Karsten Skjødt

The chicken erythrocyte-specific MHC antigen. Characterization and purification of the B-G antigen by monoclonal antibodies

Mouse monoclonal antibodies with B-G antigen (major histocompatibility complex class IV) specificity were obtained after immunization with erythrocytes or partially purified B-G antigen. The specificities of the hybridoma antibodies were determined by precipitation of B-G antigens from 125I-labeled chicken erythrocyte membranes (CEM) followed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and autoradiography. The B-G antigen had an approximate molecular mass of 46–48 kd in reduced samples, depending on the haplotype, and in unreduced samples contained either dimers (85 kd), when labeled erythrocytes were the antigen source, or trimers (130 kd), when B-G was purified and precipitated from CEM. The B-G antigen was unglycosylated as studied by (1) in vitro synthesis in the presence or absence of tunicamycin, (2) binding experiments with lectin from Phaseolus limensis, and (3) treatment of purified B-G antigen with Endoglycosidase-F or trifluoromethanesulfonic acid. Two-way sequential immunoprecipitation studies of erythrocyte membrane extracts with anti-B-G alloantisera and monoclonal antibodies revealed only one population of B-G molecules. Pulse-chase experiments have shown B-G to be synthesized as a monomer, with dimerization taking place after 20–30 min. No change in the monomers molecular mass due to posttranslational modifications was revealed. The antigen was purified from detergent extract of CEM by affinity chromatography with a monoclonal antibody, and then reduced and alkylated and affinity-purified once more. Finally, reverse-phase chromatography resulted in a pure product. The B-G antigen was identified in the various fractions by rocket immunoelectrophoresis. The final product was more than 99% pure, as estimated by SDSPAGE analysis followed by silver stain of proteins. The yield from the affinity chromatography step was 3–4 μg B-G/ml blood, calculated from Coomassie-stained SDS-PAGE of B-G using ovalbumin standards. The monoclonal antibodies were also used to identify the B-G (class IV) precipitation arc in crossed immunoelectrophoresis. No common precipitate with the B-F (class I) antigen was observed.

B-G cDNA clones have multiple small repeats and hybridize to both chicken MHC regions.

We used rabbit antisera to the chicken MHC erythrocyte molecule B-G and to the class I α chain (B-F) to screen λgt11 cDNA expression libraries made with RNA selected by oligo-dT from bone marrow cells of anemicB19 homozygous chickens. Eight clones were found to encode B-G molecules which hybridize with sequences in the chicken MHC as defined by congenic strains; the fusion proteins react with multiple immune but not preimmune sera, they select antibodies from the antisera to B-G, which then react with distinct erythrocyte B-G protein patterns, and they elicit antibodies from mice which in turn react with authentic B-G proteins. None of the clones represent a complete message, some — if not all — bear introns, and none of them match with any sequences presently stored in the data banks. The following new information did, however, emerge. At least two homologous transcripts are present in this homozygous chicken, thereby formally proving the existence of an expressed multigene family. The 3′ ends (3′UT) are simple sequences with 80% nucleotide identity between clones, while the 5′ ends (either coding or noncoding) are composed of multiple short repeats which are far less similar. These repeats could explain the bewildering variation in size of B-G proteins within and between haplotypes. Southern blots of genomic chicken DNA gave complex patterns for most probes, with many bands in common using different probes, but few bands in common between haplotypes. The sequences detected are all present in the MHC, based on the congenic lines CB and CC. Most of these sequences map into theB-G region, but some map into theB-F/B-L region as defined by the haplotypesB15, B21, and their apparently reciprocal recombinantsB21r3 andB15r1.

A simple immunoblotting method after separation of proteins in agarose gel

A simple and sensitive method for immunoblotting of proteins after separation in agarose gels is described. It involves transfer of proteins onto nitrocellulose paper simply by diffusion through pressure, a transfer which only takes about 10 min. By this method we have demonstrated the existence of multiple molecular forms of the complement factors C3 and factor B in serum from 2 species, man and chicken, after electrophoretic separation in agarose. We have also demonstrated the usefulness of the method for determining the isoelectric point of proteins after isoelectric focusing in agarose.

Isolation and characterization of chicken and turkey beta2-microglobulin

Abstract Chicken and turkey beta2-m were isolated from citrated plasma in sequential use of three Chromatographie steps: affinity chromatography, gel filtration chromatography and anion-exchange chromatography. The purified protein was identified as beta2-m by reaction with a beta2-m specific monoclonal antibody and by the ability to recombine with the chicken MHC class I heavy chain. The purity was estimated by SDS-PAGE and IEF. The pI was between 5.1 and 5.3 for chicken beta2-m and 4.7 and 4.8 for turkey beta2-m, which fact is reflected in their different electrophoretic mobilities in agarose gel (turkey migrates in the alpha and chicken migrates in the beta region). The mol. wt of both chicken and turkey beta2-m was 14,500 estimated by SDS-PAGE whereas calculations based on the amino acid compositions gave mol. wts of 11,000. E M 280 was 15.9 for chicken beta2-m and 16.4 for turkey beta2-m. The amino acid compositions and sequences of the two avian beta2-m molecules have been compared with earlier data from the literature. The sequence of the 23 N -terminal amino acids was found to be identical in our preparations from both chicken and turkey, namely DLTPKVQVYSRFPASAGTKNVLN, and is incompatible with a previously published sequence also thought to be from turkey beta2-m. Reasons for our opinion that the molecules isolated and sequenced in this paper are the correct ones are given.

Amino acid sequences and structures of chicken and turkey beta2-microglobulin

Abstract The complete amino acid sequences of chicken and turkey beta2-microglobulins have been determined by analyses of tryptic, V8-proteolytic and cyanogen bromide fragments, and by N -terminal sequencing. Mass spectrometric analysis of chicken beta2-microglobulin supports the sequence-derived M r of 11,048. The higher apparent M r obtained for the avian beta2-microglobulins as compared to human beta2-microglobulin by SDS-PAGE is not understood. Chicken and turkey beta2-microglobulin consist of 98 residues and deviate at seven positions: 60, 66, 74–76, 78 and 82. The chicken and turkey sequences are identical to human beta2-microglobulin at 46 and 47 positions, respectively, and to bovine beta2-microglobulin at 47 positions, i.e. there is about 47% identity between avian and mammalian beta2-microglobulins. The known X-ray crystallographic structures of bovine beta2-microglobulin and human HLA-A2 complex suggest that the seven chicken to turkey differences are exposed to solvent in the avian MHC class I complex. The key residues of beta2-microglobulin involved in alpha chain contacts within the MHC class I molecule are highly conserved between chicken and man. This explains that heterologous human beta2-microglobulin can substitute the chicken beta2-microglobulin in exchange studies with B-F (chicken MHC class I molecule), and suggests that the MHC class I structure is conserved over long evolutionary distances.

Variations in the cytoplasmic region account for the heterogeneity of the chicken MHC class I (B-F) molecules.

Molecular variation among major histocompatibility complex (MHC) class I (B-F) proteins from B-homozygous chickens is apparently caused by C-terminal variation. Analysis of the total B-F protein pool revealed substantial heterogeneity with two or three molecular mass constituents, each being comprised by several isoelectric focusing variants. This heterogeneity could not be reduced by enzymatic deglycosylation. By contrast, proteolytic removal of a small (Mr 1000–4000) fragment from the α chain resulted in the generation of a Mr 36 000 fragment, common to all the molecular mass variants. Unlike the parent proteins, the Mr 36 000 fragment derived from isolated variants yielded identical, simple patterns in two-dimensional gel electrophoresis and identical finger prints in peptide mapping. This, together with N-terminal amino acid sequencing, as well as comparison of hydrophobicity properties of fragments obtained by gradual proteolytic digestion, indicated that the small peptide responsible for the major B-F heterogeneity was situated in the intracellular, C-terminal part.