A unique covalent bond in basement membrane is a primordial innovation for tissue evolution

Significance

The evolution of multicellular animals from single-celled ancestors was one of the most significant transitions of life on earth. The emergence of larger, more complex animals able to resist predation and colonize new environments was enabled, in part, by a collagen scaffold, which anchors cells together to form tissues and organs. Here, we show that a unique chemical bond, a link between sulfur and nitrogen atoms called a sulfilimine bond, arose over 500 Mya, binding this scaffold together and enabling tissues to withstand mechanical forces. Peroxidasin forms the bond by generating hypohalous acids as strong oxidants, a form of bleach, which normally function as antimicrobial agents. These understandings may lead to approaches for targeting tumors and treatment of other diseases.

Fig. 1.  The sulfilimine bond stabilizes collagen IV scaffolds by the cross-linking of triple helical building block protomers.
(A) The sulfilimine bond cross-links Met93 and Hyl211 at the interface between the trimeric NC1 domains of two adjoining protomers, forming a globular hexamer structure.
(B) Dimeric subunits reflect the presence of the sulfilimine bond in human collagen IV by immunoblot (JK2 Ab) and protein stain.
(C) MS analysis of tryptic peptides derived from dimeric subunits verified the presence of the bond by a mass difference of 2.0299 between theoretical mass of uncross-linked and observed mass of cross-linked peptides and subsequent multistep CID fragmentation (MS2/MS3) analyses.

Fig. 2. Multiple sequence alignment of collagen IV NC1 domains encompassing Met93 and Hyl211 amino acid residues and Pxdn among 11 metazoan and 1 protozoan phyla.
(A) Met93 and Lys/Hyl211 (yellow) are conserved in all eumetazoans, except for the cnidarian H. magnipapillata, and they are absent in the phyla of Placozoa and Porifera and the protozoan phylum Choanozoa. All sequences belong to the collagen IV α1-like subfamily of chains, except for Drosophila (viking) and Ascaris (α2 chain).
(B) Schematic representations of Pxdn. Pxdn sequence was incomplete on both ends for Mytilus, Clytia, Trichoplax, and Monosiga and short on one end for Saccoglossus, which is indicated here by a shortened schematic representation. Sequence data were gathered from *National Center for Biotechnology Information Reference Sequence, †gathered from whole-genome shotgun/transcriptome shotgun assembly, §generated by RNA-Seq analysis of animal tissues, or ¶assembled from cDNA libraries. All National Center for Biotechnology Information GenBank accession numbers are listed in Table S1.

Fig. 3. NC1 hexamers were excised from animal basement membranes and analyzed by SDS/PAGE as shown in Fig. 1 A and B. The dimeric subunits, which indicate the presence of the bond, were found in nine major eumetazoan phyla. Among eight cnidarians investigated, only Hydra NC1 lacked dimeric subunits. All NC1s were immunoblotted against the rat monoclonal antibody, JK2, except for C. elegans (rabbit polyclonal; NW-154) and Drosophila (mouse monoclonal; 6G7). Black outlines indicate the locations of cropping for blot images. 

Fig. 4. Expression of collagen IV and Pxdn during development in zebrafish and morpholino (MO) knockdown of peroxidasin in zebrafish embryos. (A) Pxdn and collagen IV expression during zebrafish embryonic development. Real-time qPCR studies were conducted to examine expression levels of Pxdn, collagen4α1, and collagen4α2. *Student t test P value < 0.03 compared with expression at 1,000 cells. Error bars = SEM. Blue, pxdn; red, col4a; black, col4a2. (B) Control and (C) Pxdn MO groups. MO-injected embryos displayed (D) general severe defects that include cardiac edema, smaller eyes, and gross trunk patterning defects (4/45), (E) partial curved trunk (21/45), or (F) normal development (20/45). (G) SDS/PAGE analysis of Pxdn MO embryonic phenotypes at 24 hpf by Western blot. Collagenase digests were normalized for total protein load by protein stain with SYPRO-Ruby (Fig. S8).


Abstract

Basement membrane, a specialized ECM that underlies polarized epithelium of eumetazoans, provides signaling cues that regulate cell behavior and function in tissue genesis and homeostasis. A collagen IV scaffold, a major component, is essential for tissues and dysfunctional in several diseases. Studies of bovine and Drosophila tissues reveal that the scaffold is stabilized by sulfilimine chemical bonds (S = N) that covalently cross-link methionine and hydroxylysine residues at the interface of adjoining triple helical protomers. Peroxidasin, a heme peroxidase embedded in the basement membrane, produces hypohalous acid intermediates that oxidize methionine, forming the sulfilimine cross-link. We explored whether the sulfilimine cross-link is a fundamental requirement in the genesis and evolution of epithelial tissues by determining its occurrence and evolutionary origin in Eumetazoa and its essentiality in zebrafish development; 31 species, spanning 11 major phyla, were investigated for the occurrence of the sulfilimine cross-link by electrophoresis, MS, and multiple sequence alignment of de novo transcriptome and available genomic data for collagen IV and peroxidasin. The results show that the cross-link is conserved throughout Eumetazoa and arose at the divergence of Porifera and Cnidaria over 500 Mya. Also, peroxidasin, the enzyme that forms the bond, is evolutionarily conserved throughout Metazoa. Morpholino knockdown of peroxidasin in zebrafish revealed that the cross-link is essential for organogenesis. Collectively, our findings establish that the triad—a collagen IV scaffold with sulfilimine cross-links, peroxidasin, and hypohalous acids—is a primordial innovation of the ECM essential for organogenesis and tissue evolution.

Footnotes

1A.L.F., R.M.V., and S.V.C. contributed equally to this work.

2A list of The Aspirnaut coauthors can be found in Table S2. Aspirnaut is a K--20 Science, Technology, Engineering, and Math (STEM) pipeline program for diversity that partners the experiential and content expertise of Vanderbilt University with rural kindergarten through 12th grade schools and diverse high school, undergraduate, and graduate students.

3To whom correspondence should be addressed. E-mail: billy.hudson@vanderbilt.edu.

Author contributions: R.M.V., S.V.C., V.K.P., V.P.Y., M.T.I., J.K.H., and B.G.H. designed research; A.L.F., S.V.C., G.B., V.P.Y., C.L.S., K.L.R., W.H.M., T.A.C., D.-B.B., R.E.S., and T.A. performed research; G.B. contributed new reagents/analytic tools; A.L.F., R.M.V., S.V.C., V.K.P., V.P.Y., D.-B.B., and R.E.S. analyzed data; and A.L.F. and B.G.H. wrote the paper.

The authors declare no conflict of interest.

*This Direct Submission article had a prearranged editor.

Data deposition: The sequences reported in this paper have been deposited in the GenBank database (accession nos. GAMX01000001, GAMX01000002, GAND01000001, GAND01000002, GANB01000001, GANB01000002,GAMY01000001, GAMY01000002, GANA01000001, GANA01000002, GAMZ01000002, and GANC01000002).

This article contains supporting information online at 

www.pnas.org/lookup/suppl/doi:10.1073/pnas.1318499111/-/DCSupplemental.

Freely available online through the PNAS open access option. (Full Text)

ORIGINAL: PNAS

Aaron L. Fidlera,1, 

Roberto M. Vanacorea,b,1, 

Sergei V. Chetyrkina,b,1, 

Vadim K. Pedchenkoa,b,

Gautam Bhavea,b, 

Viravuth P. Yinc, 

Cody L. Stothersa, 

Kristie Lindsey Rosed,e, 

W. Hayes McDonaldd,e,f,

Travis A. Clarkg, 

Dorin-Bogdan Borzaa,b, 

Robert E. Steeleh, 

Michael T. Ivyi, 

The Aspirnautsb,2,

Julie K. Hudsonj, and 

Billy G. Hudsona,b,c,d,f,k,l,3

Author Affiliations

Edited* by Mina J. Bissell, E. O. Lawrence Berkeley National Laboratory, Berkeley, CA, and approved November 22, 2013 (received for review September 30, 2013)