Including data from the literature, the repertoire of the light-chain variable domain in pathological anti-Pr cold agglutinins exhibits a clear bias towards the use of the single germline gene-derived subgroup, Vkappa IV (eight of 17 or 47%). The association of Vkappa IV subgroup light chain-containing anti-Pr cold agglutinins with binding to alpha 2,3-, but not alpha 2,6-linked N-acetyneuraminic acid raises speculations about a possible role of subgroup-derived determinants in anti-Pr binding.

Two IgMa cold agglutinins (CAs) reacted with protease- and sialidase-resistant antigens expressed in equal strength on human adult (I), newborn (i), i adult, rabbit (I) and rhesus monkey (i) erythrocytes. The antibodies were inhibited by the linear type 2 sequence lacto-N-neotetraose and the branched type 2 sequence lacto-N-neohexaose. Endo-ß-galactosidase treatment of red cells, which splits type 2 chains from the surface, abolished CA reactivity. The CAs expressed the idiotype recognized by the anti-idiotype 9G4 specific for anti-I and anti-i CAs. The data suggest that the two CAs recognize linear (i) as well as branched (I) type 2 chains. It is proposed to term these CAs anti-j.

Three cases of IgA kappa cold agglutinins (CAs) were studied. One had anti-Pr1 specificity, one had anti-Pra, and one had anti-Sa. The CAs recognize O-glycans of glycophorins. The findings supplement previous data on anti-Pr1 specificities of four IgA kappa CAs. Because all IgA kappa CAs described recognize O-glycans of glycophorins, a close association between the CA IgA isotype and specificities for O-glycans becomes apparent. It is unlikely, however, that the striking association reflects interrelations between IgA CA structure and specificity, because anti-Sa specificity and all anti-Pr subspecificities were originally defined with IgM CAs.

Background and objectives: Pr_{1, 2, 3}, Pr^M, Sa and Sia-l1, -b1, -lb1 are sialic acid (NeuNAc)-dependent antigens recognized by human cold agglutinins. Pr and Sa antigens are the O-glycans of glycophorins containing α2,3NeuNAc (to galactose) and/or α2,6NeuNAc (to galactosamine). The antigens of the Sia-l1, -b1, -lb1 complex are gangliosides that may carry α2,3NeuNAc (to galactose) and/or α2,8NeuNAc (to NeuNAc). We studied the NeuNAc groups involved in the antigens. Materials and methods: From 74 sera with cold agglutinins against NeuNAc-dependent antigens, anti-T-free preparations were made and tested against human red cells, treated with an α2,3-specific recombinant sialidase. Results: Most (51/62) Pr antigens use α2,3NeuNAc, some (8/62) use α2,6NeuNAc and a few (3/62) use both sialyl groups as immunodominant components on glycophorins. The immunodominant component of Sa and Sia-l1, -b1, -lb1 determinants was α2,3NeuNAc in all cases. Conclusion: The red cell target structures for cold agglutinins against NeuNAc-dependent antigens have been identified. We propose a Pr nomenclature to reflect this. The binding of anti-Pr to gangliosides may be the basis for anti-Pr-induced peripheral neuropathy.