Mechanistic insights into the cellular effects of a novel FN1 variant associated with a spondylometaphyseal dysplasia

Spondylometaphyseal dysplasia (SMD) is characterized by developmental changes in long bones and vertebrae. It has large phenotypic diversity and multiple genetic causes, including a recent link to novel variants in the extracellular matrix (ECM) protein fibronectin (FN), a regulator of ECM assembly and key link between the ECM and proper cell function. We identified a patient with a unique SMD, similar to SMD with corner fractures. The patient has been followed over 19 years and presents with short stature, genu varum, kyphoscoliosis, and pectus carinatum. Radiography shows metaphyseal changes that resolved over time, vertebral changes, and capitular avascular necrosis. Whole exome sequencing identified a novel heterozygous FN1 variant (p.Cys97Trp). Using mass spectroscopy, mutant FN was detected in plasma and in culture medium of primary dermal fibroblasts isolated from the patient, but mutant protein was much less abundant than wild-type FN. Immunofluorescence and immunoblotting analyses show that mutant fibroblasts assemble significantly lower amounts of FN matrix than wild-type cells, and mutant FN was preferentially retained within the endoplasmic reticulum. This work highlights the importance of FN in skeletal development, and its potential role in the pathogenesis of a subtype of SMD.

Mammary epithelial cell interactions with fibronectin stimulate epithelial-mesenchymal transition

In the mammary gland, the stromal extracellular matrix (ECM) undergoes dramatic changes during development and in tumorigenesis. For example, normal adult breast tissue is largely devoid of the ECM protein fibronectin (FN) whereas high FN levels have been detected in the stroma of breast tumors. FN is an established marker for epithelial-mesenchymal transition (EMT), which occurs during development and has been linked to cancer. During EMT, epithelial cell adhesion switches from cell-cell contacts to mainly cell-ECM interactions raising the possibility that FN may have a role in promoting this transition. Using MCF-10A mammary epithelial cells, we show that exposure to exogenous FN induces an EMT response including up-regulation of the EMT markers FN, Snail, N-cadherin, vimentin, the matrix metalloprotease MMP2, α-smooth muscle actin, and phospho-Smad2 as well as acquisition of cell migratory behavior. FN-induced EMT depends on Src kinase and ERK/MAP kinase signaling but not on the immediate early gene EGR-1. FN initiates EMT under serum-free conditions; this response is partially reversed by a TGFβ neutralizing antibody suggesting that FN enhances the effect of endogenous TGFβ. EMT marker expression is up-regulated in cells on a fragment of FN containing the integrin-binding domain but not other domains. Differences in gene expression between FN and MG are maintained with addition of a sub-threshold level of TGFβ1. Together, these results show that cells interacting with FN are primed to respond to TGFβ. The ability of FN to induce EMT shows an active role for the stromal ECM in this process and supports the notion that the increased levels of FN observed in breast tumors facilitate tumorigenesis.

Fibronectin-fibrin cross-linking: a regulator of cell behavior

At sites of tissue injury or inflammation, extravasation of plasma proteins leads to the formation of a complex fibrillar matrix composed primarily of fibrin and plasma fibronectin (pFN). This protein meshwork serves not only to reestablish the integrity of the vascular system but also to provide a scaffold for cell migration and subsequent wound repair. The interactions between cell surface receptors and this provisional extracellular matrix (ECM) provide important cues that can modulate the cellular response at the injury site, leading to alterations in cell growth and gene expression. Key determinants of this response may lie in the structure and composition of this "injury-associated" ECM.

Roles for beta(pat-3) integrins in development and function of Caenorhabditis elegans muscles and gonads

Heterodimeric integrin receptors for extracellular matrix (ECM) play vital roles in bidirectional signaling during tissue development, organization, remodeling, and repair. The beta integrin subunit cytoplasmic domain is essential for transmission of many of these signals and overexpression of an unpaired beta tail in cultured cells inhibits endogenous integrins. Unlike vertebrates, which have at least nine beta subunit genes, the nematode Caenorhabditis elegans expresses only one beta subunit (betapat-3), and a null mutation in this gene causes embryonic lethality. To determine the functions of integrins during larval development and in adult tissues, we have taken a dominant negative approach by expression of an HA-betatail transgene composed of a hemagglutinin (HA) epitope tag extracellular domain connected to the betapat-3 transmembrane and cytoplasmic domains. Expression of this transgene in muscle and gonad, major sites of integrin expression, caused a variety of phenotypes dependent on the level of transgene expression. Abnormalities in body wall and sex muscles led to uncoordinated movement and egg-laying defects. Significant anomalies in migration and pathfinding were caused by tissue-specific expression of HA-betatail in the distal tip cells (DTC), the cells that direct gonad morphogenesis. A pat-3 gene with Tyr to Phe mutations in the cytoplasmic domain was able to rescue pat-3 null animals but also showed DTC migration defects. These results show that betapat-3 plays important roles in post-embryonic organogenesis and tissue function.

A novel fibronectin binding site required for fibronectin fibril growth during matrix assembly

Fibronectin (FN) assembly into a fibrillar extracellular matrix is a stepwise process requiring participation from multiple FN domains. Fibril formation is regulated in part by segments within the first seven type III repeats (III1-7). To define the specific function(s) of this region, recombinant FNs (recFNs) containing an overlapping set of deletions were tested for the ability to assemble into fibrils. Surprisingly, recFN lacking type III repeat III1 (FNDeltaIII1), which contains a cryptic FN binding site and has been suggested to be essential for fibril assembly, formed a matrix identical in all respects to a native FN matrix. Similarly, displacement of the cell binding domain in repeats III9-10 to a position close to the NH2-terminal assembly domain, as well as a large deletion spanning repeats III4-7, had no effect on assembly. In contrast, two deletions that included repeat III2, DeltaIII1-2 and DeltaIII2-5, caused significant reductions in fibril elongation, although binding of FN to the cell surface and initiation of assembly still proceeded. Using individual repeats in binding assays, we show that III2 but not III1 contains an FN binding site. Thus, these results pinpoint repeat III2 as an important module for FN-FN interactions during fibril growth.

Regulation of fibronectin matrix assembly by activated Ras in transformed cells

Fibronectin extracellular matrix plays a critical role in the microenvironment of cells. Loss of this matrix frequently accompanies oncogenic transformation, allowing changes in cell growth, morphology, and tissue organization. The HT1080 human fibrosarcoma cell line is deficient in formation of fibronectin matrix fibrils but assembly can be induced by the glucocorticoid dexamethasone. Here we show that fibronectin assembly can also be restored by stimulation of alpha5beta1 integrin with activating antibody or with Mn2+ suggesting that integrin activity is reduced in these cells. While dexamethasone promoted actin stress fiber formation, actin filaments remained cortical following Mn2+ treatment showing that the dexamethasone effect is not due solely to cytoskeletal changes. HT1080 cells have one activated allele of N-ras and PD98059 inhibition of signaling from Ras through ERK increased fibronectin matrix accumulation. Conversely, the p38 MAP kinase inhibitor SB203580 blocked induction of matrix and increased ERK phosphorylation. Thus, two MAP kinase pathways contribute to the control of integrin-mediated fibronectin assembly. ERK activity and fibronectin assembly were linked in three different ras-transformed cell lines but not in SV40- or RSV-transformed cells indicating that oncogenic Ras uses a distinct mechanism to down-regulate cell-fibronectin interactions.

A novel RGD-independent fibronectin assembly pathway initiated by alpha4beta1 integrin binding to the alternatively spliced V region

Fibronectin (FN) matrix assembly is a multi-step process that involves binding to integrin receptors, FN-FN interactions and connections to the actin cytoskeleton. Ultimately, FN is converted into stable matrix fibrils that are detergent-insoluble. RGD-binding integrins such as alpha5beta1 play a major role in the assembly of fibrillar FN. Here we show that alpha4beta1 binding to the alternatively spliced V (IIICS) region of FN initiates an alternative assembly pathway. Activation of alpha4beta1 with exogenous agents such as Mn(2+) or a beta1-stimulatory antibody TS2/16 was sufficient to induce initiation of FN fibrillogenesis by Ramos B lymphoma cells and by CHO(B2)alpha4 cells. Using recombinant FNs lacking specific sequences, we show that assembly is independent of the RGD sequence but requires the V25/CS-1 segment. Previously, we have characterized an activated recombinant FN (FN III(1–7)) that rapidly forms detergent-insoluble multimers upon binding to alpha5beta1 integrin. Alpha4beta1 also formed FNdeltaIII(1–7) multimers without the aid of exogenous stimulants, suggesting that an activated form of FN can override the need for activation of the integrin. In contrast to assembly by alpha5beta1, actin filaments remained largely cortical and no change in cell growth rate was observed with alpha4beta1-mediated assembly. These results show that binding sites on FN other than the RGD sequence/synergy site and distant from the cell binding domain can promote FN assembly. Thus, there appear to be multiple, integrin-specific mechanisms for assembly of FN matrix.

Fibronectin matrix composition and organization can regulate cell migration during amphibian development

Fibronectin (FN) is an adhesive extracellular matrix component that is essential for vertebrate development. It forms a fibrillar matrix at the cell surface which controls cell morphology, migration, proliferation, and other important cellular processes. To address specific functions of FN matrix structure during early vertebrate development, we introduced normal and mutant recombinant FNs (recFNs) into the blastocoel cavity of embryos of the amphibian Pleurodeles waltl. Here we show that a native recFN FN(A-B-) as well as recFNs with specific mutations in the cell-binding domain, FN(RGD-) and FN(syn-), or in a FN-binding region, FNDeltaIII(1), are assembled into fibrillar matrix. A recFN (FNDeltaIII(1-7)) that forms a structurally distinct matrix in cultured cells was assembled into aggregates at the cell periphery and was able to inhibit assembly of endogenous amphibian FN matrix in a dose-dependent manner. Cell adhesion, spreading, and migration were perturbed in vitro and in vivo on chimeric matrices containing FN(RGD-), FN(syn-), or FNDeltaIII(1-7) co-assembled with amphibian FN. Developmentally, this perturbation resulted in defects in mesoderm patterning and inhibition of gastrulation. These results indicate that FN matrix fibrillar structure and composition are important determinants of cell adhesion and migration during development.

Fibronectin fibrillogenesis: a paradigm for extracellular matrix assembly

Fibronectin matrix assembly is a regulated stepwise process. In the past year, analyses of fibronectin domains, integrin and cytoskeletal contributions, and fibril architecture have provided new insights into assembly mechanisms and matrix control of cell functions. Like fibronectin, laminin polymerization is cell-mediated. Thus a common pathway for extracellular matrix assembly is emerging.

The cartilage-specific (V+C)- fibronectin isoform exists primarily in homodimeric and monomeric configurations

Fibronectin is an extracellular-matrix glycoprotein encoded by a single gene, but with significant protein heterogeneity introduced through alternative RNA splicing and post-translational modifications. The (V+C)(-) splice variant, in which nucleotides encoding protein segments III-15 and I-10 are deleted along with the entire variable region, is unique in that expression is restricted to cartilaginous tissues. All known fibronectin splice variants retain the two C-terminal cysteine residues essential for dimerization, but cellular and/or structural constraints appear to influence homo- and heterodimerization patterns. Dimerization patterns of the (V+C)(-) isoform were studied under native conditions within canine articular cartilage and experimentally in COS-7, NIH-3T3 and CHO-K1 cell cultures. In all systems, (V+C)(-) fibronectin secretion was predominantly in a homodimeric configuration. Lower levels of (V+C)(-) monomers were also present. Heterodimers of (V+C)(-) with V(+),C(+) (V120) isoforms were not detected. Heterodimers of (V+C)(-) with V(-),C(+) (V0) subunits were detected only at low levels. Functional properties may differ significantly among monomers, homodimers and heterodimers. The unique dimerization pattern of (V+C)(-) fibronectin is consistent with this isoform having specialized functional properties in situ that are important for either the structural organization and biomechanical properties of cartilage matrix or regulation of a chondrocytic phenotype.

Requirements for alpha(5)beta(1) integrin-mediated retraction of fibronectin-fibrin matrices

Retraction of the blood clot by nucleated cells contributes both to hemostasis and to tissue remodeling. Although plasma fibronectin (FN) is a key component of the clot, its role in clot retraction is unclear. In this report, we demonstrate that the incorporation of FN into fibrin matrices significantly improves clot retraction by nucleated cells expressing the integrin alpha(5)beta(1). Further, we show that FN-fibrin clots support increased cell spreading when compared with fibrin matrices. To determine the structural requirements for FN in this process, recombinant FN monomers deficient in ligand binding or fibrin cross-linking were incorporated into fibrin clots. We show that recombinant FN monomers support clot retraction by Chinese hamster ovary cells expressing the integrin alpha(5)beta(1). This process depends on both the Arg-Gly-Asp (RGD) and the synergy cell-binding sites and on covalent FN-fibrin binding, demonstrating that cross-linking within the clot is important for cell-FN interactions. These data show that alpha(5)beta(1) can bind to FN within a clot to promote clot retraction and support cell shape change. This provides strong evidence that alpha(5)beta(1)-FN interactions may contribute to the cellular events required for wound contraction.

Beta3 integrin activation improves alphavbeta3-mediated retraction of fibrin matrices

BACKGROUND

Integrins are heterodimeric transmembrane glycoproteins that mediate cell interactions with the extracellular matrix. In vivo, integrin affinity can be modulated by intracellular signaling events. This can be simulated by a point mutation (D723R) in the cytoplasmic tail of the beta3 integrin subunit which results in constitutive activation. The effects of beta3 integrin activation on the function of alphavbeta3, an integrin which is important to the adhesive events of multiple cell types, were addressed using Chinese hamster ovary cells expressing either the wild-type alphavbeta3 integrin or the mutant alphavbeta3(D723R). The interactions of these cell lines with fibrin matrices were compared.

METHODS

Receptor expression levels were confirmed by FACS analyses using a monoclonal anti-alphavbeta3 antibody. Cell attachment to fibrin-coated dishes was determined after 1 h by fixation and crystal violet staining followed by elution of the dye and OD measurement. Fibrin clot retraction was measured by culturing cells in fibrin clots for 24 h. The clots were detached from the dish and the surface area was calculated at individual time points.

RESULTS

CHO alphavbeta3(D723R) cells displayed a greater than twofold increase in attachment to fibrinogen or to fibrin matrices when compared to wild-type transfectants. Further, CHO alphavbeta3(D723R) cell retraction of fibrin matrices was significantly greater at nearly all time points.

CONCLUSION

Activation of the beta3 integrin subunit significantly improves the interaction of alphavbeta3 with fibrin and may play a role in the integrin-mediated signaling events which occur following vascular injury.

Comparison of the fibrin-binding activities in the N- and C-termini of fibronectin

Fibronectin (Fn) binds to fibrin in clots by covalent and non-covalent interactions. The N- and C-termini of Fn each contain one non-covalent fibrin-binding site, which are composed of type 1 (F1) structural repeats. We have previously localized the N-terminal site to the fourth and fifth F1 repeats (4F1.5F1). In the current studies, using proteolytic and recombinant proteins representing both the N- and C-terminal fibrin-binding regions, we localized and characterized the C-terminal fibrin-binding site, compared the relative fibrin-binding activities of both sites and determined the contribution of each site to the fibrin-binding activity of intact Fn. By fibrin-affinity chromatography, a protein composed of the 10F1 repeat through to the C-terminus of Fn (10F1-COOH), expressed in COS-1 cells, and 10F1-12F1, produced in Saccharomyces cerevisiae, displayed fibrin-binding activity. However, since 10F1 and 10F1.11F1 were not active, the presence of 12F1 is required for fibrin binding. A proteolytic fragment of 14.4 kDa, beginning 14 residues N-terminal to 10F1, was isolated from the fibrin-affinity matrix. Radio-iodinated 14.4 kDa fibrin-binding peptide/protein (FBP) demonstrated a dose-dependent and saturable binding to fibrin-coated wells that was both competitively inhibited and reversed by unlabelled 14.4 kDa FBP. Comparison of the fibrin-binding affinities of proteolytic FBPs from the N-terminus (25.9 kDa FBP), the C-terminus (14.4 kDa) and intact Fn by ELISA yielded estimated Kd values of 216, 18 and 2.1 nM, respectively. The higher fibrin-binding affinity of the N-terminus was substantiated by the ability of both a recombinant 4F1.5F1 and a monoclonal antibody (mAb) to this site to maximally inhibit biotinylated Fn binding to fibrin by 80%, and by blocking the 90% inhibitory activity of a polyclonal anti-Fn, by absorption with the 25.9 kDa FBP. We propose that whereas the N-terminal site appears to contribute to most of the binding activity of native Fn to fibrin, the specific binding of the C-terminal site may strengthen this interaction.

Importance of the basement membrane protein SPARC for viability and fertility in Caenorhabditis elegans

The basement membrane is a specialized extracellular matrix located at epithelial-mesenchymal boundaries that supports cell adhesion, migration, and proliferation; it is highly conserved between invertebrates and vertebrates [1,2]. One of its component proteins, SPARC (osteonectin/BM-40), binds calcium and collagens, and can modulate cell-matrix interactions, so altering cell shape, growth, and differentiation [3,5]. The tissue distribution of a secreted fusion protein containing SPARC and green fluorescent protein (GFP) was analyzed in Caenorhabditis elegans. The protein localized to most basement membranes along body wall and sex muscles, and was also deposited around the pharynx and the gonad, in the spermatheca and at the distal tip cells. The contributions of SPARC to C. elegans development were determined using RNA interference, which accurately phenocopies loss-of-function defects [6-8]. A reduction in the amount of SPARC protein resulted in embryonic or larval lethality in a significant proportion of progeny. Those that survived developed a 'clear' phenotype characterized by a lack of gut granules, which made the animals appear transparent, plus small size, and sterility or reduced fecundity. No significant morphological abnormalities were observed, indicating that SPARC plays a regulatory rather than structural role in modulating cell-matrix interactions during normal development and reproduction.

Modulation of cell-extracellular matrix interactions

Changes in extracellular matrix (ECM) structure and composition, such as occur during morphogenesis, can have important regulatory effects on cell behavior. Two fibronectin (FN)-based systems have been developed to dissect how cells respond to different types of ECM. One system mimics the provisional matrix of the wound and is composed of FN cross-linked into a fibrin clot matrix. Unlike cells on FN alone, cells on an FN-fibrin matrix are smaller with cortical distribution of actin filaments and membrane ruffles. Addition of the ECM protein tenascin to the FN-fibrin matrix induces a different cell morphology. Thus, matrix composition can have profound effects on cell phenotype. Cells also interact with FN while assembling it into a fibrillar matrix. Using recombinant FNs, a domain that is required for normal progression of FN fibril formation has been identified. During assembly of this recombinant matrix, formation of actin stress fibers and focal adhesions is delayed, demonstrating that changes in FN matrix structure can affect intracellular organization and activation of signaling pathways.

Control of cell cycle progression by fibronectin matrix architecture

Developmental patterning and differentiation, maintenance of parenchymal cell function, and the size, shape, and invasiveness of tumors are all orchestrated by cell interactions with the extracellular matrix. Here we show that the fibrillar structure of fibronectin (FN) matrix encodes essential regulatory cues and controls cell proliferation and signaling through changes in matrix architecture. A matrix assembled from native FN stimulated cell growth. In contrast, a mutant FN (FNDeltaIII1-7) that contains all known cell binding motifs but forms a structurally distinct matrix inhibited progression from G0/G1 into S phase. Furthermore, FNDeltaIII1-7 suppressed the stimulatory capacity of native FN and induced different levels of tyrosine phosphorylation of pp125(FAK). The differential effects on cell growth were ablated by blocking formation of matrix fibrils. Thus, modification of matrix architecture provides a novel approach to control cell proliferation.

Dexamethasone up-regulates cadherin expression and cohesion of HT-1080 human fibrosarcoma cells

The synthetic glucocorticoid dexamethasone markedly decreases the invasiveness of HT-1080 human fibrosarcoma cells. We show here that dexamethasone treatment of HT-1080 cell aggregates more than doubles their cohesivity from 3.9 to 9.7 dyne/cm. Western blot analysis shows a corresponding increase in cadherin expression. This was accompanied by an increase in the rate of calcium-dependent aggregation. Dexamethasone-treated aggregates spread to form a monolayer in Matrigel spreading assays, but the cells remained much more contiguous than their untreated counterparts. Invasion-suppression by dexamethasone may therefore be due, at least in part, to a previously unsuspected increase in cadherin-mediated cohesion.

A single cysteine, Cys-64, is essential for assembly of tenascin-C hexabrachions

Tenascin-C is a large, multimeric extracellular matrix protein that is found in a variety of tissues and can have profound effects on cell adhesion. It is secreted from cells as a hexamer of six identical chains called a hexabrachion. Disulfide bonding among tenascin subunits mediates intracellular assembly into hexamers. The amino-terminal assembly domain consists of heptad repeats and at least six cysteine residues (Cys-64, -111, -113, -140, -146, -147) that could be involved in multimerization. We have now determined the requirements for these cysteine residues during hexamer assembly. Our results show that only Cys-64 is required to form the hexameric structure. Mutation of Cys-64 to glycine resulted in release of trimer intermediates, which probably form via the heptad repeats, but no hexamers were secreted. In contrast, individual or pairs of mutations of each of the other cysteines had no effect on tenascin hexamer formation, and inclusion of any other cysteine mutations along with C64G did not further disrupt the multimer pattern. However, when all six cysteines were mutated, monomers were the major extracellular form. Together, these results show that trimers are an intermediate of tenascin-C assembly and that Cys-64 is essential for formation of hexabrachions.

Modulatory roles for integrin activation and the synergy site of fibronectin during matrix assembly

Initiation of fibronectin (FN) matrix assembly is dependent on specific interactions between FN and cell surface integrin receptors. Here, we show that de novo FN matrix assembly exhibits a slow phase during initiation of fibrillogenesis followed by a more rapid growth phase. Mn2+, which acts by enhancing integrin function, increased the rate of FN fibril growth, but only after the initial lag phase. The RGD cell-binding sequence in type III repeat 10 is an absolute requirement for initiation by alpha5beta1 integrin. To investigate the role of the cell-binding synergy site in the adjacent repeat III9, a full-length recombinant FN containing a synergy mutation, FN(syn-), was tested for its ability to form fibrils. Mutation of this site drastically reduced FN assembly by CHOalpha5 cells. Only sparse short fibrils were formed even after prolonged incubation, indicating that FN(syn-) is defective in progression of the assembly process. These results show that the synergy site is essential for alpha5beta1-mediated accumulation of a FN matrix. However, the incorporation of FN(syn-) into fibrils and the deoxycholate-insoluble matrix could be stimulated by Mn2+. Therefore, exogenous activation of integrin receptors can overcome the requirement for FN's synergy site as well as modulate the rate of FN matrix formation.

Covalent cross-linking of fibronectin to fibrin is required for maximal cell adhesion to a fibronectin-fibrin matrix

In a blood clot, fibrin and plasma fibronectin (pFN) are covalently cross-linked by activated factor XIII (factor XIIIa) to form pFN-fibrin multimers. To determine the functional significance of covalent pFN-fibrin interactions, we have developed an in vitro model which allows the incorporation of recombinant FN (recFN) molecules into a covalently cross-linked recFN-fibrin matrix. Using the baculovirus expression system, we have expressed recFN monomers composed of the amino-terminal 70-kDa region and the first 11 type III repeats (WT) with mutations in the glutamines at positions 3 and 4 (Q2) or at 3, 4, and 16 (Q3). Examination of the covalent incorporation of these recFNs into fibrin clots confirms that glutamines 3 and 4 are major participants in FN-fibrin cross-linking as the mutation of these sites reduces cross-linking efficiency by 65%. Additional mutation of the glutamine at position 16, however, eliminates >99% of cross-linking suggesting that it also may be factor XIIIa reactive. When the Q3 recFN-fibrin clots were used as substrates for cell adhesion, there was a decrease in both cell attachment and spreading when compared with the WT recFN-fibrin clots. These data demonstrate that for maximal cell attachment to a FN-fibrin clot, FN must be cross-linked to fibrin by factor XIIIa.

Structural and functional characterization of the extracellular calcium-binding protein BM-40/secreted protein, acidic, rich in cysteine/osteonectin from the nematode Caenorhabditis elegans

Caenorhabditis elegans BM-40 (positions 19-264) and its extracellular calcium-binding domain (positions 139-264) were obtained in recombinant form from human kidney cells using an episomal expression vector. The purified proteins showed single bands of 33 kDa [BM-40-(19-264)-peptide] or 14 kDa [BM-40-(139-264)-peptide] on electrophoresis, contained internal disulfide bonds and a helices and were relatively resistant to matrix metalloproteinases. Hexosamine analysis indicated substitution by one N-linked and two O-linked oligosaccharides and recombinant BM-40 was indistinguishable in its immunological epitopes from nematode tissue-derived BM-40, suggesting that it was obtained in native form. Both recombinant C. elegans proteins showed a distinct binding activity for human collagens I and IV in solid-phase and surface-plasmon-resonance assays with an affinity (Kd = 1-2 microM), comparable to that of mammalian BM-40. However, calcium-binding studies revealed only a low-affinity site (Kd = 6.2 mM) and failed to show the characteristic conformational change upon addition of EDTA. These and a few other differences are apparently due to two extra disulfide bonds and two deletions/insertions in C. elegans BM-40 and can be partly interpreted from the X-ray structure of a large part of human BM-40. The immunological assays available and the predictions of the location of the collagen-binding epitope should facilitate a molecular and genetic approach to understand the function of BM-40 in the development of C. elegans.

Cell migration: may the force be with you

Effective linkages between the extracellular matrix and the cytoskeleton are crucial for cell migration. New insights into the formation of these links are provided by laser trap experiments and other recent advances.

Modulation of protein tyrosine phosphorylation by the extracellular matrix

Fibronectin (FN) cross-linked to fibrin following injury provides the provisional matrix required for cells to begin tissue repair. Our previous work has demonstrated that fibroblasts adherent to multimeric FN within the context of a fibrin matrix (FN-fibrin) exhibit clear phenotypic differences from those adherent to a dimeric FN-coated surface. We hypothesize that this response to multimeric FN may be mediated by altered protein tyrosine phosphatase activity following integrin activation.

METHODS

NIH 3T3 cells were plated in the presence or absence of pervanadate (PV), a phosphotyrosine phosphatase inhibitor, on wells coated with FN or FN-fibrin matrix. Spread cell areas were measured after increasing incubation times and are recorded as mean cell area (mm2) +/- SEM. Alternatively, cells were lysed and equal amounts of protein were analyzed by immunoblot using a monoclonal antibody specific for phosphotyrosine.

RESULTS

PV significantly inhibited cell spreading on FN-fibrin matrices. In contrast, PV treatment had little effect on cell area on FN alone. Analysis of cell lysates revealed that protein tyrosine phosphorylation events differ in a substrate-dependent manner.

CONCLUSION

Cell attachment to a FN-fibrin matrix induces distinct cell shape and cytoskeletal organization. Inactivation of tyrosine-specific phosphatases enhances this distinction and inhibits the spreading of cells attached to this substrate. The phosphotyrosyl protein content of treated cells on FN-fibrin matrix is also diminished. These results suggest that cell-extracellular matrix interactions affect the tyrosine phosphorylation balance of the cell, thus modifying cytoskeletal organization and related signaling events.

Coordinated regulation of fibronectin fibril assembly and actin stress fiber formation

Assembly of a fibronectin (FN) matrix is a multistep process which influences a number of cellular functions including intracellular cytoskeletal organization and signaling responses. We have previously reported on a recombinant FN (recFN), FN delta III1-7, which differs from native FN in its rate of fibril formation. To determine the intracellular consequences of a delay in assembly, we compared the distribution of cytoskeletal proteins during the formation of native and recFN matrices by immunofluorescence at various time points. CHO alpha 5 cell cytoskeleton was reorganized in response to both native and recFN matrix formation. Assembly of native FN induced a rapid reorganization of actin into stress fibers and colocalization of alpha 5 beta 1 integrin, focal adhesion kinase (FAK), vinculin, and paxillin to regions of cell-matrix contact. alpha 5 beta 1 integrins and FAK are also clustered upon binding of FN delta III1-7 to cells but actin reorganization and focal adhesion formation are delayed and appear to be dependent on the formation of FN delta III1-7 fibrils. These results suggest that the structural framework of the matrix plays an important role in the ability of FN to initiate intracellular responses.

Altered rate of fibronectin matrix assembly by deletion of the first type III repeats

The assembly of fibronectin (FN) into a fibrillar matrix is a complex stepwise process that involves binding to integrin receptors as well as interactions between FN molecules. To follow the progression of matrix formation and determine the stages during which specific domains function, we have developed cell lines that lack an endogenous FN matrix but will form fibrils when provided with exogenous FN. Recombinant FNs (recFN) containing deletions of either the RGD cell-binding sequence (RGD-) or the first type III repeats (FN delta III1-7) including the III1 FN binding site were generated with the baculovirus insect cell expression system. After addition to cells, recFN matrix assembly was monitored by indirect immunofluorescence and by insolubility in the detergent deoxycholate (DOC). In the absence of any native FN, FN delta III1-7 was assembled into fibrils and was converted into DOC-insoluble matrix. This process could be inhibited by the amino-terminal 70 kD fragment of FN, showing that FN delta III1-7 follows an assembly pathway similar to FN. The progression of FN delta III1-7 assembly differed from native FN in that the recFN became DOC-insoluble more quickly. In contrast, RGD- recFNs were not formed into fibrils except when added in combination with native FN. These results show that the RGD sequence is essential for the initiation step but fibrils can form independently of the III1-7 modules. The altered rate of FN delta III1-7 assembly suggests that one function of the missing repeats might be to modulate an early stage of matrix formation.

Changes in cell spreading and cytoskeletal organization are induced by adhesion to a fibronectin-fibrin matrix

Plasma fibronectin (pFN) cross-linked to fibrin during the injury response provides a provisional matrix required for cells to begin tissue repair. Using a synthetic matrix of pFN and fibrin as a substrate for cell adhesion and spreading, we have determined that pFN covalently cross-linked to fibrin into a complex multimer is functionally distinct from pFN immobilized onto a plastic surface. NIH-3T3 cells on a FN-fibrin matrix reach 50% of the maximal cell area of cells spread on FN-coated plastic. They neither attach nor spread on cross-linked fibrin alone. Cells on pFN-fibrin matrices form few prominent stress fibers and exhibit clear differences in membrane ruffling and filopodial extension when stained with rhodamine-labeled phalloidin. Interestingly, these differences are enhanced by upregulation of protein kinase C. These data suggest that cell-FN interactions can be modified by the molecular context of the protein within the extracellular matrix resulting in distinct cell morphology and cytoskeletal organization.

Differential targeting of recombinant fibronectins in AtT-20 cells based on their efficiency of aggregation

In pituitary-derived AtT-20 cells, recombinant fibronectin containing the N-terminal matrix assembly domain and the C-terminal half of fibronectin does not follow the regulated secretory pathway but instead concentrates in distinct organelles prior to secretion. These organelles are larger than the dense-core granules and localize to the cell body at sites that differ from lysosomes, endosomes and endoplasmic reticulum. Unlike the dense-core granules, their discharge is not stimulated by 8-bromo-cyclic-AMP or phorbol esters. The kinetics of intracellular transport and secretion of the recombinant fibronectin suggest that it is present in a post-Golgi pool that turns over more slowly than constitutive vesicles. Indeed, the fibronectin-containing organelles disappear with a half-time of 3 hours after inhibiting protein synthesis. Presence of the organelles correlates with intracellular aggregation of dimeric fibronectin polypeptides. The organelles are absent in cells expressing monomeric recombinant fibronectin (lacking C-terminal dimerization sites) or the C-terminal half of fibronectin (which dimerizes but lacks the N-terminal matrix assembly domain), both of which aggregate less efficiently than dimeric fibronectin. Instead, the latter polypeptides enter the dense-core granules. Thus while the formation of the fibronectin-containing organelles may require efficient aggregation, it may not require a specific structural signal. Moreover, efficient aggregation is not necessarily a prerequisite for following the regulated pathway.

Identification of an enhancer involved in tissue-specific regulation of the rat fibronectin gene

Fibronectin (FN) is a widely distributed extracellular matrix protein that is essential for cell adhesion in a variety of biological processes such as wound healing, tissue development and remodeling and oncogenic transformation. Appropriate FN levels are obtained by induction or repression of the FN gene in response to specific factors or circumstances in vivo. In order to identify regulatory regions involved in tissue-specific expression of FN, we have examined the transcriptional activity of overlapping fragments, within 4 kb upstream of the rat FN gene, following transfection into different cell types. Two regions conferred increases in transcription. The region between -1.08 and -2.6 displayed tissue-specificity and was active in fibroblasts but not hepatoma cells. The second region, between -3.2 and -3.9, was active in both cell types. Further characterization of the -1.08 to -2.6 segment demonstrated that it acts as an enhancer. Exonuclease III deletions of the 3' and 5' ends of the enhancer localized essential sequences between -1.5 and -1.7 and indicate that this fragment acts in concert with other sites between -1.08 and -2.6 to provide maximum enhancer activity. Gel mobility shift assays demonstrated fibroblast-specific binding of nuclear protein(s) to a 65 bp fragment within the essential region and DNase I footprinting localized this binding to a 27 bp sequence. Deletion of the sequence abolished the activity of the 1.5 kb enhancer. These studies show that a novel DNA sequence at -1688 is involved in regulating transcription of the FN gene in fibroblasts.

Rapid intracellular assembly of tenascin hexabrachions suggests a novel cotranslational process

Tenascin, an extracellular matrix protein that modulates cell adhesion, exists as a unique six-armed structure called a hexabrachion. The human hexabrachion is composed of six identical 320 kDa subunits and the structure is stabilized by inter-subunit disulfide bonds between amino-terminal segments. We have examined the biosynthesis of tenascin and its assembly into hexabrachions using pulsechase labeling of U-138 MG human glioma cells. Newly synthesized tenascin hexamers are secreted within 60 minutes of translation initiation. Intracellularly, as early as full length tenascin can be detected in pulse-labeled cell lysates, it is already in hexameric form. No precursors, such as monomers, dimers, or trimers, were identified that could be chased into hexamers. This lack of assembly intermediates suggests that nascent tenascin polypeptides associate prior to completion of translation. In contrast, fibronectin monomers in the same lysates are gradually formed into disulfide-bonded dimers. Although hexamer assembly is rapid, the rate-limiting step in secretion appears to be transport to the medial Golgi as endoglycosidase H-resistance was not detected until after a 30 minute chase. These results provide evidence for a novel co-translational mechanism of tenascin assembly which would be facilitated by its length and by the amino-terminal location of the assembly domain.

Fibronectin self-association is mediated by complementary sites within the amino-terminal one-third of the molecule

The formation of a fibrillar fibronectin (FN) extracellular matrix requires self-association of FN dimers. In this report, we show that the major sites for self-association are the amino-terminal repeats I1-5 and the first type III repeats. Recombinant FNs and fragments were generated by baculovirus expression of cysteine-rich domains and by bacterial expression of type III repeats as fusion proteins with maltose binding protein. When recombinant polypeptides were immobilized on microtiter wells, FN bound to 70-kDa amino-terminal fragment and to fusion proteins containing repeats III1-2 and III1-6 but not to other type III repeats. Similar results were obtained with a gel overlay assay. Binding was concentration-dependent and saturable. The amino-terminal binding site for III1-2 was further localized to repeats I1-5. Therefore, at least two different sites for FN-FN interaction reside near the amino terminus of the molecule. A model for the regulation of FN matrix assembly is proposed based on intramolecular interactions between these amino-terminal sites.

Interactions between fibronectin and chondroitin sulfate are modulated by molecular context

Interactions between fibronectin (FN) and glycosaminoglycans are essential for extracellular matrix morphology and cell adhesion. One of the most abundant glycosaminoglycans is chondroitin sulfate, and here we show that recombinant FNs (deminectins (DN)) containing the carboxyl-terminal cell, heparin, and fibrin domains bind specifically to chondroitin sulfate in affinity chromatography assays. Using a panel of mutant DNs, important determinants for chondroitin sulfate binding have been localized to repeats III13 and III14 within the heparin domain. In particular, mutation of an arginine pair in repeat III13 to neutral residues ablated binding to chondroitin sulfate as we previously reported for heparin (Barkalow, F.J.B., and Schwarzbauer, J.E. (1991) J. Biol. Chem. 266, 7812-7818). These results, in combination with the ability of heparin and chondroitin sulfate to compete for binding to DNs, demonstrate that these two glycosaminoglycans interact with similar or overlapping sites in FN. One important difference between FN interactions with heparin and chondroitin sulfate is that, while FN and DNs bound equally to heparin, FN bound less efficiently than DNs to chondroitin sulfate. Reduced binding to chondroitin sulfate was also observed with a larger recombinant FN lacking internal repeats III1-7 indicating that the amino-terminal region acts to limit binding to the carboxyl-terminal domain. Our results demonstrate that interactions between FN and chondroitin sulfate are modulated by molecular context.

Alternate exon usage is a commonly used mechanism for increasing coding diversity within genes coding for extracellular matrix proteins

Extracellular matrix proteins are a diverse family of secreted proteins and glycoproteins that are responsible for a variety of critical functions in different tissues. A large number of multiexon genes encode these proteins of the extracellular matrix. Over the last few years, it has become evident that the processing of the pre-mRNA from several of these genes involves alternative splicing. This review summarizes the known examples of alternative splicing in genes coding for the extracellular matrix and attempts to relate the increase in coding diversity generated by alternate exon usage to the function(s) of individual extracellular matrix proteins.

The Caenorhabditis elegans homologue of the extracellular calcium binding protein SPARC/osteonectin affects nematode body morphology and mobility

The extracellular matrix-associated protein, SPARC (osteonectin [Secreted Protein Acidic and Rich in Cysteine]), modulates cell adhesion and induces a change in cell morphology. SPARC expression in mammals is developmentally regulated and is highest at sites of extracellular matrix assembly and remodeling such as parietal endoderm and bone. We have isolated cDNA and genomic DNA clones encoding the Caenorhabditis elegans homologue of SPARC. The gene organization is highly conserved, and the proteins encoded by mouse, human, and nematode genes are about 38% identical. SPARC consists of four domains (I-IV) based on predicted secondary structure. Using bacterial fusion proteins containing nematode domain I or the domain IV EF-hand motif, we show that, like the mammalian proteins, both domains bind calcium. In transgenic nematodes expressing a SPARC-lacZ fusion gene, beta-galactosidase staining accumulated in a striated pattern in the more heavily stained muscle cells along the body. Comparison of the pattern of transgene expression to unc-54-lacZ animals demonstrated that SPARC is expressed by body wall and sex muscle cells. Appropriate levels of SPARC are essential for normal C. elegans development and muscle function. Transgenic nematodes overexpressing the wild-type SPARC gene were abnormal. Embryos were deformed, and adult hermaphrodites had vulval protrusions and an uncoordinated (Unc) phenotype with reduced mobility and paralysis.

The alternatively spliced V region contributes to the differential incorporation of plasma and cellular fibronectins into fibrin clots

During blood clot formation in vivo, plasma fibronectin (pFN) is cross-linked to fibrin by coagulation factor XIIIa. Cellular FN (cFN), which localizes to connective tissue, is distinguished from pFN by the inclusion of alternatively spliced segments. To determine if these two FNs are functionally equivalent in blood clotting, the cross-linking of rat pFN and cFN to fibrin was compared in an in vitro clotting assay. Fibrinogen and FN were incubated at physiological ratios in the presence of thrombin and factor XIIIa. Cross-linking of FN to fibrin was monitored by SDS-PAGE and immunoblotting. Over 24 h, cFN was incorporated at a significantly slower rate than pFN and was not completely cross-linked to fibrin at a temperature that favors this interaction (0 degrees C). This difference was observed with purified fibrinogens from human, rat, and bovine and with rat plasma and was maintained even after incubation of pFN with rat fibroblasts for several days. Using the same assay, purified recombinant V(+)-V0 and V(+)-V+ FN dimers resembling pFN and cFN, respectively, showed a similar difference in cross-linking kinetics. These results suggest that the asymmetric distribution of the V region among pFN dimers plays a role in regulating its incorporation into blood clots. In fibrin clots, cFN was converted into a set of cross-linked intermediates distinct from those of pFN. For example, while pFN was initially cross-linked into a pFN-fibrin alpha heterodimer, this product was not a major intermediate in clots formed with cFN. This finding, in conjunction with evidence for the formation of factor XIIIa-catalyzed cFN-cFN cross-links, indicated that cFN molecules interact with each other, and with fibrin, differently from pFN. Together, these results show an important functional distinction between pFN and cFN.

Alternative splicing of fibronectin: three variants, three functions

Fibronectin (FN) is a multi-functional extracellular matrix protein required for cell adhesion and migration, blood clotting, wound healing, and oncogenic transformation. The functional complexity is paralleled by structural diversity in that multiple forms of FN are generated by cell type-specific alternative splicing. In the rat, up to 12 different combinations of the three alternatively spliced segments (EIIIA, EIIIB, and the V region) are produced. What effects do these segments have on FN function? Recently, progress has been made in the identification of specific activities for the three variants of the V region, V120, V95, and V0. FN-mediated cell adhesion, FN synthesis and secretion, and incorporation into blood clots are differentially affected by these isoforms. These results suggest that cellular behavior is modulated by environmental cues provided by different types and proportions of alternatively spliced FN variants.

Fibronectin: from gene to protein

Recent advances in several key areas of fibronectin biology are discussed. These include its expression, from transcription to secretion of dimers, the structural requirements for several of the binding activities, potential roles for alternatively spliced segments in cell adhesion, and the assembly of a fibronectin matrix.

Identification of the fibronectin sequences required for assembly of a fibrillar matrix

During extracellular matrix assembly, fibronectin (FN) binds to cell surface receptors and initiates fibrillogenesis. As described in this report, matrix assembly has been dissected using recombinant FN polypeptides (recFNs) expressed in mammalian cells via retroviral vectors. RecFNs were assayed for incorporation into the detergent-insoluble cell matrix fraction and for formation of fibrils at the cell surface as detected by indirect immunofluorescence. Biochemical and immunocytochemical data are presented defining the minimum domain requirements for FN fibrillogenesis. The smallest functional recFN is half the size of native FN and contains intact amino- and carboxy-terminal regions with a large internal deletion spanning the collagen binding domain and the first seven type III repeats. Five type I repeats at the amino terminus are required for assembly and have FN binding activity. The dimer structure mediated by the carboxy-terminal interchain disulfide bonds is also essential. Surprisingly, recFNs lacking the RGDS cell binding site formed a significant fibrillar matrix. Therefore, FN-FN interactions and dimeric structure appear to be the major determinants of fibrillogenesis.

Localization of the major heparin-binding site in fibronectin

We have identified the major site required for the interaction of fibronectin (FN) with heparin. Affinity chromatography was used to test the binding ability of a library of truncated, monomeric forms of fibronectin (deminectins) containing deletions or two point mutations in the heparin-binding domain. This domain consists of type III repeats 12, 13, and 14. Deletions of individual repeats showed that both III13 and III14 are required for complete binding. Small deletions within these repeats localized a major site of heparin interaction to the amino-terminal half of III13. Site-directed mutagenesis of adjacent arginines within this sequence to uncharged residues reduced heparin binding by 98%, identifying these positively charged amino acids as essential for the interaction. A significant role for the flanking alternatively spliced regions and for repeat III12 was not found. We conclude that, while both repeats III13 and III14 participate in heparin binding, there is a major site of interaction in repeat III13 that accounts for nearly all of the activity. The significance of multiple heparin-binding sites within this domain is discussed and a model is proposed to account for how these sites may function in vivo.

Selective secretion of alternatively spliced fibronectin variants

We demonstrate that the alternatively spliced variable (V) region of fibronectin (FN) is required for secretion of FN dimers during biosynthesis. Alternative splicing of the V segment of the rat FN transcript generates three subunit variants (V120, V95, V0) that differ by the inclusion or omission of an additional 120 or 95 amino acids. We are exploring the functions of this segment by expressing variant cDNAs in normal and transformed fibroblasts. Like FN itself, the cDNA-encoded polypeptides (deminectins [DNs]) containing the V120 or V95 segment are efficiently secreted as disulfide-bonded homodimers. However, few homodimers of DNs lacking this region, V0 DNs, are secreted. V0 homodimers do form inside the cell, as demonstrated by biosynthetic analyses of dimer formation and secretion using pulse-chase and time course experiments, but these dimers seldom reach the cell surface and are probably degraded intracellularly. Coexpression of V0 and V120 subunits results in intracellular formation of three types of dimers, V0-V0, V0-V120, and V120-V120, but only the V120-containing dimers are secreted. This selective retention of V0 homodimers indicates that the V region is required for formation and secretion of native FN dimers. In an analogous in vivo situation, we show that plasma FN also lacks V0-V0 dimers and consists of V0-V+ and V+-V+ combinations. Dissection of V region sequences by deletion mapping localizes the major site involved in DN dimer secretion to an 18-amino acid segment within V95. In addition, high levels of dimer secretion can be restored by insertion of V into a heterologous site 10 kD COOH terminal to its normal location. We discuss the potential role of intracellular protein-protein interactions in FN dimer formation.

Organization of the fibronectin gene provides evidence for exon shuffling during evolution

We report the organization of the two ends of the rat fibronectin gene which encode the type I and II repeating units of the protein. We show that each of these modular structural units is encoded by a separate exon. Homologous type I and II repeats are known to occur in tissue plasminogen activator, factor XII and a bovine seminal plasma protein. Comparison of these sequences and the exon structures of the fibronectin and tissue plasminogen activator genes indicates that exons encoding type I and type II repeats have reassorted during evolution. We also report analyses of the extreme 5' and 3' ends of the fibronectin gene including the promoter region and the exon encoding the prepro sequence of fibronectin and we show that the gene is transcribed from a single initiation site to a single polyadenylation site. These data provide information pertinent to the transcriptional regulation of the gene, the alternative splicing of the primary transcript and the structure of the primary translation product.

Multiple sites of alternative splicing of the rat fibronectin gene transcript

We describe analyses of the structure and expression of the rat fibronectin gene with particular attention to the 40-kb stretch from the center of the gene which encodes 17 type-III repeating units. Each repeat is precisely separated from its neighbors by introns and most are encoded by pairs of exons. Three repeats are encoded precisely by single exons and two of these (EIIIA and EIIIB) are alternatively spliced in a cell type-specific fashion. A third site of alternative splicing (EIIIB) reported here is similar in expression to the previously described EIIIA segment. Both are excluded from mRNA in liver cells and are, therefore, absent from plasma fibronectin. These two alternative splices, plus a third one (V) reported previously, can occur in all possible combinations giving 12 fibronectin mRNAs from a single gene. These splicing variations account for most but not all of the known fibronectin subunit variants. We report investigations designed to detect other regions of alternative splicing. We also show that the pattern of alternative splicing is somewhat altered on oncogenic transformation.

Efficient and stable expression of recombinant fibronectin polypeptides

We describe retroviral expression vectors containing cDNAs encoding part of fibronectin preceded by the signal and "pro" sequences of parathyroid hormone. The recombinant retroviruses were used to generate NIH 3T3 cell lines stably producing functionally active fragments of fibronectin. The recombinant fibronectins (deminectins) are processed and secreted by the cells and form disulfide-bonded dimers with themselves and with endogenous fibronectin subunits. The fibronectin-deminectin heterodimers are incorporated into the extracellular matrix. We describe cell lines producing six variant forms of deminectin corresponding to variant forms of fibronectin produced by alternative splicing. In constructing fibronectin cDNAs encoding the six variant forms, we also made use of the ability of retroviral vectors to generate cDNAs by accurate splicing of cloned genomic segments. These constructs should be valuable in analyses of the structure-function relationships of fibronectins.

Cell-type-specific fibronectin subunits generated by alternative splicing

Multiple fibronectin mRNAs arise by alternative splicing of the primary transcript of a single gene. We describe analyses of the contribution of this alternative splicing to fibronectin subunit heterogeneity in three different cell types using antisera directed against specific segments of fibronectin. beta-galactosidase-fibronectin fusion proteins produced with the lambda gt11 bacterial expression vector were used as immunogens. One region of alternative splicing accounts for differences in subunit size, while a second contributes to differences between the fibronectins present in blood plasma and in fibroblastic cells. We also show, however, that these two regions of alternative splicing do not account for all detectable subunits. We have also used these segment-specific antisera to show that blood platelets contain a spectrum of fibronectin subunits distinct from that found in blood plasma.

Fibronectin: a versatile gene for a versatile protein

We have isolated cDNA and genomic clones for rat fibronectins. A single gene gives rise to three different mRNA species by alternative splicing at a complex intron-exon boundary within the coding region. The fibronectins encoded by these three mRNAs differ by the insertion of different protein segments in the C-terminal heparin-binding domain. The amino acid sequences of the cell-, heparin- and fibrin-binding domains in the C-terminal third of fibronectin were deduced from the DNA sequences. The cell- and heparin-binding regions each consist of several similar repeating sequences known as type III homologies, while the fibrin-binding region comprises three repeats of a different type (type I homologies). The sequences suggest several hypotheses for the structure-function relationships of these domains of fibronectin; these hypotheses are now being tested. Determination of intron-exon boundaries within the fibronectin gene is beginning to reveal the modular structure of the gene and its relation to the repeating structure of fibronectin.

On the origin of species of fibronectin

Multiple different subunits of fibronectin are known to occur and their origin has been unclear. Recent results showing that a single fibronectin gene can give rise to several different mRNAs by alternative splicing suggested an explanation for some of this diversity of fibronectin subunits. Because the alternative splicing events occur within the coding region, the mRNAs differ in coding potential. We have prepared recombinant phage containing a rat fibronectin cDNA segment that is present in some fibronectin mRNAs and not in others. This segment was inserted in the beta-galatosidase gene of lambda gt11, and fusion protein produced by lysogens of the recombinant phage was purified and used as immunogen. The resulting antisera recognized some subunits of rat and hamster fibronectins but not others, indicating that inclusion or removal of this segment gives rise to mRNAs that encode different fibronectin subunits. In particular, presence or absence of a 95 amino acid segment appears to account for differences in size among the subunits of plasma fibronectin, whose origin is therefore explained by alternative patterns of RNA splicing.

A single rat fibronectin gene generates three different mRNAs by alternative splicing of a complex exon

Three fibronectin mRNAs exist in rat liver, differing by the presence or absence of segments of 285 or 360 bases at a point within the coding region. We previously proposed that the three mRNAs are encoded by a single gene and arise via alternative splicing of a common transcript. In order to test this hypothesis, we have isolated clones spanning approximately half of the fibronectin gene from a Fisher rat genomic library; blot hybridization analyses reveal the presence of only one fibronectin gene in the haploid rat genome. We determined the sequence of a portion (1221 nucleotides) of this gene. This sequence shows clearly that the three fibronectin mRNAs encoded by this gene are generated by a pattern of alternative splicing in which one 5' splice site can be paired with any one of three 3' splice sites, one at the beginning of, and two within, a single complex exon.