|
|||||||||
|
|
|
|
|
|
|
|
|
|
|
Blood, 1 March 2006, Vol. 107, No. 5, pp. 1908-1915. Prepublished online as a Blood First Edition Paper on October 27, 2005; DOI 10.1182/blood-2005-05-1814.
HEMOSTASIS, THROMBOSIS, AND VASCULAR BIOLOGY
Zebrafish to humans: evolution of the
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
3 chain of type IV collagen (3(IV)NC1), the Goodpasture autoantigen. We demonstrate that basement membrane extracted NC1 domain preparations from Caenorhabditis elegans, Drosophila melanogaster, and Danio rerio do not bind Goodpasture autoantibodies, while Xenopus laevis, chicken, mouse and human 3(IV)NC1 domains bind autoantibodies. The 3(IV) chain is not present in C elegans and Drosophila melanogaster, but is first detected in the Danio rerio. Interestingly, native Danio rerio 3(IV)NC1 does not bind Goodpasture autoantibodies. Next, we cloned, sequenced, and generated recombinant Danio rerio 3(IV)NC1 domain. In contrast to recombinant human 3(IV)NC1 domain, there was complete absence of autoantibody binding to recombinant Danio rerio 3(IV)NC1. Three-dimensional molecular modeling from existing x-ray coordinates of human NC1 domain suggest that evolutionary alteration of electrostatic charge and polarity due to the emergence of critical serine, aspartic acid, and lysine residues, accompanied by the loss of asparagine and glutamine, contributes to the emergence of the 2 major Goodpasture epitopes on the human 3(IV)NC1 domain, as it evolved from the Danio rerio over 450 million years. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
-chain gene products, 1(IV)-6(IV). The 1(IV) and 2(IV) chains are ubiquitously present in most organs, while the 3(IV)-6(IV) chains exhibit more restricted distributions and are present mostly in BMs with special functional roles, such as the kidney glomerular basement membrane (GBM).1,5-7
The 3-chain of type IV collagen is found in the kidney GBM, the alveolar BM of the lung, the BMs of the inner ear and the testis,1,8 and is present in most mammals.9 Mice deficient in the 3(IV) chain die due to renal failure, as this protein is required for the proper function of the GBM.10,11 This phenotype resembles human Alport syndrome, a disease affecting kidneys and the inner ear, which results from mutations in the 3 chain, the 4 chain, or the 5 chain of type IV collagen.8,10 Pathogenic autoantibodies directed against the NC1 domain of the 3 chain of type IV collagen (Goodpasture autoantigen) are associated with human Goodpasture syndrome (GP), a disease characterized by rapidly progressive glomerulonephritis and lung hemorrhage.8,12-13 These autoantibodies are targeted to specific amino acid sequences within the NC1 domain of the 3(IV) chain.8,14 By using mutagenesis analysis, it is known that a majority of the autoantibodies are directed against 2 distinct epitopes on the 3(IV)NC1 domain, namely epitopes EA and EB.15-18 Epitope EA appears to be the immunodominant epitope, as the majority of patients' antibodies are directed against this site.18 While the production of immunoglobulin G1 (IgG1) and IgG3 subclasses of 3(IV) autoantibodies is considered to be the pathogenic feature of Goodpasture syndrome, there is also strong evidence now for the role of T cells in the initiation of the disease.19-26 Therefore, in the past few years, Goodpasture syndrome has emerged as a classic autoimmune vascular disease mediated by B cells27-29 and T cells,19-26 and immunosuppression in conjunction with plasmapheresis remains the most effective therapy.30
Since the vast majority of autoantibodies bind to the immunodominant epitope EA and epitope EB, several studies have sought to elucidate the key residues in these sites. Such studies have been primarily focused on epitope EA because this site is capable of inducing disease in rodents.14 The method most commonly used has been to recombinantly produce chimeric proteins.15-18,31-32 Many of these chimeric proteins are comprised of amino acids of the epitope EA from the 3(IV)NC1 domain with site-specific substitutions of analogous residues from the 1(IV)NC1 domain.17,31-32 Such proteins were subsequently tested for their binding to autoantibodies from Goodpasture patients, thus identifying putative critical amino acids that comprise the Goodpasture epitope.
While others have employed site-directed mutagenesis and chimeric protein analysis as a way to identify the critical amino acids within the Goodpasture epitopes, we hypothesized that evolutionary changes over 450 million years33 in the amino acids within the 3(IV)NC1 domains might offer insight into the unique amino acid residues that comprise the B-cell epitope for Goodpasture autoantibodies. Comparative type IV collagen NC1 sequence analysis between 9 species from Caenorhabditis elegans to humans, BM extraction, NC1 domain analysis from different species, cloning, and recombinant protein production were used to identify the critical amino acids within the B-cell epitopes of the Goodpasture autoantigen. This study provides a strategy to identify critical amino acids responsible for Goodpasture autoantibody binding and offers new insights into potential therapeutic alternatives. A precise identification of amino acids responsible for autoimmunity will help design cell-mediated therapies to potentially impart systemic tolerance to pathogenic epitopes associated with this devastating disease.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Wild-type C57BL/6 mouse kidney, wild-type Danio rerio strain AB* kidney, and wild-type domestic chicken kidney (Pel-Freez Clinical Systems, Rogers, AR) were used for all procedures. Wild type Xenopus laevis kidneys were the kind gift of Dr Sergei Sokol (Beth Israel Deaconess Medical Center/Harvard Medical School). Wild-type strains CS and W1118 Drosophila melanogaster were the kind gifts of Dr Mel Feany (Brigham and Women's Hospital/Harvard Medical School). Wild-type C elegans, strain NL, were the kind gift of Dr Monica Colaiacovo (Department of Genetics, Harvard Medical School).
Cloning, primers, and sequencing
Total RNA was isolated from Danio rerio embryos 24 hours after fertilization (hpf) using TRIzol reagent (Invitrogen, Carlsbad, CA). A collection of cDNAs was created using Superscript II reverse transcriptase (Invitrogen) and oligo dT primers (Invitrogen). Polymerase chain reaction (PCR) amplification of Danio rerio 3(IV)NC1 domain was performed using AccuPrime Pfx DNA polymerase (Invitrogen) and the specific primers 5' CACCAGGTGCAAAAGGTCCACAAC 3' and 5' ACGATTTTGTGTCAGGGTCAGAA 3', designed from sequence data obtained from the Project Ensembl database and software system (http://www.ensembl.org), maintained by the European Molecular Biology Laboratory-European Bioinformatics Institute and the Sanger Institute. The resulting cDNA clones were subcloned into the pCRII 4.0-TOPO vector (Invitrogen), and sequenced by the Beth Israel Deaconess Medical Center sequencing facility using the M13 (-20) forward and M13 reverse primers. Sequences were analyzed using the Mac Vector 6.0 (Oxford Molecular Group, Cambridge, United Kingdom) and BLAST programs (National Center for Biotechnology Information; http://www.ncbi.nlm.nih.gov/). The cloned sequences were deposited in GenBank under the accession number AY954909
[GenBank]
.1.
Recombinant protein production
The following PCR primers were designed to amplify a cDNA encoding the full-length Danio rerio 3(IV)NC1, which we have termed z3(IV)NC1; 5' TCA CCA GGT GCA AAG CTT CCA CAA 3' and 5' TCT GTT CTC GAG TTT GGC GGT TTG 3'. The underlined sequence indicates the HindIII and XhoI restriction sites that were used for subsequent cloning of the generated cDNA fragment. The HindIII/XhoI fragment of the amplified product was cloned into the pcBFT expression vector containing the cytomegalovirus (CMV) promoter, the BM-40 signal peptide, and a FLAG (octapeptide DYKDDDDK) tag.34 293 human embryonic kidney cells were transfected with the plasmid using the 293 Fectin-system (Invitrogen) according to manufacturer's protocol. Recombinant z3(IV)NC1 was secreted into the culture supernatant as a fusion protein with the FLAG epitope. Supernatant containing protein was collected 24 hours after transfection. Recombinant human 3(IV)NC1 production was described earlier.34
Histology and immunofluorescence
Adult Danio rerio were embedded in Tissue-Tek optimal cutting temperature (OCT) compound (Sakura Finetek, Torrance, CA) and snap-frozen on liquid nitrogen, as well as fixed in formalin and embedded in paraffin for further histologic analysis. Paraffin sections (5 µm) were stained with hematoxylin and eosin (H&E) according to routine histologic practice. Frozen sections (5 µm) were fixed with acetone, blocked with 1% bovine serum albumin (BSA) in phosphate-buffered saline (PBS), incubated with the anti-T7 antibody (raised in rabbit against the T7 peptide of human 3(IV)NC1 domain)35 at 1:75 dilution, and the anti-human 3(IV)-36mer antibody at 1:200 dilution,32 followed by an antirabbit FITC-conjugated secondary antibody at 1:100 dilution (Jackson Immunoresearch, Bar Harbor, ME). The sections were mounted using the Vectashield mounting media containing DAPI for nuclear staining (Vector Laboratories, Burlingame, CA). Histologic images were captured using a Zeiss brightfield and fluorescence microscope, a 40x/1.3 numeric aperture Plan-Neofluar objective with Immersol oil, an AxioCam HRc digital camera, and Axiovision software version 4.3 (all from Zeiss, Oberkochen, Germany).
Isolation of native type IV collagen NC1 domains
Kidneys dissected from mouse, chicken, Xenopus laevis, and Danio rerio were snap-frozen in liquid nitrogen followed by homogenization by mortar and pestle. In the case of Drosophila and C elegans, the whole organisms were subjected to homogenization. Tissue homogenates were centrifuged and resuspended in PBS containing Complete Protease Inhibitor Cocktail (Roche, Indianapolis, IN), thoroughly homogenized again, centrifuged, and resuspended in 1 M NaCl supplemented with DNase I (80 µg/mL). Following a third centrifugation, samples were resuspended in 2% deoxycholate/protease inhibitor solution and centrifuged a final time. Samples were resuspended in bacterial collagenase solution (1 U/µL; Worthington Biochemical, Lakewood, NJ) to digest proteins with collagenous regions, in order to liberate soluble NC1 domains of type IV collagen.
Western blotting and antibodies
All Western blots were performed on either 10% or 15% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE; Bio-Rad, Hercules, CA) under reducing or nonreducing conditions. Gels were transferred to nitrocellulose and incubated with sera and primary and secondary antibodies at the appropriate dilutions. The following primary antibodies were used: anti-3(IV)-36mer antibody at 1:7500 dilution;32 2 separate 3(IV)NC1 antibodies;27 3 confirmed Goodpasture patient sera at 1:20 dilution;36 and anti-FLAG M2 antibody at 1:1000 dilution (Sigma, St Louis, MO). In addition, 2 normal human sera were used as negative controls at the same dilutions as the Goodpasture sera (data not shown). Anti-human IgG and anti-rabbit IgG horseradish peroxidase (HRP; Sigma)-conjugated secondary antibodies were used at 1:1000 dilutions. Enhanced chemiluminescence (ECL) reagent for peroxidase substrate (Amersham Biosciences, Freiburg, Germany) was used to visualize antibody binding by exposure to x-ray film (Denville, Metuchen, NJ).
Molecular graphics
The structures of the [(1)22]2 NC1 hexamers from human placenta basement membranes (Protein Data Bank ID 1LI1
[PDB]
)37 were rendered using the molecular graphics visualization program YASARA (YASARA Biosciences, Graz, Austria) and analyzed to generate modeled structures using the molecular graphics software InsightII (Accelrys, San Diego, CA).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
To study the binding of Goodpasture autoantibodies, we extracted NC1 domains from BM extracts using bacterial collagenase from the following species: mouse, chicken, Xenopus laevis, Danio rerio, Drosophila, and C elegans. For the higher vertebrate organisms, kidney tissue was used as a source of BM, whereas for Drosophila and C elegans whole-organism BM extracts were used. Type IV collagen NC1 domains were successfully extracted from all organisms as shown by SDS-PAGE analysis and Western blot evaluation under reducing conditions (Figure 1A-B). When analyzing Goodpasture autoantibody reactivity to the isolated BM extracts under nonreducing conditions, positive binding of sera to 3(IV) chain dimers and monomers could be observed in mouse, chicken, and Xenopus (Figure 1C). Interestingly, binding to dimers was stronger than to monomers, as others have described9 (Figure 1C). The Goodpasture epitope is considered to be sensitive to conformational changes and thus complete reduction of disulfide bonds with 10% 
-mercaptoethanol leads to a total loss of antibody binding (Figure 1E), while incomplete reduction with 1.5% -mercaptoethanol leads to stronger binding to 3(IV)NC1 monomers in all species (Figure 1D). Reactivity with Goodpasture sera could not be observed with NC1 domains extracted from Danio rerio, C elegans, and Drosophila (Figure 1C). In Drosophila and C elegans both 1(IV)-like and 2(IV)-like genes have been described,38-40 but our genomic database analysis and cloning attempts suggest that an 3(IV) gene does not appear to be present. Interestingly, our database search indicates that Danio rerio has 6 type IV collagen chains similar to higher organisms (B.A.M., M.S., R.K., unpublished data). However, while the Danio rerio 3(IV) gene was identified in our database search, the native NC1 domain extracts do not bind to Goodpasture autoantibodies (Figure 1C).
Cloning, sequencing, and recombinant expression of the Danio rerio 3(IV) chain NC1 domain
Our results suggest that while the 3(IV) chain is present in the Danio rerio (by inference from genomic database sequences), the NC1 domain extract from BM preparations of Danio rerio renal tissue does not bind to Goodpasture autoantibodies. To further evaluate the capacity of the Danio rerio 3(IV)NC1 domain to bind these antibodies, we attempted to clone and generate recombinant protein.
In order to clone the Danio rerio 3(IV)NC1 domain (z3(IV)NC1), we isolated total RNA from Danio rerio embryos and generated a collection of cDNAs by reverse transcription (RT), using oligo dTs as primers for the reaction. These cDNAs were subsequently used as a template to obtain the z3(IV)NC1 domain by PCR. Gene-specific PCR primers were designed using sequence data information obtained from the project Ensembl database. We successfully generated an 821-bp cDNA fragment, and sequencing verified that the cloned cDNA fragment contained 23 bp of the collagenous region, 702 bp containing the entire NC1 domain, and 96 bp of noncoding sequence. The nucleotide and amino acid sequence of z3(IV)NC1 is shown in Figure 2A. The 5 exons encoding the overall nucleotide and amino acid sequence of the z3(IV)NC1, when compared with human 3(IV)NC1, have 60% and 74% homology, respectively. The 12 cysteines known to be essential for the appropriate conformational folding of the type IV collagen NC1 structure41 are all conserved in z3(IV)NC1 (Figure 2A). Interestingly, a methionine and a lysine residue, potentially involved in a novel nondisulfide covalent crosslink37 are also conserved in the z3(IV)NC1 (Figure 2A).
|
|
3(IV)NC1 using a BM-40 signal peptide to enable extracellular secretion. Recombinant FLAG-tagged z3(IV)NC1 fusion protein in the culture supernatants was analyzed by SDS-PAGE. Western blot analysis using antibodies directed against the FLAG tag and the human 3(IV)NC1 verified the recombinant expression of z3(IV)NC1 monomers at the predicted molecular weight of approximately 29.4 kDa (z3(IV)NC1+FLAG tag; Figure 2B). However, z3(IV)NC1 did not bind to Goodpasture autoantibodies from 3 different patients with clinical disease (Figure 2B), despite prolonged incubation with high concentrations of the autoantibodies to detect even weak binding. However, as shown earlier, strong binding of anti-human 3(IV)NC1 polyclonal antibodies to isolated native zNC1 domain monomers, dimers, and trimers from BM extracts were observed (Figure 2C). Under reducing conditions the higher-molecular-weight trimers were lost and monomers of Danio rerio 3(IV)NC1 revealed stronger binding (Figure 2C). However, binding with the Goodpasture sera to the isolated Danio rerio NC1 domains could be not observed, as shown earlier (Figure 2D). Together, these results demonstrate that although Danio rerio contains the encoding gene and the 3 chain of type IV collagen protein, the protein does not possess the capacity to bind Goodpasture autoantibodies.
Expression of the 3(IV) chain in Danio rerio tissues
Since this report is the first description of the 3 chain of type IV collagen in the Danio rerio, and human 3(IV) chain is localized to the GBM of the kidney and the alveolar BMs of the lung, we performed immunohistochemistry to evaluate the expression of the z3(IV)NC1 in gills and kidney tissue of Danio rerio. Immunohistochemical localization of the z3(IV) chain in tissues was performed using 2 separate antibodies directed against the human 3(IV)NC1 domain. Identical results were obtained with both antibodies (Figure 3 and data not shown). The 3(IV) chain in Danio rerio is strongly expressed in the gills and the kidney (Figure 3B-D). Gills are the fish equivalent of lungs, and the lungs and kidney are the sites of high expression of the 3(IV) chain in mouse and humans.1,5-7 On the other hand, we could not observe any expression of the 3(IV) chain in Danio rerio testis (data not shown), although in mammals it is a site with high 3(IV) expression.1,5-7,42 In the gills the 3(IV) chain appeared to be expressed in the BM underlying the epithelial cell layer (Figure 3A-B). In the Danio rerio, the kidney morphology is distinct from that of higher organisms as it is also the site of hematopoiesis, in addition to its function in ultrafiltration (Figure 3C). In Danio rerio kidney the 3(IV) chain is expressed in both the BM of the glomeruli as well as some tubules (Figure 3D). Interestingly, we could also observe strong staining of the 3(IV) chain in the hematopoietic tissue (Figure 3D). In higher organisms, the 3(IV) chain is an important component of the GBM in the kidney and is also expressed in some tubular BMs.6 However, the potential role of this molecule in hematopoiesis is unknown. As observed in humans and other mammals, we were unable to detect z3(IV)NC1 in liver and stomach basement membranes.
|
3(IV)NC1 among different organisms
Goodpasture autoantibodies bind to human and chicken 3(IV)NC1 domains, but do not bind to the Danio rerio 3(IV)NC1 domain. In order to explain observed differences in Goodpasture autoantibody binding, we compared amino acid sequences at the 2 putative B-cell Goodpasture epitope regions, EA and EB, among different species (Tables 1, 2). The Danio rerio and chicken sequence analysis reveals that while most amino acids are conserved, some critical changes can be noticed, which are further preserved in the human sequence. Our results demonstrate that Goodpasture antibody binding at epitope EA (Table 1) on the human 3(IV)NC1 domain requires conservation of threonine-17 (T17) and the presence of serine-21 (S21). Binding at epitope EB (Table 2) requires threonine-127 (T127), loss of asparagine-128 (N128), substitution of proline or alanine for glutamine-131(Q131), and lysine-141 (K141). This analysis is the first to show that T127 and K141 are important residues for antibody binding at Goodpasture epitope EB.
|
|
All of the 3(IV) sequences, whether positive or negative for Goodpasture sera binding, share the following sequence homology in epitope EA—TxxPxCPxGTxxLYx, (Table 1). Cross-species conservation indicates that this sequence motif is evolutionarily preserved for structure or function. A similar conserved sequence motif, TxxPxCPxGWxSLWx (Table 2), exists in epitope EB, regardless of Goodpasture antibody binding. The only exceptions to this sequence conservation are isoleucine (I) substituted for the N-terminal threonine (T) in Danio rerio, and aspartate (D) substituted for the middle glycine (G) in mouse. Consistent with epitope EA, conservation of the EB sequence motif indicates evolutionary preservation of the sequence for NC1 domain structure or function, and underscores the importance of the intervening nonconserved residues (x) within these conserved sequence motifs for antigenic recognition by Goodpasture autoantibodies directed against epitope EA and EB. The conserved EB sequence is distinguished from EA by tryptophan (W) residues (W136 and W140), rather than threonine (T) and tyrosine (Y), and one additionally conserved C-terminal serine residue (S138).
Molecular modeling for 3D analysis of the human, chicken, and Danio rerio 3(IV)NC1 domains
In order to spatially evaluate charge, hydropathy and side-chain polarity of the epitope EA and EB sequence in the context of the entire NC1 domain, we used molecular modeling software and the existing human type IV collagen NC1 domain crystal structure to generate models (Figure 4A-G). From crystallographic data it has been shown that the 1(IV)NC1 and 2(IV)NC1 chains are folded virtually identically, and their topologies very closely resemble one another.37 Based on the high sequence identity between all type IV collagen NC1 domains, the 3(IV)NC1 domain is expected to fold with the same topology as the 1(IV)NC1.37 Therefore, by aligning the 3(IV)NC1 and 1(IV)NC1 sequences and using the known [(1)22]2 NC1 hexamer structure (Protein Data Base ID 1LI1
[PDB]
), we have modeled amino acids that are divergent in the Danio rerio, chicken, and human EA and EB epitope regions on a backbone homology 3(IV)NC1 structure (Figure 4B-D). Amino acid differences in epitope EA (blue) and EB (green) regions in chicken compared to Danio rerio are indicated by side chain rendering in yellow, while further amino acid differences in human compared to chicken are shown by side chain rendering in red (Figure 4E-G).
|
3(IV)NC1 domain homology structures, residues that we identified as important in Goodpasture autoantibody binding at epitopes EA and EB are localized to the outer loops of the epitopes (Figure 4E-G). In addition, many of these residue changes alter electrostatic potentials in these solvent exposed epitope regions. For example, S21 and K141, described as important for epitope EA and EB binding, respectively, are glutamate residues in the immunonegative Danio rerio.
The 3D homology structures enabled us to assess conserved 3(IV) NC1 residues immediately proximal to epitope EA. Glutamine 57 (Q57) is conserved in all immunopositive proteins, but absent from all immunonegative proteins, including Danio rerio 3(IV) NC1. Surprisingly, although 26 residues away from epitope EA, in the tertiary NC1 domain structure, Q57 is immediately adjacent to epitope EA (Figure 4F-G; rendered in magenta) and provides an additional strongly hydrophilic residue. Previous work demonstrates Q57 is as important as S21 for autoantibody binding to EA. Specifically, alteration of Q57 results in 80% loss of patient sera binding17 to this discontinuous and conformational immunodominant epitope.
Our combined findings prompted us to map the amino acid character and molecular surface of species-variable residues (nonconserved) onto the modeled Danio rerio, chicken, and human EA and EB epitope regions of our backbone homology 3(IV)NC1 structures (Figure 4B-D). Using color-coded amino acid character maps, it is readily apparent that in both EA and EB electrostatic charge and polarity changes are important in Goodpasture autoantibody binding. A significant loss of electrostatic potential (positive/basic residue character) accompanies autoantibody recognition in the EA epitope region. The adjacent K27, R28 residue pair in Danio rerio is either a hydrophilic/polar residue as in chicken or hydrophobic/nonpolar residue as in human. Overall, the immunopositive chicken and human EA domains have reduced electrostatic potential and greater hydrophilic character compared with the homologous region in Danio rerio (Figure 4B-D). In contrast, the gain of a positive/basic residue in the epitope EB C-terminus is important for autoantibody binding (Figure 4A-G). K141 is 1 of 2 residue changes in epitope EB in chicken compared with Danio rerio that is present in humans; the other is T127. The substitution of threonine (T127) decreases the nonpolar character of the epitope EB N-terminus. In summary, these comparisons demonstrate that evolutionary alterations of electrostatic charge and polarity contribute to Goodpasture autoantibody binding in the 3(IV)NC1 EA and EB epitopes.
Interestingly, we found that the high hydrophobic character of both the EA and EB domains is largely conserved from Danio rerio to chicken to human. In previous studies, both V27 and P28 were shown to be important for antibody binding and critical residues of the EA autoepitope,17,31,43 through mutagenic and chimeric analysis between 1(IV)NC1 and 3(IV)NC1 sequences (in human 1(IV) these residues are lysine and isoleucine). Examination of the corresponding residues in chicken, serine and glutamine, lessens the idea that V27 and P28 are important in epitope EA for hydrophobic contributions, and suggests that the reduced antibody binding observed in mutagenic/chimeric analyses possibly reflects the effects of basic charge insertion and side-chain steric clash (Figure 4A-D). Of note, 1 variable (nonconserved) hydrophobic residue at position 131 appears important for the EB antibody recognition site (Figure 4A-D). In Danio rerio a glutamine occurs at this position, while there is a transition to alanine in chicken and proline in human. Interestingly, a nonpolar/hydrophobic residue type was conserved at position 137 (isoleucine/leucine/valine) in epitope EB between Danio rerio, chicken, and human. The immunonegative sequences of bovine 1(IV)NC1 and Drosophila and C elegans 1(IV)NC1 and 2(IV)NC1 contain glutamate or serine at position 137. This finding suggests that a nonpolar/hydrophobic residue at position 137 may be involved in autoantibody binding, but is not sufficient in the context of the Danio rerio EB region to impart antibody binding.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
3 chain of type IV collagen in the Danio rerio. This could be linked to the required homeostasis and physiology of vertebrates to exercise greater regulation of fluid and gas exchanges. Interestingly, the Goodpasture epitope is not present within the 3(IV)NC1 domain of Danio rerio. The emergence of this epitope (in frogs) coincides with the transfer from aquatic to terrestrial living. Last, our study demonstrates the importance of Danio rerio in elucidating the evolutionary changes within the 3(IV)NC1 domain, which is the only species to possess an 3 chain of type IV collagen not recognized by Goodpasture autoantibodies. This insight into the critical amino acids required for binding autoantibodies will potentially lead to cell-mediated therapies to impart systemic tolerance to the pathogenic epitopes that cause this devastating disease. Future studies will hopefully shed more light on the physiologic need for such modification in the 3 chain of type IV collagen. |
Acknowledgements |
|---|
|
Footnotes |
|---|
Prepublished online as Blood First Edition Paper, October 27, 2005; DOI 10.1182/blood-2005-05-1814.
Supported by National Institutes of Health (NIH) grants DK55001 and DK62987, and funds from the Center for Matrix Biology at the Beth Israel Deaconess Medical Center. M.S. is funded by the Sigrid Juselius Foundation, the Maud Kuistila Foundation, the Finnish Medical Society Duodecim, and the Emil Aaltonen Foundation. M.A.G. is funded by American Heart Association grant 05030348N. L.I.Z. is supported by the NIH and the Howard Hughes Medical Institute (HHMI). K.L.P. is supported by the Albert J. Ryan Fellowship.
R.K., B.A.M., M.S., and M.A.G. designed research; B.A.M., M.S., M.A.G., and K.H. performed research; K.L.P. and L.I.Z. contributed vital new reagents; R.K., B.A.M., M.S., and M.A.G. analyzed data; and R.K., B.A.M., and M.S. wrote the paper.
The publication costs of this article were defrayed in part by page charge payment. Therefore, and solely to indicate this fact, this article is hereby marked "advertisement" in accordance with 18 U.S.C. section 1734.
Reprints: Raghu Kalluri, Associate Professor of Medicine, Harvard Medical School, Director, Center for Matrix Biology, DANA 514, Beth Israel Deaconess Medical Center, 330 Brookline Ave, Boston, MA 02215; e-mail: rkalluri{at}bidmc.harvard.edu.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
3-chain of type IV collagen and emergence of the autoimmune epitopes associated with Goodpasture syndrome
