s.r. and metal-binding data were then compared with the results of a model of the combining site constructed on the basis of framework invariance in immunoglobulins [Padlan, Davies, Pecht, Givol & Wright (1976) Cold Spring Harbor Symp. Quant. Biol.41, in the press]. The overall agreement is very good. Assignments of possible chelating groups for the metal can be made.

Comparison of the dimensions of the combining sites of the dinitrophenyl-binding immunoglobulin A myeloma proteins MOPC 315, MOPC 460 and XRPC 25 by spin-label mapping.

The mouse immunoglobulin A myeloma proteins MOPC 315, MOPC 460 and XRPC 25 all possess dinitrophenyl (Dnp)-binding activity. Differences in specificities were shown by measuring the affinities of a variety of haptens. By using a series of Dnp-spin-labelled haptens, the dimensions of the binding sites of the three myeloma proteins were compared by the method described for protein MOPC 315 [Sutton, Gettins, Givol, Marsh, Wain-Hobson, Willan & Dwek (1977) Biochem. J.165, 177-197]. The dinitrophenyl ring is rigidly held in all three sites. The depths of the sites are all 1.1-1.2nm, but there are differences in the lateral dimensions at the entrance to the sites. For protein XRPC 25 these dimensions are 0.75nmx0.8nm, which may be compared with 0.85nmx1.1nm for protein MOPC 315 and >/=1.0nmx1.1nm for protein MOPC 460. The site in protein MOPC 460 is more symmetrical with respect to the plane of the dinitrophenyl ring than in either of the other two myeloma proteins and also allows greater penetration of solvent. In protein XRPC 25 a positively charged residue was located at the entrance to the site, similarly positioned to that reported for protein MOPC 315 [Sutton, Gettins, Givol, Marsh, Wain-Hobson, Willan & Dwek (1977) Biochem.J.165, 177-197]. All three proteins possess lanthanide-binding sites, but only in protein MOPC 315 is there antagonism between lanthanide and hapten binding. However, the effects of the diamagnetic La(III) on the electron-spin-resonance spectra of bound Dnp spin labels in both proteins MOPC 460 and XRPC 25 suggest an interaction between the two sites. Comparison of this effect with that caused by the addition of the paramagnetic Gd(III) enables the distance between the lanthanide- and hapten-binding sites to be calculated. In both proteins MOPC 460 and MOPC 315 the metal site is approx. 1.0nm from the nitroxide moiety of the spin-labelled hapten, but in protein XRPC 25 this distance is at least 2.0nm.

The use of gadolinium as a probe in the Fc region of a homogeneous anti-(type-III pneumococcal polysaccharide) antibody.

The binding of gadolinium [Gd(III)] to a homogeneous rabbit anti-(type-III pneumococcal polysaccharide) IgG (immunoglobulin G) and its Fab (N-terminal half of heavy and light chain) and Fc (C-terminal half of heavy-chain dimer) fragments was demonstrated by measurements of solvent-water proton relaxation rates in the appropriate Gd(III) solutions. At pH 5.5 the binding of Gd(III) to the Fc fragment is much tighter (KD approx. 5 micronM) than binding to the Fab fragment (KD approx. 250 micronM). The binding of Gd(III) to the whole IgG molecule (KD approx. 4 micronM) is very similar to that for the Fc fragment alone. This specificity of binding to the Fc region allows the use of Gd(III) as a probe of the Fc conformation. The environment of the Gd(III) in the Fc region of whole IgG is not affected by the presence of octasaccharide derived by hydrolysis of type-III pneumococcal polysaccharide, but the corresponding 28-unit saccharide does cause detectable changes. The addition of 16-unit saccharide to anti-(SIII polysaccharide) IgG in the presence of Gd(III) does not change the solvent water proton relaxation rate, although aggregation does occur. The effects of the 28-unit saccharide may be explained therefore by a change in the tumbling time of the IgG. From a study of the effect of various antigen/antibody ratios, it is concluded that the 28-unit-saccharide-induced changes in the Gd(III) environment in the Fc region are caused by the specific geometrical structure of the antigen-antibody complexes formed, and not simply by occupancy of the combining sites on the antibody.