Cell Binding Sequences in Mouse Laminin a 1 Chain*

Laminin-1, a multifunctional glycoprotein of the basement membrane, consists of three different subunits, a 1, b 1, and g 1 chains. Previously, we used synthetic peptides to screen for biologically active sequences in the laminin a 1 chain C-terminal globular domain (G domain) and identified several cell binding sequences (No-mizu, M., Kim, W. H., Yamamura, K., Utani, A., Song, S. Y., Otaka, A., Roller, P. P., Kleinman, H. K., and Yamada, Y. (1995) J. Biol. Chem. 270, 20583–20590). Here, we identify new cell binding sequences on the remainder of the laminin a 1 chain by systematic peptide screening, using 208 overlapping synthetic peptides encompassing the central and N-terminal portions of the a 1 chain. HT-1080 cell attachment activity to the peptides was evaluated using peptide-coated plastic substrates and peptide-conjugated Sepharose beads. Twenty five peptides showed cell attachment activities on either the peptide-coated plastic substrates and/or the peptide-conjugated Sepharose beads. A-13 (RQVFQVAYIIIKA) showed strongest cell attachment activity in both the assays. Cell attachment to 14 of the peptides was inhibited by heparin. EDTA and integrin antibodies inhibited cell adhesion to two of the peptides, AG-10; , apparent adhesion is low; 2 , no adhesion. Triplicate experiments

Laminin-1, a major component of the basement membrane matrix with diverse biological functions (1)(2)(3)(4), is a trimeric glycoprotein composing of ␣, ␤, and ␥ chains. At least 11 isoforms of laminin, each consisting of three different chains, have been identified to date (5,6). The most extensively characterized laminin, laminin-1 (M r ϭ 900,000) from the mouse Engelbreth-Holm-Swarm tumor, consists of ␣1, ␤1, and ␥1 chains that assemble into a triple-stranded coiled-coil structure at the long arm to form a cross-like structure (7). Laminin-1 has multiple biological activities including the promotion of cell adhesion, spreading and growth, neurite outgrowth, tumor metastasis, and collagenase IV secretion (1). Proteolytic fragments, recombinant proteins, and synthetic peptides have been used to identify and characterize functional domains in laminin (8,9). The YIGSR sequence located on the ␤1 chain has been shown to promote cell adhesion and migration and to inhibit angiogenesis and tumor metastasis (10 -12). The PDSGR and F-9 (RYVVLPR) sequences located on the ␤1 chain were also found to promote cell adhesion (13)(14)(15)). An IKVAV sequence located on the C-terminal end of the long arm of the ␣1 chain was found to promote cell adhesion, neurite outgrowth, experimental metastasis, collagenase IV induction, angiogenesis, cell growth, and tumor growth (16 -19). These synthetic peptides have been used to understand the various biological functions of laminin and to develop therapeutic agents. However, biologically functional domains/sequences of laminin-1 have not been fully characterized, since it is a large molecule with complex conformations.
Recently, we began a systematic screening of cell binding sequences from the multifunctional molecule, laminin-1, using large numbers of overlapping synthetic peptides (20,21). Several cell binding sequences in the laminin ␣1 chain C-terminal globular domain (G domain) have been identified by such an approach (20). Some of these G domain peptides have cell type-specific biological activities (22)(23)(24). AG-73, which showed the strongest cell attachment activity among the G domain peptides, promoted neurite outgrowth for some but not all laminin-responsive neuronal cells (23,25) and acinar development of salivary gland cells (26,27). In addition, this peptide promoted B16-F10 mouse melanoma cell metastasis to the liver (28,29). Additionally two other G domain peptides promoted integrin-mediated cell invasion with induction of a tumor cellspecific protease (30,31). Systematic peptide screening is a powerful and logical method for identification of bioactive peptide sequences from multifunctional large molecules such as laminin.
Here we describe a systematic peptide screening of the laminin ␣1 chain (positions 1-2110, not including the G domain) using a large set of overlapping peptides. We examined the cell attachment activities of 208 different peptides by two assay systems at the first stage of screening. Twenty five potential active sequences were identified and further evaluated for integrin binding and cell type specificity.

EXPERIMENTAL PROCEDURES
Synthetic Peptides, Laminin-1 and Collagen I-All peptides were manually synthesized by the 9-fluorenylmethoxycarbonyl (Fmoc) 1 strategy and prepared in the C-terminal amide form as described previously (20,32). For the screening, peptides were generally 12 amino acids in length and overlapped with neighboring peptides by 4 amino acids. If the N-terminal amino acid was either glutamate or glutamic * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
¶ To whom correspondence should be addressed: CDBRB, NIDR, National Institutes of Health, Bldg. 30 acid, one amino acid was extended at the N terminus to avoid pyroglutamine formation (33). Cysteine residues were omitted. The respective amino acids were condensed manually in a stepwise manner using 4-(2Ј,4Ј-dimethoxyphenyl-Fmoc-aminomethyl)-phenoxy resin (34). For condensation, diisopropylcarbodiimide/N-hydroxybenzotriazole was employed, and for deprotection of N ␣ Fmoc groups, 20% piperidine in N-methylpyrrolidone was employed. The amino acid side chain protecting groups were the same as described previously (20). The resulting protected peptide resins were deprotected and cleaved from the resin with trifluoroacetic acid/thioanisole/m-cresol/ethanedithiol/H 2 O (80:5: 5:5:5, v/v) for 3 h at room temperature. The resulting crude peptides were precipitated and washed with ethyl ether and then purified by reverse-phase high performance liquid chromatography (using Vydac 5C18 column and a gradient of water/acetonitrile containing 0.1% trifluoroacetic acid). 6 peptides (A-2, -47, -69, -75, -147, and -181) were not dissolved in aqueous solutions. These sequences were evaluated for their biological activity as peptides coupled to polystyrene beads (20). The purity of the peptides was confirmed by analytical high performance liquid chromatography. The identity of the peptides was confirmed by a Scion API IIIE triple quadruple ion spray mass spectrometer (35).
Mouse laminin-1 was prepared from the Engelbreth-Holm-Swarm tumor as described previously (36). Collagen I was purchased from Collagen Corp. (Palo Alto, CA).
Preparation of Peptide-Sepharose Beads-The synthetic peptides and laminin-1 were coupled to cyanogen bromide (CNBr)-activated Sepharose 4B (Pharmacia Biotech AB, Uppsala, Sweden) according to the manufacturer's instructions. The peptide solutions (0.3 ml, 1 mg/ml in H 2 O) were mixed with 25 mg of the activated CNBr-Sepharose beads. Ethanolamine-coupled beads were prepared as a control. Amounts of coupled peptide were determined by amino acid analysis (10 -20 mmol of peptides per 1 g of Sepharose beads) (22). If the N-terminal amino acid of peptide was proline, one glycine residue was extended at the N terminus to couple with the CNBr-activated Sepharose beads.
Antibodies-The mouse monoclonal antibodies used were as follows: P4C10, an anti-␤ 1 integrin subunit antibody; P1E6, an anti-␣ 2 integrin subunit antibody; and P1B5, an anti-␣ 3 integrin subunit antibody all purchased from Life Technologies, Inc. The rat monoclonal antibody used was GoH3, an anti-␣ 6 integrin subunit antibody purchased from AMAC (Westbrook, ME). Mouse preimmune IgG was purchased from Life Technologies Inc.
Cell Attachment Assay Using Peptide Beads-Cell attachment to peptide beads was assayed in 96-well dishes that were blocked with 0.5 ml of DMEM/well containing 1% bovine serum albumin (BSA) for 1 h at 37°C. Peptide beads (approximately 5 mg) were suspended in PBS (500 l) and centrifuged, and then the peptide beads were added to the 96-well dishes and washed with DMEM (100 l, 3 times). The HT-1080 cells (5 ϫ 10 4 ) were added to each well in a total volume of 200 l of DMEM containing 0.1% BSA and incubated for 1 h at 37°C in 5% CO 2 , 95% air. The beads were stained with crystal violet and analyzed under the microscope.
Attachment Assays Using Laminin-1 and Synthetic Peptides-Cell attachment was assayed in round bottom 96-well plates (Immulon-2, Dynatech Inc., Chantilly, VA) coated with various amounts of either synthetic peptides or laminin-1. The peptides were dissolved in Milli-Q water, and 50 l was added to each well, followed by drying overnight. The wells were blocked by the addition of 100 l of 3% BSA in DMEM (200 l) at 37°C for 1 h and then washed twice with DMEM containing 0.1% BSA. Cells, detached by 0.02% EDTA in PBS and resuspended in DMEM containing 0.1% BSA, were added (3 ϫ 10 4 /0.2 ml) to each well and incubated at 37°C for 1 h in 5% CO 2 . The attached cells were stained with 200 l of 0.2% crystal violet aqueous solution in 20% methanol for 10 min. After washing, 200 l of 1% SDS was used to dissolve the cells and the optical density at 560 nm was measured in a Titertek Multiscan. Peptide-coated wells without cells were processed simultaneously to subtract the background because some peptides at higher concentrations were stained with crystal violet (22).
In peptide inhibition experiments, wells of a 96-well Immulon-2 plate were coated at room temperature overnight with either 0.2 g of laminin-1 or 0.05 g of collagen I in H 2 O in a final volume of 50 l. The wells were then washed three times with PBS and blocked with 3% BSA in 150 l of PBS at room temperature followed by washing with PBS. HT-1080 cells (3 ϫ 10 4 cells/well) were preincubated with various amounts of peptides for 15 min in 0.02% BSA/DMEM and plated on either laminin-1 or collagen I-coated dishes. After a 20-min incubation at 37°C, the attached cells were quantitated as described above.
Cell Differentiation-Salivary gland acinar formation in the presence of 100 g/ml peptide was assessed as described previously (38). Endothelial cell tube formation in the presence of 100 g/ml peptide was assessed on Matrigel as described previously (17). PC-12 cells and neurite outgrowth were assessed as described previously (23).

RESULTS
Cell Attachment Activity on Laminin ␣1 Chain Peptide-conjugated Beads-Two hundred and eight peptides from the laminin ␣1 chain were prepared for screening cell attachment activity ( Fig. 1). First, we evaluated cell adhesion to covalently conjugated peptides on Sepharose or polystyrene beads. Two hundred and two soluble peptides were coupled with CNBractivated Sepharose beads. Six insoluble peptides were prepared directly on the polystyrene beads (20). As a positive control, the laminin ␣1 chain G domain peptides AG-10 and AG-73 were conjugated to Sepharose beads (22). As a negative control, ethanolamine-conjugated Sepharose beads were also prepared. Cell attachment activities on the peptide-Sepharose beads and peptide-polystyrene beads were tested using HT-1080 human fibrosarcoma cells ( Fig. 2 and Table I). The cells attached and spread on the AG-10-and AG-73-conjugated Sepharose beads as shown previously (22), but the cells did not attach on control ethanolamine-conjugated beads (Fig. 2). Four of peptide-Sepharose beads (A-13, -51, -99, and -112), including the RGD sequence (39,40) containing peptide, showed strong cell attachment and spreading activity comparable to that observed with the AG-73-conjugated Sepharose beads (Fig. 2). The cells attached and spread on seven of the peptide-Sepharose beads (A-10, -18, -25, -64, -167, -177, and -194) with lesser activity similar to that observed with AG-10 beads. Five of the peptide-conjugated Sepharose beads (A-3, -12, -54, -55, and -174) showed weak cell attachment activity. The remaining 191 peptide beads did not have significant cell adhesive activities in this assay (Fig. 2). None of the insoluble peptides conjugated to polystyrene beads showed cell attachment activity.
Inhibition of cell attachment to laminin-1 and collagen I by active peptides was analyzed to examine the availability of these active sites on laminin-1 and substrate specificity of their activity (Fig. 4). A-10 and A-13 significantly inhibited HT-1080 cell attachment to laminin-1 by about 90%. Inhibition of cell attachment by A-24 and A-25 was much weaker than A-10 and FIG. 1. Sequence and peptides from the laminin ␣1 chain (position ؊24 to 2108). Sequences were derived from the mouse laminin ␣1 chain (51). Locations of peptides are indicated by arrows. Active peptides are shown by a dashed line. Cell attachment activities are shown in parentheses as follows: (ϩϩ), active in both peptide-conjugated bead assay and peptide-coated plate assay; (ϩ), active in one assay, either bead assay or coated plate assay; (Ϫ), not active in any assay; *, peptide was not soluble. Cell attachment activity of the insoluble peptides was determined on peptide-conjugated polystyrene beads as described previously (20).
A-13. A-4 did not inhibit HT-1080 cell attachment on a laminin-1 substrate. All these peptides showed little inhibition of cell attachment to collagen I, indicating laminin-specific activity. AA-13, a scrambled peptide of A-13, showed no activity on either laminin-1 or collagen I (data not shown). Also A-29, A-50, A-51, A-54, and A-64 showed no activity on either substrate (data not shown). A-55 inhibited only at highest concentrations on laminin and collagen I and was not specific (data not shown). Therefore, active sites for A-10, A-13, A-24 and A-25 are likely available on the intact laminin molecule. These results suggest that conformation is important in cell recognition with these peptides.
Effects of Heparin on Cell Attachment-Next, we evaluated the effects of heparin on HT-1080 cell attachment to the 20 active peptides (Fig. 5). As a control, laminin-1 and AG-73 were used. Cell attachment to AG-73 was inhibited by heparin, whereas attachment to laminin-1 was not affected as shown previously (21). Cell attachment to 5 peptides (A-4, -10, -50, -54, and -55) was slightly inhibited by heparin, whereas attachment to the remaining 15 peptides was significantly inhibited. Thus, many of the sequences may bind to negatively charged surface molecules.
Effects of EDTA and Integrin Antibodies on Cell Attachment to the Peptides-We next focused on evaluating the 20 peptides active in the plate assay to define further their cellular interactions. The effects of EDTA on HT-1080 cell attachment to the peptide-coated plates were examined to evaluate the role of cations (Fig. 6). Cell attachment to laminin-1 and to AG-73 was inhibited by 5 mM EDTA as shown previously. Attachment to 6 peptides, A-12, -13, -25, -64, -119, and -208, was significantly inhibited by EDTA. EDTA did not have an effect on cell attachment to the other peptides tested. These results suggest that HT-1080 cell attachment to some of the peptides is mediated via cation-dependent cellular receptors such as integrins.
Next, we tested the effects of integrin antibodies on cell attachment to very active peptides, A-13, A-25, and A-64, using anti-␤1, -␣2, -␣3, and -␣6 neutralizing integrin antibodies (Fig.  7 lanes c-f, respectively). The anti-␤ 1 integrin antibody inhibited laminin-1-mediated cell attachment. None of the integrin antibodies affected AG-73-mediated cell attachment. Cell attachment to A-13 and A-25 was inhibited by the anti-␤ 1 integrin antibody, whereas cell attachment to A-64 was not affected by any of the integrin antibodies. We conclude that cell attachment to A-13 and A-25 involved integrins, whereas the A-64 peptide likely interacts with either a different integrin and/or a non-integrin receptor.
Active Core Sequence of A-13-Since A-13 is very active for cell adhesion, we examined the structural requirements for biological activity. The active core sequence of A-13 was determined using systematically truncated N-terminal and C-terminal peptides (Table II). When we examined cell attachment activity in the peptide-coated plate assay, A-13b (FQVAYII- IKA), an N-terminal truncated peptide, still retained full activity. A deletion (A-13c) of phenylalanine and glutamine from A-13b resulted in reduced activity in the plate assay but not in the bead assay. Since glutamine at the N terminus of the peptide is known to easily form a pyroglutamine (33), a deletion of phenylalanine from A-13b was not tested. A deletion (A-13d) of valine from A-13c resulted in complete loss of activity in the bead assay, but some activity was retained in the plate assay.

TABLE I Synthetic laminin ␣1 peptides and their cell attachment activities
In all cases, the biological activities were quantitated and peptide activities were evaluated relative to the activity observed with laminin-1, AG-73, and AG-10 as shown in Fig. 3. Cell attachment was evaluated on the following subjective scale: ϩϩϩ, adhesion comparable to those on laminin-1 and AG-73; ϩϩ, adhesion similar to that of AG-10; ϩ, apparent adhesion is low; Ϫ, no adhesion. Triplicate experiments gave similar results.
When the N-terminal alanine was deleted from A-13d, the peptide (A-13e) showed no activity even in the plate assay. The C-terminal deletion peptide, A-13g, showed full activity. A deletion of lysine (A-13h) reduced cell binding activity in the plate assay, but full activity was retained in the bead assay until all three isoleucines were deleted (A-13k). Further deletion (A-13m) of tyrosine and alanine from A-13k abolished activity in the plate assay as well. These results indicate that the alanine at position 7 of AG-13 is critical for cell attachment activity in the peptide-coated plate assay. In contrast, valine (position 6) and isoleucine (position 9) are important for activity in the bead assay. Based on these results from the two separate assays, we conclude that the active core sequence for A-13 is VAYI. Cell Type Specificity-We next determined if the 25 peptides active with HT-1080 cells would interact with other cell types including PC-12 cells (rat pheochromocytoma cells), HSG cells (human salivary gland cells), and human umbilical vein endothelial cells, all of which bind well to laminin-1 (Table III). Several of the peptides including A-12, -18, -25, -50, -54, -99,  -112, -177, -194, and -206 when coated on dishes were inactive with all three cell types. In contrast, A-10, -13, -55, -64, -167, -203, and -208 when coated on dishes or in solution were active with all three cell types. Surprisingly 2 peptides A-64 and A-119 were not active for endothelial cell adhesion but did reduce tube formation on Matrigel suggesting that soluble peptides could be recognized by these cells. The bound forms of these peptides were adhesive with PC-12 cells. The remaining 9 peptides, A-3, -4, -24, -51, -65, -119, -121, -174, and -203, showed some cell type specificity in their cellular interactions. Four peptides, A-24, -51, -121, and -174, were active with PC-12 and HSG cells but not with endothelial cells. Two peptides, A-3 and -65, were not active with PC-12 cells but were active with HSG and endothelial cells. A-4 was only active with HSG cells, and A-119 was active with PC-12 cells. These data demonstrate considerable cell type specificity in the cellular interactions with laminin peptides suggesting specific receptor recognition. DISCUSSION Previously we screened the G domain of the laminin ␣1 chain and identified several active sites using systematic synthetic peptides (20). Here we describe identification of cell binding sequences from the remainder of the laminin ␣1 chain (Fig. 8). We prepared 208 overlapping peptides and evaluated their cell binding activity with HT-1080 human fibrosarcoma cells using peptide-conjugated beads and peptide-coated plates. Eleven of the peptides showed cell attachment on both peptide-conjugated beads and peptide-coated plates. Five peptides showed cell attachment only to peptide-conjugated beads. These peptides may be have less binding ability to plastic plate and/or they cannot form active conformations on the plastic. In contrast, nine peptides including A-208, which contained the previously identified active IKVAV (16) sequence, were active only in the peptide-coated plate assay.
An N-terminal region of the laminin ␣1 chain, domain VI, has been shown to promote neurite outgrowth and to interact with ␣ 1 ␤ 1 integrin (41,42). We also identified several cell adhesive sequences located on the N-terminal region one of which, A-13, showed strong cell attachment activity that was inhibited by EDTA and by anti-integrin ␤ 1 antibody. In addition, several active peptides, including A-13 from domain VI, showed neural cell adhesion and neurite outgrowth activity. 2 This site may be an important biological domain on laminin-1.
Several peptides previously reported to be active with certain cells were not active in our screen further confirming cell type specificity within neuronal cells. A YFQRYLI peptide (position 1583-1589) was found previously to promote neurite outgrowth (43). A-144 and A-145 containing a part of the YFQRYLI did not show HT-1080 cell attachment activity. However, when a peptide containing the entire YFQRYLI sequence was prepared, we confirmed it was active for cell attachment with PC-12 (data not shown). Previously an LRE (Leu-Arg-Glu) sequence was found to be active for neuronal cell attachment and neurite outgrowth (44 -46). A-193 (SLAMLRESPGGM, position 2001-2012) contains an LRE sequence, but this peptide did not show HT-1080 cell attachment activity using the peptide-conjugated Sepharose beads and peptide-coated plates. It was also not active with PC-12. The LRE sequence may have 2 S. K. Powell and H. K. Kleinman, unpublished results.
cell binding activity specifically for certain neuronal cells. Such cell type specificity has been observed with some of the G domain peptides of laminin-1 (23).
The finding of additional cell type specificity in the remainder of the ␣1 chain is important. The differences observed cannot be due to coating efficiencies on the plastic or conformation of the peptides since various cells showed differences in attachment with the same peptides. The cell type specificity observed demonstrates unique cellular receptors. There may be specific receptors for each cell type. Alternatively, the observed differences could be the result of differences in the receptor levels and/or affinity to different cell types. Some of the receptors are in the integrin family, which can recognize multiple sequences (47).
We have now screened the entire laminin-1 component chains ␣1, ␤1, and ␥1 using 673 overlapping synthetic peptides (20,21). 3 Approximately 11% of the peptides (74 out of 673) were active in either the Sepharose bead assay or the peptidecoated plate assay. Of the peptides active in either assay, approximately a third (28 peptides) or 4% of the total peptides tested were active in both assays. More than half of the active peptides (31 peptides) were localized in the globular domains. The globular domains contain few cysteines and thus are relatively less conformationally stable than either the epidermal growth factor-like repeats located between some of the globular domains in the N termini or the coiled-coil domain located centrally in the ␣1 chain. The presence of the majority of the cell-binding sites in the globular domains may be due to the less rigid structures of these regions. Our data with the bead assay and the peptide-coated plate assay suggest the importance of conformations regulating cellular interactions. Since the test peptides were attached to the beads in solution versus dried on the dishes, it is likely that different conformations occurred. Inhibition of cell attachment on a laminin-1 substrate by peptides suggests that active sites of A-10, A-13, A-24, and A-25 are available on the intact molecule and that their activity is specific to laminin-1. These results confirm the importance of conformation for cell interactions.
It is possible that some active sequences in laminin are always cryptic and are not exposed with different lamininprotein interactions. Such cryptic sites may be active only when laminin is cleaved. Recently, fragments of laminin-5 generated by MMP-9 were found to induce cell migration and tumor metastases (48). Likewise, certain other proteolytic fragments of some proteins have been found to possess activities not exhibited by the intact molecule. For example, the anti-angiogenesis factors angiostatin and endostatin are fragments of plasminogen activator and collagen XVIII, respectively (49,50). Thus, while some of the active cell binding sequences may be dependent on conformation in the intact molecule, others could be cryptic and only exposed with proteolytic cleavage.