The location, number, and kinds of oligosaccharides in human IgAl and IgA2 immunoglobulins have been determined by amino acid sequence analysis of the a heavy chains. Both A2m allotypes of the a2 chain of IgA2 have two GlcN oligosaccharides that are absent in the al chain, but they lack GalN. The A2m(2) allotype has a fifth GlcN oligosaccharide. The a chains of IgA proteins also have subclass-specific and allotype-specific differences in amino acid sequence. Although other classes of human immunoglobulins differ in the number and kind of oligosaccharides, the sites are often homologous and are related to the immunoglobulin domain structure. Evolutionary preservation of the tripeptide acceptor sequence for GlcN probably indicates both a structural and biological role for carbohydrate.Although the five classes of human immunoglobulins differ greatly in their content of carbohydrate and in its distribution along the polypeptide chain, the carbohydrate often is in homologous positions and usually seems to lie in between the compact domains of the heavy chain or on the surface of a domain. With rare exceptions governed by the presence of a signal acceptor tripeptide sequence, the carbohydrate is present only on the heavy chain and is confined to the constant (C) region of the heavy chain. We earlier reported the location and nature of the five oligosaccharides of the At chain of human IgM (1, 2). In contrast, the oy chain of human IgG has a single oligosaccharide (3), and the E chain of human IgE is reported to have six (4). In all these cases the oligosaccharide contains glucosamine (GlcN), which is attached to asparagine by an N-glycosidic linkage. The IgAl subclass is unusual among glycoproteins in having two types of linkage to the polypeptide chain, the N-glycosidic linkage of GlcN to asparagine and an O-glycosidic linkage of galactosamine (GalN) to serine (5, 6). Recently we reported (7) the complete amino acid sequence of a human IgAl immunoglobulin (designated Bur), including the location of five GalN carbohydrates in the hinge region, two GlcN oligosaccharides in the Fc region, and an adventitious GlcN oligosaccharide in the variable (V) region of this al heavy chain. We have now determined the amino acid sequence of the C region of two a2 heavy chains representing different allotypes of the IgA2 subclass, namely, the A2m (2) MATERIALS AND METHODSHuman IgA was prepared from the serum of patients with multiple myeloma who produced large amounts of monoclonal IgA. The procedure involved ammonium sulfate precipitation, ion exchange chromatography, and gel filtration techniques (9, 10). The purity of the IgA was tested by immunodiffusion and immunoelectrophoresis with monospecific antisera and by gel electrophoresis and ultracentrifugation (9, 10). The purified IgA proteins were reduced at room temperature under N2 with 0.01 M dithiothreitol for 1 hr followed by alkylation with iodoacetic acid (10% molar excess) or ethylene imine. The heavy and light chains were separated on a Sephadex G-100 column. The...

The complete amino-acid sequence of the constant (C) region of the a2 heavy chain of a human IgA2 protein of the A2m(1) allotype has been determined. Excluding the hinge region and the carbohydrate content, this a2 allotype differs from the al chain in only 14 amino-acid positions; all of these are identical to the A2m(2) allotype of the a2 chain and confer subclass (or isotypic) character on the a2 chains. However, the A2m(2) allotype differs in six positions where A2m(1) and al are identical; the first two are just before the hinge and the other four are in the last (CH3) domain. The A2m allotypic character of a2 chains is attributed to several conformational factors in the sequence at positions 211-221, just before the hinge. The isoallotypic determinant shared by al chains and the A2m(1) allotype of a2 resides in the identity of their CH3 domains. Thus, the A2m(1) allotype appears to be a hybrid chain that is identical with al in the CH3 domain and identical with the A2m(2) a2 chain in the CHI and CH2 domains and in the hinge, except for the allotypic determinants arising from four structural differences from residues 211-221. The genetic origin of isotypes, allotypes, and isoallotypes of the a chain has involved several events of homologous crossing over and neutral point mutations accumulated late in the evolutionary development of IgA immunoglobulins. Since the crossing over appears to occur between CH2 and CH3, heavy chain domains may be coded for by independent units in embryonic DNA that are analogous to the variable (V) and C segments of light-chain genes.Correlation of the structure, function, evolution, and immunogenetics of antibodies is a primary goal of our studies of the amino-acid sequence of human immunoglobulins of different classes, subclasses, and allotypes. We have recently reported the complete amino-acid sequence of a human immunoglobulin A of the IgAl subclass (1) and also of the a2 heavy chain of the A2m(2) allotype of the IgA2 subclass (2). We report here the complete sequence of the constant (C) region of the a2 heavy chain of an IgA2 immunoglobulin of the A2m(1) allotype. This work has permitted the location and identification of the amino-acid substitutions and the differences in kind, number, and location of oligosaccharides that determine the structural characteristics of all the known subclasses and allotypes of human IgA. The recently discovered (3) isoallotype (or nonmarker) nA2m(2) is shown to be due to the identity in sequence of the last domain (CH3) of the a l chain and the A2m(1) allotype of the a2 chain. Like the Lepore phenomenon in human hemoglobin, the existence of hybrid a heavy chains of IgA poses an interesting question about the evolution and genetics of IgA and suggests an event of homologous crossing over early in human evolution. This may have involved reassortment of independent genetic units that code for individual domains but that are separated by untranslated sequences in embryonic DNA.

SummaryThree new 8N-alkyl analogues of roseoflavin (MM) , i.e., 8-ethylamino-(EH), 8-methylethylamino (ME) , 8-diethylamino-8-de methyl-D-riboflavin (EE), their tetraacetates , and 8-amino-8-demethyl-D riboflavin (HH) tetraacetate, were synthesized . A relation between the anti-riboflavin activity and the chemical structure of 8N-alkyl analogues (8N-methyl, ethyl) was studied by a restoration by riboflavin (RF) of inhibitory effect of the analogues on a growth of Gram positive bacteria, i.e., Sarcina lutea, Bacillus cereus, and Staphylococcus aureus. The inhibitory effect of most of the analogues was restored by RF . But in some cases, i.e., 8-methylamino-8-demethyl-D-riboflavin (MH) in Sar. lutea and MM in Staph. aureus, the effect was not completely restored . Apparently, the inhibition in early phase of growth was restored, but the maximum growth was still suppressed. The non-alkylated amino analogue (HH) showed only unrestorable suppression of maximum growth in Sar. lutea. Of restorable effect by RF of N-alkyl analogues, approximate decreasing orders of anti-RF activity were as follows. Dialkylated analogue> monoalkylated. HH showed insignificant anti-RF activity . In each group, methylated analogue>ethylated.In B. cereus monoalkylated analogues , and in Staph. aureus monoalkylated and EE showed no significant inhibitory effect. Redox potentials of the N-alkyl analogues were measured, and a definite relation between the chemical structure and the potential was found (RF =EE>ME>MM>HH>EH>MH). But the anti-RF activity of the analogues was not completely explained by the difference of the redox potential from RF.

No abstract