Faculty Bibliography

Genetic studies of human immunoglobulin variable regions have been hampered by the lack of anti-idiotypic antibodies that recognize specific heavy and light chain variable region sequences. Sixty percent of human monoclonal IgM anti-IgG autoantibodies (rheumatoid factors [RF]) from unrelated individuals share a cross-reactive idiotype (CRI) termed Wa. In previous experiments in which we used an enzyme-linked immunosorbent assay, we reported that a synthetic peptide (PSL2), corresponding to the second hypervariable region in the kappa light chain of a monoclonal IgM-RF (Sie), induced rabbit antibodies reactive with several RF paraproteins. In the present experiments, to avoid interference due to the human IgM-RF binding toward rabbit IgG, the reactivity of the anti-PSL2 antibody to the separated heavy and light chains of multiple IgM proteins and Bence-Jones proteins was assessed by the Western blot technique. The PSL2-induced anti-CRI reacted well with the separated kappa chains from 10 out of 12 IgM-RF, zero out of four light chains from IgM proteins lacking anti-IgG activity, and one out of six kappa Bence-Jones proteins. The results show that the PSL2-CRI is associated with RF and is not a kappa subgroup marker. Furthermore, a comparison of the reported light chain sequences of the PSL2-CRI-positive IgM-RF suggests that the majority of human IgM-RF light chains derive from a single germ-line VK gene or from a family of closely related VK genes that is highly conserved in the human population. Synthetic peptide-induced anti-CRI provide a potent tool for analyzing the genetic basis of CRI and abnormal autoantibody production in humans.

Light chains of the serologically and chemically defined V region sub-subgroup kappa IIIb are preferentially associated with several types of human IgM kappa (monoclonal) autoantibodies and are remarkably homologous in primary structure, as evidenced by partial amino acid sequence data. To establish the extent of homology among such proteins, we have determined the complete variable region (V) sequence of the light chains of four monoclonal IgM kappa autoantibodies, of which two (GAR and GOT) are rheumatoid factors (RFs), the third (SON) has anti-apo beta lipoprotein specificity, and the fourth (PIE) binds specifically to intermediate filaments. The region encoded by the V kappa segment gene (positions 1-95) in all four light (L) chains is virtually identical in sequence, differing by only one residue in the FR3 of protein SON and in the first CDR of protein GOT. Further, the CDR3 of kappa chain SON contains an additional residue (prolyl) located at the carboxyl-terminus of the V segment. The region encoded by the J gene (positions 96-108) is identical after position 96 for the two RFs GAR and GOT (J kappa 2), but different in proteins SON (J kappa 4) and PIE (J kappa 1). The amino acid residue at position 96, located in CDR3 at the site of combinatoriaL joining of the V kappa and J kappa gene segments and involved as a contacting residue in the hapten binding site, is different in all four light chains. These results demonstrate the extensive homology in sequence among light chains of IgM kappa autoantibodies and indicate that a particular V kappa germ line gene, kappa IIIb, is expressed as a phylogenetic response to certain self antigens or as part of a selection process by which these autoimmune responses are regulated.

There is present in rat pancreas a protein that requires an accessory factor in order to bind [3H]estradiol. To identify this accessory factor 874g of dog pancreas were acid extracted, and following selective filtration and dialysis, the low molecular weight constituents (less than 10,000) were concentrated by lyophilization. Samples of this lyophilizate were fractionated by high performance liquid chromatography (HPLC) and eluate fractions analyzed for their capacity to enhance binding of [3H]estradiol to a protein fraction from rat pancreas that had been purified relatively free of endogenous accessory factor. Such enhancement of [3H]estradiol-binding activity eluted predominantly in one peak that coincided with the elution profile of pure somatostatin (SRIF14). Analysis of eluate fractions for somatostatin-like immunoreactive material (SLIM) indicated coincidence of SLIM with the factor that enhanced binding of [3H]estradiol. It appears likely that accessory factor in pancreas is primarily somatostatin (SRIF14). Following incubation of [125I]SRIF14 and [3H]estradiol with a partially purified binding-protein fraction from rat pancreas, a complex containing labeled [125I] and [3H] was separated by Sephadex G-200 column chromatography. In the presence of 25 microM SRIF14, which activates [3H]estradiol-binding maximally in the presence of 10 nM steroid, a protein peak containing both radiolabeled ligands eluted in the void volume indicating an apparent molecular size in excess of 200,000 Daltons. At a concentration of 1 microM SRIF14, a complex eluted at a position corresponding to an apparent Mr of 120,000. Evidently, the steroid and polypeptide mutually enhance binding to this pancreatic protein, and depending on their concentrations form structures of widely varying sizes.

The carboxy terminal fragment of human plasma fibronectin has been isolated after tryptic digestion and separation by DEAE-cellulose chromatography and gel filtration on Sephadex G-50. It has a molecular weight of 6,000 which changes to 3,000 after reduction indicating that the fragment is a dimer. We have determined the amino acid sequence of the 6kDa fragment and showed that it contains 26 residues including two half-cystines which form two interchain disulfide bridges. The 6kDa fragment is not phosphorylated as in bovine fibronectin although its amino acid sequence is identical to that reported for bovine plasma fibronectin. When compared to the sequence deduced from a rat cDNA, one amino acid substitution can be found. It appears that the carboxyl end of fibronectin is highly conserved among species.

The region of the Pf3 virus genome encoding its major coat protein and its single-stranded DNA-binding protein is organized somewhat like the corresponding region of the fd (M13, f1) genome. Nevertheless, the major coat protein is unique among the major coat proteins of fd and the other filamentous phages studied in that it lacks a signal sequence and appears to be a direct translation product and in that it has fewer basic amino acid residues than its equivalent of DNA phosphates in the virion. These features are relevant to considerations of both protein insertion into membranes and DNA structure in filamentous viruses. The single-stranded DNA-binding protein also has a sequence that is different from the sequences of single-stranded DNA-binding proteins from other filamentous viruses.