No abstract

To investigate the molecular basis of the autoimmune response to the related i and I carbohydrate antigens, we studied cold agglutinins (CA) from B-cell clones and from the peripheral circulation of patients with lymphoproliferative syndromes. Sequence analyses of expressed variable region genes indicate that both anti-i and anti-I specificities from B- cell clones from two patients are encoded by the VH4.21 or a very closely related VH4 heavy chain gene, whereas the expressed light chain genes differed. The anti-i-secreting B-cells express unmutated germline- encoded VH4.21 and VKI gene sequences. The VH region gene encoding anti- I has the closest homology (97%) to the VH4.21 germline gene and differs at the protein level by only three amino acids. In contrast, while the VL region gene encoding anti-I is most homologous (96%) to the VKIII, kv328 germline gene, there are seven amino acid differences due to nonrandom replacement mutations, which suggests a role for antigen-mediated selection in the anti-I response of this individual. These studies were extended by a structural survey of 20 additional serum CA using antipeptide antibodies specific for determinants in VH and VL regions. All anti-I and anti-i CA were shown to express VH4 heavy chains, and 14 of 17 CA expressed a previously described VH4 second hypervariable region determinant, termed VH4-HV2a. We also found that 13 of 14 anti-I CA used VKIII light chains, while the anti-i CA used light chains from at least three VL families. Taken together, the data show that anti-i and anti-I CA probably both derive from the VH4.21 gene (or a closely related gene). Furthermore, the restricted VH and different VL gene use in anti-i and anti-I CA may reflect the close structural relationship of the i and I antigens.

No abstract

The erythrocyte glycoprotein-bound N-acetylneuraminic acid which determines the blood groups MN and the autoantigens Pr,/Pr, was modified and investigated with regard to antigen activities.We confirmed the results of Lisowska and Morawiecki, who inactivated MN antigenicity by acetylating the 6-amino group of lysine of erythrocyte glycoproteins, suggesting that electrostatic interactions between carboxyl groups of N-acetylneuraminic acid and side-chain lysyl amino groups ensure stable MN antigens. Further support for this concept was provided by our observation that amidation of the N-acetylneuraminic acid carboxyl groups also abolished MN antigenicity.A further differentiation between the autoantigens Pr, and Pr, was elucidated by oxidation of polyhydroxy side chains of N-acetylneuraminic acid.The immunodominant component of the myxovirus receptors, the MN blood groups [l] and the Pr,/Pr, antigens corresponding with cold agglutinins, described by Roelcke [2,3], is N-acetylneuraminic acid which is linked via a 2-0x0 to the nonreducing end of the oligosaccharide side chains of red-cellmembrane glycoproteins.According to the terminology by Reeves [4], the preferred conformation of N-acetylneuraminic acid is l C in which the large substituents are located in equatorial position (Fig. 1).The influence of the carboxyl group ((3-1) and of the polyhydroxy side chain (at (3-6) on the antigen activity of the different receptors was investigated by chemical modifications of these groups. Untreated N-acetylneuraminic acid and oxidized analogue of N-acetylneuraminic acid were determined by gas-liquid chromatography as trimethylsilyl derivatives. The samples were silylated with bis(trimethylsily1)trifluoroacetamide in acetonitril. Arabitol was used as internal standard. The chromatographic separation was performed with the liquid phase methylphenylsilicone (OV 17) 20/, (w/w) on Chromosorb G.The 6-amino groups of lysine of the glycoproteins were acetylated with acetanhydride in 0.1 M phos-

I n this paper we shall report that high titre cold antibodies can be distinguished not only with regard t o their protein substrate, but also concerning their site of action a t the erythrocyte membrane (receptor). Common-type cold agglutinins of high titre are paraproleins of the yM-ClaSS [5] showing anti-1 [6] resp. anti-i specificity [2].Recently, however, a n IgA-protein, which has not yet been shown definitely t o be a paraprotein [l], and a yA-paraprotein [3] have been described as substrates of cold agglutinins. Both IgA-proteins with their cold antibody activity did not show obvious anti-I or anti-i specificity. ROELCKE and DOROW [3] have demonstrated that the cold agglutinin, described by them, did not react with erythro-