Advertisement

Fibrillins Can Co-assemble in Fibrils, but Fibrillin Fibril Composition Displays Cell-specific Differences*

  • Noe L. Charbonneau
    Affiliations
    Department of Biochemistry and Molecular Biology, Shriners Hospital for Children, Oregon Health and Science University, Portland, Oregon 97201
    Search for articles by this author
  • Bette J. Dzamba
    Footnotes
    Affiliations
    Department of Biochemistry and Molecular Biology, Shriners Hospital for Children, Oregon Health and Science University, Portland, Oregon 97201
    Search for articles by this author
  • Robert N. Ono
    Affiliations
    Department of Biochemistry and Molecular Biology, Shriners Hospital for Children, Oregon Health and Science University, Portland, Oregon 97201
    Search for articles by this author
  • Douglas R. Keene
    Affiliations
    Department of Biochemistry and Molecular Biology, Shriners Hospital for Children, Oregon Health and Science University, Portland, Oregon 97201
    Search for articles by this author
  • Glen M. Corson
    Affiliations
    Department of Biochemistry and Molecular Biology, Shriners Hospital for Children, Oregon Health and Science University, Portland, Oregon 97201
    Search for articles by this author
  • Dieter P. Reinhardt
    Footnotes
    Affiliations
    Department of Biochemistry and Molecular Biology, Shriners Hospital for Children, Oregon Health and Science University, Portland, Oregon 97201
    Search for articles by this author
  • Lynn Y. Sakai
    Correspondence
    To whom correspondence should be addressed: Shriners Hospital for Children, 3101 S.W. Sam Jackson Park Road, Portland, OR 97201. Tel.: 503-221-3436; Fax: 503-221-3451
    Affiliations
    Department of Biochemistry and Molecular Biology, Shriners Hospital for Children, Oregon Health and Science University, Portland, Oregon 97201
    Search for articles by this author
  • Author Footnotes
    * This work was supported by grants from the Shriners Hospitals for Crippled Children (to L. Y. S. and D. R. K.), National Institutes of Health Grant AR46811 (to L. Y. S.), and Deutsche Forschungsgemeinschaft Grants Re1021/3-1 and 4-1 (to D. P. R.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
    ‡ Current address: Dept. of Cell Biology, University of Virginia, Box 800732 Health Sciences Center, Charlottesville, VA 22908.
    § Current address: Dept. of Medical Molecular Biology, University of Lübeck, D-23538 Lübeck, Germany.
Open AccessPublished:November 11, 2002DOI:https://doi.org/10.1074/jbc.M209201200
      Fibrillins are microfibril-forming extracellular matrix macromolecules that modulate skeletal development. In humans, mutations in fibrillins result in long bone overgrowth as well as other distinct phenotypes. Whether fibrillins form independent microfibrillar networks or can co-polymerize, forming a single microfibril, is not known. However, this knowledge is required to determine whether phenotypes arise because of loss of singular or composite functions of fibrillins. Immunolocalization experiments using tissues and de novo matrices elaborated by cultured cells demonstrated that both fibrillins can be present in the same individual microfibril in certain tissues and that both fibrillins can co-polymerize in fibroblast cultures. These studies suggest that the molecular information directing fibrillin fibril formation may be similar in both fibrillins. Furthermore, these studies provide a molecular basis for compensation of one fibrillin by the other during fetal life. In postnatal tissues, fibrillin-2 antibodies demonstrated exuberant staining in only one location: peripheral nerves. This surprising finding implicates distinct functions for fibrillin-2 in peripheral nerves, because a unique feature in humans and in mice mutant for fibrillin-2 is joint contractures that resolve over time.
      mAb
      monoclonal antibody
      CCA
      congenital contractural arachnodactyly
      ELISA
      enzyme-linked immunosorbent assay
      PBS
      phosphate-buffered saline
      TBS
      Tris-buffered saline
      pAb
      polyclonal antibody
      FITC
      fluorescein isothiocyanate
      EGF
      epidermal growth factor
      TRITC
      tetramethylrhodamine isothiocyanate
      Fibrillin-1 and fibrillin-2 are extracellular matrix proteins with highly homologous structures (
      • Corson G.M.
      • Chalberg S.C.
      • Dietz H.C.
      • Charbonneau N.L.
      • Sakai L.Y.
      ,
      • Zhang H.
      • Apfelroth S.D., Hu, W.
      • Davis E.C.
      • Sanguineti C.
      • Bonadio J.
      • Mecham R.P.
      • Ramirez F.
      ). Both molecules contribute to ultrastructurally identifiable fibrils, called microfibrils, present in a variety of connective tissues (
      • Zhang H.
      • Apfelroth S.D., Hu, W.
      • Davis E.C.
      • Sanguineti C.
      • Bonadio J.
      • Mecham R.P.
      • Ramirez F.
      ,
      • Sakai L.Y.
      • Keene D.R.
      • Engvall E.
      ). However, because the expression of fibrillin-2 corresponds primarily to early morphogenesis, it has been proposed that fibrillin-2 regulates the early process of elastic fiber assembly, whereas fibrillin-1 provides mostly force-bearing structural support (
      • Zhang H., Hu, W.
      • Ramirez F.
      ).
      Fibrillin-2 protein synthesis has not been well documented. Fibrillin-2, first discovered by gene cloning (
      • Lee B.
      • Godfrey M.
      • Vitale E.
      • Hori H.
      • Mattei M-G.
      • Sarfarazi M.
      • Tsipouras P.
      • Ramirez F.
      • Hollister D.W.
      ), cannot be distinguished from fibrillin-1 by migration on SDS-PAGE. However, anti-peptide antibodies have been utilized for limited tissue immunolocalization studies in fetal human (
      • Zhang H.
      • Apfelroth S.D., Hu, W.
      • Davis E.C.
      • Sanguineti C.
      • Bonadio J.
      • Mecham R.P.
      • Ramirez F.
      ), fetal mouse (
      • Zhang H., Hu, W.
      • Ramirez F.
      ), fetal bovine (
      • Mariencheck M.C.
      • Davis E.C.
      • Zhang H.
      • Ramirez F.
      • Rosenbloom J.
      • Gibson M.A.
      • Parks W.C.
      • Mecham R.P.
      ), and fetal rat (
      • Yang Q.
      • Ota K.
      • Tian Y.
      • Kumar A.
      • Wada J.
      • Kashihara N.
      • Wallner E.
      • Kanwar Y.S.
      ). A monoclonal antibody (mAb),1 JB3, has localized fibrillin-2 in the developing heart and primary axial structures of early avian embryos (
      • Wunsch A.M.
      • Little C.D.
      • Markwald R.R.
      ,
      • Rongish B.J.
      • Drake C.J.
      • Argraves W.S.
      • Little C.D.
      ). MAb 201, specific for fibrillin-1 (
      • Sakai L.Y.
      • Keene D.R.
      • Engvall E.
      ,
      • Reinhardt D.P.
      • Keene D.R.
      • Corson G.M.
      • Poschl E.
      • Bächinger H.P.
      • Gambee J.E.
      • Sakai L.Y.
      ,
      • Keene D.R.
      • Jordan C.D.
      • Reinhardt D.P.
      • Ridgway C.C.
      • Ono R.N.
      • Corson G.M.
      • Fairhurst M.
      • Sussman M.D.
      • Memoli V.A.
      • Sakai L.Y.
      ), demonstrated fibrillin-1 in and around Hensen's node and defined the primary axis of the early avian embryo (
      • Gallagher B.C.
      • Sakai L.Y.
      • Little C.D.
      ). These studies have indicated that, in early embryonic and fetal tissues, the distribution of the two fibrillins is similar, with preferential accumulation of fibrillin-2 in elastic fiber-rich tissues.
      Mutations in the fibrillin genes result in two related inherited diseases of connective tissue. Major phenotypic features of the Marfan syndrome (OMIM number 154700), caused by mutations in FBN1, appear in cardiovascular (aortic dilatation or aortic dissection), skeletal (long bone overgrowth), and ocular (ectopia lentis) tissues. The phenotype of congenital contractural arachnodactyly (CCA) (OMIM number 121050), caused by mutations in FBN2, overlaps with some of the skeletal features (arachnodactyly and scoliosis) of the Marfan syndrome, but cardiovascular and ocular features are usually absent. Together with expression studies, these genetic data suggest that fibrillin-2 performs a function comparable with fibrillin-1 in skeletal but not in other connective tissues. Therefore, expression of fibrillin-2 in postnatal tissues is expected to be restricted to skeletal connective tissues.
      Fibrillin-2 null mice (
      • Arteaga-Solis E.
      • Gayraud B.
      • Lee S.Y.
      • Shum L.
      • Sakai L.
      • Ramirez F.
      ) are viable and fertile. However, they are born with contractures of the small and large joints, which resolve a few days after birth. This common phenotype, joint contractures, demonstrated that loss of fibrillin-2 in homozygous fbn2-null mice is equivalent to the dominant-negative effects of mutations in FBN2 in heterozygous individuals with CCA. Similarly, mice deficient in fibrillin-1 display cardiovascular and skeletal features similar to those of the Marfan syndrome (
      • Pereira L.
      • Lee S.Y.
      • Gayraud B.
      • Andrikopoulos K.
      • Shapiro S.D.
      • Bunton T.
      • Biery N.J.
      • Dietz H.C.
      • Sakai L.Y.
      • Ramirez F.
      ), indicating that loss of fibrillin-1 is equivalent to the dominant-negative effects of heterozygous FBN1 mutations.
      Neither fibrillin-1 deficiency nor fibrillin-2 deficiency results in early embryonic lethality, even though both fibrillins are present in the developing embryo at gastrulation. In addition to contractures, mice lacking fibrillin-2 display a limb patterning defect, bilateral syndactyly (
      • Arteaga-Solis E.
      • Gayraud B.
      • Lee S.Y.
      • Shum L.
      • Sakai L.
      • Ramirez F.
      ). However, no other obvious early fetal or embryonic phenotype has been found, despite the restriction of fibrillin-2 expression primarily to developing tissues. And, there are also no apparent abnormalities in elastic fiber assembly or morphology, even though preferential distribution to elastic fiber-rich matrices had been originally noted in expression studies.
      These genetic investigations in humans and mice suggest that fibrillins perform both overlapping and unique functions. However, biochemical and morphological studies of the fibrillins, especially fibrillin-2, have not elucidated the distinctive or shared functional contributions of the fibrillins to microfibril assembly, to elastic fiber assembly and specific tissue architectures, or to cellular differentiation. The major obvious function of the fibrillins is microfibril assembly, and yet it is not known whether both fibrillins polymerize separately to form distinct independent microfibrils or co-polymerize to form microfibril heteropolymers. The unique functions of fibrillin-2, which result in joint contractures and bilateral syndactyly when lost, are not yet understood.
      To begin to resolve these issues, detailed analyses of fibrillin-2 in developing and postnatal tissues are required. The investigations reported here demonstrate fibrillin-2 immunolocalization in fetal and postnatal human tissues, compare fibrillin-1 and fibrillin-2 protein synthesis and fibril formation in various cell types, show that both fibrillins can be found in the same microfibril, and identify cryptic epitopes in fibrillin-2, which may indicate domains important for fibrillin-2 assembly into microfibrils or binding to ligands. Of most significance, these studies suggest that the molecular information directing fibrillin fibril formation may be similar in both fibrillins and thus provide a molecular basis for compensation of one fibrillin by the other during fetal life. In addition, localization of fibrillin-2 to postnatal peripheral nerves implicates a unique function for fibrillin-2 in peripheral nerves.

      DISCUSSION

      Microfibrils are polymers of multiple individual molecules of fibrillin. The fibrillin composition of these polymers has been unspecified so far. In theory, microfibrils could be 1) heteropolymers of both fibrillins; 2) distinct fibrillin homopolymers with similar tissue distribution and bundled together into heteropolymeric microfibrillar fibers; or 3) distinct fibrillin homopolymers with distinct tissue distributions. Each of these architectural possibilities carries important functional implications for understanding fibrillin pathophysiology.
      In postnatal tissues, where fibrillin-2 is largely absent, microfibrils are likely fibrillin-1 homopolymers. In the cell culture studies reported here, certain cell lines assembled fibrillin-1 fibrils that did not contain fibrillin-2 (Table I; Fig. 9, G andH). In other cell lines, fibrillin-2 fibrils were assembled that did not appear to contain fibrillin-1 (Fig. 9, E andF). In Xenopus tadpoles (around stage 40), microfibrils in the skin and around trunk skeletal muscle contained fibrillin-2, but not fibrillin-1 (Fig. 5 E). These data indicate that both fibrillin-1 and fibrillin-2 can form independent fibril polymers. As independent fibrils, fibrillins can be expected to perform distinct functions.
      Examination of fetal bovine and human tissues (digits, limbs, and axial skeletal tissues; ocular tissues; vascular and lung tissues) collected midway through gestation revealed no differences between fibrillin-1 and fibrillin-2 in tissue distribution or in the morphological patterns of distribution. However, temporal differences in the distribution of fibrillin-1 and fibrillin-2 were detected in lung, a finding consistent with previously published data (
      • Zhang H., Hu, W.
      • Ramirez F.
      ). A study of early fetal human tissues (5–12 weeks gestation) also found equivalent distributions of fibrillin-1 and fibrillin-2 in skin, central nervous system and meninges, heart and aorta, lung, and peripheral nervous system, but identified distinct differences in developing liver, kidney, axial skeleton, and ribs (

      Quandamatteo, F., Reinhardt, D. P., Pophal, G., Sakai, L. Y., and Herken, R. (2003) Matrix Biol., in press

      ). These tissue distribution studies indicate that fibrillins can be co-distributed as well as uniquely distributed, consistent with the notion that fibrillins can form distinct fibril polymers. However, these studies cannot rule in or rule out the formation of heteropolymeric fibrillin fibrils.
      To address whether individual microfibrils can be assembled from molecules of fibrillin-1 and fibrillin-2, double labeling experiments were performed using fetal tissues as well as fibroblast cell cultures. In tissues where fibrillins appeared to be co-distributed (fetal skin, ligament, and perichondrium), extensive immunoelectron microscopic analyses revealed microfibrils that were composed of both fibrillins. Results from fibroblast cell cultures, which allow visualization of the de novo formation of fibrillin fibrils, indicated co-polymerization of fibrillin-1 and fibrillin-2. In these latter studies, complete co-localization was demonstrated. These data together with the immunoelectron microscopic studies of microfibrils in tissues support the hypothesis that fibrillins can form heteropolymeric microfibrils.
      Co-cultures of adult dermal fibroblasts and 293 cells also indicated that fibrillin-1 and fibrillin-2 can co-assemble fibrils. In addition, these experiments showed that exogenous fibrillin-2 was co-assembled by the fibroblast as easily as the fibroblast-secreted fibrillin-1. Co-cultures utilizing transfected 293 cells identified regions in fibrillins that can inhibit fibril formation. N-terminal domains of both fibrillins were able to inhibit the formation of fibrillin-1 fibrils, whereas other domains did not affect fibrillin-1 fibril formation. These data confirm that fibrillin-1 and fibrillin-2 can co-polymerize. In addition, these data implicate N-terminal domains as crucial interactive sites in fibrillin fibril assembly and suggest the hypothesis that fibrillin-1 and fibrillin-2 may utilize homologous domains to mediate fibril assembly. We are currently using this approach to exactly specify these interactive sites and test this hypothesis.
      Until now, demonstration of fibrillin fibrils produced by cells in culture has been limited primarily to dermal fibroblasts. However, some cells, like epithelial cells, deposit fibrillin into a nonfibrillar matrix (described as “pericellular” or “punctate” in Table I) and may do so because they lack cellular factors or receptors required for fibrillin fibril formation (
      • Dzamba B.J.
      • Keene D.R.
      • Isogai Z.
      • Charbonneau N.L.
      • Juruskovska N.K.
      • Simon M.
      • Sakai L.Y.
      ). These cellular factors or receptors may be specific for fibrillin-1 or fibrillin-2. For example, SW1353 chondrosarcoma cells, which can assemble fibrillin-1 into fibrils, may lack cellular factors required for assembly of fibrillin-2 into fibrils. Thus, in contrast to fibroblasts, other cell types may display variable incorporation of fibrillins into fibrils or nonfibrillar matrices, consistent with an emerging hypothesis that fibrillin fibril formation is a cell-regulated process that can distinguish between the fibrillins.
      Fibrillin-2 contains two RGD sites, whereas fibrillin-1 contains one RGD sequence. Integrin receptors for fibrillin-1 include αvβ3 (
      • Pfaff M.
      • Reinhardt D.P.
      • Sakai L.Y.
      • Timpl R.
      ,
      • Sakamoto H.
      • Broekelmann T.
      • Cheresh D.A.
      • Ramirez F.
      • Rosenbloom J.
      • Mecham R.P.
      ). It is unknown whether integrins or other cell surface molecules are involved in the fibrillin fibril assembly process. However, recently the importance of heparan-sulfate proteoglycans to fibrillin assembly has been demonstrated (
      • Tiedemann K.
      • Batge B.
      • Müller P.K.
      • Reinhardt D.P.
      ). Our data support the hypothesis that cellular factors are required for fibrillin fibril assembly and furthermore, suggest that these cellular factors may interact differentially with fibrillin-1 and fibrillin-2. Thus, whereas fibrillins contain homologous functional domains for fibril formation, cells can be expected to perform important roles in the determination of fibril architecture in tissues.
      All mAbs generated to the recombinant fibrillin-2 polypeptide, rF37, also recognized fibrillin-2 secreted into medium by cells. This, together with the fact that mAb epitopes are dependent on intact disulfide bonds, demonstrates that the mAbs recognize naturally occurring epitopes in fibrillin-2. Although all of the mAbs immunoblotted authentic fibrillin-2 secreted into fibroblast medium, only mAbs 48 and 143 recognized fibrillin-2 in the fibroblast matrix. In fibroblast cultures, no immunofluorescence (neither fibrillar, punctate, pericellular, nor diffuse) was detected with the other antibodies. These results indicate that certain regions of fibrillin-2, which are available in the monomeric molecule, become cryptic when fibrillin-2 is deposited into the extracellular matrix. In fibroblast cultures, the process whereby these regions are made cryptic occurs immediately upon deposition into the matrix. Because assembly of fibrillin occurs quickly in the extracellular space (
      • Reinhardt D.P.
      • Gambee J.
      • Ono R.N.
      • Bächinger H.P.
      • Sakai L.Y.
      ), these cryptic epitopes may implicate certain domains in the assembly of fibrillin-2.
      Mapping the epitopes of monoclonal antibodies revealed that the 8Cys1/Gly-rich domains in fibrillin-2 are fully accessible in microfibrils, whereas other flanking domains (cbEGF 2 and EGF 4) are not available for binding. Similar studies using mAbs specific for fibrillin-1 have demonstrated that 8Cys1 in fibrillin-1 (recognized by mAb 26) is also fully accessible in tissue microfibrils. In contrast to cbEGF2 in fibrillin-2, cbEGF2 in fibrillin-1 (recognized by mAb 78) is available in tissue microfibrils (data not shown). The region (represented by rF50) around the second hybrid domain, recognized by mAbs 60, 72, 139, and 205, is also (predominantly) cryptic in tissues and fibroblast matrices. Comparable fine mapping by antibodies has not been performed in fibrillin-1.
      Our identification of domains containing cryptic epitopes is consistent with other studies of microfibril assembly. Domains around the proline-rich region of fibrillin-1 and the glycine-rich region of fibrillin-2 may direct fibrillin dimer formation (
      • Ashworth J.L.
      • Kelly V.
      • Wilson R.
      • Shuttleworth C.A.
      • Kielty C.M.
      ,
      • Trask T.M.
      • Ritty T.M.
      • Broekelmann T.
      • Tisdale C.
      • Mecham R.P.
      ). Cryptic epitopes may also result from interactions with proteins other than fibrillin. A binding site for decorin has been located near the proline-rich region of fibrillin-1, but comparable interactions near the glycine-rich region of fibrillin-2 are absent (
      • Trask B.C.
      • Trask T.M.
      • Broekelmann T.
      • Mecham R.P.
      ). The same segments between the proline-rich/glycine-rich region and the second 8 Cys domain have been reported to also mediate binding between elastin and fibrillins (
      • Trask T.M.
      • Trask B.C.
      • Ritty T.M.
      • Abrams W.R.
      • Rosenbloom J.
      • Mecham R.P.
      ). A binding site for fibulin-2 has been identified in the N-terminal region of fibrillin-1 (
      • Reinhardt D.P.
      • Sasaki T.
      • Dzamba B.J.
      • Keene D.R.
      • Chu M-L.
      • Gohring W.
      • Timpl R.
      • Sakai L.Y.
      ). No interactions have yet been reported to occur around the region of the second hybrid domain in either fibrillin. Identification of these cryptic epitopes may therefore implicate domains contained in rF50 in as yet unknown interactions.
      The postnatal expression and tissue distribution of fibrillin-2 has not been described previously. Because mutations in FBN2 result in CCA, it has been assumed that fibrillin-2 is present in affected skeletal tissues. Major features of the Marfan skeletal phenotype (long bone overgrowth accompanied by scoliosis and chest deformities) are also common in CCA, although these features are usually less severe in CCA. These genetic similarities suggest that fibrillin-2 may perform a role comparable with that of fibrillin-1 in the perichondrium and growth plate cartilage (
      • Keene D.R.
      • Jordan C.D.
      • Reinhardt D.P.
      • Ridgway C.C.
      • Ono R.N.
      • Corson G.M.
      • Fairhurst M.
      • Sussman M.D.
      • Memoli V.A.
      • Sakai L.Y.
      ). Fibrillin-1 and fibrillin-2 may impact growth of the skeleton during fetal stages when both molecules are abundant in skeletal tissues, as well as during postnatal growth when fibrillin-1 appears to be more plentiful than fibrillin-2. In this investigation, we found that, in contrast to weak expression in tissues like skin, tendon, and perichondrium, fibrillin-2 is abundant around peripheral nerves. This localization of fibrillin-2 to peripheral nerves may explain why individuals with mutations in FBN2 are affected by joint contractures that resolve over time. Congenital contractures can be caused by neuropathic abnormalities affecting peripheral nerves (
      • Hall J.G.
      ). If the fate of fibrillin-2 is to be lost with age in nerves, then one might expect the effects of mutant fibrillin-2 to be lost with time, if these effects do not result in persistent structural changes in the joint.
      Based upon studies in mice (
      • Pereira L.
      • Lee S.Y.
      • Gayraud B.
      • Andrikopoulos K.
      • Shapiro S.D.
      • Bunton T.
      • Biery N.J.
      • Dietz H.C.
      • Sakai L.Y.
      • Ramirez F.
      ), it has been proposed that aortic aneurysm and scoliosis develop when normal levels of fibrillin dip below a critical threshold. If microfibrils in certain tissues (like perichondrium) are heteropolymers of fibrillin-1 and fibrillin-2, then dominant-negative effects of a mutation in either fibrillin could result in reducing the level of fibrillin microfibrils and in similar phenotypes. This could be the case in skeletal tissues, because overlapping skeletal phenotypes occur in Marfan syndrome and CCA. In contrast, if microfibrils are distinct homopolymers, then it might be expected that loss of one fibrillin would lead to the same phenotypic features resulting from a dominant-negative allele. However, loss of fibrillin-2 in mice does not seem to result in long bone overgrowth and scoliosis (
      • Arteaga-Solis E.
      • Gayraud B.
      • Lee S.Y.
      • Shum L.
      • Sakai L.
      • Ramirez F.
      ), suggesting that these features in CCA are rather because of dominant-negative effects of fibrillin-2 on heteropolymeric fibrils containing fibrillin-1 and fibrillin-2. Hence, it can now be hypothesized that functions which, when lost, result in long bone overgrowth may be ascribed primarily to fibrillin-1 or to composite functions performed by both fibrillins within a heteropolymeric context.
      In peripheral nerves, both fibrillins are present during fetal development and in early postnatal life. And yet, mutations in human fibrillin-2 and loss of fibrillin-2 in mice result in joint contractures, whereas only certain severe cases of neonatal Marfan syndrome display joint contractures. This might suggest that in peripheral nerves, fibrillin-2 performs a distinct function, implying that microfibrils are composed primarily of fibrillin-2 polymers. In the future, tissue-specific variations in the fibrillin composition of microfibrils (homopolymers, heteropolymers, or combinations of both) should be elucidated and related to distinct or composite fibrillin functions.

      Acknowledgments

      We thank Dr. Susanne Reber-Mueller for help with the Northern analyses, Dr. Francesco Ramirez for the 3′A FBN2 plasmid, Dr. Jan Christian at Oregon Health and Science University for providing Xenopus tadpoles, and Sara Tufa for outstanding technical assistance.

      REFERENCES

        • Corson G.M.
        • Chalberg S.C.
        • Dietz H.C.
        • Charbonneau N.L.
        • Sakai L.Y.
        Genomics. 1993; 17: 476-484
        • Zhang H.
        • Apfelroth S.D., Hu, W.
        • Davis E.C.
        • Sanguineti C.
        • Bonadio J.
        • Mecham R.P.
        • Ramirez F.
        J. Cell Biol. 1994; 124: 855-863
        • Sakai L.Y.
        • Keene D.R.
        • Engvall E.
        J. Cell Biol. 1986; 103: 2499-2509
        • Zhang H., Hu, W.
        • Ramirez F.
        J. Cell Biol. 1995; 129: 1165-1176
        • Lee B.
        • Godfrey M.
        • Vitale E.
        • Hori H.
        • Mattei M-G.
        • Sarfarazi M.
        • Tsipouras P.
        • Ramirez F.
        • Hollister D.W.
        Nature. 1991; 352: 330-334
        • Mariencheck M.C.
        • Davis E.C.
        • Zhang H.
        • Ramirez F.
        • Rosenbloom J.
        • Gibson M.A.
        • Parks W.C.
        • Mecham R.P.
        Connect. Tissue Res. 1995; 31: 87-97
        • Yang Q.
        • Ota K.
        • Tian Y.
        • Kumar A.
        • Wada J.
        • Kashihara N.
        • Wallner E.
        • Kanwar Y.S.
        Dev. Biol. 1999; 212: 229-242
        • Wunsch A.M.
        • Little C.D.
        • Markwald R.R.
        Dev. Biol. 1994; 165: 585-601
        • Rongish B.J.
        • Drake C.J.
        • Argraves W.S.
        • Little C.D.
        Dev. Dyn. 1998; 212: 461-471
        • Reinhardt D.P.
        • Keene D.R.
        • Corson G.M.
        • Poschl E.
        • Bächinger H.P.
        • Gambee J.E.
        • Sakai L.Y.
        J. Mol. Biol. 1996; 258: 104-116
        • Keene D.R.
        • Jordan C.D.
        • Reinhardt D.P.
        • Ridgway C.C.
        • Ono R.N.
        • Corson G.M.
        • Fairhurst M.
        • Sussman M.D.
        • Memoli V.A.
        • Sakai L.Y.
        J. Histochem. Cytochem. 1997; 45: 1069-1082
        • Gallagher B.C.
        • Sakai L.Y.
        • Little C.D.
        Dev. Dyn. 1993; 196: 70-78
        • Arteaga-Solis E.
        • Gayraud B.
        • Lee S.Y.
        • Shum L.
        • Sakai L.
        • Ramirez F.
        J. Cell Biol. 2001; 154: 275-281
        • Pereira L.
        • Lee S.Y.
        • Gayraud B.
        • Andrikopoulos K.
        • Shapiro S.D.
        • Bunton T.
        • Biery N.J.
        • Dietz H.C.
        • Sakai L.Y.
        • Ramirez F.
        Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 3819-3823
        • Mayer U.
        • Poschl E.
        • Gerecke D.R.
        • Wagman D.W.
        • Burgeson R.E.
        • Timpl R.
        FEBS Lett. 1995; 365: 129-132
        • Reinhardt D.P.
        • Sasaki T.
        • Dzamba B.J.
        • Keene D.R.
        • Chu M-L.
        • Gohring W.
        • Timpl R.
        • Sakai L.Y.
        J. Biol. Chem. 1996; 271: 19489-19496
        • Sakai L.Y.
        • Keene D.R.
        Methods Enzymol. 1994; 245: 29-52
        • Dzamba B.J.
        • Keene D.R.
        • Isogai Z.
        • Charbonneau N.L.
        • Juruskovska N.K.
        • Simon M.
        • Sakai L.Y.
        J. Invest. Dermatol. 2001; 117: 1612-1620
        • Hollister D.W.
        • Godfrey M.
        • Sakai L.Y.
        • Pyeritz R.E.
        N. Engl. J. Med. 1990; 323: 152-159
      1. Quandamatteo, F., Reinhardt, D. P., Pophal, G., Sakai, L. Y., and Herken, R. (2003) Matrix Biol., in press

        • Pfaff M.
        • Reinhardt D.P.
        • Sakai L.Y.
        • Timpl R.
        FEBS Lett. 1996; 384: 247-250
        • Sakamoto H.
        • Broekelmann T.
        • Cheresh D.A.
        • Ramirez F.
        • Rosenbloom J.
        • Mecham R.P.
        J. Biol. Chem. 1996; 271: 4916-4922
        • Tiedemann K.
        • Batge B.
        • Müller P.K.
        • Reinhardt D.P.
        J. Biol. Chem. 2001; 276: 36035-36042
        • Reinhardt D.P.
        • Gambee J.
        • Ono R.N.
        • Bächinger H.P.
        • Sakai L.Y.
        J. Biol. Chem. 2000; 275: 2205-2210
        • Ashworth J.L.
        • Kelly V.
        • Wilson R.
        • Shuttleworth C.A.
        • Kielty C.M.
        J. Cell Sci. 1999; 112: 3549-3558
        • Trask T.M.
        • Ritty T.M.
        • Broekelmann T.
        • Tisdale C.
        • Mecham R.P.
        Biochem. J. 1999; 340: 693-701
        • Trask B.C.
        • Trask T.M.
        • Broekelmann T.
        • Mecham R.P.
        Mol. Biol. Cell. 2000; 11: 1499-1507
        • Trask T.M.
        • Trask B.C.
        • Ritty T.M.
        • Abrams W.R.
        • Rosenbloom J.
        • Mecham R.P.
        J. Biol. Chem. 2000; 275: 24400-24406
        • Hall J.G.
        J. Pediatr. Orthop. B. 1997; 6: 159-166
        • Handford P.A.
        • Downing A.K.
        • Reinhardt D.P.
        • Sakai L.Y.
        Matrix Biol. 2000; 19: 457-470