Phylogenetically associated residues within the VH3 subgroup of several mammalian species. Evidence for a "pauci-gene" basis for antibody diversity.

Immunoglobulin heavy chains from IgG pools of several mammalian species have been subjected to Edman degradation on an automated protein sequencer. The percentage of unblocked vs. blocked heavy chains was estimated from the yield of the invariant valine in the second position. Further analysis of these unblocked polypeptides unequivocally placed them in the V(HIII) subgroup on the basis of homology with known human heavy chain sequences. The mammals studied could be divided into three distinct categories on the basis of the distribution of the V(HIII) subgroup. In several species the V(HIII) subgroup could not be detected while, in others, virtually all of the heavy chains belonged to this subgroup. Several species had intermediate amounts with the level of the V(HIII) subgroup restricted to between 19 and 29% of the total pool. Within experimental error, all members of a given order had a similar V(HIII) subgroup distribution. Further amino acid sequence studies illustrated a high degree of structural homogeneity in the heavy chains of IgG isolated from pooled sera of a number of mammalian species. The very close amino acid sequence homologies of the amino terminal 24 residues of the various pools corroborated conclusions previously obtained using several myeloma proteins from some of these same species. In particular, certain phylogenetically associated residues were identifiable at characteristic positions in the pools in confirmation of their identification in the myeloma proteins. The simplest assumptions would suggest that these findings are more compatible with a pauci-gene than a multi-gene basis for the generation of antibody diversity.

The genetic control of antibody formation.

Studies of the molecular biology of lymphoid cells have markedly increased our understanding of how millions of different antibodies can be synthesized by a single animal. To date, the most detailed understanding has been achieved for the mouse, primarily because of the relatively greater experimental availability of this species. These studies, as well as those involving other species, have shown that the complete genes for antibody polypeptide chains are assembled from disparate genetic elements which are originally widely separated in the genome. The assembly process itself, together with the coding information present in the germ line genetic elements, contributes to the diversity of structure (and thus combining specificities) shown by mature antibody molecules. Specifically, the diversity of structure characteristic of antibody variable regions is due to three distinct mechanisms: innate variability of germ line genes; mismatching of individual gene segments during their somatic rearrangement leading to junctional diversity; and somatic mutation in variable region genetic material during or after the rearrangement. These processes lead to the wide array of combining specificities that permit the humoral immune system of a mature animal to interact with essentially any non-self antigen which it encounters. Complex genetic rearrangements are also responsible for the class switching phenomenon long known to be characteristic of the humoral immune response. A form of homologous recombination between constant region genes, possibly mediated by specific "switching" enzymes, is now believed to be involved in this phenomenon. It is also currently believed that the restriction of gene rearrangement processes to one of the two possible chromosomes of a diploid pair in each cell is responsible for the phenomenon of allelic exclusion that has long been associated with the normal functioning of mammalian B-cells.

Characterization of a homogeneous paraprotein from a horse with spontaneous multiple myeloma syndrome.

A novel myeloma paraprotein has been isolated from a horse with a lymphoid tumor. The protein was a euglobulin and consequently was readily isolated from serum in pure form and high yield by simple dilution in distilled water. The purified intact protein had a molecular weight of 150,000 and was composed of heavy and light chains, both of which had blocked amino-termini and were thus not susceptible to amino-terminal sequence analysis. The amino acid compositions of these respective chains corresponded to those of comparable chains from immunoglobulins of other species. Peptide maps of paraprotein light chains prepared by high pressure liquid chromatography corresponded in part to those of normal pooled equine light chains. The identification of this paraprotein as an equine AI (aggregating immunoglobulin) protein was confirmed by serological analysis using a specific antiserum. The relationship of this particular protein to other members of the immunoglobulin family was further demonstrated by the production of an anti-idiotypic antiserum individually specific for this molecule.

Immunoglobulin structure and function: genetic control of antibody diversity.

Recent studies of the molecular biological characteristics of lymphoid cells have markedly increased our understanding of how millions of different antibodies can be synthesized by an individual mammal. In particular, studies have shown how antibody genes are arranged and rearranged within B-lymphocyte clones to provide each cell clone with antibody of defined specificity for antigen. The process involves the assembly, from disparate genetic elements, of a complete antibody gene that will code for an antibody protein. The assembly process, in itself, also provides mechanisms for generating the diversity of antibody variable region structure (that part of an antibody molecule that actually binds antigen) that is essential to a full role for humoral immunity in host defense mechanisms. Specifically, the diversity of structure characteristic of mature antibodies derives from 3 distinct mechanisms: innate variability of germ-line genes; mismatching of individual gene segments during their somatic rearrangement leading to junctional diversity; and somatic mutation in variable region genetic material during or after the rearrangement. Thus, it is now clearly understood that several processes are involved in explaining the origin of the antigen-combining diversity of antibody proteins. Certain "lottery-like" aspects of these genetic processes add to the combinatorial possibilities that are characteristic of the humoral immune system.

Localization of two additional hypervariable regions in immunoglobulin heavy chains.

Cyanogen bromide fragments were isolated from the heavy chains of three human IgG myeloma proteins of the V(H)III subgroup, sequenced by an automated method, and localized to the variable region. Inspection of these sequences, together with corresponding stretches from both human and animal proteins (studied in other laboratories) led to the detection of two additional hypervariable regions characteristic of the V(H) segment of immunoglobulin heavy chains. These areas of hypervariability, involving heavy-chain residues 86-91 and 101-109, were separated by a region of relative constancy. The close relationship of these two hypervariable regions, and the previously described first heavy-chain hypervariable region (residues 31-37), to the first heavy-chain disulphide bridge implies that the three hypervariable areas might be in close steric approximation in native immunoglobulin molecules. Examination of the sequences of the terminal portion of V(H) of all these proteins (the segment from residue 95 to the beginning of homology region C(H)l) revealed that no subgroup-specific residues could be identified in this area. Thus, heavy-chain subgroup distinctions may not extend through the entire variable region.

Distribution and association of heavy and light chain variable region subgroups among human IgA immunoglobulins.

A series of randomly selected human IgA myeloma proteins were examined for the presence of the VHIII subgroup as defined by the possession of an unblocked amino terminal amino acid and characteristic linked residues along the heavy chain. Blocked heavy chains were classified as VHB proteins. The data showed that 20 of 30 such random alpha chains (67%) were classifiable as members of the VHIII subgroup. Similarly, 75% of heavy chains isolated from pools of normal serum IgA contained a VHIII variable region. The pattern stands in marked contrast to the situation in human IgG proteins where approximately 20% of heavy chains from both pools and myeloma proteins are VHIII. There was thus a clear divergence of the pattern of variable region: constant region association between these two classes of immunoglobulin. Some more limited data were obtainable for the light chains of the IgA myeloma proteins. Certain predilections for light chain subgroup:heavy chain subgroup (VHIII or VHB) associations were discernable, but more data are required for definite conclusions. Overall, this study suggests that although the pool of available variable region sequences is indeed shared among human IgA and IgG proteins, the partitioning is not exactly equivalent between the two immunoglobulin classes. The pattern is particularly apparent at present for the VHIII subgroup which comprises approximately 70% of human alpha chains and only about 20% of human gamma chains.

Structure of antibodies with shared idiotypy: the complete sequence of the heavy chain variable regions of two immunoglobulin M anti-gamma globulins.

The complete amino acid sequence of the heavy chain variable regions of two different molecules of immunoglobulin M anti-gamma globulin has been determined. These proteins, from different human patients, had independently been shown to share idiotypic specificity. Only eight sequence differences were discernible for the entire length of their heavy chain variable regions. Five of the differences occurred outside hypervariable regions, while three were placeable within such regions. A comparison of these molecules of anti-gamma globulin with the seven human V(H)III variable region sequences presently available for immunoglobulins without known antibody activity showed that the great majority of sequence differences between the two idiotypically similar antibodies and these seven proteins were confined to hypervariable regions. This study illustrates in precise terms a convergence of the distinct immunological properties of idiotypy, hypervariable region structure, and combining site specificity as they relate to the variable region of the immunoglobulin molecule. To a great degree these properties now appear to be a reflection of the same structural attributes of the variable region.

Sequence relationships among the variable regions of immunoglobulin heavy chains from various mammalian species.

Immunoglobulin heavy chains of myeloma proteins from dogs and cata have been subjected to automated sequence analysis. When the results were compared with human heavy-chain sequences, all the dog and cat proteins could be unequivocally assigned to the V(H)III subgroup. This pattern contrasts with that in human proteins in which only 25% of all heavy chains sequenced belong to this subgroup. The 30 residues at the NH(2) termini of dog, cat, and human heavy chains had sequence identities near or exceeding 90%, in contrast to established interspecies sequence homologies of constant regions of about 60%. Some genes of the immunoglobulin heavy-chain variable region thus appear to have been conserved through a considerable period of evolutionary time. The analyses also showed that the presence of certain amino acids at certain positions in these heavy chains could be correlated with the species of origin. The occurrence of such "phylogenetically associated" residues is most consistent with the presence of a restricted number of genes in the heavy-chain variable region pool.

Demonstration of antibodies to both hemagglutinin and neuraminidase antigens of H3N2 influenza A virus in domestic dogs.

Serologic evidence of infection with human (H3N2) influenza viruses of 6 of 79 dogs sampled in New York City in March-April 1973 was obtained through the use of four different methods for measurement of anti-hemagglutination and anti-neuraminidase antibody.

Variable region sequences of five human immunoglobulin heavy chains of the VH3 subgroup: definitive identification of four heavy chain hypervariable regions.

The variable regions of five human immunoglobulin heavy chains of the V(H)III subgroup have been totally sequenced. Three of the heavy chains belonged to the IgG class and two to the IgA class. Examination of these sequences, and comparison with additional published heavy chain sequences, showed that a total of four hypervariable regions is characteristic of human heavy chain variable regions. The relatively conserved character of large segments of the heavy chain variable region was very evident in these studies. The conserved segments, which are those sections located outside the hypervariable regions, comprise approximately 65% of the total heavy chain variable region. The following general structural pattern for antibody molecules emerges from this and related studies: an overall combining region superstructure is provided by the more conserved segments while the refinements of the active site specificity are a function of hypervariable regions.

Mouse alpha-macroglobulin. Structure, function and a molecular model.

Mouse alpha-macroglobulin (M-AMG) is believed to be a functional homologue of human alpha 2-macroglobulin (h-alpha 2M). The subunit composition, the tryptic cleavage pattern before and after methylamine incorporation and the two-dimensional tryptic-peptide mapping, however, indicate that these two proteins are structurally distinct. M-AMG is composed of two major types of polypeptides (Mr 163,000 and 35,000) together with a minor polypeptide (Mr 185,000), whereas h-alpha 2M has only one type of polypeptide (Mr 185,000). After incorporation of methylamine, there is no change in the normal tryptic-cleavage pattern of M-AMG; however, tryptic cleavage of h-alpha 2M is severely retarded [Hudson & Koo (1982) Biochim. Biophys. Acta 704, 290-303]. The N-terminal sequence of the 163,000-Mr polypeptide of M-AMG shows sequence homology with the N-terminal sequence of h-alpha 2M. The amino acid compositions of M-AMG and its two major polypeptide chains are compared. Thermal fragmentation studies show that the 163,000-Mr polypeptide is broken down into 125,000-Mr and 29,000-Mr fragments. Trypsin-binding studies show that M-AMG can bind two molecules of trypsin/molecule. Inactivations of the trypsin-binding property of M-AMG and h-alpha 2M with methylamine show similar kinetics of inhibition at 4 degrees C. A structural model of M-AMG is proposed, based on accumulated data.