Agrin, a basal lamina-associated proteoglycan, is a crucial nerve-derived organizer of postsynaptic differentiation at the skeletal neuromuscular junction. Because integrins serve as cellular receptors for many basal lamina components, we asked whether agrin interacts with integrins. Agrin-induced aggregation of acetylcholine receptors on cultured myotubes was completely blocked by antibodies to the β1 integrin subunit and partially blocked by antibodies to the αv integrin subunit. Agrin-induced clustering was also inhibited by antisense oligonucleotides to αv and a peptide that blocks the αv binding site. Non-muscle cells that expressed αv and β1 integrin subunits adhered to immobilized agrin, and this adhesion was blocked by anti-αv and anti-β1 antibodies. Integrin αv-negative cells that did not adhere to agrin were rendered adherent by introduction of αv. Together, these results implicate integrins, including αvβ1, as components or modulators of agrin’s signal transduction pathway.

During synapse formation at the vertebrate skeletal neuromuscular junction, the motor axon induces the formation of postsynaptic specializations in the muscle fiber (Hall and Sanes, 1993). A crucial signaling molecule in this process is agrin, a heparan sulfate proteoglycan (Tsen et al., 1995; Denzer et al., 1995) that is synthesized and secreted by motoneurons and incorporated into the basal lamina of the synaptic cleft (McMahan, 1990; Reist et al., 1992). Application of agrin to cultured myotubes stimulates formation of specialized domains that contain high density aggregates of neurotransmitter receptors (acetylcholine receptors or AChRs), associated with numerous cytoskeletal, membrane and extracellular matrix components that are synapse-specific in vivo (Denzer et al., 1996). Conversely, agrin-induced AChR clustering and other aspects of postsynaptic differentiation are grossly deficient in mutant mice lacking agrin (Gautam et al., 1996). Agrin is also expressed by other central and peripheral neurons and may play roles in synapse formation in the brain (Bowe and Fallon, 1995).

In view of its central role in synaptogenesis, it is important to identify the receptors that mediate agrin’s effects. A critical component of the agrin receptor is a muscle-specific receptor tyrosine kinase called MuSK. MuSK is selectively expressed in skeletal muscle and concentrated, along with AChRs, in the postsynaptic membrane (Valenzuela et al., 1995; Apel et al., 1997). Mutant mice lacking MuSK fail to form neuromuscular junctions, showing a phenotype similar to (but more severe than) that of agrin mutants (DeChiara et al., 1996). Moreover, MuSK-deficient myotubes are unresponsive to agrin, a dominant negative MuSK mutant inhibits agrin-induced AChR clustering, and chemical cross-linkers can attach agrin to MuSK on the myotube surface (Glass et al., 1996, 1997). However, agrin does not bind to soluble forms of MuSK or to MuSK expressed in non-muscle cells (Glass et al., 1996). It is therefore likely that there are additional receptors in muscle that link agrin to MuSK and/or act in parallel to MuSK.

Attractive candidates for the role of co-receptor are the integrins, membrane proteins that serve as signal transducing receptors for numerous extracellular matrix proteins in both neural and nonneural tissues. Integrins are heterodimers of α and β subunits; 16 α and 8 β subunits have been identified so far in vertebrates. Both α and β subunits span the membranes; their extracellular domains cooperate to bind ligands and their intracellular domains transmit signals to the cell’s interior. Heterodimers differ in what ligands they bind and in what signals they transmit, so the repertoire of integrins that a cell expresses is a major determinant of how it responds to the environment (Hynes, 1992; Sonnerberg, 1993; Schwartz et al., 1995).

Several lines of evidence raised the possibility that integrins might play a role in agrin signaling. First, integrins have been shown to transduce many of the developmentally important signals that components of the extracellular matrix provide (Hynes, 1992, 1996). Second, integrins mediate most wellcharacterized developmental effects of laminin, a basal lamina protein that shares several regions of homology with agrin (Rupp et al., 1991; Ruegg et al., 1992; Mercurio, 1995; Delwel and Sonnenberg, 1996; Engvall and Wewer, 1996). Third, integrins are abundant in both developing and adult muscle. At least eight alpha subunits (α1, α3-α7, α9 and αv) and three beta subunits (β1, β3 and β5) are expressed in embryonic muscle (Bronner-Fraser et al., 1992; Duband et al., 1992; Lakonishok et al., 1992; Rosen et al., 1992; Bao et al, 1993; Enomoto et al., 1993; George-Weinstein et al., 1993; Palmer et al., 1993; Hirsch et al., 1994; McDonald et al., 1995; Gullberg et al., 1995a,b; Blaschuk et al., 1997) and at least four of these – α3, α7, αv and β1 – are present in the postsynaptic membrane at the adult neuromuscular junction (Bozyczko et al., 1989; Martin et al., 1996), as might be expected for an agrin receptor. Finally, recent studies have shown that integrins can interact, both directly and indirectly, with other receptor systems including some that, like MuSK, are receptor tyrosine kinases (Schwartz et al., 1995; Dedhar and Hannyan, 1996; Miyamoto et al., 1996; Sastry and Horwitz, 1996). Based on these parallels, we undertook to test the possibility that integrins are involved in agrin’s signal transduction pathway.

Reagents

Sources of antibodies to integrins were as follows: JG22 (anti-chick β1; Greve and Gottlieb, 1982) and H1B4 (anti-chick α7; Bao et al., 1993) were obtained from Developmental Hybridoma Bank (Iowa City, IA). O26 (anti-rat α7; Song et al., 1992) was a gift from S. Kaufman (U. Illinois). TASC (anti-chick β1; Neugebauer and Reichardt, 1991), Chav (anti-chick αv; Neugebauer et al., 1991) and Ralph (anti-rat α3; De Freitas et al., 1995) were gifts from L. Reichard (UCSF). L230 (anti-human αv; Houghton et al., 1982), P5D2 (antihuman α3; Carter et al., 1991) and 7G2 (anti-human β3; Gresham et al., 1989) were gifts from S. Blystone and E. Brown (Washington University). P4C10 (anti-human β1; Carter et al., 1990), P1B5 (anti-human α1; Wayner et al., 1988) and GoH3 (anti-human and mouse α6; Sonnenberg et al., 1987) were purchased from GibcoBRL (Gaithersburg, MD). Antibodies were partially purified from hybridoma-conditioned medium by precipitation in 50% ammonium sulfate, followed by extensive dialysis in phosphate-buffered saline (PBS). Antibody concentrations were estimated by SDS-gel electrophoresis, using purified immunoglobulin as a standard.

Antisense oligonucleotides with thio (S) diester linkages were obtained from Oligos Etc., (Portland, OR). Sequences used were: α6: 5′GGCCATGGGCGGGGCGGTGGG; α7: 5′GGCCATGGGAT-CAACGCTCTC; αv: 5′AGCCATCGCCGGGGCGCGCGC; and reversed αv (control): 5′CGCGCGCGGGGCCGCTACCGA. Peptides (GRGDSP and GRGESP) were purchased from GibcoBRL.

Cell culture

Primary cultures of chicken muscle were prepared from embryonic day (E)11 pectoral muscles. Cleanly dissected muscle was incubated in 0.2% trypsin for 30 minutes, triturated and cultured in Dulbecco’s Modified Eagles Media (DMEM) with 10% horse serum and 5% chick embryo extract. 3-5 days prior to addition of agrin, cells were transferred to media lacking chick embryo extract. Primary cultures of rat muscle were prepared from leg muscles of neonate pups. Dissected tissue was minced and trypsinized for 30-40 minutes in 0.2% trypsin, then triturated and cultured in DMEM plus 20% fetal calf serum (FCS). Fusion was induced by transfer to DMEM with 2% horse serum. C2 myogenic cells (ATCC, Bethesda MD) were cultured as myoblasts in DMEM with 20% fetal calf serum. Cells were grown to confluence then induced to form myotubes by changing the growth medium to DMEM with 2% horse serum. QT6 cells (ATCC, Bethesda, MD) were grown in Earl’s 199 medium with 10% tryptose phosphate broth, 5% FCS, and 1% DMSO. K562 cells (provided by S. Blyestone and E. Brown, Washington University) were grown in Isocove’s modified Eagles media (IMEM) with 10% FCS. Cells were transfected with human αv and β3 cDNAs as described by Blyestone et al. (1994). NCI-H69 cells (ATCC) were grown in RPMI with 10% FCS. 293 cells (ATCC) were grown in DMEM with 10% FCS.

All media contained 50 U/ml penicillin, and 50 μg/ml streptomycin, and all cells were incubated at 37°C in a 5% CO2 atmosphere.

AChR clustering

AChR clustering was induced with a recombinant 95 kDa C-terminal agrin fragment of the x=12, y=4, z=8 form, prepared as described by Ferns et al. (1993) and generously provided by J. Campanelli and R. Scheller (Stanford University). The agrin was electrophoretically pure, and was always added at 10 ng/ml, except for the experiment shown in Fig. 2B, where it was added at 5 ng/ml. Clustering was assayed by adding 50 nM rhodamine-α-bungarotoxin (Molecular Probes, Eugene, OR) to live cells for 1 hour. Cultures were then washed in PBS, fixed in 1% paraformaldehyde, mounted in paraphenylenediamine/glycerol and viewed with an epifluorescence microscope. Clustering was quantitated as detailed by Martin and Sanes (1995).

For experiments using antibodies or peptides, the blocking reagent was added to myotubes for one hour, then agrin was added in the continued presence of blocking reagent for an additional 4-5 hours. For experiments using antisense oligonucleotides, myotubes were washed once each with PBS and OPTI-MEM serum-free medium (GibcoBRL). Nucleotides were then added at 0.2 or 2 μM in OPTI-MEM containing 6.6 μg/ml Lipofectin (GibcoBRL). Following incubation for 4 hours, cells were washed once each with OPTI-MEM and PBS, then incubated in fresh low serum media for 24 hours. Agrin was then added for 18 hours, after which AChR clustering was assayed as above.

Cell adhesion

For use in adhesion assays, cells were detached from culture dishes without enzymes by trituration in Ca-, Mg-free saline. To prepare substrata for adhesion assays, culture dishes were coated with nitrocellulose as described by Lagenaur and Lemmon (1987), then 2 μl drops of the agrin fragment described above (20 μg/ml) or of laminin-1 (20 μg/ml; GibcoBRL) were added. After 30 minutes incubation, dishes were washed with PBS and then incubated with 10 mg/ml BSA to block residual binding sites on the nitrocellulose. Cells were added (40-50,000 in 0.4-0.5 ml of saline containing Mgand Ca), except where noted, and incubated for 60-90 minutes at 37°C, with agitation every 15 minutes. Nonadherent cells were then washed off, and adherent cells were fixed in 2% paraformaldehyde and counted. Cells were counted on one-quarter of a substratum drop in each well; this corresponds to approximately 10 microscope fields. In assays on 293 cells, 20 μg/ml heparin (Sigma) was included in the buffer to reduce non-specific adhesion.

β1 integrins modulate agrin signaling

All of the eight integrin α chains known to be expressed in muscle (see Introduction) form heterodimers with the β1 subunit, and some dimerize only with β1 (Sonnerberg, 1993). We therefore began our study by asking whether JG22 (Greve and Gottlieb, 1982), an antibody that blocks all β1 integrins, affected agrin function in muscle. To this end, we assayed the ability of recombinant agrin to stimulate formation of high density AChR clusters in cultured myotubes. This assay was used to isolate agrin (McMahan, 1990) and reproduces a crucial aspect of agrin-dependent synaptogenesis in vivo (Gautum et al., 1996).

Small numbers of AChR clusters formed spontaneously on chick myotubes, but 5-to 10-fold more clusters appeared within 5 hours following application of 10 nM agrin (Fig. 1A,C). Addition of JG22 inhibited agrin-induced clustering, but not agrin-independent ‘spontaneous’ clustering (Fig. 1B,D). Inhibition was dose dependent and control antibodies were ineffective (Fig. 2A). Thus, agrin function requires β1 integrins.

Fig. 1.

Effects of agrin, anti-integrin β1 or both on AChR clustering. Cultured chick myotubes were treated with anti-integrin β1 (JG22, 100 μg/ml) and/or agrin (10 ng/ml), then labelled with rhodamine-α-bungarotoxin to visualize AChR clusters. (A) No treatment; small numbers of clusters form spontaneously. (B) Anti-integrin β1 did not affect spontaneous AChR clustering. (C) Agrin induced clustering. (D) Anti-integrin β1 inhibited agrin-induced clustering. Bar is 50 μm.

Fig. 1.

Effects of agrin, anti-integrin β1 or both on AChR clustering. Cultured chick myotubes were treated with anti-integrin β1 (JG22, 100 μg/ml) and/or agrin (10 ng/ml), then labelled with rhodamine-α-bungarotoxin to visualize AChR clusters. (A) No treatment; small numbers of clusters form spontaneously. (B) Anti-integrin β1 did not affect spontaneous AChR clustering. (C) Agrin induced clustering. (D) Anti-integrin β1 inhibited agrin-induced clustering. Bar is 50 μm.

Fig. 2.

β1 integrins modulate agrin-induced AChR clustering. (A) Effects of JG22, a blocking antibody to integrin β1, on AChR clustering in the presence (open bars) or absence (dark bars) of agrin. A similarly prepared antibody to an intracellular epitope was used as a control. (B) Effects of TASC (50 μg/ml), an activating antibody to integrin β1, on AChR clustering in the presence (open bars) or absence (dark bars) of 5 ng/ml agrin. TASC potentiates the effects of agrin. Bars show mean ± s.e.m. of values from 3-6 (A) or 5 (B) experiments.

Fig. 2.

β1 integrins modulate agrin-induced AChR clustering. (A) Effects of JG22, a blocking antibody to integrin β1, on AChR clustering in the presence (open bars) or absence (dark bars) of agrin. A similarly prepared antibody to an intracellular epitope was used as a control. (B) Effects of TASC (50 μg/ml), an activating antibody to integrin β1, on AChR clustering in the presence (open bars) or absence (dark bars) of 5 ng/ml agrin. TASC potentiates the effects of agrin. Bars show mean ± s.e.m. of values from 3-6 (A) or 5 (B) experiments.

The inhibition of agrin’s effects by JG22 suggested that integrins are required specifically for agrin signaling, but it was also possible that blockade of integrin perturbed myotube stability more generally, and thereby led indirectly to an inability to form AChR clusters. For example, adhesion of muscle cells to substrata is integrin-mediated (Neff et al., 1982) and myotubes detached from the gelatin-coated culture dishes at high JG22 antibody concentrations (⩾200 μg/ml; not shown). However, blockade of agrin-dependent AChR clustering occurred at lower concentrations of antibody, which did not perturb adhesion (Fig. 2A). Moreover, JG22 had no effect on agrin-independent clustering, indicating that β1 integrins are not required for AChR aggregation per se (Fig. 2A). As an additional means of distinguishing direct from indirect effects, we tested an ‘activating’ antibody to β1 integrin, TASC, which does not block integrin function but rather serves to enhance the sensitivity of integrins to ligands (Neugebauer and Reichardt, 1991). TASC had no effect on AChR clustering when added alone or in the presence of a saturating dose of agrin (data not shown). However, TASC more than doubled the number of AChR clusters induced by a submaximal amount of agrin (Fig. 2B). The interaction of TASC and agrin places β1 integrins on agrin’s signal transduction pathway.

αv integrins modulate agrin signaling

We next asked which integrin α chains were associated with the agrin-responsive β1 integrins. We focused on the α3, α6, α7 and αv chains, for the following reasons: (a) All four chains are present in embryonic muscle at the time when synapses are forming (see Introduction). (b) All four bind laminin (Delwel and Sonnenberg, 1996), which contains several regions of homology to agrin (Rupp et al., 1991; Ruegg et al., 1992; Engvall and Wewer, 1996). (c) The α3, α7 and αv subunits are present in the postsynaptic membrane at adult neuromuscular junctions (Martin et al., 1996). (d) The α3 and αv subunits are upregulated during muscle regeneration (J. R. S., unpublished observations) on fibers known to be competent to receive synapses.

Cultured myotubes were treated with blocking antibodies to α3, α6, α7 or αv integrins. In each case, myotubes were cultured from the species for which antibodies were available: rat for α3 and α7, chick for αv, and mouse for α6. Antibodies to the α3, α6 and α7 subunits had no significant effect, but antiintegrin αv inhibited agrin-induced AChR clustering by ∼40% (Fig. 3A). This result suggests that αvβ1 is one, but perhaps not the only, integrin that modulates agrin signaling.

Fig. 3.

αv integrins modulate agrin-induced AChR clustering. Individual integrin subunits were blocked with antibodies (A), peptides (B) or oligonucleotides (C). (A) Blocking antibodies to the integrin α3, α6, α7 or αv subunits were added to agrin-treated myotubes from the appropriate species and their ability to block AChR clustering was assessed. Anti-α3 (Ralph) was added to rat myotubes at 200 μg/ml, anti-α6 (GoH3) to mouse myotubes at 100 μg/ml, anti-α7 (O26) to rat myotubes at 250 μg/ml and anti-αv (Chav) to chick myotubes at 200 μg/ml. Only anti-αv blocked clustering. (B) GRGDSP (80 μg/ml), which blocks the ligand binding site of αv, inhibited agrin-induced AChR clustering in C2 cells. The control peptide, GRGESP, was without effect. (C) Antisense oligonucleotides complementary to integrin α6, α7 and αv mRNAs were added at 200 nM to mouse C2 cells. The reverse of the αv sequence was used as a control. Only the antisense αv sequence inhibited agrin-induced clustering. Antisense α7 and control oligonucleotides were also without effect and α6 oligonucleotide was slightly stimulatory, at 2 μM. Bars show mean ± s.e.m. of values from 4-5 (A), 4 (B), or 5-6 (C) independent experiments.

Fig. 3.

αv integrins modulate agrin-induced AChR clustering. Individual integrin subunits were blocked with antibodies (A), peptides (B) or oligonucleotides (C). (A) Blocking antibodies to the integrin α3, α6, α7 or αv subunits were added to agrin-treated myotubes from the appropriate species and their ability to block AChR clustering was assessed. Anti-α3 (Ralph) was added to rat myotubes at 200 μg/ml, anti-α6 (GoH3) to mouse myotubes at 100 μg/ml, anti-α7 (O26) to rat myotubes at 250 μg/ml and anti-αv (Chav) to chick myotubes at 200 μg/ml. Only anti-αv blocked clustering. (B) GRGDSP (80 μg/ml), which blocks the ligand binding site of αv, inhibited agrin-induced AChR clustering in C2 cells. The control peptide, GRGESP, was without effect. (C) Antisense oligonucleotides complementary to integrin α6, α7 and αv mRNAs were added at 200 nM to mouse C2 cells. The reverse of the αv sequence was used as a control. Only the antisense αv sequence inhibited agrin-induced clustering. Antisense α7 and control oligonucleotides were also without effect and α6 oligonucleotide was slightly stimulatory, at 2 μM. Bars show mean ± s.e.m. of values from 4-5 (A), 4 (B), or 5-6 (C) independent experiments.

In light of the incomplete blockade by anti-αv, we used two additional methods to assess the role of this subunit in agrindependent signaling. First, we tested a blocking peptide that contains the sequence RGD, which is a crucial determinant of the binding site in ligands of a subset of integrins, including αv integrins (Ruoshlahti, 1996). Although agrin does not contain an RGD sequence, it has been shown that αv integrins bind both RGD-containing and non-RGD-containing ligands, and that RGD peptides inhibit interactions of αv integrins with both types of ligands (Vogel et al., 1993; Piali et al., 1995; Buckley et al., 1996; Sakamoto et al., 1996). GRGDSP inhibited agrin-induced AChR clustering in mouse myotubes by ∼40%, whereas an inactive peptide, GRGESP, had no effect (Fig. 3B). Second, we attenuated integrin expression with antisense oligonucleotides complementary to the translation initiation sites of the integrin α6, α7 and αv subunits. The oligonucleotides were introduced into C2 myotubes, then agrin was added and AChR clustering assayed. Anti-α7 and control oligonucleotides had no effect on clustering at concentrations up to 2 μM, and anti-α6 oligonucleotides enhanced the effect of agrin slightly (Fig. 3C and data not shown). In contrast, the oligonucleotide complementary to the αv sequence completely inhibited agrin-induced AChR clustering at a concentration of 0.2 μM (Fig. 3C). Thus, three types of reagents directed against αv integrins – antibodies, peptides, and antisense oligonucleotides – all inhibited agrin-dependent AChR clustering, strongly implicating these integrins in agrin’s signaling mechanism.

Integrin αvβ1 mediates cell adhesion to agrin

Results presented so far suggest that integrins are agrin receptors. However, it was also possible that integrins bound some distinct factor in the culture medium or on the substratum that was required for agrin to exert its effects. We therefore conducted short-term cell adhesion assays in saline to seek a direct interaction of agrin with integrins. Myotubes are poorly suited for adhesion assays, but we found that cells of the better suited quail QT-6 fibroblast line expressed both αv and β1 integrins subunits (Galileo et al., 1992 and data not shown). Accordingly, we asked whether QT-6 cells adhered to substrata coated with the purified, recombinant agrin fragment that we had used to induce AChR clustering. In fact, QT-6 cells adhered to agrin-coated substrata in a magnesium-dependent manner (Fig. 4A). This dependence on magnesium is characteristic of integrin-mediated adhesion, because binding of ligands to integrin requires a divalent cation (Hynes, 1992). Adhesion was completely inhibited by JG22, the anti-β1 antibody that blocked agrin-dependent AChR clustering in chick myotubes, and partially inhibited by Chav, the anti-αv antibody that partially blocked clustering (Figs 2A, 3A, 4A). Adhesion was also partially blocked by the GRGDSP peptide, but was unaffected by control antibodies or peptides (Fig. 4A). Thus, integrins containing αv and/or β1 subunits are likely to bind directly to agrin.

Fig. 4.

αv and β1 integrins mediate adhesion of cells to immobilized agrin. (A) QT6 cells, which express β1 and αv integrins, bind to immobilized agrin. Adhesion required Mgand was blocked by a β1 integrin blocking antibody (JG22, 100 μg/ml). Anti-integrin αv (Chav; 200 μg/ml) and the αv-blocking peptide (GRGDSP, 80 μg/ml) blocked adhesion ~50% whereas control antibodies and peptides were without effect. (B) 293 cells express a variety of β1 integrins, including αvβ1, but no αv integrins other αvβ1. Adhesion of these cells to agrin, like that of QT6 cells, required Mgand was blocked by anti-integrin β1 (P4C10; 20 μg/ml). Adhesion was partially inhibited by anti-integrin αv (L230; 20 μg/ml) or the αv blocking peptide (GRGDSP, 80 μg/ml). Antibodies to integrins α3 (P1B5; 20 μg/ml) or α6 (GoH3; 20 μg/ml), both of which are expressed by 293 cells, or anti-integrin β3 (7G2; 20 μg/ml) had no effect. Bars show mean ± s.e.m. of 4-5(A) or 2-5 (B) experiments. Mean numbers of cells per well counted in controls (100% values) were 870 in A and 383 in B.

Fig. 4.

αv and β1 integrins mediate adhesion of cells to immobilized agrin. (A) QT6 cells, which express β1 and αv integrins, bind to immobilized agrin. Adhesion required Mgand was blocked by a β1 integrin blocking antibody (JG22, 100 μg/ml). Anti-integrin αv (Chav; 200 μg/ml) and the αv-blocking peptide (GRGDSP, 80 μg/ml) blocked adhesion ~50% whereas control antibodies and peptides were without effect. (B) 293 cells express a variety of β1 integrins, including αvβ1, but no αv integrins other αvβ1. Adhesion of these cells to agrin, like that of QT6 cells, required Mgand was blocked by anti-integrin β1 (P4C10; 20 μg/ml). Adhesion was partially inhibited by anti-integrin αv (L230; 20 μg/ml) or the αv blocking peptide (GRGDSP, 80 μg/ml). Antibodies to integrins α3 (P1B5; 20 μg/ml) or α6 (GoH3; 20 μg/ml), both of which are expressed by 293 cells, or anti-integrin β3 (7G2; 20 μg/ml) had no effect. Bars show mean ± s.e.m. of 4-5(A) or 2-5 (B) experiments. Mean numbers of cells per well counted in controls (100% values) were 870 in A and 383 in B.

In muscle cells, the integrin αv subunit forms heterodimers not only with β1 but also with other β subunits including β3 and β5 (Rajaraman, 1993; Gullberg et al., 1995b; Blaschuk et al., 1997). To ask whether αvβ1 heterodimers interact with agrin, we used a human kidney epithelial cell line, 293, which is unusual in having αvβ1 as its only αv-containing heterodimer (Hu et al., 1995). Like QT6 cells, 293 cells exhibited magnesium-dependent adhesion to agrin, and this adhesion was completely inhibited by anti-integrin β1 and partially blocked by anti-αv or the GRGDSP peptide (Fig. 4B). Antibodies to the α3 and α6 subunits were ineffective, even though 293 cells express both α3β1 and α6β1 (Hu et al., 1995). Thus, the αvβ1 heterodimer can mediate adhesion to agrin.

We also tested the human NCI-H69 cell line, which expresses the laminin-binding α3β1 and α6β1 heterodimers, but does not express detectable αv (Elices et al., 1991). These cells adhered to laminin, as expected, but did not adhere to agrin (data not shown). Thus, of three laminin-binding β1 integrins tested – α3β1, α6β1, and αvβ1 – only αvβ1 mediated adhesion to agrin.

Forced expression of integrin αv renders cells adhesive to agrin

As an additional test of the ability of αv integrins to mediate adhesion to agrin, we asked whether addition of αv to a cell’s repertoire of integrin subunits could endow it with the ability to bind agrin. For this experiment, we used human K562 cell, which express the α5 and β1 subunits but not αv, α1, α3, α4, α6 or β3 (Blystone et al., 1994, 1995). These cells did not adhere to agrin under our assay conditions (Fig. 5A). In initial experiments, we transfected K562 cells with an αv expression vector. However, K562 cells were unable to make αvβ1 heterodimers, and transfection of αv did not lead to the expression of αvβ1 on the cell surface (data not shown and S. Blystone, personal communication). When K562 cells are transfected with αv and β3 integrin, however, αvβ3 heterodimers are expressed on the surface, (Blyestone et al., 1995), and it has been shown that αvβ1 and αvβ3 integrins have similar ligand specificity (Felding-Habermann and Cheresh, 1993; Hu et al., 1995). We therefore transfected K562 cells with both αv and β3 expression vectors, and selected αv-expressing cells by fluorescence-activated cell sorting. The αv-positive cells were strongly adherent to agrin-coated substrata (Fig. 5A). Adhesion was magnesium-dependent (Fig. 5A) and was blocked by antibodies to the αv and β3 subunits as well as by GRGDSP peptide, but not by an antibody to β1 integrin or GRGESP peptide (Fig. 5B). Thus, expression of the αv subunit can render cells adherent to agrin. Moreover, αvβ3 as well as αvβ1 integrins can bind to agrin.

Fig. 5.

Introduction of integrin αv renders cells adhesive to agrin. K562 cells, which do not express integrin αv, do not adhere to agrin, either in the presence (open bars) or the absence (dark bars) of magnesium. Following transfection with integrin αv and β3 expression vectors, however, the cells adhered to agrin. Adhesion required magnesium (A), and was blocked by antibodies to integrins αv (L230, 20 μg/ml), or β3 (7G2; 20 μg/ml), or by the blocking peptide (GRGDSP, 80 μg/ml). Anti-integrin β1 (P4C10; 20 μg/ml) or a control peptide (GRGESP; 80 μg/ml) had no effect (B). K562 cells express integrin β1 but are unable to form αvβ1 dimers; therefore coexpression of β3 was necessary to obtain significant levels of αv on the cell surface. Bars show mean ± s.e.m. of 5 (A) or 3 (B) experiments.

Fig. 5.

Introduction of integrin αv renders cells adhesive to agrin. K562 cells, which do not express integrin αv, do not adhere to agrin, either in the presence (open bars) or the absence (dark bars) of magnesium. Following transfection with integrin αv and β3 expression vectors, however, the cells adhered to agrin. Adhesion required magnesium (A), and was blocked by antibodies to integrins αv (L230, 20 μg/ml), or β3 (7G2; 20 μg/ml), or by the blocking peptide (GRGDSP, 80 μg/ml). Anti-integrin β1 (P4C10; 20 μg/ml) or a control peptide (GRGESP; 80 μg/ml) had no effect (B). K562 cells express integrin β1 but are unable to form αvβ1 dimers; therefore coexpression of β3 was necessary to obtain significant levels of αv on the cell surface. Bars show mean ± s.e.m. of 5 (A) or 3 (B) experiments.

We have provided four lines of evidence that agrin interacts with integrins First, agrin-induced AChR clustering on avian and mammalian myotubes is inhibited by reagents (antibodies, peptides, and antisense oligonucleotides) that interfere with αv or β1 integrin activity or expression. Second, an antibody that activates β1 potentiates the effect of agrin on AChR clustering. Third, two cell lines (quail fibroblasts and human epithelial cells) that express αv and β1 integrins bind to an immobilized agrin fragment, and this adhesion is inhibited by antibodies and peptides that bind αv or β1. Fourth, cells that do not adhere to agrin and do not express αv can be rendered adherent by forcing expression of αvβ3 on their surface. Together, these results suggest that αv and β1 integrins, probably including αvβ1, can serve as agrin receptors.

That integrins might be agrin receptors is not unexpected, given the known abilities of integrins to bind many components of basement membranes (e.g. laminins, collagens, entactin, tenascin and fibronectin) and to transmit numerous signals from the extracellular matrix to the cell’s interior (Hynes, 1992; Venstrom and Reichardt, 1993; Schwartz et al. 1995). To our knowledge, however, the present report provides the first experimental support for this proposition. Ma et al. (1993; see also Bowe and Fallon, 1995) noted that RGD peptides, which inhibit many but not all integrin-based interactions, do not affect the binding of agrin to cells. However, the binding being assayed was to α-dystroglycan (discussed below), and interactions with integrins might not have been detected under the assay conditions used. Moreover, at least one investigator found that RGD inhibited agrin-induced AChR clustering in chick myotubes by 31% (Table 1 in Wallace, 1988), but did not comment on this datum in the text. Thus, the impression may have arisen that integrins do not bind agrin, but we know of no data inconsistent with this possibility.

Agrin has previously been shown to interact, directly or indirectly, with several components of the myotube surface, including α-dystroglycan (Gee et al., 1994; Bowe et al., 1994; Sugiyama et al., 1994; Campanelli et al., 1994), the neural cell adhesion molecule (N-CAM; Cole and Halfter, 1996), heparin/heparan sulfates (Campanelli et al., 1996; O’Toole et al., 1996), N-acetylgalactosamine-terminated glycocongujates (Martin and Sanes, 1995), the heparin-binding growth associated molecule HB-GAM/pleitrophin (Daggett et al., 1996) and MuSK (Glass et al., 1996). We now add integrins to this list. In evaluating this multiplicity of potential receptors, lessons learned from a more intensively studied protein, laminin, may be informative. Both laminin and agrin are multidomain proteins of the basal lamina that provide critical developmental signals to neurons and muscles. Moreover, although the two proteins show no global similarity in primary sequence or predicted secondary structure, they share several discrete regions of significant homology. Most important in the present context, it has been clearly demonstrated that laminin binds to multiple receptors, including integrins, α-dystroglycan, cell adhesion molecules, heparan sulfate proteoglycans and other glycoconjugates (Mercurio, 1995; Mecham and Hinek, 1996).

One striking parallel is the presence of G repeats in both agrin and laminin α chains. All five of the known laminin α chains (laminins are heterotrimers of α, β and γ chains) contain a large C-terminal globular (G) domain, composed of five ∼200-amino-acid-long repeats, called G1-5 (Sasaki et al., 1988; Engvall and Wewer, 1996; Miner et al., 1997). Much (but by no means all) of the bioactivity of laminin heterotrimers resides in the G domain, which has been called the ‘business end’ of the molecule (Engvall and Wewer, 1996). Isolated G domains mimic several effects of laminin, including promotion of neurite outgrowth and myoblast adhesion (Edgar et al., 1984; Yurchenco et al., 1993). Binding sites for heparin/heparan sulfate, α-dystroglycan and several integrins have been mapped to distinct G repeats, and much has been learned about the ways in which these receptors mediate laminin’s effects (Mercurio, 1995). Three similar repeats (G1-G3), separated from each other by EGF-like repeats, form much of the C-terminal half (∼95 kDa) of agrin, which is fully active in AChR clustering (Ferns et al., 1993; Denzer et al., 1996). Deletion of the most N-terminal G repeat, G1, yields a ∼50 kDa fragment that is about half as active as the 95 kDa fragment. Deletion of G2 yields a ∼20 kDa fragment, which is <1% as active as the 95 kDa fragment (Hoch et al., 1994; Gesseman et al., 1995). These results suggest that, although G3 is crucial for activity, G1 and G2 are also important. In fact, α-dystroglycan and heparin bind to distinct but overlapping sites within G2 (Gesseman et al, 1996; Campanelli et al., 1996; O’Toole et al., 1996; Hopf and Hoch, 1996). Thus, for agrin as for laminin, distinct G repeats may bind to multiple receptors to generate a signal.

Second, α-dystroglycan and integrins are both crucial laminin receptors, but they play different roles. α-Dystroglycan was isolated independently as a laminin-binding protein from brain and as a component of the dystrophin-glycoprotein complex that links the cytoskeleton to the basal lamina in muscle cells. The precise function of α-dystroglycan remains undefined, but available evidence favors the idea that it is primarily important for assembly and maintenance of the basal lamina (Henry and Campbell, 1996). Likewise, although it was originally suspected that α-dystroglycan might act as a signaling receptor for agrin (Gee et al., 1994; Bowe et al., 1994; Campenelli et al., 1994), attempts to block agrin function with antibodies to α-dystroglycan have given inconsistent results (Gee et al., 1994; Campanelli et al., 1994; Sugiyama et al., 1994), and alternatively spliced forms of agrin that do not cluster AChRs bind at least as tightly to α-dystroglycan as fully active forms (Sugiyama et al., 1994; Gesseman et al., 1996; Hopf and Hoch, 1996). Thus, for agrin as for laminin, α-dystroglycan may be more involved in assembly and stabilization than in signaling. In contrast, there is abundant evidence that integrins are major signaling receptors for laminins, and the same may be the case for agrin.

Third, laminins signal through multiple integrins. At least 10 integrin dimers (α1β1, α2β1, α3β1, α6β1, α6β4, α7β1, α9β1, αvβ3, αIIbβ3 and α?β8), bind to various domains on laminins, (Delwel and Sonnenberg, 1996). Individual cells sometimes express several laminin-binding integrins, providing for cooperativity, specificity and diversity in signaling functions (Mercurio, 1995). Likewise, although our strongest evidence implicates αvβ1 integrin as an agrin receptor, blockade of β1 consistently inhibited agrin activity (clustering or adhesion) more completely than blockade of αv. Moreover, αvβ1 and αvβ3, both of which are expressed by muscle (Gullberg et al., 1995a,b; Blaschuk et al., 1997), can mediate adhesion to agrin. It is therefore possible that agrin, like laminin, makes use of multiple integrins, including β1 integrins that do not contain αv and αv integrins that do not contain β1.

Finally, laminins are potent promoters of neurite outgrowth and these effects are largely integrin-mediated (Venstrom and Reichardt, 1993). Intramuscular nerve branching patterns are greatly perturbed in agrin-deficient mutant mice (Gautam et al., 1996) and agrin inhibits neurite outgrowth in vitro (Campagna et al., 1995; Chang et al., 1997). For laminins, integrinmediated outgrowth correlates with integrin-mediated adhesion. The integrin-dependent adhesion to agrin described here therefore raises the possibility that agrin’s effects on neurons, like its effect on muscle, involve integrins.

A major question that remains is the relationship of integrins to the best candidate agrin receptor, the tyrosine kinase MuSK (see Introduction). Recent studies in several systems have revealed previously unsuspected relationships between integrins and tyrosine kinases (Schwartz et al., 1995; Dedhar and Hannigan, 1996; Miyamoto et al., 1996), and provide precedents for a MuSK-integrin interaction. One possibility is that integrin and MuSK interact physically. The observation that MuSK alone cannot bind agrin is consistent with this idea; MASC, the hypothesized binding subunit of the MuSK receptor complex (Glass et al., 1996), might be an integrin. Second, integrins could act downstream of MuSK. The observations that spontaneous (agrin-independent) clustering requires MuSK (Glass et al., 1996; DiChiara et al., 1996) but not integrin (this study) favors this possibility. Third, integrins and MuSK might provide parallel signaling pathways for agrin, with the MuSK pathway being essential and the integrin pathway facilitatory. In other systems, growth factors and adhesive ligands play synergistic roles in developmental processes (Sastry and Horwitz, 1997); synaptic differentiation might require agrin to exert both roles but via different pathways. The availability of mutant mice lacking MuSK (DeChiara et al., 1996; Apel et al., 1997) and several integrin subunits (Hynes, 1996) should facilitate tests of these and other possibilities.

We thank R. Scheller and J. Campanelli (Stanford U.), L. Reichardt (UCSF), S. Kaufman (U. Illinois), S. Blystone, E. Brown and E. Johnson (Washington University) for generous gifts of antibodies and reagents. We are also grateful to S. Blystone and E. Brown for FACS purification of transfected K562 cells. Some antibodies were obtained from the Developmental Hybridoma Bank (Iowa City, IA). This work was sponsored by grants from the NIH and MDA to J. R. S. and a postdoctoral fellowship from the NIH to P. T. M.

Apel
,
E. D.
,
Glass
,
D. J.
,
Moscoso
,
L. M.
,
Yancopoulos
,
G. D.
and
Sanes
,
J. R.
(
1997
).
Rapsyn is required for MuSK signaling and recruits synaptic components to a MuSK-containing scaffold
.
Neuron
18
,
623
635
.
Bao
,
Z. Z.
,
Lakonishok
,
M.
,
Kaufman
,
S.
and
Horwitz
,
A. F.
(
1993
).
α7β1 is a component of the myotendinous junction on skeletal muscle
.
J. Cell Sci
.
106
,
579
590
.
Blaschuk
,
K. L.
,
Guérin
,
C.
and
Holland
,
P. C.
(
1997
).
Myoblast αvβ3 integrin levels are controlled by transcriptional regulation of expression of the β3 subunit and down-regulation of β3 subunit expression is required for skeletal muscle cell differentiation
.
Dev. Biol
.
184
,
266
277
.
Blyestone
,
S. D.
,
Graham
,
I. L.
,
Lindberg
,
F. P.
and
Brown
,
E. J.
(
1994
).
Integrin αvβ3 differentially regulates adhesive and phagocytic functions of the fibronectin receptor α5β1
.
J. Cell Biol
.
127
,
1129
1137
.
Blyestone
,
S. D.
,
Lindberg
,
F. P.
,
La Flamme
,
S. E.
and
Brown
,
E. J.
(
1995
).
Integrin cytoplasmic tail is necessary and sufficient for regulation of α5β1 phagocytosis by αvβ3 and integrin-associated protein
.
J. Cell Biol
.
130
,
745
754
.
Bowe
,
M. A.
,
Deyst
,
K. A.
,
Leszyk
,
J. D.
and
Fallon
,
J. R.
(
1994
).
Identification and purification of an agrin receptor from Torpedo postsynaptic membranes: A heteromeric complex related to the dystroglycans
.
Neuron
12
,
1173
1180
.
Bowe
,
M. A.
and
Fallon
,
J. R.
(
1995
).
The role of agrin in synapse formation
.
Ann. Rev. Neurosci
.
18
,
443
62
.
Bozyczko
,
D.
,
Decker
,
C.
,
Muschler
,
J.
and
Horwitz
,
A. F.
(
1989
).
Integrin on developing and adult skeletal muscle
.
Exp. Cell Res
.
183
,
72
91
.
Bronner-Fraser
,
M.
,
Artinger
,
M.
,
Muschler
,
J.
and
Horwitz
,
A. F.
(
1992
).
Developmentally regulated expression of α6 integrin in avian embryos
.
Development
115
,
197
211
.
Buckley
,
C. D.
,
Doyonnas
,
R.
,
Newton
,
J. P.
,
Blystone
,
S. D.
,
Brown
,
E. J.
,
Watt
,
S. M.
and
Simmons
,
D. L.
(
1996
).
Identification of αvβ3 as a heterotypic ligand for CD31/PECAM-1
.
J. Cell Sci
.
109
,
437
445
.
Campagna
,
J. A.
,
Ruegg
,
M. A.
and
Bixby
,
J. L.
(
1995
).
Agrin is a differentiation-inducing ‘stop signal’ for motorneurons in vitro
.
Neuron
15
,
1365
1374
.
Campanelli
,
J. T.
,
Roberds
,
S. L.
,
Campbell
,
K. P.
and
Scheller
,
R. H.
(
1994
).
A role for dystrophin-associated glycoproteins and utrophin in agrin-induced AChR clustering
.
Cell
77
,
663
674
.
Campanelli
,
J. T.
,
Gayer
,
G. G.
and
Scheller
,
R. H.
(
1996
).
Alternative RNA splicing that determines agrin activity regulates binding to heparin and alpha dystroglycan
.
Development
122
,
1663
1672
.
Carter
,
W. G.
,
Wayner
,
E. A.
,
Bouchard
,
T. S.
and
Kauer
,
P.
(
1990
).
The role of integrins alpha 2 beta 1 and alpha 3 beta 1 in cell-cell and cell-substrate adhesion of human epidermal cells
.
J. Cell Biol
.
110
,
1387
1404
.
Carter
,
W. G.
,
Ryan
,
M. C.
and
Gahr
,
P. J.
(
1991
).
tEpiligrin, a new cell adhesion ligand for integrin α3β1 in epithelial basement membranes
.
Cell
65
,
599
610
.
Chang
,
D.
,
Woo
,
J. S.
,
Campanelli
,
J.
,
Scheller
,
R. H.
and
Ignatius
,
M. J.
(
1997
).
Agrin inhibits neurite outgrowth but promotes attachment of embryonic motor and sensory neurons
.
Dev. Biol
.
181
,
21
35
.
Cole
,
G. J.
and
Halfter
,
W.
(
1996
).
Agrin – an extracellular matrix heparan sulfate proteoglycan involved in cell interactions and synaptogenesis
.
Perspectives on Dev. Neurobiol
.
3
,
359
371
.
Daggett
,
D. F.
,
Cohen
,
M. W.
,
Stone
,
D.
,
Nikolics
,
K.
,
Rauvala
,
H.
and
Peng
,
H. B.
(
1996
).
The role of an agrin-growth factor interaction in acetylcholine receptor clustering
.
Mol. Cell. Neurosci
.
8
,
272
285
.
DeChiara
,
T. M.
,
Bowen
,
D. C.
,
Valenzuela
,
D. M.
,
Simmons
,
M. V.
,
Poueymirou
,
W. T.
,
Thomas
,
S.
,
Kinetz
,
E.
,
Compton
,
D. L.
,
Rojas
,
E.
,
Park
,
J. S.
,
Smith
,
C.
,
DiStefano
,
P. S.
,
Glass
,
D. J.
,
Burden
,
S. J.
and
Yancopoulos
,
G. D.
(
1996
).
The receptor tyrosine kinase MuSK is required for neuromuscular junction formation in vivo
.
Cell
85
,
501
512
.
Dedhar
,
S.
and
Hannigan
,
G. E.
(
1996
).
Integrin cytoplasmic interactions and bidirectional transmembrane signalling
.
Curr. Opin. Cell Biol
.
8
,
657
669
.
DeFreitas
,
M. F.
,
Yoshida
,
C. K.
,
Frazier
,
W. A.
,
Mendrick
,
D. L.
,
Kypta
,
R. M.
and
Reichardt
,
L. F.
(
1995
).
Identification of integrin α3β1 as a neuronal thrombospondin receptor mediating neurite outgrowth
.
Neuron
15
,
333
343
.
Delwel
,
G. O.
and
Sonnenberg
,
A.
(
1996
).
Adhesion Receptors as Theraputic Targets (ed
.
M.A.
Horton
). pp.
9
36
.
Boca Raton FL: CRC Press
.
Denzer
,
A. J.
,
Gesemann
,
M.
,
Schumacher
,
B.
and
Ruegg
,
M. A.
(
1995
).
An amino-terminal extension is required for the secretion of chick agrin and its binding to extracellular matrix
.
J. Cell Biol
.
131
,
1547
1560
.
Denzer
,
A. J.
,
Gesemann
,
M.
and
Ruegg
,
M.A.
(
1996
).
Diverse functions of the extracellular matrix molecule agrin
.
Sem. in The Neurosciences
8
,
357
366
.
Duband
,
J.-L
,
Belkin
,
A. M.
,
Syfrig
,
J.
,
Thiery
,
J. P.
and
Koteliansky
,
V. E.
(
1992
).
Expression of α1 integrin, a laminin-collagen receptor, during myogenesis and neurogenesis in the avian embryo
.
Development
116
,
585
600
.
Edgar
,
D.
,
Timpl
,
R.
and
Thoenen
,
H.
(
1984
).
The heparin-binding domain of laminin is responsible for its effects on neurite outgrowth and neuronal survival
.
EMBO J
.
3
,
1463
1468
.
Elices
,
M. J.
,
Urry
,
L. A.
and
Hemler
,
M. E.
(
1991
).
Receptor functions for the integrin VLA-3; fibronectin, collagen, and laminin are differentially influenced by Arg-Gly-Asp peptide and by divalent cations
.
J. Cell Biol
.
112
,
169
181
.
Engvall
,
E.
and
Wewer
,
U. M.
(
1996
).
Domains of laminin
.
J. Cell. Biochem
.
61
,
493
501
.
Enomoto
,
M. I.
,
Boettiger
,
D.
and
Menko
,
A. S.
(
1993
).
α5 integrin is a critical component of adhesion plaques during myogenesis
.
Dev. Biol
.
155
,
180
197
.
Felding-Habermann
,
B.
and
Cheresh
,
D. A.
(
1993
).
Vitronectin and its receptors
.
Curr. Opin Cell Biol
.
5
,
864
868
.
Ferns
,
M. J.
,
Campanelli
,
J. T.
,
Hoch
,
W.
,
Scheller
,
R. H.
and
Hall
,
Z.W.
(
1993
).
The ability of agrin to cluster AChRs depends on alternative splicing and on cell surface proteoglycans
.
Neuron
11
,
491
502
.
Galileo
,
D. S.
Majors
,
J.
,
Horwitz
,
A. F.
and
Sanes
,
J. R.
(
1992
).
Retrovirally introduced antisense integrin RNA inhibits neuroblast migration in vivo
.
Neuron
9
,
1117
1131
.
Gautum
,
M.
,
Noakes
,
P. G.
,
Moscoso
,
L.
,
Rupp
,
F.
,
Scheller
,
R. H.
,
Merlie
,
J. P.
and
Sanes
,
J. R.
(
1996
).
Defective neuromuscular synaptogenesis in agrin-deficient mutant mice
.
Cell
85
,
525
535
.
Gee
,
S. H.
,
Montanaro
,
F.
,
Lindenbaum
,
M. H.
, and
Carbonetto
,
S.
(
1994
).
Dystroglycan-α, a dystrophin-associated glycoprotein, is a functional agrin receptor
.
Cell
77
,
675
686
.
George-Weinstein
,
M.
,
Foster
,
R. F.
,
Gerhart
,
J. V.
and
Kaufman
,
S. J.
(
1993
).
In vitro and in vivo expression of α7 integrin and desmin define the primary and secondary myogenic lineages
.
Dev. Biol
.
156
,
209
229
.
Gesseman
,
M.
,
Denzer
,
A. J.
and
Ruegg
,
M. A.
(
1995
).
Acetylcholine receptor-aggregating activity of agrin isoforms and mapping of the active site
.
J. Cell Biol
.
128
,
625
636
.
Gesseman
,
M.
,
Cavalli
,
V.
,
Denzer
,
A. J.
,
Brancaccio
,
A.
,
Schumacher
,
B.
and
Ruegg
,
M. A.
(
1996
).
Alternative splicing of agrin alters its binding to heparin, dystroglycan, and the putative agrin receptor
.
Neuron
16
,
755
767
.
Glass
,
D. J.
,
Bowen
,
D. C.
,
Stitt
,
T. N.
,
Radziejewski
,
C.
,
Bruno
,
J.
,
Ryan
,
T. E.
,
Gies
,
D. R.
,
Shah
,
S.
,
Mattsson
,
K.
,
Burden
,
S. J.
,
DiStefano
,
P. S.
,
Valenzuela
,
D. M.
,
DeChiara
,
T. M.
and
Yancopoulos
,
G. D.
(
1996
).
Agrin acts via a MuSK receptor complex
.
Cell
85
,
513
523
.
Glass
,
D. J.
,
Apel
,
E. D.
,
Shah
,
S.
,
Bowen
,
D. C.
,
DeChiara
,
T. M.
,
Stitt
,
T. N.
,
Sanes
,
J. R.
and
Yancopoulos
,
G. D.
(
1997
)
Kinase domain of the muscle-specific receptor tyrosine kinase (MuSK) is sufficient for phosphorylation but not clustering of acetylcholine receptors: Required role for the MuSK ectodomain
.
Proc. Natl. Acad. Sci. USA
94
,
8848
8853
.
Gresham
,
H. D.
,
Goodwin
,
J. L.
,
Anderson
,
D. C.
and
Brown
,
E. J.
(
1989
).
A novel member of the integrin receptor family mediates Arg-Gly-Asp stimulated neutrophil phagocytosis
.
J. Cell Biol
.
108
,
1935
1943
.
Greve
,
J. M.
and
Gottlieb
,
D. I.
(
1982
).
Monoclonal antibodies which alter the morphology of cultured chick myogenic cells
.
J. Cell Biochem
.
18
,
221
229
.
Gullberg
,
D.
,
Velling
,
T.
,
Sjoberg
,
G.
and
Sejersen
,
T.
(
1995a
).
Upregulation of a novel integrin alpha-chain (alpha mt) on human fetal myotubes
.
Dev. Dyn
.
204
,
57
65
.
Gullberg
,
D.
,
Sjoberg
,
G.
,
Velling
,
T.
and
Sejersen
,
T.
(
1995b
).
Analysis of fibronectin and vitronectin receptors on human fetal skeletal muscle cells upon differentiation
.
Exp. Cell Res
.
220
,
112
123
.
Hall
,
Z. W.
and
Sanes
,
J. R.
(
1993
).
Synaptic structure and development: the neuromuscular junction
.
Cell
72/
Neuron10(
Suppl
.),
99
121
.
Henry
,
M.D.
and
Campbell
,
K.P.
(
1996
).
Dystroglycan: an extracellular matrix receptor linked to the cytoskeleton
.
Curr. Opin. in Cell Biol
.
8
,
625
631
.
Hirsch
,
E.
,
Gullberg
,
D.
,
Balzac
,
F.
,
Altruda
,
F.
,
Silengo
,
L.
and
Tarone
,
G.
(
1994
).
αv integrin subunit is predominantly located in nervous tissue and skeletal muscle during mouse development
.
Develop. Dyn
.
201
,
108
20
.
Hoch
,
W.
,
Campanelli
,
J. T.
,
Harrison
,
S.
, and
Scheller
,
R. H.
(
1994
).
Structural domains of agrin required for clustering of nicotinic acetylcholine receptors
.
EMBO J
.
13
,
2814
2821
.
Hopf
,
C.
and
Hoch
,
W.
(
1996
).
Agrin binding to α-dystroglycan
.
J. Biol. Chem
.
271
,
5231
5236
.
Houghton
,
A. N.
,
Eisinger
,
M.
,
Albino
,
A. P.
,
Cairncross
,
J. G.
and
Old
,
L. J.
(
1982
).
Surface antigens of melanocytes and melanomas. Markers of melanocyte differentiation of melanoma subsets
.
J. Exp. Med
.
156
,
1755
1766
.
Hu
,
D. D.
,
Lin
,
E. C. K.
,
Kovach
,
N. L.
,
Hoyer
,
J. R.
, and
Smith
,
J. W.
(
1995
).
A biochemical characterization of the binding of osteopontin to integrins αvβ1 and αvβ5
.
J. Biol. Chem
.
270
,
26232
26238
.
Hynes
,
R. O.
(
1992
).
Integrins: versatility, modulation, and signaling in cell adhesion
.
Cell
69
,
11
25
.
Hynes
,
R.O.
(
1996
).
Targeted mutations in cell adhesion genes: what have we learned from them? Dev
.
Biol
.
180
,
402
412
.
Lagenaur
,
C.
and
Lemmon
,
V.
(
1987
).
An L1-like molecule, the 8D9 antigen, is a potent substrate for neurite extention
.
Proc. Natl. Acad. Sci. USA
84
,
7753
7757
.
Lakonishock
,
M.
,
Muschler
,
J.
and
Horwitz
,
A. F.
(
1992
).
The α5β1 integrin associates with a dystrophin-containing lattice during muscle development
.
Dev. Biol
.
152
,
209
220
.
Ma
,
J.
,
Nastuk
,
M.A.
,
McKechnie
,
B.A.
and
Fallon
,
J.R.
(
1993
).
The agrin receptor
.
J. Biol. Chem
.
268
,
25108
25117
.
Martin
,
P. T.
and
Sanes
,
J. R.
(
1995
).
Role for a synapse-specific carbohydrate in agrin-induced clustering of acetylcholine receptors
.
Neuron
14
,
743
754
.
Martin
,
P. T.
,
Kaufman
,
S. J.
,
Kramer
,
R. H.
, and
Sanes
,
J. R.
(
1996
).
Synaptic integrins in developing, adult, and mutant muscle: Selective association of α1, α7A, and α7B integrins with the neuromuscular junction
.
Dev. Biol
.
174
,
125
139
.
McDonald
,
K. A.
,
Lakonishok
,
M.
, and
Horwitz
,
A. F.
(
1995
).
αv and α3 integrin subunits are associated with myofibrils during myofibrillogenesis
.
J. Cell Sci
.
108
,
975
983
.
McMahan
,
U. J.
(
1990
).
The agrin hypothesis
.
Cold Spring Harbor Symp. Quant. Biol
.
50
,
407
418
.
Mecham
,
R.P.
and
Hinek
,
A.
(
1996
).
Non-integrin laminin receptors
.
In, The Laminins
(ed.
P.
Ekblom
and
A.
Itinek
). pp.
159
183
.
Harwood Academic
.
Mercurio
,
A. M.
(
1995
).
Laminin receptors – achieving specificity through cooperation
.
Trends Cell Biol
.
5
,
419
423
.
Miner
,
J.H.
,
Patton
,
B.L.
,
Lentz
,
S.I.
,
Gilbert
,
D.J.
,
Snider
,
W.D.
,
Jenkins
,
N.A.
,
Copeland
,
N.G.
, and
Sanes
,
J.R.
(
1997
)
The laminin α chains: Expression, developmental transitions, and chromosomal locations of 1-5, identification of hetrotrimeric laminins 8-11, and cloning of a novel α3 isoform
.
J. Cell Biol
.
137
,
685
701
.
Miyamoto
,
S.
,
Teramoto
,
H.
,
Gutkind
,
J.S.
, and
Yamada
,
K.M.
(
1996
).
Integrins can collaborate with growth factors for phosphorylation of receptor tyrosine kinases and MAP kinase activation: roles in integrin aggregation and occupancy of receptors
.
J. Cell Biol
.
135
,
1633
1642
.
Neugebauer
,
K. M.
and
Reichardt
,
L. F.
(
1991
).
Cell-surface regulation of β1-integrin activity on developing retinal neurons
.
Nature
350
,
68
71
.
Neugebauer
,
K. M.
,
Emmett
,
C. J.
,
Ventstrom
,
K. A.
, and
Reichardt
,
L. F.
(
1991
).
Vitronectin and thrombospondin promote retinal neurite outgrowth: developmental regulation and role of integrins
.
Neuron
6
,
345
358
.
Neff
,
N. T.
,
Lowrey
,
C.
,
Decker
,
C.
,
Tovar
,
A.
,
Damsky
,
C.
,
Buck
,
C.
, and
Horwitz
,
A. F.
(
1982
).
A monoclonal antibody detaches embryonic skeletal muscle from extracellular matrices
.
J. Cell Biol
.
95
,
654
666
.
O’Toole
,
J. J.
,
Deyst
,
K. A.
,
Bowe
,
M. A.
,
Nastuk
,
M. A.
,
McKechnie
,
B. A.
, and
Fallon
,
J. R.
(
1996
).
Alternative splicing of agrin regulates its binding to heparin, α-dystroglycan, and the cell surface
.
Proc. Natl. Acad. Sci. USA
93
,
7369
7374
.
Palmer
,
E. L.
,
Ruegg
,
C.
,
Ferrando
,
R.
,
Pytela
,
R.
, and
Sheppard
,
D.
(
1993
).
Sequence and tissue distribution of the α9 integrin subunit, a novel partner of β1 that is widely distributed in epithelia and muscle
.
J. Cell Biol
.
123
,
1289
1297
.
Piali
,
L.
,
Hammel
,
P.
Uherek
,
C.
,
Bachmann
,
F.
,
Gisler
,
R. H.
,
Dunon
,
D.
, and
Imhof
,
B. A.
(
1995
).
CD31/PECAM-1 is a ligand for αvβ3 integrin involved in adhesion of leukocytes to endothelium
.
J. Cell Biol
.
130
,
451
460
.
Rajaraman
,
R.
(
1993
).
Avian integrin αv subunit associates with multiple β subunits and shows tissue-specific variation in its association
.
Exp. Cell Res
.
205
,
25
31
.
Reist
,
N. E.
,
Werle
,
M. J.
, and
McMahan
,
U. J.
(
1992
).
Agrin released by motor neurons induces the aggregation of acetylcholine receptors at neuromuscular junctions
.
Neuron
8
,
865
868
.
Rosen
,
G. D.
,
Sanes
,
J. R.
,
LaChance
,
R.
,
Cunningham
,
J. M.
,
Roman
,
J.
, and
Dean
,
D. C.
(
1992
).
Roles for the integrin VLA-4 and its counter receptor VCAM-1 in myogenesis
.
Cell
69
,
1107
1119
.
Ruegg
,
M. A.
,
Tsim
,
K. W. K.
,
Horton
,
S. E.
,
Kroger
,
S.
Escher
,
G.
,
Gensch
,
E. M.
and
McMahan
,
U. J.
(
1992
).
The agrin gene codes for a family of basal lamina proteins that differ in function and distribution
.
Neuron
8
,
691
699
.
Ruoslahti
,
E.
(
1996
).
RGD and other recognition sequences for integrins
.
Ann. Rev. Cell Dev. Biol
.
12
,
697
715
.
Rupp
,
F.
,
Payan
,
D. G.
,
Magill-Solc
,
C.
,
Cowan
,
D. M.
and
Scheller
,
R. H.
(
1991
).
Structure and expression of a rat agrin
.
Neuron
6
,
811
823
.
Sakamoto
,
H.
,
Broekelmann
,
T.
,
Cheresh
,
D. A.
,
Ramirez
,
F.
,
Rosenbloom
,
J.
, and
Mecham
,
R. P.
(
1996
).
Cell type specific recognition of RGD- and non-RGD-containing cell binding domains in fibrillin-1
.
J. Biol. Chem
.
271
,
4916
4922
.
Sasaki
,
M.
,
Kleinman
,
H.K.
,
Huber
,
H.
,
Deutzmann
,
R.
and
Yamada
,
Y.
(
1988
).
Laminin, a multidomain protein
.
J. Biol. Chem
.
263
,
16536
16544
.
Sastry
,
S. K.
and
Horwitz
,
A. F.
(
1996
).
Adhesion-growth factor interactions during differentiation: An integrated biological response
.
Dev. Biol
.
180
,
455
467
.
Schwartz
,
M. A.
,
Schaller
,
M. D.
and
Ginsberg
,
M. H.
(
1995
).
Integrins: emerging paradigms of signal transduction
.
Annu. Rev. Cell Dev. Biol
.
11
,
549
599
.
Song
,
W.-K.
,
Wang
,
W.
,
Foster
,
R. F.
,
Bielser
,
D. A.
and
Kaufman
,
S. J.
(
1992
).
H36-α7 is a novel integrin alpha chain that is developmentally regulated during skeletal myogenesis
.
J. Cell Biol
.
117
,
643
657
.
Sonnenberg
,
A. H.
,
Janssen
,
F.
,
Hogervorst
,
J.
,
Calafat
J.
and
Hilgers
,
J.
(
1987
)
A complex of platelet glycoproteins Ic and IIa identified by a rat monoclonal antibody
.
J. Biol. Chem
.
262
,
10376
10383
.
Sonnenberg
,
A.
(
1993
).
Integrins and their ligands
.
Curr. Top. Microbiol. Immunol
.
184
,
7
35
.
Sugiyama
,
J.
,
Bowen
,
D. C.
and
Hall
,
Z. W.
(
1994
).
Dystroglycan bind nerve and muscle agrin
.
Neuron
13
,
103
115
.
Tsen
,
G.
,
Halfter
,
W.
,
Kroger
,
S.
and
Cole
,
G. J.
(
1995
).
Agrin is a heparan sulfate proteoglycan
.
J. Biol. Chem
.
270
,
3392
3399
.
Valenzuela
,
D. M.
,
Stitt
,
T. N.
,
DiStefano
,
P. S.
,
Rojas
,
E.
,
Mattsson
,
K.
,
Compton
,
D. L.
,
Nuñez
,
L.
,
Park
,
J. S.
,
Stark
,
J. L.
,
Gies
,
D. R.
,
Thomas
,
S.
,
LeBeau
,
M. M.
,
Fernald
,
A. A.
,
Copeland
,
N. G.
,
Jenkins
,
N. A.
,
Burden
,
S. J.
,
Glass
,
D. J.
and
Yancopoulos
,
G. D.
(
1995
).
Receptor tyrosine kinase specific for skeletal muscle lineage: expression in embryonic muscle, at the neuromuscular junction, and after injury
.
Neuron
15
,
573
584
.
Venstrom
,
K. A.
and
Reichardt
,
L. F.
(
1993
).
Extracellular matrix 2: Role of extracellular matrix molecules and their receptors in the nervous system
.
FASEB J
.
7
,
996
1003
.
Vogel
,
B. E.
,
Lee
,
S. J.
,
Hildebrand
,
A.
,
Craig
,
W.
,
Pierschbacher
,
M. D.
,
Wong-Staal
,
F.
and
Ruoslahti
,
E.
(
1993
).
A novel integrin specificity exemplified by binding of alpha v beta 5 integrin to the basic domain of the HIV Tat protein and vitronectin
.
J. Cell Biol
.
121
,
461
468
.
Wallace
,
B. G.
(
1988
).
Regulation of agrin-induced acetylcholine receptor aggregation by Ca++ and phorbol ester
.
J. Cell Biol
.
107
,
267
278
.
Wayner
,
E. A.
,
Carter
,
W. G.
,
Piotrowicz
,
R. S.
, and
Kunicki
,
T. J.
(
1988
).
The function of multiple extracelular matrix receptors in mediating cell adhesion to extracellular matrix: preparation of monoclonal antibodies to the fibronectin receptor that specificaly inhibit cell adhesion to fibronecting and react with platelet glycoproteins Ic-Ila
.
J. Cell Biol
.
107
,
1881
1891
.
Yurchenco
,
P. D.
,
Sung
,
U.
,
Ward
,
M. D.
,
Yamada
,
Y.
, and
O’Rear
,
J. J.
(
1993
)
Recombinant laminin G domain mediates myoblast adhesion and heparin binding
.
J. Biol. Chem
.
268
,
8356
8365
.