This paper deals solely with idiotypic determinants, the configurations of which are modified when the antibody bearing them interacts with its ligand. This phenomenon is measured as an inhibition of the reaction between anti-idiotype and idiotype.
Two points are made: (a) The assay for ligand-modifiable determinants can be used to determine the "size" of the combining site. This is illustrated here with the anti-α(1 → 6) dextran mouse myeloma immunoglobulin W3129. Whether the interaction between a homologous series of α(1 → 6) oligosaccharide ligands and the combining site of W3129 is measured by inhibition of precipitation with α(1 → 6) dextran (4) or of binding of W3129 to anti-W3129 idiotype, the finding is the same. The order of inhibition is isomaltohexaose = isomaltopentaose >> isomaltotetraose > isomaltotriose >>> isomaltose. The combining site is optimally complementary to isomaltopentaose.
(b) Cross-idiotypic specificity is closely correlated with cross-combining specificity; the converse is not true. This is illustrated here with three groups of mouse myeloma immunoglobulin, each specific for α(1 → 3) dextran, α(1 →6) dextran, ß(2 → 1) or ß(2 → 6) levan. If a given anti-idiotypic serum cross-reacted with several myeloma proteins, they always had similar combining specificity. Thus the three proteins, J558, MOPC 104E, and UPC 102, which cross-react with anti-J558 have combining specificity for α(1 → 3) dextran; cross-reacting W3082, UPC 61, and Y5476 have specificity for levan; and cross-reacting W3129 and W3434 have specificity for α(1 → 6) dextran. This extends previous studies with proteins specific for phosphorylcholine (7) or γ-globulin (8). As expected, the converse is not true, for proteins may have combining specificity for α(1 → 6) dextran e.g. QUPC 52, or levan e.g. J606, UPC 10 and yet not carry the above-mentioned reference idiotypes.
The correlation between cross-idiotypic and combining specificity breaks down when idiotypic determinants which are not modifiable by ligand are studied. The implications of this are pointed out since most investigations deal with ligand-nonmodifiable determinants.