Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

Integrins are heterodimeric proteins consisting of 1 - and 1 -subunit. More than 16 - and 8 -subunits have been identified that can form more than 20 functional combinations. Integrins are cell surface receptors permitting cell-cell and cell-extracellular matrix (ECM) interactions. These receptors thus contribute to the state of cells through their adhesion to ECM and to cellular dynamic changes such as mobility, growth, and proliferation (for reviews see Meredith et al., 1996; Giancotti, 1997). Integrins are subdivided into different families based on their structural composition, their expression in specific cell types, or their affinity toward certain groups of ECM proteins. Among them, several integrins, namely, _IIb3, ₅₁, ₈₁, _v1, _v3, _v5, _v6, _v8, and, under special conditions, ₃₁, ₄₁, ₂₁, and ₁₁, have been documented to bind through the RGD motif present in proteins such as fibronectin, fibrinogen, von Willbrand factor, vitronectin, osteopontin, and others (Ruoslahti, 1997).

Whereas the functional importance of integrins in cellular growth has been well established, there are few means of investigating their presence and functionality on the cell surface. Snake venom contains toxins among which are peptides with the potential to inhibit platelet aggregation by interacting and inhibiting _IIb3-integrin. Therefore, they have been termed disintegrins (Gould et al., 1990). All disintegrins possess in their central core an RGD motif framed by disulfide bridges and conserved sequences surrounding this motif. Disintegrins have been used in radioligand-binding assays to evaluate the density of _IIb3-integrin on platelets or _v3 on vascular smooth muscle cells (McLane et al., 1994; Marcinkiewicz et al., 1996; Kumar et al., 1997). However, this type of assay cannot discriminate between different RGD-dependent integrins if several of them are present on the cell surface. Although disintegrins have the potential to interact with all RGD-dependent integrins, their use as probes to study cell surface integrins has not been extensively exploited. The present study demonstrates that ¹²⁵I-echistatin, through the formation of SDS-stable disintegrin-integrin complexes, can be an effective pharmacological tool to evaluate the presence of individual RGD-dependent integrins on the surface of cardiac fibroblasts.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. Echistatin, GRGDSP, GRGDTP, acPenRGDC, cycloGRGDSPA, cycloRGDdFV, GRGDdSP, and GRGESP were purchased from Bachem California, Inc. (Torrance, CA). Elegantin, flavoridin, and flavostatin were purified from Trimeresurus elegans and Trimeresurus flavoviridis snake venom, bought from Miami Serpentarium (Miami, FL). Purification was performed by Bio-Gel P-30 and by reverse-phase high-performance liquid chromatography (Scarborough et al., 1993; Maruyama et al., 1997) with purity and concentration being assessed by amino acid analysis and sequencing. Antibodies against rat integrins were generously provided by Dr. R.O. Hynes (Howard Hughes Medical Institute, Cambridge, MA; anti-₃, no. 8-4; anti-₁, no. 130), by Dr. Lynn M. Schnapp (Department of Medicine, Mount Sinai School of Medicine, NY; anti-₈), by Dr. Lou Reichardt (Howard Hughes Medical Institute, University of California, San Francisco, CA; anti-₈, no. 2415), or were bought from Chemicon International, Inc. (Temecula, CA; anti-₁, AB1934; anti-₂, AB1936; anti-₅, AB1928; anti-_v, AB1930; anti-₄, MAB1396 and MAB1383; anti-₅, AB1926) or from Pharmingen Canada (Mississauga, ON, Canada; anti-₃, F11; anti-₁, Ha2/5). Horseradish peroxidase-coupled anti-mouse or anti-rabbit Ig antibody was from Bio-Rad Laboratories (Hercules, CA). Streptavidin-horseradish peroxidase and Hybond ECL were from Amersham Canada, Ltd. (Oakville, ON). X-OMAT AR5 films were from Eastman Kodak Co. (Rochester, NY).

Cardiac Fibroblasts. Fibroblasts were obtained by collagenase digestion of cardiac ventricles from 200- to 250-g Sprague-Dawley rats as described previously (Fareh et al., 1997). Digested cells were plated in plastic dishes in Dulbecco's modified Eagle's medium supplemented with 0.1% bovine serum albumin and 10% fetal bovine serum. The cells were grown for 7 to 10 days after seeding and used at confluency. Only primary cultures were used. Characterization of cultured cells with histological markers indicated that more than 95% of them were indeed fibroblasts.

Iodination of Echistatin and Radioligand-Binding Filtration Assay. Echistatin was iodinated by the lactoperoxidase method with 15 µg of echistatin in the presence of 1 mCi of Na¹²⁵I. The monoiodinated product was purified by reverse-phase high-performance liquid chromatography on a C₄ Vydac column (The Separations Group, Hesperia, CA) with a CH₃CN gradient in 0.1% trifluoroacetic acid. Between 8 and 12 × 10⁸ cpm were usually collected with a specific activity of 800 to 1200 cpm/fmol.

13

5

13

5

RGD Affinity Chromatography. A GRGDSP affinity column was prepared by coupling 25 mg of GRGDSP to 2.5 ml of Affi-Gel 10 matrix (Bio-Rad Laboratories) according to the manufacturer's instructions. Proteins (2 mg) of NP-40-solubilized fibroblasts were incubated for 2 h with the gel. After the sample settled in a 10-ml column, the gel was washed with 20 ml of 0.05 M HEPES, pH 7.4, 5 mM MnCl₂, and 0.1% NP-40. Proteins were eluted with 10 ml of 0.05 M HEPES, pH 7.4, 10 mM EDTA, and 0.1% NP-40. Fractions (1 ml) were collected and subsequently concentrated in Centricon-30 (Amicon, Bedford, MA). Each fraction was either analyzed for ¹²⁵I-echistatin binding as described previously or boiled and subjected to a 7.5% gel SDS-PAGE for the immunoblotting analysis of integrin subunits.

SDS-PAGE and Immunoblotting. Proteins samples were separated by SDS-PAGE according to the method of Laemmli (1970) in a Mini-Protean II cell system (Bio-Rad Laboratories). Prestained calibration protein standards were from Life Technologies (Burlington, ON, Canada; catalog no. 10748-010). Electrophoresis was run until the red protein marker (~60-65 kDa) reached the end of the gel. Proteins were then stained with Coomassie R-250. Dried gels were subjected to autoradiography for 1 to 3 h on X-OMAT AR5 film. Radioactive bands were quantitated by a PhosphorImager system (Molecular Dynamics, Sunnyvale, CA) or by gamma counting after cutting out. In the case of immunoblotting, immediately after SDS-PAGE, proteins were transferred to Hybond ECL (Amersham Canada Ltd.) in Tris-glycine buffer containing 20% methanol. The nitrocellulose membrane was dried and exposed to a X-OMAT film. After exposure, the membrane was recovered and used for immunoblotting. It was saturated for 60 min in blotting buffer (0.05 M NaPO₄, pH 7.4, 0.154 M NaCl, 0.05% Tween 20, 0.1% polyvinylpyrrolidone 40) supplemented with 1% polyvinylpyrrolidone 40. The membrane was then incubated for 90 min in the presence of appropriate diluted antibodies. After the membrane was washed, antibody binding was visualized either by anti-mouse (or anti-rabbit) Ig antibody coupled to peroxidase or by streptavidine-peroxidase conjugate, depending on the primary antibody. Peroxidase activity was revealed by chemiluminescence with ECL (Amersham Canada Ltd.).

Immunoneutralization. Solubilized fibroblast proteins (0.5 µg) were incubated for 2 h at room temperature in the presence of 2 µl of integrin antiserum in a volume of 18 µl in 0.05 M HEPES, pH 7.4, 5 mM MnCl₂, and 1% NP-40. ¹²⁵I-Echistatin (25,000 cpm, 2 µl) was then added to every tube and, after a 90-min incubation, SDS sample buffer was added to obtain a final SDS concentration of 0.6%. Proteins, without heating and reducing agents, were separated on a 6% gel by SDS-PAGE. The gels were then stained, dried, and autoradiograms were obtained.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Binding Properties of ¹²⁵I-Echistatin to Cardiac Fibroblasts. Fibroblasts in suspension were used to test whether ¹²⁵I-echistatin can bind to the cell surface in an RGD-dependent fashion. Initial experiments were designed to optimize basal conditions resulting in a maximum signal. ¹²⁵I-Echistatin bound to fibroblasts demonstrated the presence of binding sites with several properties attributed to receptors. Specific binding was 1) proportional to cell number, 2) time dependent, reaching a plateau after 90 min of incubation, 3) partially reversible, because addition of 10⁶ M echistatin after reaching equilibrium was able to displace the labeled tracer by 30% in 90 min, 4) saturable, and 5) RGD dependent, because peptides with an RGD motif (echistatin, cycloGRGDSPA, GRGDTP, and GRGDSP) were able to displace labeled echistatin, whereas peptides such as GRGESP with an RGE motif, or unrelated peptides such as atrial natriuretic peptide, angiotensin II, and endothelin-1, had no effect (results not shown). The saturability of fibroblasts with ¹²⁵I-echistatin is shown in Fig. 1. Transformation of the saturation curve by Scatchard analysis indicated that the equilibrium affinity constant of ¹²⁵I-echistatin binding to fibroblast integrins was 1.24 nM with a density of 160,000 sites/cell. A correlation coefficient of 0.98 suggested that ¹²⁵I-echistatin bound to a single site.

View larger version (22K): [in this window] [in a new window]   Fig. 1.   Saturation analysis of 125I-echistatin binding to cardiac fibroblasts in suspension as assessed by filtration. Twenty thousand cells were incubated for 90 min in the presence of increasing amounts of 125I-echistatin. Nonspecific binding was determined in the presence of 10 mM EDTA. After incubation, samples were filtered and washed through glass fiber paper. The inset shows the Scatchard transformation of specific binding data.

Formation of SDS-Stable ¹²⁵I-Echistatin-Protein Complexes. Although we attempted to chemically cross-link ¹²⁵I-echistatin to binding sites on the cell surface, the resulting complexes could not be clearly resolved by SDS-PAGE (results not shown). However, we observed that ¹²⁵I-echistatin, once bound to fibroblast proteins, formed complexes resistant to 0.6% SDS in the absence of heating. Figure 2a shows that incubation of ¹²⁵I-echistatin with fibroblasts resulted in the appearance of three radioactive bands (designated f1, f2, and f3). These bands were SDS-resistant unless the samples were heated at high temperatures or a reducing agent (-mercaptoethanol) was added. The sensitivity to reduction was not unexpected because echistatin and integrins possess several disulfide bridges. These radioactive bands, compared with protein standards, have molecular masses of 220, 210, and 180 kDa. Their molecular masses are compatible with the binding of ¹²⁵I-echistatin to heterodimeric integrins rather than to their  or -subunits individually. The molecular masses of - and -subunits of RGD-dependent integrins, under nonreduced conditions, range from 80 to 180 kDa (Hemler, 1990; Schnapp et al., 1995).

View larger version (56K): [in this window] [in a new window]   Fig. 2.   a, analysis of 125I-echistatin binding to cardiac fibroblasts by SDS-PAGE. Fifteen µg of protein from NP-40-solubilized cardiac fibroblasts were incubated with 12.5 × 106 cpm/ml 125I-echistatin for 90 min. After incubation, SDS sample buffer (with or without 0.7 M -mercaptoethanol) was added to each sample and heated as indicated. Samples were then separated on 8% SDS-PAGE, stained, dried and exposed to X-ray film. b, cation dependence of 125I-echistatin binding. Ten µg of NP-40-solubilized fibroblasts were incubated for 90 min with 12.5 × 106 cpm/ml 125I-echistatin in the presence of either EDTA, MnCl2, MgCl2 or CaCl2. After incubation, proteins were separated on 6% SDS-PAGE and an autoradiogram was obtained.

6

Identification of the Components of SDS-Stable Complexes as Integrins. RGD peptides like echistatin have been documented to bind to a subfamily of integrins. To verify the identity of these proteins as RGD-dependent integrins, the following strategy was used. RGD-dependent integrins were purified on a GRGDSP affinity column. and the EDTA-eluted fractions were analyzed for ¹²⁵I-echistatin binding and for integrin subunit immunoreactivity. As illustrated in Fig. 3, the three radioactive protein bands were still observed in the eluted fractions when incubated with ¹²⁵I-echistatin, although EDTA caused a differential elution of these bands. This elution pattern probably reflects the affinity of the proteins for GRGDSP peptide. Immunoblotting of eluted fractions, which were heated before separation by SDS-PAGE, revealed the presence of ₁-, ₅-, ₃-, _v-, and ₈-subunits of expected molecular mass under nonreducing conditions. The ₁-subunit, which may form ₁₁-integrin, a non-RGD-dependent integrin that binds collagen, could not be detected in the eluted fractions, although its presence in fibroblast extract was well detected as a 190-kDa band. The anti-₃ antibody (F11) failed to recognize the 90-kDa denatured ₃-subunit and bound only the native ₃-heterodimer (see also Fig. 4b). These results suggest that ¹²⁵I-echistatin bind only to RGD-dependent integrins.

View larger version (69K): [in this window] [in a new window]   Fig. 3.   Analysis of 125I-echistatin binding and identification of integrin subunits following affinity chromatography purification of integrins. NP-40-solubilized fibroblasts were applied to a GRGDSP affinity matrix, and proteins were eluted with 10 mM EDTA. Samples from each EDTA-eluted fraction were analyzed for 125I-echistatin binding (autoradiography) or were heated in SDS sample buffer and western blotted with anti-integrin subunit antibodies. Arrow heads indicate the position of integrin subunits, and arrows the protein standard. At the right end of the anti-1 lane a fibroblast extract is shown as a positive control.

View larger version (66K): [in this window] [in a new window]   Fig. 4.   Identification of 125I-echistatin-integrin complexes by western blotting. NP-40-solubilized fibroblasts (15 µg) were incubated with increasing concentrations of 125I-echistatin (1.5 (a), 3 (b), 6.2 (c), 12.5 (d) x 106 cpm/ml, as in a, c, e and g). Alternatively, different NP-40-solubilized fibroblasts (1, 2 and 3) (as in b) or samples in duplicate (as in d and f) were used and incubated with 12.5 × 106 cpm/ml 125I-echistatin. Samples were separated on 6% SDS-PAGE and transferred to nitrocellulose, from which autoradiograms were obtained (on the left panels), and incubated with the antibody at the appropriate dilution. Horseradish-coupled second antibodies were revealed by chemiluminescence (right panels).

View larger version (50K): [in this window] [in a new window]   Fig. 5.   Analysis of 3, 5 and v-subunit immunoreactivity by 2-dimensional SDS-PAGE. NP-40-solubilized fibroblasts, incubated with 125I-echistatin, were separated on 6% SDS-PAGE. Three different gel lanes were frozen, cut in 0.5 mm slices and counted in a gamma counter. Slices corresponding to each band were pooled, heated at 100°C, and reapplied to 7.5% SDS-PAGE. After separation and transfer, the blots were incubated with specific antisera. The molecular mass of the immunoreactive bands was 150 kDa.

View larger version (32K): [in this window] [in a new window]   Fig. 6.   NP-40-solubilized fibroblasts (0.5 µg) were first incubated with 2 µl of antiserum for 120 min, after which 2 µl of 125I-echistatin (25,000 cpm) was added for another 90 min. Proteins were separated on a 6% gel SDS-PAGE and autoradiograms were obtained. Binding of the antibodies to integrins can either block 125I-echistatin binding or, under mild SDS-PAGE conditions, cause a shift of the 125I-echistatin-echistatin complexes to a higher molecular mass (supershift). cont stands for control.

Analysis of the Binding Properties of RGD Peptides to Fibroblast RGD-Dependent Integrins. To evaluate whether or not other RGD-containing disintegrins and peptides are potential ligands of fibroblast RGD-dependent integrins, we examined, by radioligand-binding filtration assay, the ability of four different disintegrins (echistatin, elegantin, flavostatin, and flavoridin) and of three synthetic RGD peptides (acPenRGDC, cycloRGDdFV, and GRGDdSP) to compete with ¹²⁵I-echistatin from fibroblasts. As shown in Fig. 7, disintegrins were about 1000 times more potent than synthetic RGD peptides in displacing ¹²⁵I-echistatin, because the most potent RGD peptide (acPenRGDC) presented an IC₅₀ of 1280 nM, whereas disintegrins have IC₅₀ values ranging from 0.044 to 1.1 nM. Analysis of the competition curves demonstrated that the slope factor of each curve, with the exception of echistatin (slope factor of 0.93), was below unity, suggesting that these peptides may interact with different affinities to more than one site. Consequently, competition curves were also analyzed by SDS-PAGE, and each band was quantitated by PhosphorImager (Fig. 8 and Table 1). As suggested by the previous analysis, RGD peptides behave differently with each integrin, the strongest interaction being with _v3, although the affinities ranged from 10 to 50,000 nM. GRGDdSP had a very low affinity toward ₈₁ and a higher affinity for _v3. AcPenRGDC had a stronger interaction with _v3 than cycloRGDdFV. Among the disintegrins, elegantin and flavostatin had the highest affinities toward fibroblast integrins, and flavoridin showed a greater selectivity toward _v3 than for other integrins. As expected from Scatchard analysis (Fig. 1) and from the competition curve (Fig. 7), echistatin presented similar affinities for all integrins.

View larger version (20K): [in this window] [in a new window]   Fig. 7.   Analysis of RGD peptide affinities by radioligand-binding filtration assay. Fibroblasts in suspension (25,000) were incubated with 12.5 × 106 cpm/ml 125I-echistatin for 90 min in the presence of increasing concentrations of disintegrins and RGD peptides. The reaction was stopped by filtration through glass fiber paper and washing. Competition curves were fitted according to Hill analysis. Slope factors below 1 indicated that the ligand may bind with different affinities to more than 1 site. Data are the means ± S. D. of 3 different experiments.

View larger version (61K): [in this window] [in a new window]   Fig. 8.   Analysis of RGD peptide affinities by SDS-PAGE. NP-40-solubilized fibroblasts (15 µg) were incubated with 12.5 × 106 cpm/ml 125I-echistatin for 90 min and increasing concentrations of disintegrins or RGD peptides. After separation on 6% SDS-PAGE, an autoradiogram was obtained and each radioactive band was subsequently quantitated by PhosphorImager. Each competition curve was fitted by the Hill equation, assuming the presence of a single site for each band. Only the results for echistatin, elegantin, flavoridin and acPenRGDC are illustrated. Each competition curve represents the means ± S. D. of 3 experiments. A representative autoradiogram of a competition experiment for each peptide is shown in the inset.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

Because they possess an RGD sequence, disintegrins mimic the binding motif of some ECM matrix proteins and can thus interact with several integrins for which this RGD sequence motif is a key binding element. ¹²⁵I-Echistatin has been used to evaluate the affinity of activated _IIb3-integrin on platelets by conventional radioligand experiments (McLane et al., 1994). More recently, it has been recognized that radioiodinated echistatin can interact with _v3- and ₅₁-integrins (McLane et al., 1994, 1998; Pfaff et al., 1994; Marcinkiewicz et al., 1996). In the present study, the binding of ¹²⁵I-echistatin to rat cardiac fibroblasts was investigated as a tool to detect the presence of individual RGD-dependent integrins. By radioligand-binding assay, it was observed that ¹²⁵I-echistatin binding presented all the characteristics of a ligand to a receptor, including specificity, reversibility, high affinity (in the nanomolar range), and limited density (saturability). It was also evident that RGD peptides displaced ¹²⁵I-echistatin from more than one site with relatively similar affinities, because the competition curves could not be resolved into their individual components, suggesting the presence of several integrins. SDS-PAGE analysis of unheated samples in the absence of reducing agents resulted in the detection of three distinct radioactive bands.

Based on the cross-reactivity of these bands with selective anti-- and -subunit antibodies, their dependence on divalent cations, in particular Mn²⁺, their displacement by RGD peptides, and their molecular masses, it is suggested that ¹²⁵I-echistatin binds to _{81, 51}, ₃₁, _v1, and _v3 and forms a complex, in the presence of Mn²⁺, with each of these integrins resistant to SDS under mild conditions. Therefore, ¹²⁵I-echistatin-integrin complexes can allow identification of individual RGD-dependent integrins. Because of the presence of Mn²⁺, these complexes may not reflect the actual binding state of integrins in vivo. The present experiments also show that ¹²⁵I-echistatin does not bind ₁₁-integrin, a non-RGD-dependent integrin, but they do not rule out that ¹²⁵I-echistatin failed to recognize some RGD-dependent integrins or that some ¹²⁵I-echistatin-RGD-dependent integrin complexes were unstable in SDS. Further experiments with cell types harboring other RGD-dependent integrins are required to verify these possibilities.

The presence of these integrins, as revealed by SDS-PAGE, on cardiac fibroblasts is in partial agreement with the literature. Indeed, the presence of ₃₁-, ₅₁-, and _v3-integrins has already been repoted (Gullberg et al., 1990; Wilke et al., 1996; MacKenna et al., 1998). The finding of ₈₁-integrin on cardiac fibroblasts is novel and unexpected. Schnapp et al. (1995) could not detect its presence by immunohistochemistry in the rat cardiac ventricle, but found it on lung alveolar myofibroblasts. Whether or not culturing of cardiac fibroblasts amplifies the signal or augments the expression of this integrin remains to be determined.

Using SDS-PAGE to resolve echistatin-integrin complexes, affinities of RGD-containing peptides were determined simultaneously on ₈₁- and _v3-integrins. However, because the f2 band consists of ₃₁-, ₅₁-, and _v1-integrins, affinity can hardly be resolved on these individual integrins. The results demonstrate that disintegrins bind with high affinity to RGD-dependent integrins. Elegantin and flavostatin, with an affinity lower than 1 nM, appear to be the most potent disintegrins, and interact with the same potency to _v3 and ₈₁. These 2 peptides possess a similar sequence around the RGD motif with only 3 amino acid differences (Table 1) (Gould et al., 1990). Flavoridin had a 10- to 20-fold lower avidity for ₈₁- and _3/5/v1-integrin mixture and a comparable affinity for _v3. This implies that other residues beyond the immediate vicinity of the RGD motif may influence the 3-dimensional structures of the binding site.

Potency of short RGD peptides was also analyzed in competition experiments. These peptides were selected according to their ability to interact selectively with RGD-dependent integrins. The cyclic peptide acPenRGDC preferentially binds _IIb3 (Bugosky et al., 1993), whereas cycloRGDdFV may interact more specifically with _v3 (Hammes et al., 1996). GRGDdSP may inhibit cell adhesion to fibronectin and platelet aggregation (Leven and Tablin, 1992). The present results demonstrate that all 3 peptides preferentially bind to _v3-integrin, and acPenRGDC presents the highest selectivity toward _v3, with a K_i of 8 nM.

The existence of SDS-stable integrin-echistatin complexes thus offers a powerful means to detect, visualize and quantitate the presence of different RGD-dependent integrins on the cell surface. Radioligand-binding filtration assay allows quantitation of the total amount of RGD-dependent integrins on fibroblasts, but does not discriminate between them because echistatin binds tightly to fibroblast integrins. To date, there is no ligand reported to be specific for each of these integrins. In the absence of selective agonists or antagonists, conventional radioligand-binding assay cannot discriminate between the different RGD-dependent integrins. By using SDS-PAGE and autoradiography, the present study demonstrates, firstly, that each integrin has a high affinity state, but with a limited density, and, secondly, that some RGD-containing peptides possess greater or lesser avidity for each of these integrins. The first finding will allow examination of the regulation of expression of RGD-dependent integrins on cardiac fibroblasts and other cell types. Indeed, we also observed SDS-resistant echistatin-integrin complexes on rat cardiomyocytes, vascular smooth muscle cells, platelets, vascular tissue, human monocytes, and human skin fibroblasts (unpublished observations). Regarding the second finding, the potency of RGD peptides interacting with integrins has usually been evaluated by their efficacy in inhibiting cell adhesion or by their ability to bind to purified integrin in a solid phase assay. The alternative approach presented here offers the possibility of an assay to evaluate and to rapidly and directly compare the binding properties of RGD peptides on several integrins.

Another interesting observation relates to the signal of several integrin subunits (see Fig. 4, a, c, d, e, and f) that was detected by western blotting. The intensity of the signal related to the integrin subunits before and after addition of ¹²⁵I-echistatin indicated that a variable fraction of each integrin subunit appears to be involved in the formation of disintegrin-integrin complexes. This may suggest that each subunit may also be involved in the formation of other integrins, like ₁₁ and ₂₁, and therefore not recognized by echistatin. Alternatively, only a variable percentage of each integrin heterodimer may be in an active, high affinity state and able to bind echistatin. Inside-out signals have been recognized as a mechanism of activating integrins (Humphries, 1996). Disintegrins may detect changes in the affinity of activated platelet _IIb3-integrins (McLane et al., 1994). Whether or not they can be used to investigate the activation of other RGD-dependent integrins deserves further study.

In summary, the present experiment demonstrate for the first time that ¹²⁵I-echistatin forms SDS-stable complexes with molecular mass of 180-220 kDa when incubated with solubilized or intact fibroblast membranes. Because these complexes can be recognized by antibodies specific to integrin subunits, are divalent cation-dependent, and are displaceable by RGD peptides and disintegrins, they reflect the interaction of ¹²⁵I-echistatin with ₈₁-, ₅₁-, ₃₁-, _v1-, and _v3-integrins. This novel approach can be applied to identify the presence and regulation of RGD-dependent integrins on the cell surface, to investigate the potency of newly-designed RGD peptides and mimetics for interaction with RGD-dependent integrins, and to evaluate the functional state of RGD-dependent integrins.