Extraction of basement membrane, bacterial collagenase solubilization and the analysis of native type IV collagen NC1 domains from different species. (A) Native protein (60 μg per lane) from the 6 different species and 250 ng of recombinant human α3(IV)NC1, (rh-α3(IV)NC1)35 was run on a 15% SDS-PAGE under extreme reducing (10% β-mercaptoethanol) conditions, and incubated with the anti-36mer antibody.23 Monomer forms (M) of NC1s are detected in all lanes. Although subjected to extreme reduction, significant amounts of the isolated NC1 remain as dimers (D). (B) C elegans NC1 in dimeric form (D) was visualized after increased film exposure. (C) The same samples were run on a 15% SDS-PAGE under nonreducing conditions and incubated with 3 different Goodpasture sera (1:20 dilution). A representative blot is shown. As expected, sera from Goodpasture patients bound recombinant human α3(IV)NC1 (lane 1). Mouse, chicken, and frog kidney BM extracts show strong reactivity with Goodpasture sera (lanes 2-4). NC1 in monomer form (M) could only be detected for chicken without the use of a reducing agent, whereas for the other species the Goodpasture sera detected dimers (D) under nonreducing conditions. Interestingly, no reactivity to Danio rerio, Drosophila, and C elegans NC1 was found with any of the analyzed Goodpasture sera (lanes 5-7). (D) Reduction of the samples eliminates the conformational structure of the Goodpasture epitope. With slight reduction (1.5% β-mercaptoethanol) some monomer (M) from each of the 3 positive species (Mus, Gallus, Xenopus) native NC1s was recognized by the Goodpasture sera. (E) Under extreme reducing conditions (10% β-mercaptoethanol), binding to all monomer and dimer forms of NC1 by Goodpasture sera was eliminated, as expected and as previously described.17

Cloning and sequencing of the Danio rerio α3(IV)NC1 and characterization of recombinant Danio rerio α3(IV)NC1 domain. (A) Nucleotide and peptide sequence of the NC1 domain of Danio rerio type IV collagen α3-chain, (zα3(IV)NC1). The 2 regions homologous to Goodpasture epitope, E_A and E_B, are boxed. The conserved methionine and lysine (diamonds) residues, potentially involved in a conserved novel nondisulfide covalent crosslink,27 are shown. The circles indicate the 12 conserved cysteines. (B) FLAG-tagged Danio rerio α3(IV)NC1 was recombinantly produced and an expected band at 29.4 kDa is recognized by the anti-FLAG antibody (lane 1), as well as with the anti-human α3(IV) NC1 antibody (lane 2). However, no binding with Goodpasture sera is observed (lane 3), despite prolonged exposure of film. (C) Native Danio rerio kidney extract (30 μg per lane) digested with bacterial collagenase was run on 15% SDS-PAGE under both reducing and nonreducing (NR) conditions. Under nonreducing conditions, only dimers (D) and trimers (T) can be observed with the anti-human α3(IV) NC1 antibody antibody (lane 1). Under reducing conditions, native α3(IV)NC1 monomers (M) are observed at the expected molecular weight (lane 3). FLAG-tagged recombinant human α3(IV)NC1 (rh-α3(IV)NC1; 250 ng) was used as positive control (lane 2).35 (D) No reactivity with Goodpasture sera with native Danio rerio kidney extract can be seen (lane 1). FLAG-tagged rh-α3(IV)NC1 was used as positive control (lane 2).

Expression of type IV collagen α3 chain in the Danio rerio. Adult Danio rerio tissues were stained with anti-human α3(IV) antibodies and H&E. (A-B) In Danio rerio the gills are the equivalent of the lung. Strong expression of α3(IV) can be seen in the Danio rerio gills (green). DAPI staining is used to stain the nuclei of cells (blue). (C-D) In Danio rerio the kidney is the main hematopoietic organ and thus morphologically distinct from that of higher organisms (C). Strong expression of α3(IV) can be seen in both the GBM (G) and the tubular basement membranes (T) as well as in the hematopoietic tissue (HP) (D).

Species-specific sequence changes in the emergent E_A and E_B epitope regions of the α3(IV)NC1 domain. (A) Alignment of epitopes and E_A and E_B of Danio rerio, chicken, and human. The 6 variable (nonscaffold) residues of epitope E_A and the 5 variable (nonscaffold) residues of E_B are numbered. The color coding of residues (which is identical to that in panels B-D) denotes the chemical properties of residues, and is explained in the colored key at right. The numbered keys delineate the emergence of these epitopes as they arose from immunonegative Danio rerio to immunopositive chicken and human by defining the evolutionarily induced changes in chemical properties and their effects on Goodpasture autoantibody binding. (B-G) Using the known [(α1)₂α2]₂ NC1 hexamer crystal structure (PDB ID 1LI1), we have modeled the species-specific amino acids in the E_A and E_B regions for Danio rerio (B,E), chicken (C, F) and human (D,G) onto a α3(IV)NC1 backbone-homology structure. (B-D) The molecular surface and amino acid character of nonconserved residues in both the E_A and E_B epitope regions for Danio rerio (B), chicken (C), and human (D) are rendered in a color-coded, semitransparent surface representation: neutral (gray), strongly hydrophilic (green), moderately hydrophilic (cyan), basic (blue), and acidic (red) amino acid character. The remainder of the α3(IV)NC1 domain backbone is shown in ribbon representation (gray). (E-G) The peptide backbone from residues 17-31 (using human α3(IV) numbering) in the α3(IV) chain corresponding to the E_A epitope region is colored blue, and from residues 127-141 corresponding to the E_B epitope region is colored green. Nonconserved residues in E_A and E_B are shown with rendered side chains, and identified by single-letter residue code. Sequence differences between the Danio rerio and chicken E_A and E_B epitopes are highlighted in yellow on the chicken homology α3(IV)NC1 domain structure. Additionally, sequence differences between the chicken and human E_A and E_B epitope regions are highlighted further in red on the human homology α3(IV)NC1 domain structure. The Q57 critical in forming the conformational, discontinuous epitope E_A is colored magenta.