Hans Nowack

Preparation and use in immunohistology of antibodies specific for type I and type III collagen and procollagen

Antibodies to bovine type I and type III collagen and their precursor form procollagen were produced in rabbits and rendered specific for the immunizing antigen by immunoadsorption. These purified antibodies showed distinct immunofluorescence staining on frozen sections of both bovine and human connective tissue at concentrations as low as 1-10 mug/ml. Antibodies to type III collagen and procollagen reacted with reticulin in liver and spleen, with fascicles around tendons and with the upper portion of the dermis. Antibodies to type I collagen and procollagen reacted with skin and fiber bundles in tendon but did not stain reticulin. No reaction was observed with cartilage collagen or with kidney glomerular basement membrane.

Specificity of human antibodies against galα1-3gal carbohydrate epitope and distinction from natural antibodies reacting with galα1-2gal or galα1-4 gal

Antibodies against galactosyl-α1-3-galactose epitopes were characterized in normal and patient sera by radioimmunoassay binding to mouse laminin and oligosaccharide inhibition. Binding was strictly dependent on α-linked galactose in a terminal position. Reduced affinities were observed for digalactoses with α(1-2)-, α(1-6)- and α(1-4)-linkages and for the blood group B epitope, Galα1-3(Fucα1-2)Gal. Conformational models of various active and inactive oligosaccharides provided a clearer picture of the epitope requirements for the observed antibody specificity. Some antibody heterogeneity was detected by comparing individual sera and by hapten elution from a laminin adsorbent. New assays were developed with synthetic Galα1-3Gal-albumin conjugates and these were shown to be more sensitive than assays with mouse laminin. Two more ubiquitous human antibodies could be detected with Galα1-2Gal and Galα1-4Gal conjugates. They were distinct from Galα1-3Gal-specific antibodies as shown by carbohydrate inhibition. This demonstrates a considerable diversity in the recognition of α-linked galactose epitopes by natural antibodies.

Radioimmunoassay for the aminoterminal peptide of procollagen pα1(I)-chain

Abstract Peptides derived from the aminoterminal portion of the pα1(I)-chain of calf and sheep procollagen were labeled with iodine-125. Despite changes in electrophoretic homogeneity after labeling, reaction of the labeled peptide with antisera to unlabeled peptide was retained. Antisera to procollagen or the isolated procollagen peptide showed high titers for the native peptide, a much weaker binding with the reduced and alkylated peptide and little or no reaction with collagen. Antisera to collagen showed strong binding with collagen and a weaker but distinct reaction with the procollagen peptide. Evidence was obtained that a minor contaminant of procollagen peptide was present in the acid-extracted collagen and that there were no shared antigenic determinants. Bovine serum and amniotic fluid contained 1–10 μg/ml reactive antigen. These results indicate that the labeled peptides can be used as a specific, sensitive and accurate assay for the amino-terminal portion of procollagen in biological samples.

Antigenic structure of the aminoterminal region in type I procollagen: Characterization of sequential and conformational determinants

Abstract Antibodies to calf or sheep skin procollagen, to its consitituent pα1(I) and pα2 chain or to peptide fragments from the aminoterminal region in pα1(I) were studied by radioimmune assays using various 125 0-labeled antigens. The highest antibody response was obtained against the globular region in pα1(I). The adjacent precursor-specific collagenous and non-helical domains may modulate the antigenicity of the globular determinants but are themselves only weak antigens. No antigenic determinants were detected in the pα2 chain. Antibodies to pα1(I) chain did not cross-react with pα2 or pα1(III) chains. Antisera against the globular procollagen peptide showed no difference between the peptide and procollagen indicating that the release of the globular region from pα1(I) by collagenase is not accompanied by large conformational alterations. Complete cross-reaction was observed between precursor-specific peptides obtained from calf or sheep procollagen. Reduction and S-carboxymethylation of the five disulfide bridges in the globular region largely destroyed antigenicity. However, a minor fraction of the antibodies against the native peptide could still react with the unfolded peptide. Determinants recognized in this reaction are apparently shared by the native and unfolded peptide and were stable towards trypsin. On the other hand, antibodies prepared against the reduced and alkylated procollagen peptide did not react with the native peptide. The data indicate that that globular regions in pα1(I) chain possesses both conformational and sequential determinants which differ considerably in their immunogenic capacity.

Immune response to calf collagen type I in mice: A combined control of Ir-1A and non H-2 linked genes

The genetically controlled immune response to calf skin collagen type I in mice could be demonstrated to be governed by at least two genes. One is linked to theH-2 complex and located within theIA subregion. High-responder alleles areH-2b,H-2f, andH-2s. The other gene(s) is not linked to theH-2 complex and high-responder allele(s) are found in the genome of B10 mice but not in the genome of DBA mice. There are strong indications that theIr-1A gene controls the response at the T-cell level, whereas it is assumed that the background gene(s) control the immune response at a different level.

Genetic control and carrier and suppressor effects in the antibody response of mice to procollagen.

The response potential of inbred mouse strains against bovine type I procollagen and its separated structural domains, i.e., the globular procollagen peptide and the triple-helical collagen moiety, was compared by passive hemagglutination and radioimmune assays. Studies with congenic and recombinant lines and with backcross populations showed that the antibody response to procollagen peptide is under the control of a gene located in theIA, IB subregion of theH-2 complex. The strain distribution pattern of high response to the procollagen peptide is not identical to that of the high response to collagen. Both the procollagen peptide and procollagen are thymus-dependent antigens, since no response was observed in nude mice. Low response to procollagen peptide and/or collagen could be corrected in some, but not all mouse strains by using procollagen as immunogen. Strains which show low response against collagen but high response against procollagen were injected with collagen prior to a challenge with procollagen. This treatment reduced the antibody response to the procollagen peptide but not to collagen. Carrier and suppressor effects in the response to procollagen are apparently more complex than those observed in the response to synthetic peptides.

Immunochemistry of bovine fibrinogen—I. Immunogenic activity and diversity of antigenic determinants of reduced and carboxymethylated α,β and γ chains

Abstract Antisera were produced in rabbits against carboxymethylated α, β and γ chains of bovine fibrinogen. Quantitative precipitation results indicated a comparable immunogenicity of α chains, β chains and fibronogen but γ chains were less immunogenic. Distinct antibody fractions could be prepared from each antiserum by successive immunoadsorption with different antigens. Antibodies bound onto and eluted from a fibronogen adsorbent showed in a hemagglutination-inhibition assay equal reactivity for fibronogen and the respective carboxymethylated or aminoethylated component chains. Subsequent adsorption on reduced and aminoethylated fibronogen demonstrated a second antibody fraction in the sera which reacted with carboxymethylated or aminoethylated chains but not with fibronogen. Final adsorption on reduced and carboxymethylated fibronogen and/or albumin gave rise to a third antibody fraction which reacted only with carboxymethylated chains. The results were interpreted as indicating three groups of antigenic determinants in the polypeptide chains of fibrinogen: (1) surface determinants of fibronogen; (2) chain determinants which are masked in fibrinogen; and (3) determinants probably located around the alkylated cysteine residues. Antibodies to determinants shared by fibronogen predominated in antisera to α and β chains. Antisera to γ chains contained mainly antibodies to the latter two types of determinants.

Immunochemistry of bovine fibrinogen—II.: Analysis by immunoadsorption of the antibody response against three large cyanogen bromide peptides

Abstract The specificity of rabbit antisera against three large cyanogen bromide peptides which account for 80% of the disulfide-bonded regions in bovine fibrinogen were characterized by immunoadsorption and passive hemagglutination. Antibodies to the α chain derived peptide F-CB3 showed complete cross-reaction with fibrinogen. The antigenicity of F-CB3 was not changed after reduction of the single disulfide bridge. Antibodies to the N-terminal, multi-chain peptide F-CB1 either cross-reacted with fibrinogen or were directed to antigenic determinanrs unique to reduced, unfolded chain constituents of fibrinogen. However, part of the antibodies against determinants shared by native fibrinogen were also able to react with its component chains, particularly with β and γ chains. Two antibody fractions with a similar specificity as in antisera to F-CB1 could be detected in antisera to the multichain peptide F-CB2. The β chain constituent in F-CB2 was important for antigenic activity. Antisera to F-CB2 contained in addition antibodies which did not cross-react with native or reduced fibrinogen.

Immunologie des Kollagens: Genetische Regulation der Immunantwort bei Inzuchtmäusen

Autoantikorper gegen strukturelle Elemente des Interzellularraumes haben bei der Diskussion der Pathogenese krankhafter Vorgange im Bindegewebe immer wieder eine Rolle gespielt. Die grundsatzliche Frage nach der Fruchtbarkeit eines solchen Konzeptes konnte bisher deshalb nicht gelost werden, weil ein geeignetes tierexperimentelles Modell fehlte. Bis 1954 galt das Kollagen sogar als nicht immunogen, und erst in den letzten Jahren ist es gelungen, die Antikorperantwort auf natives und denaturiertes Kollagen bei verschiedenen Tierspezies zu charakterisieren [1], unterstutzt durch erhebliche Fortschritte bei der Aufklarung der Struktur dieses ubiquitaren Bindegewebsproteins [1 bis 4]. Dabei ergaben sich starke artbedingte und individuelle Unterschiede, die einen komplexen Regulationsmechanismus der Immunantwort vermuten liesen. Die Maus kann als das am besten etablierte Tiermodell zum Studium der Regulation von Immunantworten gelten [5 bis 8]. Wir haben uns fur 2 Fragen interessiert, indem wir uns der Maus als Immuntier bedienten: