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

J. Biol. Chem., Vol. 275, Issue 31, 23589-23595, August 4, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/31/23589    most recent M003526200v1 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 Takahashi, H. Articles by Saito, Y. Search for Related Content PubMed PubMed Citation Articles by Takahashi, H. Articles by Saito, Y. Social Bookmarking                      What's this?

Tissue Transglutaminase, Coagulation Factor XIII, and the Pro-polypeptide of von Willebrand Factor Are All Ligands for the Integrins ₉₁ and ₄₁^*

Hiroo Takahashi^§, Takashi Isobe^§, Shiho Horibe, Junichi Takagi, Yasuyuki Yokosaki^¶, Dean Sheppard, and Yuji Saito^**

From the  Department of Biological Sciences, Graduate School of Bioscience and Biotechnology, Tokyo Institute of Technology, 4259 Nagatsuta-cho, Midori-ku, Yokohama 226-8501, Japan, the ^¶ Departments of Internal Medicine and Laboratory Medicine, National Hiroshima Hospital, 513 Jike, Saijoh, Higashi-Hiroshima 739-0041, Japan, and the  Lung Biology Center, Department of Medicine, University of California, San Francisco, California 94143-0854

Received for publication, May 16, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

We previously reported that MOLT-3 human lymphocyte-like leukemia cells adhere to tissue-type transglutaminase (tTG) through the integrin ₄₁. We now report that G-361 human melanoma cells also adhere to tTG, although they do not express ₄₁. G-361 cells utilize two additional integrins, ₉₁ and ₅₁ to adhere to tTG. Furthermore, blood coagulation factor XIII (FXIII), another member of the transglutaminase family that is highly homologous to tTG, and propolypeptide of von Willebrand factor (pp-vWF) also promoted cell adhesion through ₉₁ or ₄₁ in G-361 or MOLT-3 cells, respectively. In the case of pp-vWF, ₉₁ and ₄₁ both bind to the same site, comprised of 15 amino acid residues and designated T2-15. Moreover, SW480 human colon cancer cells stably transfected to express ₉₁, but not mock transfectants, adhered to tTG, FXIII, pp-vWF, and T2-15/bovine serum albumin conjugate. These data identify tTG, FXIII, and pp-vWF as shared ligands for the integrins ₉₁ and ₄₁. This report is the first to unambiguously show that these two integrins share the same cell adhesion site within one protein and provides strong support for classifying _91- and ₄-integrins as functionally related members of an integrin subfamily.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Integrins are a family of heterodimeric transmembrane receptors that mediate cell-extracellular matrix and cell-cell interactions and play important roles in a wide variety of cellular events (1-6). Each integrin is composed of noncovalently associated  and  subunits, and the combination of  and  subunits generates many different receptors with different ligand specificity.

Integrin  subunits can be grouped into subfamilies based on sequence similarity, and these subfamilies generally define integrins that share common ligands. Thus,  subunits can be divided into five groups (6-8): the first group (₁, ₂, ₁₀, and ₁₁) recognizes collagen, the second group (₃, ₆, and ₇) recognizes laminin, the third group (₅, ₈, _v, and _IIb) recognizes RGD-containing sequences, and the fourth group (_L, _M, _X, and _D) recognizes ICAM-1. ₄ and ₉ are the only members of the fifth group (9). Somewhat surprisingly, the initial ligands identified for ₉ and ₄-containing integrins did not appear to overlap. Thus, for example, the integrin ₄₁ was found to recognize fibronectin (10, 11) and the vascular cell adhesion molecule-1 (VCAM-1)¹ as ligands (12), whereas the integrin ₉₁ was reported to recognize tenascin-C, (13, 14), and osteopontin (15, 16). However, recently Bayless et al. (17) reported that ₄₁ recognizes osteopontin as a ligand, and Taooka et al. (18) reported that ₉₁ recognizes VCAM-1. It thus appears that the _41- and ₉₁-integrins, like other integrins that are related based on  subunit sequence homology, do share at least some common ligands.

In the present study, we demonstrate that three additional proteins are ligands for both ₉₁ and ₄₁. The first is a tissue-type transglutaminase (tTG). This protein belongs to a family of transglutaminases (EC 2.3.2.13) that catalyze -(-glutamyl)lysine cross-link formation between specific substrate proteins (19-21) and are distributed widely in various tissues. The second is blood coagulation factor XIII (FXIII). This protein is also a member of the transglutaminase family and has an important role in the final stage of the blood coagulation cascade. The last is the propolypeptide of von Willebrand factor (pp-vWF). This protein is obtained from a large precursor of von Willebrand factor by specific cleavage during biosynthesis and is stored in granules of both endothelial cells and platelets (22, 23). In the case of pp-vWF, we have mapped the recognition sequence for both integrins to the same cell adhesion site, a 15-residue linear sequence that we have previously shown is required for ₄₁-mediated adhesion to this protein (24).

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- tTG was purified from guinea pig liver using an anti-tTG monoclonal antibody (mAb) 8D, which was a gift from Dr. K. Ikura (Kyoto Institute of Technology, Kyoto, Japan), raised against guinea pig liver tTG as described (25). Human placental FXIII was kindly provided from Hoechst Japan Co. (Tokyo, Japan) and was further purified as described (26). pp-vWF was purified from washed bovine platelets by immunoaffinity chromatography, as described (27). Porcine vitronectin was a gift from Dr. M. Hayashi (Ochanomizu University, Tokyo, Japan). Mouse laminin, RGD peptide (GRGDSP), and RGE peptide (GRGESP) were purchased from Iwaki Glass (Tokyo, Japan).

Antibodies-- Mouse mAb 4B4, which recognizes human ₁-integrin, was a gift from Dr. C. Morimoto (Dana Farber Cancer Institute, Boston, MA). The hybridoma cell line secreting mAb TS2/16, which recognizes and activates human ₁-integrin (28), was obtained from the American Type Culture Collection. This mAb was obtained in ascites and purified by ammonium sulfate precipitation and anion exchange chromatography. Mouse anti-human 2 subunit mAb 6F1 was a gift from Dr. B. S. Coller (Mount Sinai School of Medicine, New York, NY). Mouse anti-human ₃ subunit mAb, P1B5, mouse anti-human ₄ subunit mAb HP2/1, and mouse anti-human ₅ subunit mAb, P1D6, were purchased from Telios Pharmaceuticals, Inc. (La Jolla, CA), Cosmo Bio Co., LTD. (Tokyo, Japan), and Life Technologies, Inc. (Tokyo, Japan), respectively. Rat anti-human and -mouse ₆ subunit mAb, GoH3, was a gift from Dr. A. Sonnenberg (Central Laboratory of the Netherlands Red Cross Blood Transfusion Service, Amsterdam, The Netherlands). Mouse anti-human ₉ subunit mAb, Y9A2, was generated and characterized as described previously (29). Polyclonal antibody against human vitronectin receptor (_v3) was purchased from Telios Pharmaceuticals Inc.

Cell Lines and Cell Culture-- G-361 human melanoma cells and MOLT-3 human lymphoma cells were provided from the Japanese Cancer Research Resources Bank (Tokyo, Japan) and cultured in RPMI 1640 medium containing 10% fetal calf serum and nonessential amino acids. ₉-transfected or mock transfected SW480 human colon cancer cells were previously described in Ref. 13. These cells were cultured in Dulbecco's modified Eagle's medium containing 10% fetal calf serum and 1 mg/ml neomycin analog, G418 (Life Technologies, Inc.).

Peptide Synthesis-- Peptides of CS-1 (EILDVPST), located in an alternatively spliced segment of fibronectin, and TNfnIII3 (AEIDGIELTYG) located in the third fibronectin type III repeat of tenascin-C, were manually synthesized by the Fmoc (N-(9-fluorenyl)methoxycarbonyl)-based solid phase method. T2-15 peptide (DCQDHSFSIVIETVQ) was synthesized as described previously (24). Bovine serum albumin (BSA) conjugated with CS-1 peptide, TNfnIII3 peptide, or T2-15 peptide (CS-1/BSA, TNfnIII3/BSA, or T2-15/BSA, respectively) was prepared in our laboratory.

Cell Adhesion Assay-- Cells were washed three times and suspended in serum-free RPMI 1640 or Dulbecco's modified Eagle's medium containing 0.25 mM MnCl₂ prior to the cell adhesion assay. In the case of MOLT-3 cells, in addition to 0.25 mM MnCl₂, 5 µg/ml TS2/16, a ₁-integrin activating mAb, was added to the cell suspension. Cell adhesion assays were performed according to the method described previously (30). In brief, 6-mm square chips cut from bacteriologic plastic dishes (Falcon 1029) were coated for 16 h at 4 °C with 50 µl of solution containing adhesive proteins at the following concentrations diluted in 10 mM Tris-HCl (pH 7.4) and 150 mM NaCl. Laminin and vitronectin were 5 µg/ml; tTG and pp-vWF were 10 µg/ml; TNfnIII3/BSA, CS-1/BSA, and T2-15/BSA were 20 µg/ml; FXIII was 30 µg/ml. More than 80% of ligands were absorbed on the surface under these conditions. After blocking nonspecific protein binding sites by incubation with 1% BSA at room temperature for 1 h, chips were placed in 48-well tissue culture dishes (Costar 3548) and overlaid with 2-5 × 10⁴ cells in 50 µl of serum-free medium. After incubation at 37 °C for 90 min, chips were picked up, rinsed in phosphate-buffered saline to remove nonadherent cells, and fixed with 1% glutaraldehyde in phosphate-buffered saline. Cells on chips were stained, when necessary, with 0.5% crystal violet in 20% methanol. Adherent cells were either photographed or counted using a light microscope with a calibrated grid marked on the ocular lens. When inhibitory antibodies were used in adhesion assays, cells were first incubated with antibody at 37 °C for 30 min. Concentrations of synthetic peptides and inhibitory antibodies used in this study, except Y9A2, were 1 mM and 10 µg/ml, respectively. Y9A2 was used as a 1:50 dilution of hybridoma supernatant.

Flow Cytometry-- Flow cytometry was performed according to the method described previously (31). Cells were washed as described above and treated with an appropriate primary antibody in serum-free medium for 30 min on ice. After removal of unbound antibody, cells were incubated with 1:100 diluted fluorescein isothiocyanate-conjugated anti-mouse IgG for 30 min on ice, washed, and analyzed on a FACScan flow cytometer (Beckton Dickinson and Co.).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

G-361 Human Melanoma Cells Adhere to tTG through ₁-Integrin(s) but Not ₄₁-- In a previous study (31), we reported that tTG promoted cell adhesion of MOLT-3 cells mediated by the integrin ₄₁. To determine whether adhesion of all cells to tTG is dependent on ₄₁, we performed adhesion assays with a variety of cell lines in the presence and absence of anti-₄ blocking antibody. Although adhesion of most cells was inhibited by an anti-₄ mAb, there was one notable exception. Adhesion of G-361 human melanoma cells to tTG was not inhibited at all by this mAb (Fig. 1). Furthermore, this adhesion was not inhibited by CS-1 peptide, which inhibits ₄₁-mediated adhesion. However, G-361 cell adhesion to tTG was completely inhibited by anti-₁ antibody. These results suggest that one or more ₁-containing integrin, other than ₄₁, is responsible for G-361-mediated adhesion to tTG.

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Adhesion of G-361 cells to tTG is not mediated by 41-integrin. Cell adhesion was determined in the presence of no mAb (control), an anti-1 mAb 4B4 (1), an anti-4 mAb HP2/1 (4), or CS-1 peptide (CS-1). The data represent the means ± S.E. of one of the representative experiments in which triplicate determinations were made.

Expression of ₉₁ and ₅ Subunit on G-361 Cells-- The ₁-integrin family includes at least 12 heterodimers composed of ₁ paired with each of 12 distinct  subunits (₁-₁₁ and _v). Akimov et al. (32) have recently demonstrated that the integrins ₅₁ and _IIb3 can mediate adhesion to tTG. In addition, Palmer et al. (9) reported that the ₉ subunit had high homology with ₄. Yokosaki et al. (14) demonstrated that the ligand binding sequence of tenascin-C, which is recognized by ₉₁-integrin, is homologous to the ₄₁ binding sequence in VCAM-1, and Taooka et al. (18) have shown that ₉₁ recognizes VCAM-1 as a ligand. _IIb3 is not generally expressed in cells other than platelets, but ₅₁ and ₉₁ are more widely expressed. We therefore explored the possibility that G-361 cells express ₉₁ and/or ₅₁ and that one or more of these integrins mediates adhesion to tTG. As depicted in Fig. 2, G-361 cells expressed ₉₁and a low level of ₅ but minimal ₄. In contrast, MOLT-3 cells, which adhere to tTG through ₄₁ (31), expressed a high level of ₄ but not ₉₁,

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Flow cytometry of integrins expressed on G-361 cells and MOLT-3 cells. Cells were reacted with no IgG (control), an anti-human 1 mAb TS2/16 (1 subunit), an anti-human 9 mAb Y9A2 (9 subunit), an anti-human 4 mAb HP2/1 (4 subunit), or an anti-human 5 mAb P1D6 (5 subunit).

Adhesion of G-361 Cells to tTG Is Mediated by ₉₁- and ₅₁-Integrins-- To determine whether ₉₁ is involved in adhesion of G-361 cells to tTG, we investigated the effect of anti-₉₁ blocking mAb, Y9A2, on cell adhesion to tTG. TNfnIII3 peptide conjugated with BSA (TNfnIII3/BSA) was used as a positive control ligand for ₉₁ (14). As shown in Fig. 3A, adhesion to the conjugate was completely inhibited by this mAb, as was expected. Although adhesion of G-361 cells to tTG was inhibited only modestly by the anti-₉-integrin mAb, the antibody dramatically inhibited cell spreading (Fig. 3B). As expected, anti-4 mAb had no effect on either adhesion or spreading of these cells. These findings suggested that G-361 cell adhesion to tTG is at least partly mediated by ₉₁. However, because anti-₉₁ only partially inhibited adhesion, whereas anti-₁ antibody completely inhibited adhesion, our results suggest a role for other ₁-containing integrins (e.g. ₅₁) in this process. To determine which other integrins were involved, we examined the combined effects of anti-₉₁ and other anti-integrin antibodies and peptides in cell adhesion to tTG. As depicted in Fig. 4, anti-₅ mAb or RGD peptide dramatically augmented the inhibitory effect of the anti-₉₁ mAb on adhesion of G-361 cells to tTG, whereas antibodies against ₂, ₃, ₄, or ₆ were without effect. Taken together, these results indicate that both ₉₁ and ₅₁ mediate adhesion of G-361 cells to tTG. These results, together with our previous report that ₄₁ is a receptor for tTG on MOLT-3 cells (31), suggest that tTG is an additional shared ligand for ₉₁ and ₄₁.

View larger version (69K): [in this window] [in a new window]   Fig. 3.   The effect of anti-9 subunit mAb on G-361 cell adhesion to tTG. A, adhesion of G-361 cells to TNfnIII3/BSA or tTG was determined in the presence of no mAb (black columns, c), an anti-4 mAb HP2/1 (gray columns, 4), or an anti-9 mAb Y9A2 (hatched columns, 9). The data represent the means ± S.E. of one of the representative experiments in which triplicate determinations were made. TNfnIII3/BSA, TNfnIII3 peptide-BSA conjugate. B, photographs show morphology of G-361 cells adhering to tTG in the presence of mAbs.

View larger version (38K): [in this window] [in a new window]   Fig. 4.   Identification of integrins involved in cell adhesion of G-361 to tTG. The effect of mAbs and peptides on adhesion of G-361 cells to tTG was determined in the absence (black columns, ) or presence (gray columns, +) of anti-9 mAb Y9A2. The following mAbs or peptides were used: control, without mAb; 2, an anti-2 mAb 6F1; 3, an anti-3 mAb P1B5; 4, an anti-4 mAb HP2/1; 5, an anti-5 mAb P1D6; 6, an anti-6 mAb GoH3; RGE, RGE peptide (GRGESP); RGD, RGD peptide (GRGDSP). The data represent the means ± S.E. of one of the representative experiments in which triplicate determinations were made.

Adhesion of G-361 Cells to FXIII Is Mediated by ₉₁-Integrin-- In a previous study (33), we reported that the active form of FXIII promoted cell adhesion of G-361 cells. We neither analyzed whether the activation of FXIII was necessary for the adhesion of G-361 cells nor determined the kind of integrins involved. In the present study, we show that the nonactivated form of FXIII also promotes G-361 cell adhesion. Both tTG and FXIII belong to the transglutaminase family and are highly homologous to each other. However, the recent study by Akimov et al. (32) reported that FXIII, in contrast to tTG, did not interact with ₅₁ (or any other integrins expressed on WI-38 human fibroblasts or human erythroleukemia cells, HEL). We speculated that the adhesion of G-361 cells to FXIII might be mediated by ₉₁, because this integrin is not expressed on WI-38 cells or HEL.² To confirm this, we investigated the effect of anti-₉₁ mAb on adhesion of G-361 cells to FXIII. As shown in Fig. 5A, adhesion of G-361 cells to FXIII was almost completely inhibited either by mAb against ₁ or ₉₁, and the few cells that did adhere did not spread (Fig. 5B). The addition of mAbs against ₅- and ₄-integrins did not have any effect on adhesion or spreading. This result strongly indicates that the integrin, ₉₁, is the receptor for FXIII on G-361 cells.

View larger version (43K): [in this window] [in a new window]   Fig. 5.   Identification of integrins involved in cell adhesion of G-361 to FXIII. A, adhesion of G-361 cells to FXIII was determined in the presence of no mAb (control), an anti-1 mAb 4B4 (1), an anti-9 mAb Y9A2 (9), an anti-5 mAb P1D6 (5), or an anti-4 mAb HP2/1 (4). The data represent the means ± S.E. of at least two independent experiments in which triplicate determinations were made. B, photographs show morphology of G-361 cells adhering to FXIII in the presence of mAbs against various  subunits of integrin.

Adhesion of MOLT-3 Human Lymphocyte-like Leukemia Cells to FXIII Is Mediated by ₄₁-Integrin and Not by ₉₁-Integrin-- We have previously reported that MOLT-3 human lymphocyte-like leukemia cells did not adhere to FXIII (33). But at that time, we made no effort to activate integrins during the cell adhesion assays. Masumoto and Hemler (34) described that ₄₁-mediated adhesion of leukocyte cell lines required activation of the integrin by divalent cations, stimulatory mAbs, or both. In fact, we previously showed that MOLT-3 cells adhered to tTG only when integrins were activated (31). We therefore reexamined adhesion of MOLT-3 cells to FXIII under conditions where integrins would be activated. As shown in Fig. 6A, integrin activation was needed for MOLT-3 cells to adhere to CS-1 peptide conjugated with BSA (CS-1/BSA). This adhesion was inhibited by anti-₄-integrin mAb, as expected. As shown in Fig. 6B, MOLT-3 cell adhesion to FXIII also required activation, and this adhesion was completely inhibited by the mAb against ₄. These results indicate that MOLT-3 cells adhere to FXIII through ₄₁. Taken together, these results indicate that FXIII, as well as tTG, serves as a ligand for both ₉₁- and ₄₁-integrins.

View larger version (18K): [in this window] [in a new window]   Fig. 6.   41-integrin-dependent adhesion of MOLT-3 cells to FXIII. Adhesion of MOLT-3 cells to CS-1/BSA (A) or FXIII (B) was determined in the presence of following mAbs: control, without mAb; 4, an anti-4 mAb HP2/1; 9, an anti-9 mAb Y9A2. Assay was carried out in the absence (striped columns, ) or presence (black columns, +) of MnCl2 and TS2/16 for stimulating integrins. The data represent the means ± S.E. of one of the representative experiments in which triplicate determinations were made.

Adhesion of G-361 Cells to pp-vWF Is Also Mediated by ₉₁-- Previously, we reported that pp-vWF also mediated cell adhesion through ₄₁ (24). We also reported that G-361 cells adhered to pp-vWF (30) and that this adhesion was not inhibited at all by an anti-₄ mAb, but we were unable to define the responsible integrin(s). By analogy with tTG and FXIII, we have suspected that this adhesion might be mediated by ₉₁. As shown in Fig. 7, adhesion of G-361 cells to pp-vWF was completely inhibited by anti-₉₁ mAb. As expected, adhesion to TNfnIII3/BSA was also completely inhibited by this mAb. Anti-₉₁, however, did not inhibit adhesion of the cells to laminin, which is mainly mediated by ₆₁-integrin. This result indicates that the integrin ₉₁ is the receptor G-361 cells use for adhesion to pp-vWF. Thus, pp-vWF is also a ligand for both ₉₁- and ₄₁-integrins.

View larger version (48K): [in this window] [in a new window]   Fig. 7.   Adhesion of G-361 cells to pp-vWF is mediated by 91-integrin. Adhesion of G-361 cells to laminin, TNfnIII3/BSA, or pp-vWF was determined in the presence of no mAb (black columns, c), an anti-9 mAb Y9A2 (striped columns, 9), or an anti-4 mAb HP2/1 (gray columns, 4). The data represent the means ± S.E. of at least two independent experiments in which triplicate determinations were made.

₉₁- and ₄₁-Integrins Share the Same Cell Adhesion Site in pp-vWF-- In a previous study (24), we demonstrated by experiments using synthetic peptides that an amino acid sequence within pp-vWF, DCQDHSFSIVIETVQ (designated as T2-15) was responsible for ₄₁-mediated adhesion of MOLT-3 cells to this protein. To determine whether ₄₁ and ₉₁ might share the same recognition sequence in pp-vWF, we performed cell adhesion assays in the presence of T2-15 or the irrelevant peptide, GRGDSP. As shown in Fig. 8A, adhesion of G-361 cells to pp-vWF was completely inhibited by the addition of T2-15 peptide, but the RGD peptide had no effect. On the other hand, adhesion to vitronectin, which is known to be mediated by RGD-dependent integrins, was inhibited by the RGD peptide, but not by T2-15 peptide. Furthermore, as depicted in Fig. 8B, G-361 cells adhered to T2-15 peptide conjugated with BSA (T2-15/BSA), and this adhesion was also completely inhibited by anti-₉₁ mAb. This complete inhibition was also observed when cells adhered to intact pp-vWF. These results strongly suggest that ₉₁ and ₄₁-integrins share the same cell adhesion sequence within pp-vWF.

View larger version (36K): [in this window] [in a new window]   Fig. 8.   Identification of the essential adhesion site of pp-vWF recognized by 91-integrin. A, adhesion of G-361 cells to pp-vWF or vitronectin was determined in the absence (black columns, 1) or presence of T2-15 peptide (white columns, 2), RGD peptide (striped columns, 3), or RGE peptide (gray columns, 4). B, adhesion of G-361 cells to pp-vWF or T2-15/BSA was determined in the presence of no mAb (black columns, c), an anti-9 mAb Y9A2 (striped columns, 9), or an anti-4 mAb HP2/1 (gray columns, 4). The data are the means ± S.E. of one of the representative experiments in which triplicate determinations were made. T2-15/BSA, T2-15 peptide-BSA conjugate.

₉ Subunit-transfected SW480 Human Colon Cancer Cells Adhere to tTG, FXIII, and pp-vWF-- Using a single cell system, G-361 cells, we have defined tTG, FXIII, and pp-vWF as ligands for the integrin, ₉₁. To determine whether other cells could also use this integrin for adhesion to each of these ligands, we employed SW480 colon carcinoma cells that were stably transfected with either an 9-expression plasmid or empty vector (₉-transfected SW480 cells and mock transfected SW480 cells, respectively). We determined by flow cytometry that only ₉-transfected SW480 cells expressed ₉₁, that neither SW480 cell line expressed ₄, and that both expressed similar levels of ₅ and ₁ (Fig. 9A). As shown in Fig. 9B, only ₉-transfected SW480 cells adhered to TNfnIII3/BSA, FXIII, pp-vWF, and T2-15/BSA. In each case, adhesion was inhibited almost completely by anti-₉₁ mAb but not by anti-₄ mAb. Mock transfected SW480 cells and ₉-transfected SW480 cells adhered equally well to laminin, indicating that all these cells were equally viable and that heterologous expression of ₉ did not nonspecifically augment cell adhesion. In the case of tTG, ₉-transfected SW480 cells adhered to the ligand, but mock transfected SW480 cells also adhered. The inhibitory effect of the anti-₉-integrin mAb on adhesion of ₉-transfected SW480 cells to tTG was not complete. In fact, the extent of the remaining adhesion activity was similar to that observed with mock transfected SW480 cells. Complete inhibition of adhesion of ₉-transfected SW480 cells to tTG was obtained by the simultaneous addition of anti-₉- and anti-₅-integrin mAb. The mock transfected SW480 cells did not adhere to tTG in the presence of the anti-₅-integrin mAb. The most plausible explanation would be that ₅₁-integrin is also involved in cell adhesion to tTG, as was depicted in Figs. 3 and 4 using G-361 cells. These results indicate unambiguously that tTG, FXIII, and pp-vWF are adhesive ligands recognized by both ₉₁- and ₄₁-integrins, and that T2-15 within pp-vWF is the cell adhesion site that is recognized by ₉₁ and ₄₁.

View larger version (28K): [in this window] [in a new window]   Fig. 9.   Adhesion of 9-transfected SW480 cells to tTG, FXIII, and pp-vWF. A, flow cytometry of integrins expressed on mock transfected and 9-transfected SW480 cells. Cells were reacted with no IgG (control), an anti-human 1 mAb TS2/16 (1 subunit), an anti-human 4 mAb HP2/1 (4 subunit), anti-human 5 mAb P1D6 (5 subunit), or an anti-human 9 mAb Y9A2 (9 subunit). B, mock transfected SW480 cells (white columns, 1) or 9-transfected SW480 cells were subjected to adhesion assays using various ligands. Adhesion of 9-transfected SW480 cells to these ligands was assessed in the presence of no mAb (black columns, 2), an anti-4 mAb HP2/1 (gray columns, 3), or an anti-9 mAb Y9A2 (hatched columns, 4). In the case of tTG, mock transfected cells were also treated with the anti-5 mAb (column 1a) and 9-transfected cells were also treated with anti-9 and anti-5 mAbs simultaneously (column 4a). The data represent the means ± S.E. of one of the representative experiments in which triplicate determinations were made.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Previous reports identified three ligands for the integrin ₉₁, VCAM-1 (18), tenascin-C (13, 14), and osteopontin (15, 16). In this report, we describe three additional ligands, tTG, FXIII, and pp-vWF. Among the previously identified ₉₁ ligands, VCAM-1 and osteopontin have both been reported to be recognized by both ₉₁ and the structurally related integrin ₄₁ (17, 35). In this report, we show that each of the three new ₉₁ ligands we identified is also a ligand for ₄₁.

Although it might seem reasonable to assume that ₉₁ and ₄₁ recognize the same adhesive sites in shared ligands, this is not necessarily the case. There are numerous examples of different integrins recognizing distinct sites in the same ligand. For example, at least five integrins can recognize the third fibronectin type III repeat in tenascin-C as a ligand, but four of them recognize the RGD site in the F-G loop, whereas ₉₁ clearly recognizes a different sequence, AEIDGIEL, in the B-C loop present on the same face of the protein (14). Similarly, both the integrins _v3 and ₉₁ recognize a thrombin-cleaved N-terminal fragment of osteopontin as a ligand (15), but _v3 binds to an RGD site, whereas ₉₁ binds to the adjacent sequence, SVVYGLR (16). In fact, prior to the present report, there was no definitive evidence that ₉₁ and ₄₁ could recognize the same site in a shared ligand. The evidence presented in this report that both integrins recognize the same linear peptide, T2-15, within pp-vWF is to our knowledge the first conclusive evidence that these integrins can recognize the same cell adhesion site within one molecule (Table I). Interestingly, the sequence of this peptide, DCQDHSFSIVIETVQ, bears little resemblance to the previously identified ₉₁-recognition sequences in tenascin-C and osteopontin. This observation provides additional evidence that ₉₁ can recognize a surprisingly broad array of adhesive ligands. Taken together, the accumulated evidence that ₉₁ and ₄₁ share five ligands and, at least in one instance, recognize the same linear peptide within a ligand confirms the utility of  subunit sequence comparisons for predicting ligand binding specificity of integrin heterodimers.

View this table: [in this window] [in a new window]   Table I Various adhesion ligands that are recognized by 91 and/or 41 integrin Adhesion to various ligands is illustrated according to the results obtained by ourselves and other investigators. , recognition of the integrin; ×, no recognition of the integrin; ?, no determination. The proposed essential cell adhesion sequences are shown in parentheses.

The interaction between tTG and integrins is likely to be biologically significant. tTG directly binds to a number of components of the extracellular matrix, including osteopontin and tenascin-C, where it plays an important role in matrix protein cross-linking. A recent report demonstrated that tTG bound to the integrins ₅₁ and _IIb3 within the secretory apparatus prior to appearance of the integrins on the cell surface, thereby suggesting a mechanism for localization of tTG to its extracellular targets (32). Furthermore, the same study showed that the interaction of integrins with tTG enhanced cell attachment and spreading on fibronectin. The findings in the present study greatly expand the complexity of these potential interactions and suggest that similar mechanisms could be involved in cellular interactions with additional sites in fibronectin (e.g. the CS-1 site) and with a broad array of extracellular matrix ligands. Recently, as noted above, Akimov and co-workers reported that tTG could interact with several members of the integrin family, including ₅₁ (as we also report in the current study), but also ₁₁, ₃₁, _v3 and _IIb3 (32). Interestingly, the report by Akimov et al. demonstrated that the related transglutaminase family member, FXIII, was not a ligand for ₅₁, and thus suggested that the interactions between integrins and transglutaminases was specific for tTG. Although that report and this one both described interactions between ₅₁ and tTG, there were several substantial differences. For example, Akimov et al. found no effect of either GRGDSP peptide or anti-₁ antibody on the interaction between tTG and integrins they describe, whereas in this report we found that GRGDSP completely eliminated the residual adhesion of G-361 cells to tTG once ₉₁-mediated adhesion was inhibited, and we also found complete inhibition with anti-₁ antibody. Furthermore, we found no role for integrins other than ₄₁, ₅₁, or ₉₁ in the adhesion of MOLT-3 or G-361 cells to tTG. We have checked the expression of various integrin subunits other than those shown in Fig. 2. As is clearly depicted in Fig. 10, G-361 cells also expressed ₂, ₃, ₆, and _v3. MOLT-3 cells, however, did not express these integrins substantially. Results obtained with G-361 cells further substantiate the notion that these integrins are not involved in the adhesion of G-361 cells to tTG. There are several possible explanations for these differences. One obvious difference is that Akimov et al. were examining the effects of tTG on cell adhesion to the 42-kDa fragment of fibronectin, and in their case the tTG was supplied to integrins within the secretory apparatus of the same cell, whereas the present study examined adhesion to immobilized tTG. Under the conditions examined by Akimov et al., it is conceivable that blocking antibodies and GRGDSP peptide would not be able to displace already bound tTG, whereas these reagents were perfectly capable of binding to unligated integrin under the conditions used in the present study, thereby blocking subsequent interactions with tTG. Such an explanation is plausible because it is often more difficult to detach already adherent cells with integrin-blocking reagents than to block the attachment of suspended cells. Alternatively, the WI-38 cells and HEL cells used by Akimov et al. and the MOLT-3 cells and G-361 cells used in the present study could interact with tTG through different mechanisms.

View larger version (21K): [in this window] [in a new window]   Fig. 10.   Flow cytometry of integrins expressed on G-361 cells and MOLT-3 cells. Cells were reacted with no IgG (control), an anti-human 2 mAb 6F1 (2 subunit), an anti-human 3 mAb P1B5 (3 subunit), an anti-human and -mouse 6 mAb GoH3 (6 subunit), or anti-human v3 polyclonal antibodies (v3).

The results of the present study also clearly demonstrate that although ₅₁ is not a receptor for inactive FXIII, both ₄₁ and ₉₁ are. These findings appear to differ from our previous report suggesting that ₅₁ did contribute to adhesion of TIG-1 cells to FXIII. However, the adhesion observed in those experiments required that the FXIII be activated. The results of the present study confirm that ₅₁ on G-361 cells can also mediate adhesion to activated FXIII (data not shown). We are uncertain whether the requirement for activation in this case is due to a conformational change in FXIII upon activation that unmasks an ₅₁ binding site or to a requirement for enzymatic activity to induce ₅₁-mediated adhesion

The interaction of ₄₁ and ₉₁ with FXIII could be biologically significant. Both ₄₁ and ₉₁ are highly expressed on specific populations of leukocytes, where they play a prominent role in transendothelial leukocyte migration. As a member of the coagulation cascade, FXIII is likely to be enriched at sites of vascular injury and inflammation, where its interaction with ₄₁ and ₉₁ could promote leukocyte extravasation.

We speculated previously about the function of pp-vWF as an emergency tag for targeting of ₄₁-expressing leukocytes, together with VCAM-1, to sites of inflammation and injury (24). Such a mechanism would be expected to be most relevant to lymphocytes, monocytes, and eosinophils, all of which express high levels of ₄₁ (36). The recent report by Taooka et al. (18) that human neutrophils express ₉₁ and utilize this integrin for migration on VCAM-1 and through activated endothelial monolayers extends this possible function to all populations of circulating leukocytes.

	ACKNOWLEDGEMENTS

We thank Drs. K. Ikura, M. Hayashi, C. Morimoto, B. S. Coller, and A. Sonnenberg for valuable gifts.

	FOOTNOTES

^* This work was supported in part by a research grant from the Cosmetology Research Foundation (to Y. S.), by Research Fellowship of the Japan Society for the Promotion of Science for Young Scientists 094491 (to T. I.), by Research Grant for Cardiovascular Diseases 11C-1, by Grant-in-Aid for Cancer Research 11-12 from the Ministry of Health and Welfare, by a grant from the Tsuchiya Memorial Foundation for Medical Research (to Y. Y.), and by National Institutes of Health Grant HL/AI33259 (to D. S.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^§ These authors contributed equally to this work.

^** To whom correspondence should be addressed. Tel.: 81-45-924-5728; Fax: 81-45-924-5808; E-mail: ysaito@bio.titech.ac.jp.

Published, JBC Papers in Press, May 17, 2000, DOI 10.1074/jbc.M003526200

² H. Takahashi, T. Isobe, S. Horibe, J. Takagi, Y. Yokosaki, D. Sheppard, and Y. Saito, unpublished observations.

	ABBREVIATIONS

The abbreviations used are: VCAM-1, vascular cell adhesion molecule-1; tTG, tissue-type transglutaminase; FXIII, blood coagulation factor XIII; pp-vWF, propolypeptide of von Willebrand factor; mAb, monoclonal antibody; BSA, bovine serum albumin.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES