HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 279, Issue 23, 24774-24784, June 4, 2004

This Article Abstract Full Text (PDF) All Versions of this Article: 279/23/24774    most recent M400670200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kariya, Y. Articles by Miyazaki, K. Search for Related Content PubMed PubMed Citation Articles by Kariya, Y. Articles by Miyazaki, K. Social Bookmarking                      What's this?

Characterization of Laminin 5B and NH₂-terminal Proteolytic Fragment of Its 3B Chain

PROMOTION OF CELLULAR ADHESION, MIGRATION, AND PROLIFERATION^*

Yoshinobu Kariya, Chie Yasuda, Yukiko Nakashima¶, Kumiko Ishida¶, Yoshiaki Tsubota, and Kaoru Miyazaki¶||

From the Division of Cell Biology, Kihara Institute for Biological Research, Yokohama City University, Yokohama 244-0813, the ^¶Graduate School of Integrated Sciences, Yokohama City University, 641-12 Maioka-cho, Totsuka-ku, Yokohama 244-0813, and the Kihara Memorial Yokohama Foundation for the Advancement of Life Sciences, 641-12 Maioka-cho, Totsuka-ku, Yokohama 244-0813, Japan

Received for publication, January 21, 2004 , and in revised form, March 18, 2004.

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
Various laminin isoforms have specific biological functions depending on their structures. Laminin 5A, which consists of the three truncated chains 3A, 3, and 2, is known to have strong activity to promote cell adhesion and migration, whereas a laminin 5 variant consisting of a full-sized 3 chain (3B) and the 3 and 2 chains, laminin 5B, has not been characterized yet. In the present study, we for the first time cloned a full-length human laminin 3B cDNA and isolated the human laminin 5B protein. The molecular size of the mature 3B chain (335 kDa) was approximately twice as large as the mature 3A chain in laminin 5A. Laminin 5B had significantly higher cell adhesion and cell migration activities than laminin 5A. In addition, laminin 5B potently stimulated cell proliferation when added into the culture medium directly. Furthermore, we found that the 3B chain undergoes proteolytic cleavage releasing a 190-kDa NH₂-terminal fragment. The 190-kDa fragment had activities to promote cellular adhesion, migration, and proliferation through its interaction with integrin ₃₁. These activities of the NH₂-terminal structure of the 3B chain seem to contribute to the prominent biological activities and the physiological functions of laminin 5B.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
Laminins are an important family of extracellular matrix proteins, which are mainly localized in the basement membranes of various tissues and regulate various cellular functions including cell adhesion and migration (1, 2). The different combinations of five , three , and three chains give rise to at least 15 laminin heterotrimers with the well known cross-shaped structure (1). Of the three laminin subunits, the chain is most important in determining the tissue-specific distribution and specific biological activity of the laminin isoforms (3). The laminin chains are classified into two groups, "full-sized" (1, 2, 3B, and 5) and "truncated" (3A and 4) (1). The amino-terminal region of the full-sized chains contains three globular domains (VI, IVb, and IVa) and three rodlike domains (V, IIIb, and IIIa) consisting of laminin type epidermal growth factor repeats. The laminin NH₂-terminal domain VI, which is completely absent in the truncated chains (3A and 4), is thought to be essential for the self-assembly and co-polymerization of laminins into large noncovalent networks in the basement membranes and have several cell-binding sites (1). On the other hand, all chains contain a globular (G) domain consisting of five laminin G modules (G1 to G5) at the COOH terminus. The COOH-terminal G domain of the chains plays major roles in the interaction with cell surface receptors such as integrins, syndecans, and dystroglycan (4-6).

Typical laminins are produced by the assembly of different chains with the laminin 1 or 2 and 1 chains, whereas laminin 5 (LN5)¹ is produced by the combination of 3A, 3, and 2 chains. The 3 and 2 chains are found only in LN5, and all the three LN5 subunits are truncated in the short arms (NH₂-terminal regions). Consistent with the unique structure, LN5 has unique biological activities. LN5 strongly promotes cellular scattering, adhesion, and migration of various types of cells compared with other laminins and extracellular matrix proteins (7-9). These activities are mediated by the interaction with integrin receptors, ₃₁, ₆₁, and ₆₄ (10, 11). In the skin, the association of LN5 with integrin ₆₄ is critical to form the hemidesmosome structure, which supports the stable adhesion of basal keratinocytes to the underlying connective tissues (12-14). Therefore, functional defects of LN5 cause a lethal skin disease, Herlitz's junctional epidermolysis bullosa (15, 16). On the other hand, the cell migration-promoting activity of LN5 is thought to contribute to wound healing (17, 18) and tumor invasion (19).

Cloning of the laminin 3 chain cDNA has revealed two distinct mRNA transcripts encoding the truncated, 3A chain (3EpA) and the full-sized, 3B chain (3EpB) (17, 20). The expression of the two transcripts is regulated by alternative splicing and distinct promoters (21). The complete sequence of the 3B chain has recently been deduced from the sequences of partial cDNA clones and genomic DNA clones in mice (3, 20, 22) and humans (23, 24). However, neither complete cDNA clones for the 3B chain nor the 3B-containing laminins have been isolated so far. The 3B chain is expected to associate with the laminin 3 and 2 chains and with the laminin 1/2 and 1 chains to produce laminin 5B (LN5B) and laminin 6B/7B, respectively, because the 3A and 3B chains share the common I/II domain, which is essential for the coiled-coil structure of the three subunits. LN5B consists of the full-sized 3 chain (3B) and the truncated 3 and 2 chains, and thereby it is expected to have a long rodlike structure. This type of laminin isoform is not known apart from LN5B. Because the laminin 3B chain is more widely and more intensively expressed than the 3A chain in mouse and human tissues (20, 24), LN5B seems to be a major LN5 isoform in adult tissues.

In the present study, we cloned a human laminin 3B chain cDNA and for the first time isolated and characterized the human LN5B protein from HEK293 cells, which had been transfected with the cDNAs for the three LN5B subunits. The results demonstrate unique biological activities of this novel LN5 isoform and an NH₂-terminal, proteolytic fragment of the 3B chain. To distinguish the two forms of LN5, we designate the well characterized LN5 consisting of the 3A, 3, and 2 chains as LN5A in this report.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
Materials—Human laminin 10/11 was purchased from Invitrogen (Carlsbad, CA). Mouse EHS laminin 1, human laminin 2/4 (merosin), and human fibronectin were purchased from Chemicon (Temecula, CA). Human recombinant LN5A was purified as described previously (25). Mouse monoclonal antibodies against the NH₂-terminal regions of human laminin 3A chain (Lsc3) and 2 chain (D4B5) have been described previously (26, 27). A monoclonal antibody against the human laminin 3 chain (kalinin B1) was purchased from Transduction Laboratories (Lexington, KY). Function-blocking anti-integrin antibodies used are the anti-₂-integrin antibody (P1E6), the anti-₃-integrin antibody (P1B5), the anti-₅-integrin antibody (P1D6), and the anti-₁-integrin antibody (6S6) from Chemicon, and the anti-₆-integrin antibody (GoH3) from Pharmingen (San Diego, CA).

Cells and Culture—The human embryonic kidney cell line HEK293 (ATCC CRL-1573) was purchased from American Type Culture Collection (ATCC) (Rockville, MD). The human bladder carcinoma cell line EJ-1 and the Buffalo rat liver-derived epithelial cell line BRL have been used in previous studies (26). A spontaneously immortalized human keratinocyte cell line, HaCaT, was a generous gift from Dr. N. E. Fuseing (Deutsches Krebsforschungszentrum, Heidelberg, Germany). These four cell types were maintained in DMEM/F-12 medium (Invitrogen) supplemented with 10% fetal calf serum (FCS), penicillin, and streptomycin sulfate. The human mammary epithelial cell line MCF-10A (ATCC CRL-10317) was obtained from ATCC, and cultured in DMEM/F-12 supplemented with 20 ng/ml epidermal growth factor (EGF), 100 ng/ml cholera toxin, 0.01 mg/ml insulin, 500 ng/ml hydrocortisone, and 5% horse serum.

Expression Vector Constructions—The expression vectors of human laminin 3A, 3, and 2 chains have been described previously (25). A human laminin 3B chain cDNA was constructed as follows. Because approximately half of the 3B sequence is identical to the 3A sequence, we used the 3A cDNA LS/CX encoding the COOH-terminal sequence of the 3B chain (28). The remaining 3B cDNA sequence was deduced from a previously reported partial sequence (GenBank^TM accession no. AF005258 [GenBank] ) and the sequences of two genomic clones, RP11-609K12 and RP11-666022 (GenBank^TM accession nos. AC067796 [GenBank] and AC090366 [GenBank] , respectively). To obtain overlapping cDNA clones, a human lung 5'-Stretch Plus cDNA library (Clontech, CA) was screened and amplified by the ECL direct nucleic acid labeling/detection system and PCR with Ex Taq polymerase (Takara, Tokyo, Japan), respectively. A cDNA fragment of 300 bp in the 5' region was synthesized as three oligonucleotides and amplified by PCR. All PCR-derived cDNA fragments were cloned into the pGEM T-Easy vector (Promega, Madison, WI), and their sequences were verified. All primers used for PCR are listed in Table I. The different cDNA fragments that overlap each other were ligated. The full-length laminin 3B cDNA consisted of 10,002 bp and had an open reading frame encoding 3,333 amino acids (GenBank^TM accession no. AB107369 [GenBank] ). To prepare an expression vector, we constructed a truncated 3B cDNA of 8,832 bp that had an open reading frame encoding 2,944 amino acid residues without the COOH-terminal G4-G5 domain. This truncated 3B cDNA was inserted into the EcoRV sites of pcDNA3.1/Hygro (+) mammalian expression vector (Invitrogen) in correct orientation and used as LN3B#3 pcDNA3.1 Hygro (+).

For purification of LN5B, the serum-free conditioned media from 32-HEK293 cells and LN5B-HEK293 cells were collected in roller bottles, which had been precoated with poly-L-lysine, and proteins were precipitated by 80% saturated ammonium sulfate. The precipitate was dissolved in and dialyzed against 20 mM Tris-HCl buffer (pH 7.5) containing 0.5 M NaCl, 0.005% (w/v) Brij35, and 0.1% (w/v) CHAPS, and then applied to molecular sieve chromatography on a Sepharose 4B column (Amersham Biosciences) pre-equilibrated with the same buffer. Fractions containing LN5B were pooled and applied to a gelatin-Sepharose 4B column to remove fibronectin. LN5B in the unbound fractions from the gelatin column was purified by immunoaffinity chromatography with the anti-laminin 3 monoclonal antibody LSc3. Bound proteins were eluted from the affinity column with 0.05% (v/v) trifluoroacetic acid and immediately neutralized to pH 7.0 to 7.5 with a small volume of 1 M Tris-HCl (pH 8.0). The LN5B protein thus purified was stored in the presence of 0.005% Brij35 and 0.1% CHAPS. Protein concentrations were determined using a Bio-Rad protein assay kit with bovine serum albumin (BSA) as a standard.

Purification of 190-kDa NH₂-terminal Fragment of Laminin 3B Chain—The serum-free conditioned medium from LN5B-HEK293 cells was fractionated by molecular-sieve chromatography on a Sepharose 4B column as described above. Fractions containing the 190-kDa, proteolytic fragment of laminin 3B chain were pooled, dialyzed against 20 mM Tris-HCl buffer (pH 7.5) containing 0.05 M NaCl, 0.005% (w/v) Brij35 and 0.1% (w/v) CHAPS, and applied to a Q-Sepharose FF anion-exchange column (Amersham Biosciences) pre-equilibrated with the same buffer. Bound proteins were eluted with a 0.05-0.5 M NaCl gradient. The 190-kDa fragment was eluted at 0.15-0.2 M NaCl, and the protein was further purified by molecular sieve chromatography on a Superdex 200 HPLC column (Amersham Biosciences). To avoid a possible contamination of the sample with LN5B, the purified 190-kDa fragment was passed through an immunoaffinity column with the anti-laminin 2 monoclonal antibody D4B5.

Immunofluorescence Microscopy—Two hundred microliters of the cell suspension (2 x 10⁵ cells/ml) in DMEM/F-12 medium supplemented with 10% FCS was inoculated per well of Lab-Tek 8-well chamber slides (Nunc, Naperville, IL). After incubation for 2 days, the cells were washed with PBS (-) and then fixed in 3.7% (w/v) formaldehyde in PBS (-) for 15 min. For permeabilization, the cells were treated with 0.2% (v/v) Triton X-100 in PBS (-). The fixed cells were blocked with 1.2% BSA in PBS (-) for 1 h and then incubated with a primary antibody diluted in the BSA-containing buffer for 1 h. A fluorescein isothiocyanate-coupled secondary antibody (Vector Laboratories, Burlingame, CA) was used for detection. F-actin was stained using rhodamine phalloidin (Molecular Probes). Fluorescence images were obtained using a fluorescence microscope (BX50-FLA; Olympus, Tokyo, Japan) equipped with 100 x/1.35 UPlan-Apochromat oil immersion objectives.

Cell Adhesion Assay—Cell adhesion assay was performed as described previously (26). Briefly, each well of 96-well enzyme-linked immunosorbent assay plates (Costar, Cambridge, MA) was coated with a substrate protein and then blocked with 1% BSA. Cells (2 x 10⁴ cells) were inoculated per each well containing serum-free DMEM/F-12 medium, and incubated at 37 °C for 1 h. After nonadherent cells were removed, adherent cells were fixed and stained with Hoechst 33432. The fluorescent intensity of each well of the plates was measured using a CytoFluor 2350 fluorometer (Millipore, Bedford, MA). For inhibition assay, the cell suspension was incubated with function-blocking anti-integrin antibodies for 20 min at room temperature before inoculation.

Assays of Cell Scattering and Migration—The cell-scattering and cell migration activities of LN5A and LN5B proteins were assayed as reported previously (26). BRL cells were suspended in DMEM/F-12 plus 1% FCS and inoculated into each well of 24-well plates at a density of 7 x 10³ cells/well. Each test sample was directly added to the cultures and incubated at 37 °C for approximately 40 h. For the cell migration assay, EJ-1 cells (2.0 x 10⁴ cells in DMEM/F-12 plus 1% FCS) were inoculated per well of 24-well plates precoated with a test protein. After the preincubation for 1.5 h at 37 °C, cell movement was monitored using a time-lapse video equipment for 8 h.

SDS-PAGE and Immunoblotting—SDS-PAGE was performed on 5% gels or 4.0-7.5% gradient gels under reducing or nonreducing conditions. Separated proteins were stained with a Wako silver staining kit II (Wako, Osaka, Japan) or Coomassie Brilliant Blue (CBB). For immunoblotting analyses, proteins resolved by SDS-PAGE were transferred to nitrocellulose membranes and detected by the ECL detection reagents (Amersham Biosciences).

Determination of Amino Acid Sequence—Proteins were separated by SDS-PAGE and transferred to polyvinylidene difluoride membranes (Nippon Genetics, Tokyo, Japan). The blotted membranes were stained with Coomassie Brilliant Blue G, and the NH₂-terminal protein sequences were determined using a 492 Procise protein sequencer (Applied Biosystems, Foster City, CA).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
Secretion of LN5B by 32-HEK Cells—The laminin 3B chain has been reported to be expressed in HEK293 cells (24). We detected none of the laminin 3B, 3, and 2 chains in the culture medium of HEK293 cells by immunoblotting analysis (data not shown). However, when HEK293 cells were transfected with the cDNA expression vectors for the 3 and 2 chains, the conditioned medium of the HEK293 transfectant (32-HEK) showed a weak cell-scattering activity toward BRL cells, suggesting that the heterotrimer of LN5 or LN5B was secreted to the culture medium (data not shown). We purified this cell-scattering factor from the serum-free conditioned medium of 32-HEK cells by immunoaffinity chromatography with a monoclonal antibody against the laminin 3 chain. The purified protein had activities to promote cell adhesion, migration, and scattering (data not shown). SDS-PAGE under nonreducing conditions resolved the purified protein into four major bands at 630, 580, 435, and 385 kDa (Fig. 1A, lane 2), of which the latter two bands were similar in size to the two LN5A forms (Fig. 1A, lane 1). Under reducing conditions, the purified protein was resolved into five major bands of 335, 150, 145, 135, and 105 kDa (Fig. 1A, lane 4). These bands were reactive to one of the antibodies against the three LN5 subunits in immunoblotting analysis, and the three smaller protein bands were also found in LN5A (data not shown). Because the 335-kDa band was recognized by the anti-laminin 3 antibody, the purified protein was thought to be LN5B. However, the yield of LN5B was too low for further characterization.

View larger version (29K): [in this window] [in a new window]   FIG. 1. SDS-PAGE analyses of LN5A or LN5B secreted by 32-HEK, LN5-HEK, and LN5B-HEK cell lines. A, SDS-PAGE of LN5B (lanes 2 and 4) purified from 32-HEK cells. LN5A (lanes 1 and 3) purified from LN5-HEK cells was analyzed as a control on the same gel. Proteins were separated on a 4.0-7.5% gradient gel under nonreducing (-ME; lanes 1 and 2) and reducing (+ME; lanes 3 and 4) conditions and stained with silver. B, SDS-PAGE of conditioned media (CM; lanes 1-4) and extracellular matrix proteins (ECM; lanes 5-7) prepared from parent HEK cells (P; lanes 1 and 5), LN5-HEK cells (L5A; lanes 2 and 6), and LN5B-HEK cells (L5B; lanes 3, 4, and 7). Proteins were stained with CBB. -ME (lanes 1-3), nonreducing SDS-PAGE; +ME (lanes 4-7), reducing SDS-PAGE. Major protein bands and their approximate molecular sizes in kDa are shown by bars. Other experimental conditions are described under "Experimental Procedures."

View larger version (17K): [in this window] [in a new window]   FIG. 2. Comparison of domain structures among laminin 3B, 3A, 1, and 5 chains. Three globular domains (VI, IVb, and IVa) in the NH2-terminal region of the chains are indicated by ovals, and three rodlike EGF repeats (V, IIIb, and IIIa) are shown as vertical rectangles (1). The COOH-terminal globular domain consists of five globular modules (G1-G5). Domain II/I is the coiled coil-forming domain. The NH2-terminal amino acid sequences for the 335-kDa 3B chain (ATARDPGA), the 190-kDa fragment chain (ATARDPGA), and the 145-kDa 3B and 3A chains (DSSPAEE) were determined by amino acid sequencing and are indicated with bold letters. The last amino acid residue number in each domain is indicated above the schematic diagram of the 3B chain. Arrows indicate proteolytic cleavage sites of the 3B and 3A chains to produce the 145-kDa 3B/A chains. Arrowheads indicate the cleavage sites within the G domain.

View larger version (52K): [in this window] [in a new window]   FIG. 3. Comparison of parent HEK, 32-HEK, LN5-HEK, and LN5B-HEK cell lines. A, cell morphology. Parent HEK (HEK), 32-HEK, LN5B-HEK, and LN5-HEK cell lines were cultured in DMEM/F-12 medium supplemented with 10% FCS for 2 days, and the cell morphology was examined under a phase-contrast microscope. Original magnification, x300. B, deposition of LN5A and LN5B. LN5-HEK and LN5B-HEK cells were cultured in DMEM/F-12 supplemented with 10% FCS for 2 days, fixed, and permeabilized. LN5A and LN5B deposited in the cultures were immunostained with the anti-laminin 2 antibody as described under "Experimental Procedures" (2; upper panels). F-actin was stained with rhodamine phalloidin (Actin; lower panels). Asterisks (unoccupied space) and # marks (cell body) indicate the same position in the respective cultures. Arrowheads indicate a position of treelike LN5A deposition. C, growth curves of three cell lines. Parent HEK cells (open triangles), LN5-HEK cells (open circles), and LN5B-HEK cells (closed squares) were inoculated at a density of 2 x 104 cells/60-mm dish in DMEM/F-12 medium containing 10% FCS, and incubated for the indicated lengths of time with medium change every 2 days. Grown cells were harvested with trypsin and counted with a hemocytometer. Each point represents the average ± S.D. of the cell numbers in triplicate dishes. D, comparison of cell growth rates of parent HEK, 32-HEK, and three clones of LN5B-HEK. Parent HEK cells, 32-HEK cells, and LN5B-HEK cell clones (clones 2, 7, and 9) were plated at a density of 2 x 104 cells/35-mm dish in DMEM/F-12 medium containing 10% FCS, and incubated for 6 days. Medium change was done 3 days later. Each point represents the average ± S.D. of the cell numbers in triplicate dishes.

Deposition of the LN5 isoforms was compared between LN5B-HEK and LN5-HEK cell lines. When the extracellular matrix proteins deposited by LN5B-HEK cells were analyzed by reducing SDS-PAGE, the LN5B subunits found in the conditioned medium, but not the 190-kDa protein, were also detected as major bands by the CBB staining (Fig. 1B, lane 7). There was no apparent difference in the efficiency of the LN5B/5A deposition between LN5B-HEK and LN5-HEK cell lines. When LN5B and LN5A were subjected to SDS-PAGE under nonreducing conditions, these proteins did not migrate from the top of the gel, suggesting that they formed large protein complexes in the matrix (data not shown).

The deposition of LN5B and LN5A on the matrix was further analyzed by immunofluorescent microscopy using the anti-2 antibody (Fig. 3B). LN5B-HEK cells and LN5-HEK cells showed distinct distribution of the LN5 isoforms. LN5A was highly deposited at peripheral regions of cell bodies forming treelike patterns. LN5A was also deposited on cell-free spaces. In contrast, LN5B was deposited under cell bodies producing cloudlike high density spots, which were co-localized with the marginal actin accumulation. LN5B was scarcely deposited on cell-free spaces.

We also found that expression of LN5B in HEK293 cells has a striking effect on the cell growth (Fig. 3C). Parent HEK293 cells very poorly grew in a standard culture medium containing 10% FCS, whereas LN5B-HEK cells rapidly grew in the same conditions. LN5-HEK cells also grew well but at a slightly lower rate than LN5B-HEK cells. The two other clones, 2 and 7 of LN5B-HEK cells, also showed high growth rates similar to that of clone 9 (Fig. 3D). Unexpectedly, the growth rate of 32-HEK cells was comparable with that of LN5B-HEK cells (Fig. 3D), although the two cell types showed different morphology (Fig. 3A). These results suggested that the secreted LN5B and LN5A stimulated the proliferation of the producing cells.

The LN5B production in 32-HEK cells was estimated to be less than 1% of that in LN5B-HEK cells. Therefore, it is considered that a high LN5B production is required for the morphological change, but a very low level of LN5B is enough for the growth stimulation.

Purification of Recombinant LN5B—To purify recombinant LN5B, the conditioned medium of LN5B-HEK was applied to molecular sieve chromatography and then to immunoaffinity chromatography with the anti-laminin 3 monoclonal antibody. The purified LN5B was analyzed by SDS-PAGE under both reducing and nonreducing conditions (Fig. 4). The purified LN5B was separated into two major bands of 630 and 580 kDa and two minor bands of 435 and 385 kDa under nonreducing conditions (Fig. 4A, lane 1). The relative weight ratio of the 435/385-kDa LN5B to the 630/580-kDa form was estimated to be approximately 1/4. Under reducing conditions, it was separated into five major bands of 335, 150, 145, 135, and 105 kDa (Fig. 4A, lane 3). These electrophoretic profiles were essentially the same as those of the LN5B purified from 32-HEK cells (Fig. 1A) and similar to the CBB staining profile of the conditioned medium from LN5B-HEK cells (Fig. 1B). Immunoblotting analysis identified the 335- and 145-kDa bands as the 3B chain, the 135-kDa band as the 3 chain, and the 150- and 105-kDa bands as the 2 chain (Fig. 4B).

View larger version (25K): [in this window] [in a new window]   FIG. 4. Electrophoretic analyses of purified LN5B and 190-kDa fragment. LN5B and the 190-kDa 3B fragment (3Bnt) purified from the conditioned medium of LN5B-HEK cells, as well as purified LN5A as a control, were separated by SDS-PAGE on 4.0-7.5% gradient gels under nonreducing conditions (-ME) or reducing conditions (+ME). A, CBB staining of separated LN5B (lanes 1 and 3), LN5A (lanes 2 and 4), and the 190-kDa fragment (3Bnt) (lane 5). Bars indicate major protein bands with their approximate molecular sizes in kDa. Subunits were identified by the immunoblotting in B. The 145-kDa 3A/B chains and the 105-kDa 2 chain are the proteolytically cleaved forms. B, immunoblotting of purified LN5B (lanes 1, 3, and 5) and LN5A (lanes 2, 4, and 6) with the antibodies against the 3 chain (lanes 1 and 2), the 3 chain (lanes 3 and 4), and the 2 chain (lanes 5 and 6). In lane 1, the relatively high ratio of the 145-kDa 3B chain to the 335-kDa form is an artifact because of the low transfer efficiency of the 335-kDa form.

To confirm the conversion of the 335-kDa 3B chain to a 145-kDa form, we purified the 190-kDa protein found in the conditioned medium of LN5B-HEK cells to a single protein as described under "Experimental Procedures" (Fig. 4A, lane 5; also see Fig. 1B, lane 4). We analyzed the NH₂-terminal amino acid sequences of the 190-kDa protein, the 335- and 145-kDa 3B bands, and the 145-kDa 3A band (Fig. 2). Both NH₂-terminal amino acid sequences of the 335- and 190-kDa proteins were determined to be ATARDPGA (amino acid numbers 34-41), indicating that the 190-kDa protein was an NH₂-terminal fragment of the 3B chain and they had been cleaved after the signal peptide sequence. Both 145-kDa 3B and 3A bands had an NH₂-terminal amino acid sequence of DSSPAEE (amino acid numbers 1812-1818), which was a common sequence between LN5A and LN5B. These results indicate that the 335-kDa 3B chain, which does not contain the COOH-terminal G4-G5 domain, is further converted to the 145-kDa form by releasing the NH₂-terminal 190-kDa fragment. This conversion is thought to be mediated by a metalloproteinase of the BMP-1 family, which selectively cleaves X-D (Asp) peptide bonds (31).

Cell Adhesive Activity of LN5B and Receptor Identification—LN5A strongly promotes adhesion of a variety of cultured cell lines (8, 9). Because the 3B chain of LN5B contains almost the whole sequence of the 3A chain in LN5A, both LN5 isoforms are expected to have similar cell adhesion activities. To confirm this, their activities were compared using the Buffalo rat liver-derived epithelial cell line BRL and the human keratinocyte cell line HaCaT. LN5B supported the attachment of BRL cells to plastic plates at significantly lower concentrations than LN5A (Fig. 5A). The half-maximal effective concentration (ED₅₀) in weight/ml for LN5B was approximately half of that for LN5A. This means that LN5B has approximately 3 times higher cell adhesion activity than LN5A, because LN5B is 1.4 times larger than LN5A. A similar difference was reproduced in the assay with HaCaT keratinocytes, which secrete LN5A endogenously (Fig. 5B).

View larger version (13K): [in this window] [in a new window]   FIG. 5. Cell adhesion activity of LN5B and LN5A. BRL cells (A) or HaCaT cells (B) were suspended in serum-free DMEM/F-12 medium and placed into each well of 96-well plates precoated with the indicated concentrations of purified LN5A (open circles) or LN5B (closed squares). After incubation for 1 h, the relative number of adherent cells was determined by fluorescent intensity. Each point represents the mean of triplicate assays. Other experimental conditions are described under "Experimental Procedures."

View larger version (28K): [in this window] [in a new window]   FIG. 6. Inhibitory effects of various integrin antibodies, heparin, and EDTA on adhesion of EJ-1 cells to LN5A or LN5B substrate. EJ-1 cells, which had been pretreated with the indicated, function-blocking integrin antibodies, heparin (Hep.), or EDTA for 20 min, were incubated for 1 h on 96-well plates precoated with 50 µl of 2 µg/ml LN5A (A) or LN5B (B). The relative number of adherent cells in the presence of mouse IgG was taken 100%. Each point represents the mean ± S.D. for triplicate assays.

View larger version (30K): [in this window] [in a new window]   FIG. 7. Effects of LN5B on cell scattering and migration. A, effects of LN5A and LN5B on morphology of BRL cells. Five hundred microliters of BRL cell suspension (1.4 x 104 cells/ml in DMEM/F-12 plus 1% FCS) was inoculated per well of 24-well plates. Purified LN5B (center) and LN5A (right) were individually added to the culture of BRL cells to make a final concentration of 0.02 µg/ml. Cell morphology was examined under a phase-contrast microscope after incubation for 40 h. None (left), cells incubated without test sample. B, effects of various concentrations of LN5A and LN5B on cell scattering. BRL cells were incubated with the indicated concentrations of purified LN5A (open circles) or LN5B (closed squares) for 40 h, and scattered cells were counted in three randomly selected microscopic fields. Each point represents the mean of the percentages of scattered cells in triplicate cultures. C, effects of LN5B, 3Bnt (the 190-kDa 3B chain fragment), and five other matrix proteins on migration of EJ-1 cells. Five hundred microliters of EJ-1 cell suspension (3 x 104 cells/ml in DMEM/F-12 plus 1% FCS) was inoculated per well of 24-well plates precoated with 10 µg/ml fibronectin (FN), 10 µg/ml laminin 1 (LN1), 10 µg/ml laminin 2/4 (LN2/4), 2 µg/ml LN5A, 1 µg/ml LN5B, 2 µg/ml laminin 10/11 (LN10/11), or 10 µg/ml 3Bnt. After preincubation for 1.5 h to allow cell attachment, the migration on each substrate was monitored by video microscopy for 8 h. Each bar represents the mean ± S.D. of the migration speeds of 8 cells.

Because LN5B-HEK cells grew much faster than the parent HEK293 cells (Fig. 3C), LN5B and two other laminins were examined for growth-stimulatory activity toward the parent HEK293 cells. When directly added to a culture medium containing 1% FCS, LN5B strongly stimulated the growth of HEK293 cells. LN5A and laminin 1 appeared to stimulate the cell growth but at far lower levels than LN5B (Fig. 8, A and B). When the LN5A-producing, nontransformed mammary epithelial cell line MCF-10A was used, LN5B also stimulated the cell growth most effectively. In contrast, laminin 1 and LN5A showed little growth stimulation toward this cell line (Fig. 8C). This suggests that soluble LN5B has a growth factor-like activity.

View larger version (22K): [in this window] [in a new window]   FIG. 8. Effects of LN5B, other laminin isoforms, and 3Bnt on HEK293 cell proliferation. A, HEK293 cells (5 x 103 cells in DMEM/F-12 plus 1% FCS) were inoculated per 35-mm dish. Each culture was added with 0.05 µg/ml (0.08 nM) LN5B (closed squares), 0.05 µg/ml (0.11 nM) LN5A (open circles), or 0.25 µg/ml (0.31 nM) laminin 1 (LN1; closed triangles), and incubated for the indicated lengths of time with medium change every 2 days. Grown cells were harvested by trypsinization at the indicated points and counted with a hemocytometer. Each point represents the mean ± S.D. of the cell numbers in triplicate cultures. Open triangles, none; cells incubated without test sample. B and C, HEK293 cells (B; 4 x 104 cells in DMEM/F-12 plus 1% FCS) and MCF-10A cells (C; 4 x 104 cells in DMEM/F-12 plus 1% horse serum) were inoculated per 35-mm dish. After incubation for 5 h, the cultures were added with the indicated protein samples and incubated for another 2 days. After the incubation, the number of cells was counted as described above. Values in parentheses indicate the concentrations of the test samples (nM). One nM is equivalent to 0.8 µg/ml for laminin 1, 0.445 µg/ml for LN5A, 0.62 µg/ml for LN5B, and 0.19 µg/ml for 3Bnt.

LN5B and LN5A, when precoated at 2 µg/ml on culture plates, supported rapid attachment and spreading of EJ-1 cells and HT1080 cells (Fig. 9A). 3Bnt supported attachment of both EJ-1 and HT1080 cell lines to the plates at concentrations higher than 5 µg/ml, although it could not support cell spreading (Fig. 9, A and B). The cell attachment activity of 3Bnt was effectively blocked by pretreating the cells with the anti-integrin ₃ or ₁ antibody (Fig. 9C). The anti-integrin ₆ antibody weakly inhibited the cell attachment. EDTA inhibited the cell attachment activity of 3Bnt, but not heparin at all. These results indicate that 3Bnt supports weak cell adhesion through the interaction with integrin ₃₁ or ₆₁. 3Bnt seems not to interact with heparan sulfate proteoglycans on the cell surface.

View larger version (81K): [in this window] [in a new window]   FIG. 9. Properties of the 190-kDa 3B NH2-terminal fragment 3Bnt. A, morphology of EJ-1 cells and HT1080 cells adhered to 3Bnt, LN5A, and LN5B. EJ-1 cells in serum-free medium were placed on plates precoated without (None) or with 15 µg/ml 3Bnt, 2 µg/ml LN5A, or 2 µg/ml LN5B. After 1 h of incubation, the cell morphology was examined under a phase-contrast microscope. Original magnification, x300. B, effect of various concentrations of 3Bnt on attachment of EJ-1 cells. EJ-1 cells were incubated on plates precoated with the indicated concentrations of 3Bnt for 1 h. Other experimental conditions are described in Fig. 5 and under "Results." C, inhibitory effect of various integrin antibodies, heparin, and EDTA on attachment of EJ-1 cells to 3Bnt. The experiment was done as described for Fig. 6 using plates pretreated with 50 µl of 15 µg/ml 3Bnt. The relative number of adherent cells in the presence of mouse IgG was taken as 100%. Each point represents the mean ± S.D. for triplicate assays.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
In the present study, we for the first time cloned a full-length human laminin 3B cDNA and characterized human LN5B consisting of the 3B, 3, and 2 chains. The molecular size of the mature 3B chain (without G4-G5 domain) was 335 kDa, approximately twice as large as the mature 3A chain (160 kDa) (Fig. 2). Comparison of the biological activities of the two LN5 isoforms demonstrated that LN5B had significantly higher cell adhesion and cell migration activities than LN5A. In addition, LN5B potently stimulated cell proliferation when added to the culture medium directly. It has been generally accepted that integrin-mediated signals cooperate with growth factor signals (33). Indeed, there are some reports showing that laminin-derived signals support cell proliferation (34, 35). However, there is little information about growth stimulation by soluble laminin. In the present study, LN5B stimulated cell growth as well as cell migration in soluble conditions. Recently, we found that LN5A, but not other laminins, stimulates cell migration as a soluble factor.² The association of soluble LN5A with integrin ₃₁ on the apical cell surface induces protein kinase C-mediated growth signal. Such growth factor-like activities of LN5B and LN5A suggest their contribution to the cellular growth and migration in some physiological and pathological conditions. This was supported by the findings that LN5B-HEK cells showed highly motile cell morphology and a very high growth rate as compared with the parent HEK cells.

Laminins are large complex molecules, which have many functional domains in their three subunits (1, 2). They interact with cell surface receptors such as integrins, syndecans, and dystroglycan and matrix proteins such as nidogen, collagens, and perlecan through different domains. It has been generally accepted that the NH₂-terminal regions of the three subunits are essential for the assembly to the matrix, whereas the COOH-terminal G domain of chains mediates the interaction with cell surface receptors including integrins (1). In LN5A, the G1-G3 domain of the 3A chain, especially G3, is critical for the expression of the biological activity (6, 26, 36). The COOH-terminal G4-G5 fragment of the 3A chain, which is proteolytically released from the immature LN5A after secretion (29), supports weak cell adhesion by interacting with cell surface syndecans (37). Thus, the G (G1-G5) domain of 3 chain plays a primary role in the interaction with cell surface receptors and the resulting regulation of cellular functions. In some full-sized laminin chains, however, their NH₂-terminal regions are known to contain binding sites to integrins and heparan sulfate proteoglycans. The recombinant domain VI-V proteins of the 1 chain (38-40) and the 2 chain (41) mediate cell adhesion through the interaction with integrins ₁₁ and ₂₁. Both integrin-binding site and heparin-binding site are localized in domain VI but not domain V. A recent study has shown that the domain VI-V of mouse laminin 5 chain exhibits a cell binding activity, which is mainly mediated by integrin ₃₁ and heparan sulfate proteoglycans on the cell surface (42). This study also identified two amino acid sequences in domain VI, which are responsible for the cell binding. These sequences are highly conserved in the domain VI of human laminin 3B chain. In the present study, we found that a 190-kDa NH₂-terminal fragment of the 3B chain, named 3Bnt, was released from LN5B presumably by the action of BMP-1-related metalloproteinases. Consistent with the studies about the recombinant domain VI-V of other laminin chains, 3Bnt had activities to promote cell adhesion through the interaction with integrin ₃₁. Unlike other chains, however, the cell adhesion activity of 3Bnt was not inhibited by heparin, suggesting that cell surface heparan sulfate proteoglycans were not involved in the cell adhesion. In addition, 3Bnt stimulated cell migration and proliferation. These activities have not been reported in any NH₂-terminal recombinant proteins of other chains. It is interesting that 3Bnt promotes cell migration without spreading cells. These activities of the NH₂-terminal structure of the 3B chain may contribute to the elevated biological activity of LN5B compared with LN5A, which lacks most part of the sequence of 3Bnt. Although the cell adhesion activity of 3Bnt was approximately one-fiftieth of those of LN5A and LN5B, its activity was close to that of laminin 1. Furthermore, the growth-stimulatory activity of 3Bnt toward HEK cells was rather higher than or comparable with those of laminin 1 and LN5A. It is also noted that the activity of 3Bnt was similar to or higher than the reported activities of the recombinant NH₂-terminal domains of mouse laminin 1, 2, and 5 chains (40-42). Taken together, it is expected that the biological activity of 3Bnt through the interaction with integrin ₃₁ has some biological significance. It seems possible that the NH₂-terminal structure of the 3B chain cooperates with its COOH-terminal structure to regulate cellular functions. In vivo, the released 3Bnt may cooperate with laminins and other matrix proteins to regulate cellular adhesion and migration. On the other hand, the proteolytic processing of the 3B chain converts LN5B to the LN5A with the 145-kDa 3 chain. The functional conversion of LN5B seems to have some physiological significance.

Laminin polymerization is an important process to produce the basement membrane scaffold (1). The NH₂-terminal domains of the full-sized chains, as well as those of laminin 1, 2, and 1 chains, are involved in the self-assembly and co-assembly of laminins (43, 44). It is known that the genetic deletion of the NH₂-terminal domain in the 2 chain causes muscular dystrophy (45). Because LN5A lacks the NH₂-terminal domains in the three subunits, it is assembled into the basement membrane scaffold through making a complex with laminin 6/7 (3, 1/2, 1) (14, 46). A recent report has shown that a recombinant NH₂-terminal fragment (domain VI/V) of mouse laminin 3B chain has a strong activity to bind to domain VI/V of laminin 1 and 1 chains and a moderate self-binding activity (22). This suggests that LN5B, unlike LN5A, may be assembled into laminin networks. The matrix assembly of LN5B may be very different from that of the full-sized laminins (1/2/5-laminins), because LN5B has a very long rodlike structure with the NH₂-terminally truncated forms of / chains (3 and 2). The NH₂-terminal domain VI can associate not only with other laminins and presumably other matrix proteins but also with integrins. Although we found no significant difference in the efficiency of matrix deposition of LN5B and LN5A, their deposition patterns were clearly different from each other as analyzed by immunofluorescent microscopy. This suggests that LN5B and LN5A interact with different extracellular matrix molecules and have different effects on the producing cells. It is conceivable that in vivo LN5B links epithelial cells on the basement membrane with the underlying stromal cells by binding the former cells through the COOH-terminal G domain of the 3B chain and the latter through the NH₂-terminal domain VI. Further studies are needed to clarify the structural and functional roles of LN5B in the basement membranes. It seems also important to characterize two unidentified full-sized laminins, laminin 6B/7B (3B11 and 3B21), that are expected to form non-covalent networks by self-polymerization and co-polymerization with other full-sized laminins.

It has been reported that the laminin 3B chain is more widely and more intensively expressed than the 3A chain. For example, in human tissues 3B chain is exclusively expressed in the uterus, lung, liver, brain, and heart, whereas 3A is predominantly expressed in the placenta and salivary gland (24). In mouse skin, 3B and 3A display different distribution (20, 22). On the other hand, human cancer cell lines almost exclusively express the 3A transcript (28). The differential expression of 3B and 3A is regulated by distinct two promoters (21). Acute phase reactant consensus sequences and NFIL-6 binding sites are present in the two 3B/3A promoters. The present study demonstrates that LN5B has higher activities to stimulate cell migration and cell growth than LN5A and other laminins. Together with the past studies, our results suggest that LN5B may be involved in some pathological processes such as inflammation and wound healing. The differential expression and biological activity between LN5A and LN5B demonstrate that they play differential physiological roles.

In conclusion, this study provides new information on the unique laminin isoform LN5B. The recombinant LN5B and the 190-kDa NH₂-terminal fragment 3Bnt produced in this study seem useful to elucidate their physiological and pathological roles.

	FOOTNOTES

 
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank^TM/EBI Data Bank with accession number(s) AB107369 [GenBank] .

^* This work was supported in part by grants-in-aid from the Ministry of Education, Culture, Sports, Science and Technology and from the Ministry of Welfare and Labor of Japan. 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: Division of Cell Biology, Kihara Inst. for Biological Research, Yokohama City University, 641-12 Maioka-cho, Totsuka-ku, Yokohama 244-0813, Japan. Tel.: 81-45-820-1905; Fax: 81-45-820-1901; E-mail: miyazaki{at}yokohama-cu.ac.jp.

¹ The abbreviations used are: LN5, laminin 5; LN5A, laminin 5A; LN5B, laminin 5B; BSA, bovine serum albumin; CBB, Coomassie Brilliant Blue; FCS, fetal calf serum; DMEM, Dulbecco's modified Eagle's medium; PBS, phosphate-buffered saline; EGF, epidermal growth factor; HEK, human embryonic kidney; CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid.

² Kariya, Y., and Miyazaki, K. (2004) Exp. Cell Res., in press.

	ACKNOWLEDGMENTS

 
We thank Drs. H. Yasumitsu, S. Higashi, T. Hirosaki, and T. Ogawa for helpful suggestions and discussions.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS<