B82L mouse fibroblasts respond to fibronectin or vitronectin via a syndecan-1-mediated activation of the αvβ5 integrin. Cells attached to syndecan-1-specific antibody display only filopodial extension. However, the syndecan-anchored cells extend lamellipodia when the antibody-substratum is supplemented with serum, or low concentrations of adsorbed vitronectin or fibronectin, that are not sufficient to activate the integrin when plated alone. Integrin activation is blocked by treatment with (Arg-Gly-Asp)-containing peptides and function-blocking antibodies that target αv integrins, as well as by siRNA-mediated silencing of β5 integrin expression. In addition, αvβ5-mediated cell attachment and spreading on high concentrations of vitronectin is blocked by competition with recombinant syndecan-1 ectodomain core protein and by downregulation of mouse syndecan-1 expression by mouse-specific siRNA. Taking advantage of the species-specificity of the siRNA, rescue experiments in which human syndecan-1 constructs are expressed trace the activation site to the syndecan-1 ectodomain. Moreover, both full-length mouse and human syndecan-1 co-immunoprecipitate with the β5 integrin subunit, but fail to do so if the syndecan is displaced by competition with soluble, recombinant syndecan-1 ectodomain. These results suggest that the ectodomain of the syndecan-1 core protein contains an active site that assembles into a complex with the αvβ5 integrin and regulates αvβ5 integrin activity.

Extracellular matrix proteins interact with a number of cell surface adhesion receptors, stimulating signaling cascades that regulate gene expression, proliferation, differentiation, cell shape and motility (Lukashev and Werb, 1998). Most matrix proteins contain binding domains for members of the integrin family of heterodimeric adhesion receptors in tandem with heparin-binding domains (HBDs) that mediate interactions with cell surface proteoglycans. Although much study has focused on understanding the roles of integrins in transmitting signals from the extracellular matrix to the cell interior (Giancotti and Ruoslahti, 1999), relatively little is known about the roles of cell surface proteoglycans in these processes. However, several reports suggest that the proteoglycans, and especially the syndecans, collaborate with integrins to form adhesion signaling complexes (Beauvais et al., 2004; Beauvais and Rapraeger, 2003; Beauvais and Rapraeger, 2004; Couchman et al., 2001; Thodeti et al., 2003; Woods and Couchman, 2001).

Members of the syndecan family of cell surface proteoglycans are expressed on all adherent cells and engage matrix proteins including fibronectin (FN), vitronectin (VN), thrombospondin, laminin and fibrillar collagens via heparan sulfate (HS)-glycosaminoglycan (GAG) chains attached to the syndecan ectodomains (Couchman et al., 2001). This engagement communicates to the cell via functional domains within the syndecan core protein (Beauvais et al., 2004; Beauvais and Rapraeger, 2004; Couchman, 2003; Couchman et al., 2001; Tkachenko et al., 2005). Several studies have characterized functional sequences within the syndecan cytoplasmic domains. Molecules that interact with the syndecan cytoplasmic domains include signaling molecules such as PKCα, phosphatidylinositol (4,5)-bisphosphate and Src (Kinnunen et al., 1998; Oh et al., 1997a) and scaffolding proteins, including ezrin, syntenin and CASK (Granes et al., 2000; Grootjans et al., 1997; Hsueh et al., 1998). Other work has demonstrated that the syndecan extracellular and transmembrane domains also play important roles in the regulation of cell shape (Beauvais and Rapraeger, 2003; Kusano et al., 2004; Liu et al., 1998; McQuade and Rapraeger, 2003; Munesue et al., 2002; Park et al., 2002; Tkachenko and Simons, 2002). The mechanisms by which these domains regulate cell morphology are unknown, although they presumably interact with partners that modulate cellular signaling pathways.

One technique for testing syndecan-dependent regulation of cell morphology is to plate cells on syndecan-specific antibodies. This technique targets a single syndecan family member, compared with plating cells on matrix ligands that engage multiple families of cell surface receptors, including multiple syndecans. When transfected to express syndecan-1 (Sdc1), proteoglycan-deficient Raji cells gain the ability to bind to Sdc1 antibody and spread on Sdc1 antibody-coated surfaces in two phases that depend on the syndecan transmembrane and ectodomains, respectively (Lebakken et al., 2000; Lebakken and Rapraeger, 1996; McQuade and Rapraeger, 2003). Colon carcinoma cells spread when plated on antibodies specific to syndecan-2 (Sdc2) (Park et al., 2002). When grown on serum-coated substrata, these cells exhibit a highly spread morphology, but revert to a rounded morphology when treated with recombinant Sdc2 ectodomain, suggesting an important role for this region of the core protein. Similarly, T lymphocytes extend filopodia when plated on a substratum comprised of antibody directed against the syndecan-4 ectodomain (Yamashita et al., 1999). Thus, antibody ligation of syndecans can trigger signaling leading to cell spreading, although the mechanisms of signaling are not clear.

Some studies demonstrate that the cellular response to the extracellular matrix involves engagement and cooperative signaling between the integrins and syndecans. One of the first descriptions of this came from fibroblasts plated on fragments of FN. Although cells spread when plated on the cell-binding (e.g. integrin-binding) domain of FN, which binds the α5β1 integrin, they fail to form focal adhesions unless syndecan-4 (Sdc4) is also ligated by either adding the HBD-FN or Sdc4-specific antibody (Saoncella et al., 1999; Woods and Couchman, 2000; Woods et al., 1986). Engagement of Sdc4 appears to stimulate Sdc4 oligomerization, and activation of PKCα, a signal required for the assembly of focal contacts and stress fibers (Oh et al., 1997a; Oh et al., 1997b; Oh et al., 1998). Recent work indicates that this mechanism involves Sdc4 working in concert with the α5β1 integrin (Mostafavi-Pour et al., 2003). In a similar mechanism, mesenchymal cells bind the disintegrin domain of ADAM-12 via β1 integrins and interact with the cysteine-rich domain of ADAM-12 via Sdc4 (Iba et al., 2000); cells that overexpress Sdc4 display increased levels of activated β1 integrin (Thodeti et al., 2003), an activity that requires the Sdc4 cytoplasmic domain and PKCα.

Sdc2 also cooperates with integrins during focal adhesion assembly. Lewis lung carcinoma cells require both α5β1 integrin and a cell surface proteoglycan to form stress fibers and focal adhesions when plated on FN. When Sdc2 expression levels are reduced by treatment with anti-sense RNA, the ability of these cells to form focal adhesions is impaired (Kusano et al., 2000). Although the mechanism by which Sdc2 stimulates the formation of focal adhesion is unknown, this suggests that Sdc2 and Sdc4 share overlapping roles and/or cooperate in the regulation of focal adhesion formation (Kusano et al., 2004).

Recent work from our laboratory has demonstrated a crucial link between Sdc1 and αvβ3 integrin on carcinoma cells (Beauvais et al., 2004; Beauvais and Rapraeger, 2003). Ligation of Sdc1 via its ectodomain to either an antibody substratum or a matrix ligand leads to activation of the αvβ3 integrin. Anchorage solely via the syndecan to antibody causes the cells to spread utilizing signaling from the activated αvβ3 integrin despite the fact that the integrin itself is not engaged by a ligand. On an αvβ3-specific matrix ligand, such as VN, αvβ3-dependent cell spreading and cell migration is blocked by short-interfering RNA (siRNA) blockade of Sdc1 expression, or by competition with recombinant Sdc1 ectodomain (S1ED). The activity of other integrins, such as the α5β1 or αvβ1 integrins, is not affected by these treatments. Moreover, the siRNA blockade of endogenous human Sdc1 activity can be reversed by expression of a glycosylphosphatidylinositol-linked mouse Sdc1 ectodomain (mS1ED), albeit the ectodomain must retain its HS chains, presumably to engage the matrix. This suggests that this domain of the core protein of the proteoglycan mediates its association into a cell surface complex that regulates αvβ3 activity.

Here, we describe similar techniques to examine the role of Sdc1 in adhesion signaling of mouse B82L fibroblasts. We find that anchorage of Sdc1 by syndecan-specific antibody primes the cells to respond to low concentrations of VN or FN, concentrations that would otherwise fail to trigger a cellular response. Although integrin inhibitory peptides and antibodies demonstrate a role for an αv-containing integrin, the cells do not express the αvβ3 integrin. Rather, they express αvβ5 - a closely related but distinct family member. siRNA blockade of β5-subunit expression confirms that αvβ5 is the target of Sdc1. Regulation of αvβ5 integrin activity does not require the HS-GAG chains of Sdc1 - something that distinguishes Sdc1-mediated regulation of the αvβ5 integrin from its regulation of αvβ3. However, interactions of the ectodomain of the core protein of Sdc1 are required for the regulation of both integrin heterodimers because activity is blocked by competition with S1ED protein and by selective downregulation of Sdc1 expression by siRNA. These data extend the emerging role of Sdc1 as a regulator of integrin activation.

Sdc1 mediates spreading in B82L fibroblasts

B82L mouse fibroblasts express approximately equal amounts of Sdc1 and Sdc4, and a lesser amount of Sdc2 (Ott and Rapraeger, 1998). In this work, we questioned the potential signaling activity of Sdc1 during B82L-cell adhesion and spreading. Although Sdc1 binds to matrix ligands via its HS-GAG chains, these matrix ligands are likely to bind other syndecans and integrins, thus confusing the adhesion assay. Therefore, the cells were plated on a nitrocellulose substratum coated with mouse Sdc1-specific antibody mAb 281.2. Within 15-20 minutes of plating, 60-90% of the adherent cells extend numerous filopodia up to 20 μm from the cell body (Fig. 1C). This response is specific for Sdc1, because cells plated on antibody against Sdc4, which is expressed at equivalent levels on these cells, bind via Sdc4 but fail to spread (Fig. 1E).

The filopodial extension observed upon Sdc1 ligation is only a partial spreading response compared with the response of cells grown in serum. Indeed, adding 10% serum to the plating medium induces cells adherent to mAb 281.2 to spread more extensively (Fig. 1D). This response is specific for Sdc1, because cells adhering to Sdc4-specific antibody (Fig. 1F) or plated on non-specific antibody (Fig. 1B) fail to spread in response to serum.

Sdc1 ligation enhances cell spreading on low levels of VN and FN

Although the nitrocellulose substratum coated with antibody is blocked with BSA, it appears that small amounts of matrix ligand in serum are sufficiently adsorbed to the substratum to stimulate signaling through an integrin that is activated when Sdc1 is ligated. Indeed, when antibody-coated and blocked substrata are pre-incubated with serum and then washed prior to the addition of cells, complete spreading still occurs (data not shown), duplicating the spreading response seen in the presence of serum (cf. Fig. 1D).

Fig. 1.

Sdc1 mediates cell spreading in B82L fibroblasts. B82L cells were plated on nitrocellulose coated with 60 μg/ml non-immune mouse IgG (A,B), 60 μg/ml mAb 281.2 (C,D) or 200 μg/ml S4ED pAb (E,F). Fetal bovine serum (FBS) (10%) was added to cells plated in B, D and F. Cells were incubated for 2 hours then fixed with paraformaldehyde and stained with Rhodamine-conjugated phalloidin. Bar, 50 μm.

Fig. 1.

Sdc1 mediates cell spreading in B82L fibroblasts. B82L cells were plated on nitrocellulose coated with 60 μg/ml non-immune mouse IgG (A,B), 60 μg/ml mAb 281.2 (C,D) or 200 μg/ml S4ED pAb (E,F). Fetal bovine serum (FBS) (10%) was added to cells plated in B, D and F. Cells were incubated for 2 hours then fixed with paraformaldehyde and stained with Rhodamine-conjugated phalloidin. Bar, 50 μm.

To directly test whether small amounts of VN and FN are capable of modulating cell spreading, B82L fibroblasts were added to wells on which Sdc1 antibody was co-coated with limiting dilutions of VN (Fig. 2) and FN (see Fig. S1 in supplementary material). In the absence of Sdc1-specific antibody, B82L-cell binding and spreading requires relatively high plating concentrations of VN (5 μg/ml) or FN (60 μg/ml). Even a twofold reduction in this amount completely abolishes binding (data not shown). However, cells in which Sdc1 is engaged by plating on mAb 281.2 (which normally induces filopodial extension) will spread with a fusiform morphology when as little as 0.2 μg/ml VN or 1 μg/ml FN, a level that by itself is insufficient to sustain adhesion, is co-plated with the antibody.

One possible explanation for increased sensitivity to these matrix ligands is that the Sdc1 bound to antibody simply tethers cells to the substratum and facilitates their interaction with low levels of matrix ligands to which cells would normally not adhere. To test this possibility, cells were plated on substrata co-coated with an antibody directed against the ectodomain of Sdc4, a syndecan that is expressed at equal levels to Sdc1 (Ott and Rapraeger, 1998). Because of their anchorage to the antibody, cell binding is seen either with antibody alone or with the mixed antibody and matrix ligand substrata (Fig. 2 and supplementary material Fig. S1). However, the cells fail to spread on Sdc4 antibody alone or when the antibody is supplemented with low concentrations of VN or FN. They spread only when VN or FN reaches a concentration that promotes adhesion and spreading on its own, e.g. 5 μg/ml VN and 60 μg/ml FN (Fig. 2 and supplementary material Fig. S1). This demonstrates that anchorage to the substratum by itself is insufficient for the matrix-dependent spreading response, and suggests that Sdc1 ligation is required to `activate' B82L cells to interact with and respond to low levels of VN and FN.

Response to VN and FN requires the αvβ5 integrin

The classical receptors for VN and FN are members of the integrin family of heterodimeric adhesion receptors. Interactions between integrins and these extracellular matrix proteins are mediated largely by Arg-Gly-Asp (RGD)-sequences found within many matrix proteins. Indeed, the spreading of B82L cells on either VN or FN alone, or on a mixed substratum with Sdc1-specific antibody, is blocked by RGD-containing peptides, including the cycloRGDfV peptide (Aumailley et al., 1991; Brooks et al., 1996) known to specifically target αv-containing integrins (see Fig. S2 in supplementary material). Similar results (data not shown) are obtained by treating cells with mAb H9.2B8, an antibody that blocks mouse αv integrin function (Moulder et al., 1991). The αvβ1 and αvβ3 integrins are expressed on fibroblasts and are well-known to act as VN or FN receptors (Sanders et al., 1998). The αvβ5 integrin has also been shown to be involved in fibroblast cell spreading on VN and FN (Pasqualini et al., 1993; van Leeuwen et al., 1996), although its recognition of FN is more controversial. Analysis of integrin expression of the B82L cells by flow cytometry shows that they express the αv integrin subunit, but relatively low amounts of the β1 integrin subunit and little or no β3 (Fig. 3A). Furthermore, B82L fibroblasts treated with β1 or β3 integrin inhibitory antibodies, mAb HMβ1-1 (Noto et al., 1995) or mAb 2C9.G2 (Yasuda et al., 1995), respectively, or with both antibodies together, show no effect on adhesion or spreading on Sdc1 antibody plus VN or FN or on high concentrations of matrix ligand alone (data not shown).

Fig. 2.

Vitronectin induces complete spreading of B82L fibroblasts. B82L cells were plated on wells co-coated with increasing amounts of VN and 60 μg/ml mAb 281.2, 150 μg/ml S4ED pAb or in the absence of antibody. Cells were allowed to spread 2 hours before fixation and labeling with Rhodamine-phalloidin. Bar, 50 μm.

Fig. 2.

Vitronectin induces complete spreading of B82L fibroblasts. B82L cells were plated on wells co-coated with increasing amounts of VN and 60 μg/ml mAb 281.2, 150 μg/ml S4ED pAb or in the absence of antibody. Cells were allowed to spread 2 hours before fixation and labeling with Rhodamine-phalloidin. Bar, 50 μm.

These data suggest the αvβ5 integrin as a candidate for Sdc1 regulation. Unfortunately, neither inhibitory antibodies to the mouse β5 subunit nor antibodies amenable for use in flow cytometry are currently available. However, Ab1926 binds the cytoplasmic domain of the β5 subunit and western blot analysis of B82L cell lysates shows expression of this integrin subunit (Fig. 3B). To test the role of the αvβ5 integrin in Sdc1-regulated cell spreading, an siRNA oligonucleotide specific for the mouse β5 subunit was used to block αvβ5 integrin expression. Transfection of cells with this siRNA reduces expression of αv-containing integrin by approximately 90%, as shown by monitoring αv integrin expression by flow cytometry (Fig. 3A). This correlates with a similar reduction in β5 subunit expression upon treatment with a range of siRNA concentrations on western blots (Fig. 3B,C). The siRNA has no effect on expression of mouse Sdc1 or that of other integrin β-subunits (Fig. 3A). Finally, it is observed that cells treated with 800 nM siRNA to block expression of β5 integrin fail to respond to either VN or FN when plated on these ligands together with Sdc1 antibody, even at high concentrations of VN or FN that are sufficient to promote cell spreading without Sdc1 antibody (Fig. 3D and supplementary material Fig. S3). Cells plated on Sdc1 antibody alone, which normally extend filopodia (Fig. 1C), continue to do so despite the blockade of β5 integrin expression (inset, Fig. 3D), which indicates that the αvβ5 integrin is not required for this spreading response.

Fig. 3.

siRNA blockade of β5-subunit expression blocks syndecan-induced cell spreading. (A) Suspended cells are analyzed by flow-cytometry with antibodies capable of recognizing mouse β1 (HMβ1-1), β3 (2C9.G2) or αv (H9.2B8) integrin subunits, mAb 281.2 specific for mouse Sdc1, or nonspecific IgG control (gray fill). Cells treated with β5-integrin-specific or control siRNA are compared. (B) Representative western blot of lysates of cells treated with 0, 200, 400, 600 or 800 nM β5-specific siRNA and probed for expression of β5-integrin subunit. FAK expression levels are shown as a loading control. (C) Quantification (± s.e.) of relative β5 integrin subunit expression from duplicate blots as described in (B). (D) B82L cells were plated on wells coated with 60 μg/ml mAb 281.2 and increasing amounts of VN after treatment with β5-integrin-specific or control siRNA. Cells were allowed to spread 2 hours before fixation and labeling with Rhodamine-phalloidin. Bar, 50 μm.

Fig. 3.

siRNA blockade of β5-subunit expression blocks syndecan-induced cell spreading. (A) Suspended cells are analyzed by flow-cytometry with antibodies capable of recognizing mouse β1 (HMβ1-1), β3 (2C9.G2) or αv (H9.2B8) integrin subunits, mAb 281.2 specific for mouse Sdc1, or nonspecific IgG control (gray fill). Cells treated with β5-integrin-specific or control siRNA are compared. (B) Representative western blot of lysates of cells treated with 0, 200, 400, 600 or 800 nM β5-specific siRNA and probed for expression of β5-integrin subunit. FAK expression levels are shown as a loading control. (C) Quantification (± s.e.) of relative β5 integrin subunit expression from duplicate blots as described in (B). (D) B82L cells were plated on wells coated with 60 μg/ml mAb 281.2 and increasing amounts of VN after treatment with β5-integrin-specific or control siRNA. Cells were allowed to spread 2 hours before fixation and labeling with Rhodamine-phalloidin. Bar, 50 μm.

αvβ5-dependent cell attachment and spreading requires the Sdc1 ectodomain but not its HS-GAG chains

The loss of Sdc1-regulated cell spreading in cells transfected with β5 siRNA indicates that the αvβ5 integrin is the αv-bearing integrin targeted by the syndecan. To test what domain(s) of the syndecan is required for this activity, endogenous mouse Sdc1 expression was silenced with mouse-specific siRNA (directed against the 3′-untranslated region) and subsequently replaced either by expression of human Sdc1 constructs or a mouse Sdc1 construct, comprised of GPI-linked mouse Sdc1 ectodomain alone (GPI-mS1ED) that lacks the siRNA-targeting sequence (Fig. 4 and supplementary material Fig. S4). Transfection with siRNA efficiently silences endogenous mouse Sdc1 by ∼98% as indicated by FACS (Fig. 4A and supplementary material Fig. S4A) and western blot analysis (Fig. 4F). Importantly, the mouse Sdc1 siRNA does not affect the expression of αvβ5, as indicated by western blotting (data not shown) and FACS analysis of the αv integrin subunit (Fig. 4B), and does not affect the expression of endogenous mouse Sdc4 (Fig. 4C) or the ectopic human Sdc1 (Fig. 4D-E and supplementary material Fig. S4B-C) and GPI-mS1ED (Fig. 4F) constructs. These results suggest that the siRNA is both species and syndecan-type specific.

Fig. 4.

Downregulation of mouse Sdc1 expression by siRNA blocks αvβ5-dependent cell attachment and spreading on vitronectin. FACS or immunoblot analysis for (A,F) mouse Sdc1 (mAb 281.2), (B) αv integrin subunit (mAb H9.2B8), (C) mouse Sdc4 (mAb KY8.2), (D) human Sdc1 (mAb B-B4) and (E) FcRecto-hS1 chimera (FITC-conjugated hIgG) expression against isotype IgG controls (red fill) in vector NEO (A-C), human Sdc1 (D), FcRecto-hS1 (E) and GPI-mS1ED (F) expressing B82L cells 48 hours after transfection with either 600 nM control (Control) or mouse Sdc1-specific siRNA (siRNA). (G) B82LNEO empty vector-transfected control cells and B82L cells expressing human Sdc1, the FcRecto-hS1 or GPI-mS1ED chimera were transfected with control or mouse Sdc1-specific siRNA and seeded on wells coated with either 5 μg/ml VN alone or a mixed substratum of VN plus 60 μg/ml of antibody directed against mouse Sdc1 (281.2), human Sdc1 (B-B4) or the FcRecto-hS1 chimera (hIgG). Cells were incubated at 37°C for 2 hours, fixed and stained with Rhodamine-conjugated phalloidin. Bar, 50 μm.

Fig. 4.

Downregulation of mouse Sdc1 expression by siRNA blocks αvβ5-dependent cell attachment and spreading on vitronectin. FACS or immunoblot analysis for (A,F) mouse Sdc1 (mAb 281.2), (B) αv integrin subunit (mAb H9.2B8), (C) mouse Sdc4 (mAb KY8.2), (D) human Sdc1 (mAb B-B4) and (E) FcRecto-hS1 chimera (FITC-conjugated hIgG) expression against isotype IgG controls (red fill) in vector NEO (A-C), human Sdc1 (D), FcRecto-hS1 (E) and GPI-mS1ED (F) expressing B82L cells 48 hours after transfection with either 600 nM control (Control) or mouse Sdc1-specific siRNA (siRNA). (G) B82LNEO empty vector-transfected control cells and B82L cells expressing human Sdc1, the FcRecto-hS1 or GPI-mS1ED chimera were transfected with control or mouse Sdc1-specific siRNA and seeded on wells coated with either 5 μg/ml VN alone or a mixed substratum of VN plus 60 μg/ml of antibody directed against mouse Sdc1 (281.2), human Sdc1 (B-B4) or the FcRecto-hS1 chimera (hIgG). Cells were incubated at 37°C for 2 hours, fixed and stained with Rhodamine-conjugated phalloidin. Bar, 50 μm.

In contrast to parental or control siRNA-transfected cells, B82L vector-control cells (NEO) transfected with mouse Sdc1-siRNA fail to attach and spread to wells coated with either 5 μg/ml of VN alone (Fig. 4G, right column) or a mixed substratum of 5 μg/ml of VN plus 60 μg/ml of mAb 281.2 (Fig. 4G, left column). The failure of these cells to even engage the mixed substratum clearly indicates the efficient blockade of mouse Sdc1 expression by siRNA. It is unlikely that the mouse Sdc1-siRNA treatment has any non-specific cellular effects since spreading on VN alone or on a mixed substratum of VN plus Sdc1 antibody (60 μg/ml mAb B-B4) is specifically rescued by the expression of full-length human Sdc1 (hS1, Fig. 4G). Notice that B82L-hS1 cell spreading on VN alone is indistinguishable from either parental or NEO cells. Intriguingly, similar rescue was observed in cells expressing human Sdc1 mutants, hS1pLeuTM and hS1Δcyto (see Fig. S4D in supplementary material), indicating that neither the transmembrane nor the cytoplasmic domain of the syndecan is required for the spreading response. This was further confirmed with cells that express GPI-linked Sdc1 ectodomain alone (GPI-mS1ED, Fig. 4G). Cells expressing this chimera recover spreading on a VN substratum, either in the presence or absence of Sdc1 antibody. However, cells expressing the FcRecto-hS1 chimera - a construct in which the ectodomain of syndecan is replaced by that of the human Fcγ receptor Ia (FcRecto-hS1, Fig. 4G) - fail to recover spreading, regardless of whether the cells are plated on VN alone or VN is supplemented with 60 μg/ml of hIgG to engage the FcRecto-hS1 construct. These results suggest that αvβ5 integrin activity depends on Sdc1 expression and that the ectodomain of syndecan is both necessary and sufficient to regulate such activity.

The regulation of αvβ3 integrin activity by Sdc1 on matrix ligands has been shown to require syndecan engagement of the matrix via its HS chains (Beauvais et al., 2004). To test the GAG requirement for αvβ5 activity, cells were pretreated for 2 hours with HS-specific and chondroitin-sulfate-specific lyases and plated in the presence of these enzymes (Fig. 5). Treated B82L cells fail to bind HBD-FN (used as a test for the efficacy of GAG removal) whereas untreated cells bind and rapidly extend filopodia, assuming morphologies similar to that seen when cells are bound to a Sdc1 antibody substratum (Fig. 5A,B). However, GAG removal has no effect on the filopodial extension or complete cell spreading observed either on Sdc1 antibody alone (Fig. 5C-E), antibody mixed with low concentrations of VN (Fig. 5F-H) or FN (Fig. 5H), or high concentrations of VN (Fig. 5I-K) or FN alone (Fig. 5K).

The failure of GAG removal to affect the Sdc1-dependent cell spreading suggests that signaling leading to spreading relies on an interaction with the syndecan core protein. To test this, B82L cells were plated on 5 μg/ml VN in the presence of a recombinant GST fusion protein containing either the ectodomain of mouse Sdc1 (GST-mS1ED) or human Sdc1 (GST-hS1ED; data not shown). Because the fusion protein is derived from bacteria, it does not contain attached GAG chains. GST-mS1ED and hS1ED display similar activity; competition occurs at the lowest concentration tested (1 μM) with increasing blockade of cell attachment and spreading over a concentration range of 1-30 μM of S1ED (Fig. 6A,B). Similar results are obtained with GST-mS1ED and B82L-hS1 cells attached to a mixed substratum of mAb B-B4 and 1 μg/ml VN (data not shown). Notice that, mS1ED is not recognized by the human specific mAb B-B4 and thus does not compete for human Sdc1 engagement of the antibody substratum. Moreover, the fusion protein has no effect on the ability of these cells to extend filopodia when plated on an antibody substratum alone (data not shown). Competition with 30 μM GST alone is without effect in all cases, as is competition with identical concentrations of recombinant GST-mS4ED indicating that competition for syndecan-regulated αvβ5 activity is S1ED-specific.

Fig. 5.

GAG chains are not required for filopodial extension or complete B82L cell spreading. B82L cells were detached using EDTA and treated in suspension either without (A,C,F,I) or with a combination of heparinases I and III and chondroitin ABC lyases (B,D,G,J) for 2 hours before plating on 200 μg/ml HBD-FN (A,B), 60 μg/ml mAb 281.2 (C,D), 60 μg/ml plus 1 μg/ml VN (F,G) or 5 μg/ml VN (I,J). Quantification of treated (white bars) or untreated (gray bars) cell extension of filopodia on 60 μg/ml mAb 281.2 (E) or complete cell spreading on mAb 281.2 supplemented with 1 μg/ml VN or 3 μg/ml FN (H) or complete cell spreading on 5 μg/ml VN or 60 μg/ml FN (K) is also shown. Bar, 50 μm.

Fig. 5.

GAG chains are not required for filopodial extension or complete B82L cell spreading. B82L cells were detached using EDTA and treated in suspension either without (A,C,F,I) or with a combination of heparinases I and III and chondroitin ABC lyases (B,D,G,J) for 2 hours before plating on 200 μg/ml HBD-FN (A,B), 60 μg/ml mAb 281.2 (C,D), 60 μg/ml plus 1 μg/ml VN (F,G) or 5 μg/ml VN (I,J). Quantification of treated (white bars) or untreated (gray bars) cell extension of filopodia on 60 μg/ml mAb 281.2 (E) or complete cell spreading on mAb 281.2 supplemented with 1 μg/ml VN or 3 μg/ml FN (H) or complete cell spreading on 5 μg/ml VN or 60 μg/ml FN (K) is also shown. Bar, 50 μm.

Fig. 6.

B82L-cell spreading on VN is blocked by recombinant S1ED. (A) B82L fibroblasts were plated on 5 μg/ml VN in the absence of other treatment, or in the presence of 30 μM GST, 1-30 μM GST-mS1ED or 30 μM GST-mS4ED (inset), then fixed and stained with Rhodamine-phalloidin for visualization. (B) Quantification of cell adhesion in triplicate samples (± s.e.) plated either with no additions, or concentrations of GST-mS1ED ranging from 0-30 μM. Bar, 50 μm.

Fig. 6.

B82L-cell spreading on VN is blocked by recombinant S1ED. (A) B82L fibroblasts were plated on 5 μg/ml VN in the absence of other treatment, or in the presence of 30 μM GST, 1-30 μM GST-mS1ED or 30 μM GST-mS4ED (inset), then fixed and stained with Rhodamine-phalloidin for visualization. (B) Quantification of cell adhesion in triplicate samples (± s.e.) plated either with no additions, or concentrations of GST-mS1ED ranging from 0-30 μM. Bar, 50 μm.

Fig. 7.

Co-immunoprecipitation of β5 integrin with Sdc1 requires the Sdc1 ectodomain. Western blots probed with rabbit polyclonal β5 integrin (A,B), pan-syndecan or S1ED (C) antibody for detection of β5 integrin and Sdc1, respectively, in immune complexes isolated after immunoprecipitation of full-length mouse Sdc1 (mAb 281.2), human Sdc1 constructs (mAb B-B4) and FcRecto-hS1 chimera (mAb 10.1) from pre-cleared B82L whole-cell lysates. In S1ED-competition experiments (A), 30 μM soluble GST, GST-mS1ED (with mAb B-B4) or GST-hS1ED (with mAb 281.2) was added to the reaction mixture. Provided as a reference is a methanol precipitation (MeOH) of 300 μg of whole-cell lysate. β5 integrin immunoblotting reveals a 110 kDa band, under reduced conditions, detectable in the mouse Sdc1 and select human Sdc1 (hS1, pLTM, Δcyto), but not FcRecto-hS1 chimera immunoprecipitates or immunoprecipitates isolated with antibody isotype control IgG: mouse IgG1 (mIgG) for mAbs B-B4 and 10.1 and rat IgG2A (rIgG) for mAb 281.2.

Fig. 7.

Co-immunoprecipitation of β5 integrin with Sdc1 requires the Sdc1 ectodomain. Western blots probed with rabbit polyclonal β5 integrin (A,B), pan-syndecan or S1ED (C) antibody for detection of β5 integrin and Sdc1, respectively, in immune complexes isolated after immunoprecipitation of full-length mouse Sdc1 (mAb 281.2), human Sdc1 constructs (mAb B-B4) and FcRecto-hS1 chimera (mAb 10.1) from pre-cleared B82L whole-cell lysates. In S1ED-competition experiments (A), 30 μM soluble GST, GST-mS1ED (with mAb B-B4) or GST-hS1ED (with mAb 281.2) was added to the reaction mixture. Provided as a reference is a methanol precipitation (MeOH) of 300 μg of whole-cell lysate. β5 integrin immunoblotting reveals a 110 kDa band, under reduced conditions, detectable in the mouse Sdc1 and select human Sdc1 (hS1, pLTM, Δcyto), but not FcRecto-hS1 chimera immunoprecipitates or immunoprecipitates isolated with antibody isotype control IgG: mouse IgG1 (mIgG) for mAbs B-B4 and 10.1 and rat IgG2A (rIgG) for mAb 281.2.

Further, immunoprecipitation of Sdc1 (Fig. 7C) with monoclonal antibodies directed against endogenous mouse Sdc1 (281.2) or ectopically expressed human Sdc1 constructs (B-B4) reveals β5 integrin (Fig. 7A,B) in the immune complexes isolated from B82L-NEO, B82L-hS1, B82L-hS1pLeuTM and B82L-hS1Δcyto cell lysates. Similar results were obtained with an affinity-purified pan-syndecan antibody that recognizes the ten C-terminal amino acids of Sdc1 with the exception of hS1Δcyto, which lacks these amino acids (data not shown). β5 integrin is not detectable in mAb B-B4 isolated immune complexes from B82L-NEO cells that are not expressing human Sdc1 (data not shown) nor in isotype IgG controls (Fig. 7A,B, mIgG and rIgG). Moreover, β5 integrin is not present in immune complexes isolated from B82L-NEO cell lysates using mAb KY8.2 (Fig. 7A) - an antibody directed against mouse Sdc4 (Yamashita et al., 1999). These results indicate that Sdc1 and the β5 integrin are present in a cell surface complex and Sdc1-specific association within this complex is conserved between mouse and human. By contrast, β5 integrin is not detectable with immunoprecipitated FcRecto-hS1 chimera isolated by mAb 10.1 (Fig. 7B,C), human IgG or pan-syndecan antibody (data not shown) from FcRecto-hS1-expressing cells, but is associated with the endogenous mouse Sdc1 immunoprecipitated with mAb 281.2 (Fig. 7B). This supports the conclusion that association of the syndecan with the β5 integrin depends on the Sdc1 ectodomain. To test this more directly, the immunoprecipitations were conducted with species-specific Sdc1 antibody in the presence of 30 μM GST-S1ED protein (competitive inhibitor of the integrin activation) from the opposite species (e.g. mS1ED with human-specific B-B4 and hS1ED with mouse-specific 281.2 to avoid recognition by the antibody). This demonstrates that GST-S1ED protein efficiently disrupts the association of the β5 integrin with Sdc1 (Fig. 7A, S1ED). Addition of GST alone is without effect (Fig. 7A, GST). These results suggest that the inhibition observed in the VN cell spreading assays upon treatment with soluble S1ED protein (Fig. 6) is due to perturbation of the association of αvβ5 integrins with cell surface Sdc1 leading to a loss in integrin activity.

The αvβ5 integrin is expressed on a variety of tissues and cell types, including endothelia, epithelia and fibroblasts (Felding-Habermann and Cheresh, 1993; Pasqualini et al., 1993). It is closely related to the αvβ3 integrin (56.1 % identity and 83.5% homology between the two integrin β-subunits) but is distinguished from the αvβ3 by divergent sites near its ligand-binding domain and within the C-terminus of its cytoplasmic domain (McLean et al., 1990). It has a role in matrix adhesion to VN, FN, SPARC and bone sialoprotein (Plow et al., 2000) and is implicated in the invasion of gliomas and metastatic carcinoma cells (Brooks et al., 1997; Jones et al., 1997; Tonn et al., 1998), the latter especially to bone (De et al., 2003). A second major role is in endocytosis, including endocytosis of VN (Memmo and McKeown-Longo, 1998; Panetti and McKeown-Longo, 1993; Panetti et al., 1995), the engulfment of apoptotic cells by phagocytes (Albert et al., 2000) and participation in the internalization of shed outer rod segments in the retinal pigmented epithelium (Finnemann, 2003a; Finnemann, 2003b; Hall et al., 2003). A third major role is in growth-factor-induced angiogenesis, where cooperative signaling by the αvβ5 integrin and growth factors regulates endothelial cell proliferation and survival. Angiogenesis promoted by VEGF and TGFα in human umbilical-vein endothelial cells relies on signaling together with the αvβ5 integrin, whereas FGF-2 and tumor necrosis factor-α collaborate with the αvβ3 integrin (Eliceiri and Cheresh, 1999; Friedlander et al., 1995).

Several studies have demonstrated that members of the syndecan family of cell adhesion receptors cooperate with integrins to mediate signals that regulate cytoskeletal rearrangements and cell shape. Work from this laboratory has described a prominent role for Sdc1 in regulating the activity of the αvβ3 integrin in MDA-MB-231 and MDA-MB-435 breast carcinoma cells (Beauvais et al., 2004; Beauvais and Rapraeger, 2003). Although the mechanism remains unclear, a region of the Sdc1 ectodomain appears to regulate the active state and signaling of the integrin. Experiments presented here describe a similar mechanism by which Sdc1 regulates the activity of the αvβ5 integrin in murine B82L fibroblasts. These cells are particularly useful for this study because they express a limited repertoire of integrin receptors, dominated by the αvβ5 integrin.

Ligation of Sdc1 alone, using a substratum consisting of syndecan-specific antibody or the HBD-FN, leads to B82L-cell adhesion but incomplete cell spreading, e.g. the extension of filopodia, suggesting that ligation of the syndecan alone generates a signal. Prior work with Sdc1 expressed in Raji lymphoid cells demonstrated that ligation of Sdc1 generates two phases of signaling. The first results in formation of a broad lamellipodium and depends on the Sdc1 transmembrane domain; the second induces cell polarization, an activity that traces to the Sdc1 ectodomain (McQuade and Rapraeger, 2003). Like the Raji cell-signaling responses, the Sdc1-mediated filopodial extension seen here does not require the Sdc1 cytoplasmic tail (data not shown) and appears to be integrin-independent. It is not blocked by treatment of the cells with RGD peptides, soluble S1ED protein, which perturbs a syndecan-integrin interaction, or siRNA-targeted silencing of integrin expression. However, providing trace amounts of VN or FN to B82L cells already anchored to a Sdc1 antibody reveals an integrin-related activity of the syndecan, namely, extensive cell spreading via syndecan-regulated signaling of the αvβ5 integrin. These data might help explain the apparent link between Sdc1 and αvβ5-dependent turnover of VN (Wilkins-Port et al., 2003). The regulation of integrin activity by Sdc1 might occur either by altering integrin activation and/or by altering integrin signaling in response to the ligand (i.e. post-receptor occupancy events). Integrin activation or `priming' (Carman and Springer, 2003) is classically defined as adopting the conformation necessary to bind ligand, often a response to inside-out signaling. However, the integrin can also be regulated by lateral interactions with other membrane proteins. Examples of receptors that regulate integrin activation through lateral interactions include uPAR/CD87 and IAP/CD47, which regulate αvβ3- and αvβ5-binding to VN and adhesion-dependent signaling events, tetraspanins, and chondroitin sulfate proteoglycan NG2, which interacts with the α4β1 integrin (Dedhar, 1999; Iida et al., 1998; Iida et al., 2001; Kugler et al., 2003; Porter and Hogg, 1998). Integrin activation is also enhanced by its binding to matrix ligand, which stimulates `outside-in' signals necessary for cell spreading. These include activation of FAK, Rho GTPases, PI3-kinase and other pathways (Carman and Springer, 2003; Dedhar, 1999; Liddington and Ginsberg, 2002).

In the B82L fibroblasts, it is envisioned that the syndecan assembles into a cell surface signaling complex that is necessary for αvβ5 integrin signaling (Fig. 8), although it is not entirely clear what other receptors, if any, are in the complex. What are the features of this complex? One feature is that anchorage of the syndecan to the substratum appears to lower the threshold for integrin activation by VN or FN. Thus, if Sdc1 is engaged by antibody, then low concentrations of matrix ligand appear sufficient to activate the integrin and lead to signaling. The simplest explanation would be that the syndecan simply anchors the cell to the substratum so that the integrin can engage the limited amounts of matrix ligand. However, this explanation appears to be ruled out by the fact that engagement of Sdc4 with antibody, which also anchors the cells to the substratum, does not result in αvβ5 integrin-dependent signaling. Alternatively, it is possible that the specificity arises from Sdc1 and αvβ5 integrin being in a complex together, such that anchorage of Sdc1 clusters the integrin as well as bringing it into close apposition to the matrix ligands - a contention strongly supported by the immunoprecipitation data presented here. As such, cells adhering via Sdc4 are presumably not primed to bind VN and FN, because Sdc4, which possesses a very different ectodomain (both in size and sequence) relative to Sdc1, fails to interact with the integrin.

Fig. 8.

Model of Sdc1-regulated αvβ5-integrin signaling.

Fig. 8.

Model of Sdc1-regulated αvβ5-integrin signaling.

A second feature is that the syndecan HS chains are not required for integrin activation, either on syndecan antibody or on matrix ligand. The syndecan HS chains, at least on the B82L cells, do not appear to bind VN or FN sufficiently well for the cells to strongly adhere, and cell binding occurs only at sufficiently high matrix concentrations for the integrin to become engaged. Here, the high concentration of matrix ligand can presumably bind and activate the integrin and trigger outside-in signaling. Nonetheless, this signaling also requires Sdc1, because the third feature of this complex is that the integrin-mediated cell adhesion and spreading on these matrix ligands is blocked by recombinant S1ED and by selective downregulation of Sdc1 expression with siRNA. Integrin-mediated cell spreading on VN is rescued in mouse Sdc1-siRNA transfected cells by expression of full-length human Sdc1 or GPI-mS1ED, which contains only the Sdc1 ectodomain, but not by a Sdc1 mutant lacking its ectodomain (FcRecto-hS1). Moreover, immunoprecipitation of the syndecan brings down β5 integrin with full-length Sdc1, either mouse or human, but not with FcRecto-hS1, despite the fact that this chimera retains the human Sdc1 transmembrane and cytoplasmic domains.

These features suggest that the syndecan and the integrin are in a complex together and that interactions of the Sdc1 ectodomain within the complex, which can be disrupted by soluble S1ED or by siRNA-mediated removal of the syndecan from the cell surface, are necessary for αvβ5-integrin signaling. Whether the interaction of the syndecan with the integrin is direct or indirect is not yet known, but it is known that sequences within the β5 ectodomain regulate integrin post-ligand-binding signaling events (Filardo et al., 1996). This suggests that αvβ5-dependent cellular responses depend on signals transmitted by the syndecan as a laterally associated protein, to include the activation of PKC, which is required for αvβ5-dependent cell spreading and migration (Klemke et al., 1994; Lewis et al., 1996; Yebra et al., 1995) and endocytosis of VN (Panetti et al., 1995).

This model is similar to the regulation of the αvβ3 integrin in mammary carcinoma cells (Beauvais et al., 2004), but has two significant differences. First, activation of the αvβ3 integrin by Sdc1 requires that the Sdc1 ectodomain is engaged, either directly to antibody or via its HS chains to a matrix ligand. Surprisingly, ligation of the syndecan alone is sufficient to activate αvβ3, but activation of the integrin is specific for Sdc1, because it cannot be recapitulated by adhesion of the cells via Sdc4. Heparinase removal of the HS chains (unpublished data) or expression of a Sdc1 mutant lacking HS chains prevents αvβ3 signaling when mammary carcinoma cells are plated on VN. By contrast, αvβ5 activation in B82L cells still occurs on VN after removal of the HS chains of syndecan with heparinases. In this case, ligation of the syndecan is rendered moot for the simple reason that Sdc1 is probably constitutively associated with the β5 integrin and it is this association that appears important for αvβ5-dependent signaling. Therefore, regardless of whether one or both receptors engage the substratum, engagement of one is sufficient to localize the other into the signaling complex via an ectodomain interaction. In the case of αvβ3-integrin signaling, however, engagement of the syndecan to the substratum is probably required to bring the syndecan and integrin together in a complex. Second, anchorage of the Sdc1 alone to the substratum, e.g. to a Sdc1-specific antibody, is sufficient to activate signaling from the αvβ3 integrin in mammary carcinoma cells, whereas the αvβ5 integrin on B82L fibroblasts does not signal upon syndecan engagement unless the integrin is also provided with a ligand. This suggests that the αvβ3 integrin in epithelial cells does not need to perform an adhesion role per se, but transduces signals when activated by syndecan engagement. The integrin could contribute by directly assembling into a signaling complex with the syndecan or could act as a downstream effector that relays signals required to reorganize the cytoskeleton. An opposite result is observed for the αvβ5 integrin in B82L cells. Here, the syndecan does not need to perform an adhesion role, but does need to be in a complex with the αvβ5 integrin in order for the integrin to engage the matrix and transduce signals. Association of the integrin with the syndecan via the ectodomain may be required for the integrin to adopt an active conformation, to localize the integrin to a particular membrane microdomain and/or to associate with certain downstream effectors. Although these differences may be subtle, they may provide insight into different mechanisms by which the syndecan regulates these two related integrins.

Cell culture and molecular biology

B82L mouse fibroblasts (provided by Paul Bertics, University of Wisconsin-Madison) were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% FBS, 4 mM L-glutamine and antibiotics, as previously described (Ott and Rapraeger, 1998). Human Sdc1 cDNA was provided by Markku Jalkanen (University of Turku, Finland) in pBGS. The coding sequence was PCR amplified from pBGS, cloned into the KpnI and XhoI sites of pcDNA3 (Invitrogen) and verified by sequencing. Human Sdc1 mutants in pcDNA3, including FcRecto-hS1 (a chimera comprised of the ectodomain of human IgG Fcγ receptor Ia/CD64 fused to the transmembrane and cytoplasmic domains of human Sdc1), hS1Δcyto (lacking the 33 C-terminal amino acids) and hS1pLeuTM (transmembrane domain replaced with leucine residues), were constructed with human-specific primers and/or methods previously described (McQuade and Rapraeger, 2003). GPI-mS1ED (Liu et al., 1998), a chimera comprised of the ectodomain of mouse Sdc1 fused to the glycosylphosphatidylinositol (GPI) tail of rat glypican-1, in pcDNA3 was generously provided by Ralph Sanderson (University of Arkansas for Medical Sciences, Little Rock, AR). Cells were transfected with pcDNA3 alone, human Sdc1 expression constructs or GPI-mS1ED using LipofectAMINE (Invitrogen) and the highest 10% of cells immunoreactive for mAb B-B4 (full-length hS1, hS1Δcyto and hS1pLeuTM), mAb 281.2 (GPI-mS1ED) and mAb 10.1 (FcRecto-hS1) were sorted by flow cytometry. After sorting, the cells were maintained in medium containing 300 μg/ml geneticin (Gibco BRL). Cells were passaged 1:20 every 3 days and grown to 60-80% confluency for experiments.

Cell-spreading assays

Cell-spreading assays were performed with modification to our previous procedures (Lebakken and Rapraeger, 1996). Briefly, ligands were applied to nitrocellulose-coated ten-well slides (Erie Scientific) and incubated for 1-2 hours at 37°C. Ligands used in this study include the Sdc1-specific mAb 281.2 (Jalkanen et al., 1985) and mAb B-B4 (Serotec), an affinity-purified rabbit polyclonal antibody (S4ED pAb) generated against the mouse Sdc4 ectodomain (amino acids 1-120) fused to the C-terminus of glutathione-S-transferase (GST-mS4ED; (McFall and Rapraeger, 1998), human plasma VN and FN and recombinant HBD-FN. A GST fusion protein consisting of either mouse or human S1ED (mS1ED, amino acids 1-233 or hS1ED, amino acids 1-232) was also used as a competitor in cell adhesion studies (Beauvais and Rapraeger, 2003; McFall and Rapraeger, 1998). Slides were blocked with 1% heat-denatured bovine serum albumin (hdBSA) for a minimum of 30 minutes at 37°C. B82L cells were detached from the substratum using 5 mM EDTA in Tris-buffered saline and resuspended in HEPES-buffered culture medium containing 0.1% hdBSA or, in appropriate experiments, 10% FBS. Cells were plated at a density of 15,000 cells per well and allowed to attach and spread for 2 hours prior to fixation in 4% paraformaldehyde in CMF-PBS. For fluorescence microscopy, fixed cells were permeabilized in 0.2% Triton X-100, labeled with Rhodamine-conjugated phalloidin and analyzed with a Nikon Microphot-FX microscope (Nikon, Inc.) equipped with a cooled CCD camera and Image-Pro Plus software (Media Cybernetics). Cells extending five or more fingerlike projections were scored as extending filopodia, whereas cells spreading with a diameter of at least 25 μm and without filopodia were scored as completely spread. Spreading was quantified from a minimum of triplicate wells and is shown as the mean ± standard error (s.e.). For antibody and mS1ED inhibition experiments, cells were pre-incubated 10 minutes before plating in the presence of the inhibitor.

siRNA design and transfection

siRNAs against the mouse β5 integrin subunit (GenBank™ accession number NM_010580.1, nucleotide annotation 269CAGGGCTCAACATATGCACTA289) and mouse Sdc1 (GenBank™ accession number NM_011519.1, nucleotide annotation 1660GAGGTCTACTTTAGACAACTT1680) were designed by Ambion, Inc. (Austin, TX) in accordance with a Cenix BioScience algorithm. For transfection, mouse β5 siRNA at 200-800 nM or mouse Sdc1 siRNA at 600 nM were added to 2.5×105 cells plated in 35-mm wells using LipofectAMINE2000 at a ratio of 1:1 (μg siRNA: μl LipofectAMINE2000) for the β5 siRNA or a ratio of 1:4 for the mouse Sdc1 siRNA diluted in Opti-MEM I transfection medium (Invitrogen) lacking serum and antibiotics. Control cells were transfected with a control oligonucleotide provided by the siRNA manufacturer. At 4 hours after transfection, each well was supplemented with 3 ml of complete growth medium; at 24 hours post-transfection the cells were lifted in trypsin and expanded in 100-mm tissue-culture plates. Cells were harvested 48 hours after transfection and 5×104 cell equivalents in Laemmli sample buffer electrophoresed per lane under non-reducing conditions on a 7.5% Laemmli gel, transferred to ImmobilonP (Millipore) and probed with 1:1000 rabbit polyclonal β5 antibody (Ab1926, Chemicon) or 1:200 rabbit anti-FAK antibody (FAK A-17, Santa Cruz Biologicals) followed by an AP-conjugated anti-rabbit secondary antibody. Alternatively, cohorts were detached with EDTA, resuspended in 100 μl HEPES-buffered DME supplemented with 10% FBS and subjected to FACS analysis using anti-integrin (BDBiosciences) mAb H9.2B8 to detect αv, HMβ1-1 to detect β1 and 2C9.G2 to detect β3 followed by a FITC-conjugated anti-Armenian hamster secondary antibody or anti-syndecan mAbs 281.2 or KY8.2 with an Alexa-Fluor-488-conjugated anti-rat secondary, mAb B-B4 with an Alexa-Fluor-488-conjugated anti-mouse secondary IgG and FITC-conjugated human IgG (hIgG) to detect mouse Sdc1, mouse Sdc4, human Sdc1 constructs and the FcRecto-hS1 chimera, respectively. Cells were analyzed at the University of Wisconsin Comprehensive Cancer Center Flow Cytometry Facility using a FACSCalibur benchtop cytometer (BDBiosciences). Cell-scatter and propidium-iodide (Sigma, 1 μg/sample) staining profiles were used to gate live, single-cell events for data analysis.

Co-immunoprecipitation assays

Antibodies used to isolate Sdc1 include an affinity-purified pan-syndecan antibody generated in rabbit against the ten C-terminal amino acids of Sdc1 (Reiland et al., 1996) and mAbs 281.2 (Jalkanen et al., 1985) and B-B4 (Wijdenes et al., 1996) specific for the ectodomains of mouse and human Sdc1, respectively. Monoclonal Ab KY8.2 (generously provided by Paul W. Kincade, Oklahoma Medical Research Foundation) was used to isolate mouse Sdc4. Monoclonal Ab 10.1 (Santa Cruz Biotechnology) or human IgG (Jackson ImmunoResearch), which recognize the ectodomain of CD64/FcγRIa, was used to isolate the FcRecto-hS1 chimera. Cells (3-5×106/ml) were washed and then lysed in lysis buffer containing 50 mM HEPES, 150 mM NaCl, 10 mM EDTA, 1% Triton X-100, 0.25% sodium deoxycholate and 1:100 dilution of protease inhibitor cocktail set III (Calbiochem) for 20 minutes on ice. Insoluble cell debris was removed by centrifugation at 20,000 g for 15 minutes at 4°C. Cell lysates (2 mg protein per reaction determined by Pierce BCA Assay per reaction) were pre-cleared using 50 μg/ml antibody isotype-matched IgG (rabbit IgG for pan-syndecan, rat IgG2A for 281.2 and KY8.2 and mouse IgG1 for B-B4 and 10.1) and 100 μl of GammaBind Sepharose (Amersham Biosciences). Pre-cleared lysates were then incubated at 4°C overnight with either 10 μg/ml of anti-syndecan antibody (pan, mAbs 281.2, B-B4 or KY8.2) or 30 μg/ml anti-FcγRIa antibody (mAb 10.1 or human IgG). For S1ED-competition experiments, 30 μM soluble GST (negative control) or GST-S1ED protein was added to the pre-cleared lysates in conjunction with anti-Sdc1 antibody. Immune complexes were precipitated with 50 μl of GammaBind Sepharose, washed with lysis buffer lacking detergents and extracted in Laemmli sample buffer. For methanol precipitates, 300 μg of total protein was precipitated overnight at -20°C in 2.5 volumes of methanol. Precipitates were washed once with 0.5 ml acetone (chilled to -20°C) and allowed to dry for 15 minutes at room temperature. Soluble material was resuspended in 50 μl of heparitinase buffer (50 mM HEPES, 50 mM NaOAc, 150 mM NaCl, 5 mM CaCl2, pH 6.5) with 2.4×10-3 IU/ml heparitinase (IBEX Technologies, Inc.) and 0.1 conventional units/ml chondroitin ABC lysase (ICN Biochemicals) for 4 hours at 37°C (with fresh enzymes added after 2 hours) to remove GAG side chains. Samples were resolved by electrophoresis under reduced conditions on a 7.5% Laemmli gel, transferred to ImmobilonP and probed with rabbit polyclonal β5 integrin (1:1000), mS1ED or pan-syndecan (1 μg/ml) antibody followed by an AP-conjugated anti-rabbit secondary. Visualization of immunoreactive bands was performed using ECF reagent (Amersham Pharmacia) and scanned on a Storm PhosphoImager (Molecular Dynamics).

This work was supported by NIH grants HD21881 and CA109010 to A.R. and a training grant stipend (T32 GM08688) to D.M.B. It was aided by the core facilities of the University of Wisconsin Comprehensive Cancer Center, supported by NIH P30-CA14520. The authors thank Deane Mosher and Donna Pesciotta-Peters for generously providing matrix ligands used in this study. The technical support and purification of antibodies provided by Andrea McWhorter is gratefully acknowledged.

Albert, M. L., Kim, J. I. and Birge, R. B. (
2000
). alphavbeta5 integrin recruits the CrkII-Dock180-rac1 complex for phagocytosis of apoptotic cells.
Nat. Cell Biol.
2
,
899
-905.
Aumailley, M., Gurrath, M., Muller, G., Calvete, J., Timpl, R. and Kessler, H. (
1991
). Arg-Gly-Asp constrained within cyclic pentapeptides. Strong and selective inhibitors of cell adhesion to vitronectin and laminin fragment P1.
FEBS Lett.
291
,
50
-54.
Beauvais, D. M. and Rapraeger, A. C. (
2003
). Syndecan-1-mediated cell spreading requires signaling by alphavbeta3 integrins in human breast carcinoma cells.
Exp. Cell Res.
286
,
219
-232.
Beauvais, D. M. and Rapraeger, A. C. (
2004
). Syndecans in tumor cell adhesion and signaling.
Reprod. Biol. Endocrinol.
2
,
3
.
Beauvais, D. M., Burbach, B. J. and Rapraeger, A. C. (
2004
). The syndecan-1 ectodomain regulates alphavbeta3 integrin activity in human mammary carcinoma cells.
J. Cell Biol.
167
,
171
-181.
Brooks, P. C., Stromblad, S., Sanders, L. C., von Schalscha, T. L., Aimes, R. T., Stetler-Stevenson, W. G., Quigley, J. P. and Cheresh, D. A. (
1996
). Localization of matrix metalloproteinase MMP-2 to the surface of invasive cells by interaction with integrin alpha v beta 3.
Cell
85
,
683
-693.
Brooks, P. C., Klemke, R. L., Schon, S., Lewis, J. M., Schwartz, M. A. and Cheresh, D. A. (
1997
). Insulin-like growth factor receptor cooperates with integrin alpha v beta 5 to promote tumor cell dissemination in vivo.
J. Clin. Invest.
99
,
1390
-1398.
Carman, C. V. and Springer, T. A. (
2003
). Integrin avidity regulation: are changes in affinity and conformation underemphasized?
Curr. Opin. Cell Biol.
15
,
547
-556.
Couchman, J. R. (
2003
). Syndecans: proteoglycan regulators of cell-surface microdomains?
Nat. Rev. Mol. Cell Biol.
4
,
926
-937.
Couchman, J. R., Chen, L. and Woods, A. (
2001
). Syndecans and cell adhesion.
Int. Rev. Cytol.
207
,
113
-150.
De, S., Chen, J., Narizhneva, N. V., Heston, W., Brainard, J., Sage, E. H. and Byzova, T. V. (
2003
). Molecular pathway for cancer metastasis to bone.
J. Biol. Chem.
278
,
39044
-39050.
Dedhar, S. (
1999
). Integrins and signal transduction.
Curr. Opin. Hematol.
6
,
37
-43.
Eliceiri, B. P. and Cheresh, D. A. (
1999
). The role of alphav integrins during angiogenesis: insights into potential mechanisms of action and clinical development.
J. Clin. Invest.
103
,
1227
-1230.
Felding-Habermann, B. and Cheresh, D. A. (
1993
). Vitronectin and its receptors.
Curr. Opin. Cell Biol.
5
,
864
-868.
Filardo, E. J., Deming, S. L. and Cheresh, D. A. (
1996
). Regulation of cell migration by the integrin beta subunit ectodomain.
J. Cell Sci.
109
,
1615
-1622.
Finnemann, S. C. (
2003a
). Focal adhesion kinase signaling promotes phagocytosis of integrin-bound photoreceptors.
EMBO J.
22
,
4143
-4154.
Finnemann, S. C. (
2003b
). Role of alphavbeta5 integrin in regulating phagocytosis by the retinal pigment epithelium.
Adv. Exp. Med. Biol.
533
,
337
-342.
Friedlander, M., Brooks, P. C., Shaffer, R. W., Kincaid, C. M., Varner, J. A. and Cheresh, D. A. (
1995
). Definition of two angiogenic pathways by distinct alpha v integrins.
Science
270
,
1500
-1502.
Giancotti, F. G. and Ruoslahti, E. (
1999
). Integrin signaling.
Science
285
,
1028
-1032.
Granes, F., Urena, J. M., Rocamora, N. and Vilaro, S. (
2000
). Ezrin links syndecan-2 to the cytoskeleton.
J. Cell Sci.
113
,
1267
-1276.
Grootjans, J. J., Zimmermann, P., Reekmans, G., Smets, A., Degeest, G., Durr, J. and David, G. (
1997
). Syntenin, a PDZ protein that binds syndecan cytoplasmic domains.
Proc. Natl. Acad. Sci. USA
94
,
13683
-13688.
Hall, M. O., Abrams, T. A. and Burgess, B. L. (
2003
). Integrin alphavbeta5 is not required for the phagocytosis of photoreceptor outer segments by cultured retinal pigment epithelial cells.
Exp. Eye Res.
77
,
281
-286.
Hsueh, Y. P., Yang, F. C., Kharazia, V., Naisbitt, S., Cohen, A. R., Weinberg, R. J. and Sheng, M. (
1998
). Direct interaction of CASK/LIN-2 and syndecan heparan sulfate proteoglycan and their overlapping distribution in neuronal synapses.
J. Cell Biol.
142
,
139
-151.
Iba, K., Albrechtsen, R., Gilpin, B., Frohlich, C., Loechel, F., Zolkiewska, A., Ishiguro, K., Kojima, T., Liu, W., Langford, J. K. et al. (
2000
). The cysteine-rich domain of human ADAM 12 supports cell adhesion through syndecans and triggers signaling events that lead to beta1 integrin-dependent cell spreading.
J. Cell Biol.
149
,
1143
-1156.
Iida, J., Meijne, A. M., Oegema, T. R., Jr, Yednock, T. A., Kovach, N. L., Furcht, L. T. and McCarthy, J. B. (
1998
). A role of chondroitin sulfate glycosaminoglycan binding site in alpha4beta1 integrin-mediated melanoma cell adhesion.
J. Biol. Chem.
273
,
5955
-5962.
Iida, J., Pei, D., Kang, T., Simpson, M. A., Herlyn, M., Furcht, L. T. and McCarthy, J. B. (
2001
). Melanoma chondroitin sulfate proteoglycan regulates matrix metalloproteinase-dependent human melanoma invasion into type I collagen.
J. Biol. Chem.
276
,
18786
-18794.
Jalkanen, M., Nguyen, H., Rapraeger, A., Kurn, N. and Bernfield, M. (
1985
). Heparan sulfate proteoglycans from mouse mammary epithelial cells: localization on the cell surface with a monoclonal antibody.
J. Cell Biol.
101
,
976
-984.
Jones, J., Watt, F. M. and Speight, P. M. (
1997
). Changes in the expression of alpha v integrins in oral squamous cell carcinomas.
J. Oral Pathol. Med.
26
,
63
-68.
Kinnunen, T., Kaksonen, M., Saarinen, J., Kalkkinen, N., Peng, H. B. and Rauvala, H. (
1998
). Cortactin-Src kinase signaling pathway is involved in N-syndecan-dependent neurite outgrowth.
J. Biol. Chem.
273
,
10702
-10708.
Klemke, R. L., Yebra, M., Bayna, E. M. and Cheresh, D. A. (
1994
). Receptor tyrosine kinase signaling required for integrin alpha v beta 5-directed cell motility but not adhesion on vitronectin.
J. Cell Biol.
127
,
859
-866.
Kugler, M. C., Wei, Y. and Chapman, H. A. (
2003
). Urokinase receptor and integrin interactions.
Curr. Pharm. Des.
9
,
1565
-1574.
Kusano, Y., Oguri, K., Nagayasu, Y., Munesue, S., Ishihara, M., Saiki, I., Yonekura, H., Yamamoto, H. and Okayama, M. (
2000
). Participation of syndecan 2 in the induction of stress fiber formation in cooperation with integrin alpha5beta1: structural characteristics of heparan sulfate chains with avidity to COOH-terminal heparin-binding domain of fibronectin.
Exp. Cell Res.
256
,
434
-444.
Kusano, Y., Yoshitomi, Y., Munesue, S., Okayama, M. and Oguri, K. (
2004
). Cooperation of syndecan-2 and syndecan-4 among cell surface heparan sulfate proteoglycans in the actin cytoskeletal organization of Lewis lung carcinoma cells.
J. Biochem.
135
,
129
-137.
Lebakken, C. S. and Rapraeger, A. C. (
1996
). Syndecan-1 mediates cell spreading in transfected human lymphoblastoid (Raji) cells.
J. Cell Biol.
132
,
1209
-1221.
Lebakken, C. S., McQuade, K. J. and Rapraeger, A. C. (
2000
). Syndecan-1 signals independently of beta1 integrins during Raji cell spreading.
Exp. Cell Res.
259
,
315
-325.
Lewis, J. M., Cheresh, D. A. and Schwartz, M. A. (
1996
). Protein kinase C regulates alpha v beta 5-dependent cytoskeletal associations and focal adhesion kinase phosphorylation.
J. Cell Biol.
134
,
1323
-1332.
Liddington, R. C. and Ginsberg, M. H. (
2002
). Integrin activation takes shape.
J. Cell Biol.
158
,
833
-839.
Liu, W., Litwack, E. D., Stanley, M. J., Langford, J. K., Lander, A. D. and Sanderson, R. D. (
1998
). Heparan sulfate proteoglycans as adhesive and anti-invasive molecules. Syndecans and glypican have distinct functions.
J. Biol. Chem.
273
,
22825
-22832.
Lukashev, M. E. and Werb, Z. (
1998
). ECM signalling: orchestrating cell behaviour and misbehaviour.
Trends Cell Biol.
8
,
437
-441.
McFall, A. J. and Rapraeger, A. C. (
1998
). Characterization of the high affinity cell-binding domain in the cell surface proteoglycan syndecan-4.
J. Biol. Chem.
273
,
28270
-28276.
McLean, J. W., Vestal, D. J., Cheresh, D. A. and Bodary, S. C. (
1990
). cDNA sequence of the human integrin beta 5 subunit.
J. Biol. Chem.
265
,
17126
-17131.
McQuade, K. J. and Rapraeger, A. C. (
2003
). Syndecan-1 transmembrane and extracellular domains have unique and distinct roles in cell spreading.
J. Biol. Chem.
278
,
46607
-46615.
Memmo, L. M. and McKeown-Longo, P. (
1998
). The alphavbeta5 integrin functions as an endocytic receptor for vitronectin.
J. Cell Sci.
111
,
425
-433.
Mostafavi-Pour, Z., Askari, J. A., Parkinson, S. J., Parker, P. J., Ng, T. T. and Humphries, M. J. (
2003
). Integrin-specific signaling pathways controlling focal adhesion formation and cell migration.
J. Cell Biol.
161
,
155
-167.
Moulder, K., Roberts, K., Shevach, E. M. and Coligan, J. E. (
1991
). The mouse vitronectin receptor is a T cell activation antigen.
J. Exp. Med.
173
,
343
-347.
Munesue, S., Kusano, Y., Oguri, K., Itano, N., Yoshitomi, Y., Nakanishi, H., Yamashina, I. and Okayama, M. (
2002
). The role of syndecan-2 in regulation of actin-cytoskeletal organization of Lewis lung carcinoma-derived metastatic clones.
Biochem. J.
363
,
201
-209.
Noto, K., Kato, K., Okumura, K. and Yagita, H. (
1995
). Identification and functional characterization of mouse CD29 with a mAb.
Int. Immunol.
7
,
835
-842.
Oh, E. S., Woods, A. and Couchman, J. R. (
1997a
). Multimerization of the cytoplasmic domain of syndecan-4 is required for its ability to activate protein kinase C.
J. Biol. Chem.
272
,
11805
-11811.
Oh, E. S., Woods, A. and Couchman, J. R. (
1997b
). Syndecan-4 proteoglycan regulates the distribution and activity of protein kinase C.
J. Biol. Chem.
272
,
8133
-8136.
Oh, E. S., Woods, A., Lim, S. T., Theibert, A. W. and Couchman, J. R. (
1998
). Syndecan-4 proteoglycan cytoplasmic domain and phosphatidylinositol 4,5-bisphosphate coordinately regulate protein kinase C activity.
J. Biol. Chem.
273
,
10624
-10629.
Ott, V. L. and Rapraeger, A. C. (
1998
). Tyrosine phosphorylation of syndecan-1 and -4 cytoplasmic domains in adherent B82 fibroblasts.
J. Biol. Chem.
273
,
35291
-35298.
Panetti, T. S. and McKeown-Longo, P. J. (
1993
). The alpha v beta 5 integrin receptor regulates receptor-mediated endocytosis of vitronectin.
J. Biol. Chem.
268
,
11492
-11495.
Panetti, T. S., Wilcox, S. A., Horzempa, C. and McKeown-Longo, P. J. (
1995
). Alpha v beta 5 integrin receptor-mediated endocytosis of vitronectin is protein kinase C-dependent.
J. Biol. Chem.
270
,
18593
-18597.
Park, H., Kim, Y., Lim, Y., Han, I. and Oh, E. S. (
2002
). Syndecan-2 mediates adhesion and proliferation of colon carcinoma cells.
J. Biol. Chem.
277
,
29730
-29736.
Pasqualini, R., Bodorova, J., Ye, S. and Hemler, M. E. (
1993
). A study of the structure, function and distribution of beta 5 integrins using novel anti-beta 5 monoclonal antibodies.
J. Cell Sci.
105
,
101
-111.
Plow, E. F., Haas, T. A., Zhang, L., Loftus, J. and Smith, J. W. (
2000
). Ligand binding to integrins.
J. Biol. Chem.
275
,
21785
-21788.
Porter, J. C. and Hogg, N. (
1998
). Integrins take partners: cross-talk between integrins and other membrane receptors.
Trends Cell. Biol.
8
,
390
-396.
Reiland, J., Ott, V. L., Lebakken, C. S., Yeaman, C., McCarthy, J. and Rapraeger, A. C. (
1996
). Pervanadate activation of intracellular kinases leads to tyrosine phosphorylation and shedding of syndecan-1.
Biochem. J.
319
,
39
-47.
Sanders, R. J., Mainiero, F. and Giancotti, F. G. (
1998
). The role of integrins in tumorigenesis and metastasis.
Cancer Invest.
16
,
329
-344.
Saoncella, S., Echtermeyer, F., Denhez, F., Nowlen, J. K., Mosher, D. F., Robinson, S. D., Hynes, R. O. and Goetinck, P. F. (
1999
). Syndecan-4 signals cooperatively with integrins in a Rho-dependent manner in the assembly of focal adhesions and actin stress fibers.
Proc. Natl. Acad. Sci. USA
96
,
2805
-2810.
Thodeti, C. K., Albrechtsen, R., Grauslund, M., Asmar, M., Larsson, C., Takada, Y., Mercurio, A. M., Couchman, J. R. and Wewer, U. M. (
2003
). ADAM12/syndecan-4 signaling promotes beta 1 integrin-dependent cell spreading through protein kinase Calpha and RhoA.
J. Biol. Chem.
278
,
9576
-9584.
Tkachenko, E. and Simons, M. (
2002
). Clustering induces redistribution of syndecan-4 core protein into raft membrane domains.
J. Biol. Chem.
277
,
19946
-19951.
Tkachenko, E., Rhodes, J. M. and Simons, M. (
2005
). Syndecans: new kids on the signaling block.
Circ. Res.
96
,
488
-500.
Tonn, J. C., Wunderlich, S., Kerkau, S., Klein, C. E. and Roosen, K. (
1998
). Invasive behaviour of human gliomas is mediated by interindividually different integrin patterns.
Anticancer Res.
18
,
2599
-2605.
van Leeuwen, R. L., Yoshinaga, I. G., Akasaka, T., Dekker, S. K., Vermeer, B. J. and Byers, H. R. (
1996
). Attachment, spreading and migration of melanoma cells on vitronectin. The role of alpha V beta 3 and alpha V beta 5 integrins.
Exp. Dermatol.
5
,
308
-315.
Wijdenes, J., Vooijs, W. C., Clement, C., Post, J., Morard, F., Vita, N., Laurent, P., Sun, R. X., Klein, B. and Dore, J. M. (
1996
). A plasmocyte selective monoclonal antibody (B-B4) recognizes syndecan-1.
Br. J. Haematol.
94
,
318
-323.
Wilkins-Port, C. E., Sanderson, R. D., Tominna-Sebald, E. and McKeown-Longo, P. J. (
2003
). Vitronectin's basic domain is a syndecan ligand which functions in trans to regulate vitronectin turnover.
Cell Commun. Adhes.
10
,
85
-103.
Woods, A. and Couchman, J. R. (
2000
). Integrin modulation by lateral association.
J. Biol. Chem.
275
,
24233
-24236.
Woods, A. and Couchman, J. R. (
2001
). Syndecan-4 and focal adhesion function.
Curr. Opin. Cell Biol.
13
,
578
-583.
Woods, A., Couchman, J. R., Johansson, S. and Hook, M. (
1986
). Adhesion and cytoskeletal organisation of fibroblasts in response to fibronectin fragments.
EMBO J.
5
,
665
-670.
Yamashita, Y., Oritani, K., Miyoshi, E. K., Wall, R., Bernfield, M. and Kincade, P. W. (
1999
). Syndecan-4 is expressed by B lineage lymphocytes and can transmit a signal for formation of dendritic processes.
J. Immunol.
162
,
5940
-5948.
Yasuda, M., Hasunuma, Y., Adachi, H., Sekine, C., Sakanishi, T., Hashimoto, H., Ra, C., Yagita, H. and Okumura, K. (
1995
). Expression and function of fibronectin binding integrins on rat mast cells.
Int. Immunol.
7
,
251
-258.
Yebra, M., Filardo, E. J., Bayna, E. M., Kawahara, E., Becker, J. C. and Cheresh, D. A. (
1995
). Induction of carcinoma cell migration on vitronectin by NF-kappa B-dependent gene expression.
Mol. Biol. Cell
6
,
841
-850.

Supplementary information