Nidogen

EXPERIMENTAL CELL RESEARCH ARTICLE NO.

241, 36–45 (1998)

EX984016

Entactin-2: A New Member of Basement Membrane Protein with High Homology to Entactin/Nidogen Naoki Kimura,*,† Tomoko Toyoshima,* Tetsuo Kojima,* and Miyuki Shimane*,1 *Gene Search Program, Chugai Research Institute for Molecular Medicine, Nagai Niihari Ibaraki, Japan; and †Department of Molecular Cell Biology, Medical Research Institute, Tokyo Medical and Dental University, Tokyo, Japan

quence, was performed [2, 3] on this cell line, originally established from the bone marrow of female C3H/He mice [1]. As a consequence, we discovered a clone revealing a novel protein, with high homology to entactin/ nidogen, which we named entactin-2. Entactin/nidogen is an important basement membrane component, which reportedly promotes cell attachment [4], neutrophil chemotaxis [5], trophoblast outgrowth [6], and angiogenesis [7]. It was identified as a sulfated and glycosylated 148-kDa protein [8, 9], with 85% of its amino acid sequence being identical in both human [10] and mouse [11, 12]; therefore it is suggested that its molecular structure has been highly conserved during evolution. Elucidation of recombinant entactin/nidogen, by ultrastructural studies [9], revealed three globular domains (G1,G2,G3) and provided a model for this molecule. G1 and G2 are connected by a thread-like structure, whereas that between G2 and G3 is rod-like. Each domain contains particular motifs: in G2 it is similar to a cysteine-rich epidermal-growth factor (EGF) [13, 14]; the rod-like structure contains four EGF-like motifs and one which is thyroglobulin-like [156]; G3 possesses an EGF-like motif and a region homologous to the low density lipoprotein (LDL) receptor [16]. It was shown that entactin/ nidogen could potentially bind to laminin through its G3 domain [9, 17] and to collagen IV and proteoglycan through G2 [18], resulting in formation of ternary complexes [4]; collagen type IV and laminin are important matrix components which can maintain the differentiated phenotype, both in vivo and in vitro [19, 20]. For cell attachment, two major sites within the molecule have been suggested: first, the RGD sequence localized in EGF-like repeat 2 of the rod-like domain [4, 11, 21, 22], which is recognized by the avb3 integrin receptor [23]; secondly, the G2 domain of entactin, which is recognized by a3b1 [24]. In this paper, we undertook several studies to further characterize this new gene for ‘‘entactin-2’’ and compared it with entactin. We found that its structure and ability to bind to the cell surface were closely related to those of entactin. Furthermore, both the tissue distribution and the change of expression level of entactin2 gene during osteoblast differentiation were similar

Using the new signal sequence trap (SST) method, we isolated several clones encoding secreted and transmembrane proteins from KUSA cells, a murine osteoblast-like cell line. One isolated novel clone, termed entactin-2, exhibited a high similarity to mouse entactin/nidogen, a basement membrane protein. Although deduction of the amino acid sequence of entactin-2 revealed only 27.4% homology to entactin, many structural similarities were seen between both proteins. Entactin-2 contains five EGF-like and two thyroglobulin-like motifs, which are both cysteine-rich. Comparison of both proteins clearly revealed that entactin-2 also contains related domain structures. The rod-like domain of entactin-2, containing the RGD integrin recognition sequence, fused to glutathione-S transferase (GST), revealed a cell surface-binding activity similar to that of entactin. In addition, the tissue distribution of entactin-2 mRNA resembled that of entactin. Furthermore, mRNA expression of both genes decreased as osteoblastic differentiation progressed. These results suggest that entactin-2 is a member of the entactin gene family, may have entactin-related functions, and might act as a basement membrane component. q 1998 Academic Press

INTRODUCTION

One unique property of KUSA cells, a murine osteoblast-like cell line, is their ability to differentiate into mature bone tissue and induce trilineage hematopoiesis upon transplantation into athymic Balb/c nu/nu mice both in vitro and in vivo [1]. However, information on the types of cell factors expressed by this cell line is limited. In order to identify novel factors secreted from KUSA cells which might affect the bone formation, the signal sequence trap method, which isolates cDNA encoding secreted proteins and/or type I transmembrane proteins by virtue of N-terminal signal se1 To whom correspondence and reprint requests should be addressed at Gene Search Program, Chugai Research Institute for Molecular Medicine, 153-2, Nagai Niihari Ibaraki 300-41, Japan. Fax: 81-298-30-6270.

36

0014-4827/98 $25.00 Copyright q 1998 by Academic Press All rights of reproduction in any form reserved.

AID

ECR 4016

/

6i31$$$321

04-24-98 11:09:44

ecal

ENTACTIN-2, A NEW BASEMENT MEMBRANE PROTEIN

to those of the entactin gene. These finding suggest that entactin-2 is a new member of the entactin/nidogen family; has structural similarities to entactin, which is a basement membrane protein; and might be involved in the process of osteoblast differentiation. MATERIALS AND METHODS Cells and cell culture. For cell culture, all media were supplemented with 10% (v/v) fetal calf serum (FCS). KUSA cells [1] were maintained in Isocove’s modified Dulbecco’s medium. MC3T3-E1 cells, another osteoblast-like cell line established from newborn mouse calvaria [34, 35], were maintained in MEM a, in the presence or absence of 10 mM b-glycerophosphate. P5, a mouse stromal cell line, was maintained in RPMI 1640 medium (GIBCO BRL). COS-7, NIH3T3, and human fibrosarcoma HT1080 cells were cultured in Dulbecco’s modified Eagle’s medium (GIBCO BRL). Molecular cloning of Entactin-2 cDNA. Cloning of the cDNA that encoded secreted or membrane-associated protein was performed by a signal sequence trap (SST) strategy [2]. At first, the expression vector, pSRa-Tac was constructed as follows. Tac gene (human interleukin-2 receptor a chain) was excised with EcoRI and Eco47III from pKCR-Tac2A containing full length Tac cDNA (Riken DNA Bank) and inserted downstream of the SRa promoter of pcD-SRa [25]. The resulting vector, pSRa-Tac contained a cloning site between the SRa promoter and the coding region of the Tac gene without a signal sequence, previously cleaved off by digestion with SacI. The SST expression library, enriched for the 5* end of cDNA, was made by treating 5 mg of KUSA cell mRNA with reverse transcriptase (Superscript II, GIBCO BRL) and a random primer. After the first strand cDNA synthesis, a dC tail was attached at the 5* end of cDNA using a terminal deoxynucleotidyl transferase (GIBCO BRL). The second strand was synthesized with Taq polymerase (Takara) primed with 5*-GCGGCCGCGAATTCTGACTAACTGAC(dG)17, containing an EcoRI site. After sonication and blunting of this cDNA (DNA blunting kit, Takara), the SacI linkers, 5*-CCGCGAGCTCGATATCAAGCTTGTAC and 3*-GGCGCTCGAGCTATAGTTCGAACATGGAG, were ligated (DNA ligation kit, version 2, Takara). Subsequently, the cDNA fragments were amplified by PCR with two primers, 5*-GAGGTACAAGCTTGATATCGAGCTCGCG and 5*-GCCGCGAATTCTGACTAACTGAC. Ten cycles of PCR reaction (947C, 30 min; 557C, 1 min; 727C, 1 min) were carried out with a thermal cycler (Perkin Elmer). Amplified cDNA fragments were loaded on a 1.5% agarose gel; those of 400 bp were isolated, digested with EcoRI and SacI, and then ligated into these restriction enzyme sites in pSRa-Tac. After transformation, the library, consisting of 49 clones, was divided into several pools. Plasmid DNAs purified from each of the pools (Qiagen plasmid kit) were transfected into COS-7 cells (Lipofectamine, GIBCO BRL). Afterward, cells were cultured in medium with reduced serum (Opti-mem I, GIBCO BRL) for 2 days, harvested, and then incubated with mouse anti-Tac monoclonal antibody (Oncogene Science), and FITC-conjugated goat anti-mouse IgG monoclonal antibody (Becton Dickinson) at 47C. After immunostaining, cells transfected with each pool of cDNAs were analyzed by flow cytometry (Elite, Coulter). After a pool was selected as positive, these screening steps were repeated until a single positive clone was identified. Cloning of full length cDNA of entactin-2 gene. After obtaining total RNA by the method of Chirgwin et al. [26], poly(A) RNA was purified (mRNA purification kit, Pharmacia Biotech). An oligo (dT)primed cDNA library was constructed from 5 mg of KUSA mRNA (ZAP-cDNA Synthesis Kit, Stratagene). This cDNA library was then screened by the plaque hybridization method: 3 1 105 plaques were spread on the NZY plates and DNA was transferred to nylon membranes (hybond-N/, Amersham); these were screened by hybridization with an entactin-2 cDNA fragment labeled with [a32P]dCTP (Megaprime DNA labeling system, Amersham) in buffer (50% formamide, 5X SSPE, 0.5% SDS, 5X Denhart’s solution, and 0.1 mg/ml

AID

ECR 4016

/

6i31$$$321

04-24-98 11:09:44

37

salmon sperm DNA) at 427C for 16 h; posthybridization membranes were washed in 2X SSC containing 0.1% SDS and then in 0.1% SSC containing 0.1% SDS. After the positive clone was excised into a pBluescriptII vector (ExAssist helper phage, Stratagen), deletion mutants of this clone were produced (Kilo-sequence deletion kit, Takara). Deletion mutants containing various sizes of the entactin-2 gene’s cDNA insert were sequenced (Automated Fluorescence-Based Sequencing System 377, ABI; ABI Prism Dye Terminator Cycle Sequencing Ready Reaction Kit, Perkin Elmer). Expression and purification of GST-fusion protein. The cDNA fragment of entactin-2 coding for amino acids 754 to 1067 and that for its full sequence, contained in pBluescriptII, were amplified by the polymerase chain reaction with the specific oligonucleotide primers, 5*-GGGAATTCGAATCCTTGTTACGACGGAAGC and 5*-CCCTCGAGTCTCTGCCATCCTTGTCCACAC. After the PCR product was digested with EcoRI and XhoI and purified through an agarose gel, the isolated DNA fragment was subcloned into the pGEX-4T-3 GST expression vector (Pharmacia) between the EcoRI and the SacI site. E. coli host strain, JM109, was then transformed with either this construct or the original pGEX-4T-3 as a control and grown at 377C to a density with an A600 value of between 0.6 and 0.8. Protein expression was induced with 0.1 mM IPTG at 307C for 20 h. For purification of GST, bacteria were harvested by centrifugation at 47C and lysed on ice for 30 min with 1 mM PMSF in sonication buffer (PBS-1% Tween 20); for the GST-fusion protein purification, according to Dong et al. [24], lysozyme buffer (25 mM Tris, pH 8.0, and 10 mM EDTA containing 1 mg/ml lysozyme) was used instead of sonication buffer. Subsequently, bacteria from each respective lysis were further disrupted by brief sonication on ice then centrifuged at 15000 rpm for 30 min at 47C. Glutathione Sepharose 4B gel (Pharmacia) was then added to the resulting supernatants, incubated overnight at 47C, and washed three times by centrifugation and the addition of 10 bed volumes of cold PBS (0–17C). Proteins bound to the matrix were eluted with 10 mM reduced glutathione (Pharmacia) and their concentration was determined (Protein Assay Kit, BioRad). Purity of the final product was assessed by SDS-PAGE and Western blotting; a single band with the expected molecular weight was detected by Coomassie blue and immunostaining with anti-GST protein antibody (data not shown). Assessment for binding to cell surface. In this binding assay, cells were diluted in FACS buffer (PBS containing 2% FCS and 0.02% NaN3). All incubation steps were performed in a total volume of 50 ml at 47C. KUSA, P5, NIH3T3, and HT1080 cells were harvested with 0.05% EDTA, washed once, incubated at 1 1 106 cells with 20 mg/ml of either GST-rod (2397–3337) fusion protein or GST protein for 1 h, and then washed again. Subsequently, cells were incubated with anti-GST mouse IgG for 30 min, further washed, stained with FITC-conjugated goat anti-mouse IgG (Becton Dickinson) for another 30 min, and then evaluated by flow cytometry (Elite, Coulter). RT-PCR. For entactin-2, entactin, osteocalcin, and G3PDH, 1 mg of respective total RNA, prepared with a kit (RNeasy total RNA extraction kit, Qiagen), was mixed with 0.5 mg of oligo(dT) primer and diluted with distilled water to 11 ml. The resulting mixture was heated at 707C for 10 min and chilled quickly, and then the following were added sequentially: 4 ml of 5X first strand buffer (GIBCO BRL), 2 ml of 0.1 M DTT, 1 ml of dNTP (10 mM of each dNTP), and 1 ml (200 units) of RT (SuperScript, GIBCO BRL). The resulting solution was incubated at 427C for 1 h and the reaction terminated by raising the temperature to 707C for 15 min. Immediately afterward, 30 ml of distilled water and 0.5 mg of glycogen were added and samples centrifuged through individual Chroma Spin-100 columns (Clontech) for 5 min at 700g. The respective first strand cDNAs obtained were then used directly for amplification of the cDNA mouse gene by PCR. Reaction mixtures for PCR were as follows: 1 ml of the first strand cDNAs, 5 ml of 10 1 PCR reaction buffer (Toyobo), 4 ml of 2.5 mM dNTP mixture, 36 ml of distilled water, 0.5 ml of 5 unit/ml Taq gold DNA polymerase (Toyobo), and 2 ml of each primer, stored as 20 mM solution, for sets related to

ecal

38

KIMURA ET AL.

each respective cDNA. The following are sequences for each set of related primers, designated to a particular gene. Entactin:

ENT-1 primer, 5* GCTGTAGTTGGGCGATGTGG (28-47); ENT-2 primer, 5* CAGGCCATCGGTTGTGTCCA (343-362). Entactin-2: EN2-1 primer, 5* AGGCCACAGCAGAATTGATG (1054-1073); EN2-2 primer, 5* GGTCCTCCAGTCCTATCACA (1501-1520). Osteocalcin: BGP-1 primer, 5* TCTCTGCTCACTCTGCTGGC (61-80); BGP-2 primer, 5* GCAGCACAGGTCCTAAATAG (329-345). G3PDH: G3PDH-1 primer, 5* CCGGTGCTGAGTATGTCGTG (309-328); G3PDH-2 primer, 5* GTCCTCAGTGTAGCCCAAGA (858-857).

The PCR reaction was performed (Perkin-Elmer thermal cycler) at 947C for 12 min, followed by between either 25 and 30 or 35 and 40 cycles of 947C, 30 s; 587C, 30 s; 727C, 30 s. Reaction products were resolved on 1% agarose gel and stained with ethidium bromide. Northern blot analysis. A mouse multiple-tissue Northern blot (Clontech) containing 2 mg of poly(A)/ RNA was hybridized consecutively with a-32P-labeled 1.3-kb cDNA fragment of entactin-2 (sequence position 1–1295) and a-32P-labeled 1.1-kb cDNA fragment of entactin cDNA (sequence position 847–1966). Hybridization was carried out in ExpressHyb hybridization solution (Clontech) at 687C for 2 h. Filters were washed in 2X SSC containing 0.1% SDS at room temperature with several changes of buffer, then for 40 min at 507C in 0.1X SSC containing 0.1% SDS. The images were analyzed using a BAS 2000II bio-imaging analyzer (Fujix). Alkaline phosphatase activity in osteoblastic cells. Cells were washed twice with PBS, harvested by the addition of homogenization buffer (50 mM Tris-HCl, pH 7.2, 0.1% Triton X-100), and frozen at 0207C for storage. Immediately after thawing, cell lysates were briefly sonicated and then centrifuged, and supernatants obtained were assayed for total protein, using a Bio-Rad kit, and for alkaline phosphatase (ALP) activity as follows. In brief, 2 mg (KUSA cells) or 20 mg (MC3T3-E1 cells) of cell lysate in homogenization buffer was mixed with equal volume of incubation buffer (0.1 M 2-amino-2methyl-1-propanol, pH 10.5, 2 mM MgCl2) containing 20 mM p-nitrophenyl phosphate, followed by an incubation at 377C for 30 min. After the addition of 0.1 M NaOH, the amount of p-nitrophenol color was measured with a spectrophotometer.

RESULTS

Molecular Cloning of Novel cDNA Clone Encoding Entactin-2 In the signal sequence trap technique, the 5* ends of the cDNAs were inserted in place of the original signal peptide sequence of IL-2 receptor (tac), which had been deleted in the cloning vector. Inserts containing a signal sequence in-frame with the tac gene could therefore be expressed on the surface of transfected cells as tac fusion proteins. After, construction of a library, transfection into cos-7 cells, then screening for tac expression on their surface, 10 positive clones were obtained by screening approximately 600 clones prepared from KUSA cell mRNA. Seven clones were identified as known genes containing the authentic signal sequences; these were TA1 [27], Sgp1 [28], pigment epithelium-differentiation factor (PEDF) [29], osteonectin [30], collagen type I [31], biglycan [32], and cystatin C [33]. Also obtained were three novel clones. One of these (clone 7E5) obtained had a novel open reading frame with an insert size of 400 bp, therefore it was used as a probe to screen an oligo(dT)-primed KUSA cDNA library to obtain full length cDNA of this clone.

AID

ECR 4016

/

6i31$$$321

04-24-98 11:09:44

The nucleotide sequence of full length clone, 7E5, was determined and shown in Fig. 1; it consisted of 5104 bp containing a single open reading frame and a 3* untranslated region of 4209 and 853 bp, respectively, which included the polyadenylation signal, AATAAA, at the 3* end. The number of amino acids in the deduced sequence was 1403, 30 of which were in a hydrophobic stretch that resembled a typical protein secretary leader sequence, located adjacent to the putative initiation codon. The deduced amino acid sequence of this clone was compared for homology with amino acid sequences using the BLAST search program. The gene with most similarity (approximately 71%) to full length clone 7E5 was a novel human osteoblast mRNA named osteonidogen, which was cloned from human osteoblast by Ohno et al. (unpublished) and submitted to the DDBJ/EMBL/GenBank databases(D86425). The second most similar gene (approximately 27%) was of mouse entactin/nidogen; some motifs found in entactin were also conserved in clone 7E5. These results led us to conclude that our clone represented a member of the entactin/nidogen gene family. We thus named it entactin-2 and, based on similarities of motif, considered osteonidogen as its human counterpart. A Structural Comparison between Entactin-2 and Entactin An amino acid comparison among entactin, entactin2, and osteonidogen showed that entactin-2 and osteonidogen contained many similarities; thus we presumed that osteonidogen was its human counterpart. Furthermore, entactin-2 also has domains similar to entactin. A structural comparison in each domain between entactin-2 and entactin showed a number of similarities as follows: There was high homology between their individual G2 and G3 domains (31.2 and 38.4%, respectively), and rod-like structures (33.7%), which were all rich in cysteine residues (Fig. 2). Within the G2 domain of both entactin-2 and entactin, the EGFlike motif was found at their N-termini. The rod-like structure of both molecules contained four EGF-like motifs. The EGF precursor/LDL receptor-homologous region was found in the G3 domain of both molecules (Fig. 1 and Fig. 3). When counting these EGF-like motifs from the N-terminus of both molecules, the third and fifth for entactin-2 contained a consensus sequence for calcium binding [41–43] as was the case for entactin (Fig. 2). Moreover, both the second for entactin and the fifth for entactin-2 contained the cell surface receptor recognition sequence for integrin, RGD, located within their rod-like structure. There were some differences (Fig. 3): Entactin-2 had two thyrogloblin-like motifs at its rod-like structure, whereas, entactin had only one. Spanning across all domains and interconnecting structures, entactin contained six EGF-like motifs, whereas, entactin-2 had

ecal

ENTACTIN-2, A NEW BASEMENT MEMBRANE PROTEIN

39

FIG. 1. The nucleotide sequence of the entactin-2 cDNA and the deduced amino acid sequence. Hydrophobic core of the predicted signal sequence is written in italic letters. The sequences of the five EGF-like motifs are highlighted, and two thyroglobulin-like motifs are shaded. The LDL receptor homologous region is underlined with a waved line, and the RGD site is in bold italic letters. The polyadenylation signal, AATAAA, is underlined.

AID

ECR 4016

/

6i31$$4016

04-24-98 11:09:44

ecal

40

KIMURA ET AL.

FIG. 2. Multiple amino acid sequence alignment of entactin, entactin-2, and osteonidogen. The amino acid sequence among the entactin (upper sequence), the entactin-2 (middle sequence), and the osteonidogen (lower sequence) are compared at G2 domain, rod-like domain, and G3 domain in each molecule, respectively. The identical sequences are shadowed. Regions of EGF-like repeats are in boxes. The consensus sequences based on calcium-binding studies with similar repeats from coagulation factors [40–42] are shown by a star symbol.

04-24-98 11:09:44

ecal

ENTACTIN-2, A NEW BASEMENT MEMBRANE PROTEIN

41

FIG. 3. Structural domains in entactin and entactin-2. Each domain is shown based on the recombinant protein [13]. The numbered boxes represent the EGF-like motifs, the open boxes represent the thyroglobulin-like motifs, and the hatched boxes represent the LDL receptor homologous region. The potential Ca2/-binding sites are marked by a star symbol, and RGD sequence are indicated with a arrow. The scale shows the residue numbers.

only five; entactin-2 did not contain the extra EGFlike motif at its carboxyterminus. Low homology was associated with their G1 domains and thread-like structures (24 and 8%, respectively, data not shown). Another two potential calcium-binding sites existing in this N-terminal region of entactin were not found in entactin-2. Tissue Distribution of Entactin-2 and Entactin In order to compare the tissue distribution of both genes, Northern blot analysis was then performed using a 1.3-kb cDNA fragment of entactin-2 (sequence position, 1–1300) and a 1.1-kb cDNA fragment of entactin (sequence position, 847–1985) as probes. Two transcripts of entactin-2 were markedly detected in heart, lung, skeletal muscle, kidney, testis, and liver, whereas lower levels were observed in brain and spleen (Fig. 4B). Tissue distribution of the entactin-2 gene closely resembled that of entactin (Fig.4A). Interestingly, two transcripts hybridized to cDNA for entactin2 and three to that for entactin. We thought that the longer mRNA species corresponding to entactin-2 were the products of cloned cDNA, whereas the two signals at 6 and 4 kb corresponding to entactin were most probably as a consequence of the efficiency of the polyadenylation signal; in contrast, this was not the case with shorter signals at approximately 4.3 kb (entactin-2) and 2.4 kb (entactin).

idues conferred the same function to entactin-2, a GSTfusion protein consisting of its related region between amino acid residues 754–1067 (GST-rod) was synthesized in E. coli and purified according to Dong et al. and then assessed in a number of cell lines, as described under Materials and Methods. All cell lines, except for HT1080, bound to the GSTrod protein and not to the control fusion protein alone (Fig. 5). Especially prominent was the amount of GSTrod protein attached to the surface of KUSA cells.

Binding of Entactin-2 to Cell Surface Dong and co-workers revealed the two distinct cell attachment sites in entactin using a series of GSTfusion proteins containing the four major subdomains of entactin [24]. To determine whether the observed structural similarities of entactins cell attachment res-

AID

ECR 4016

/

6i31$$$321

04-24-98 11:09:44

FIG. 4. Tissue distribution of entactin-2 and entactin transcript. Mouse multiple tissue northern (MTN) blots (Clontech) were subjected to a32P-labeled entactin cDNA (A), and subsequently to a a32Plabeled entactin-2 cDNA (B), as probes. Molecular size markers are shown on the left.

ecal

42

KIMURA ET AL.

FIG. 5. Cell attachment of the RGD sequence in the rod-like domain of entactin-2. Purified GST-fusion protein of the rod domain in entactin-2 (GST-rod) and GST protein were used. KUSA cells (A), NIH3T3 cells (B), P5 cells (C), and HT1080 cells (D) were harvested with 0.05% EDTA and were incubated with 2 mg of GST protein (solid line) or GST-rod protein (dotted line) in FACS buffer (PBS containing 2% FCS and 0.02% NaN3) at 47C. Cells were immunostained with FITC anti-GST and then subjected to flow cytometry (Elite Coulter).

These results clearly demonstrated that the rod-like domain of our recombinant entactin-2 had cell attachment activity. Expression Analysis of Entactin-2 Gene Expression of entactin-2 and expression of entactin were readily observed in MC3T3-E1 and NIH3T3, whereas, KUSA expressed slightly lower levels of the former and not the latter (Fig. 6A). The increase of alkaline phosphatase activity in KUSA cells was about 1.6 times after 8 days of culture when cells had reached confluence (from 1.34 { 0.16 to 2.18 { 0.04 mmol/min/ mg of protein), as was the case for MC3T3-E1 cells [34, 35] cultured for several days after confluence. We therefore examined the expression levels in KUSA cells harvested immediately after confluence and at 8 days after this event; surprisingly, expression was decreased in the older cultures (Fig. 6B). Since a similar result was predicted for differentiated MC3T3-E1 cells, we induced them to differentiate by culturing them for a long term in the presence of 10 mM b-glycerophosphate. The differentiation of MC3T3-E1 cells was confirmed by assessing increased expression of osteocalcin gene, a differentiation marker of osteoblasts [36, 37]. Expression of the entactin-2 gene decreased concurrently with the onset of differentiation (Fig. 6C); the entactin gene also exhibited almost the same behavior as the entactin-2 gene. DISCUSSION

In the present study, we isolated a new gene, termed entactin-2, from the osteoblast-like cell line, KUSA

AID

ECR 4016

/

6i31$$$321

04-24-98 11:09:44

cells, using a signal sequence trap strategy. Deduced amino acid sequence of this clone revealed high similarity to osteonidogen, which had been cloned from human osteoblast and submitted to the DDBJ/EMBL/GenBank databases (D86425). Thus our entactin-2 was considered as the mouse counterpart for human osteonidogen. Although osteonidogen was submitted as a homologous gene to entactin and nidogen, more extensive characterization has not yet been done. With respect to entactin-2, we attempted to elucidate on how this gene should be characterized by comparing it to entactin and nidogen, in the context of its structure, tissue distribution, cell attachment activity, and gene expression during osteoblast differentiation. Whereas the amino acid sequence deduced from entactin-2 cDNA showed only approximately 27% sequence identity with entactin, both genes were closely related to entactin in terms of structure. This was especially true for both G2 and G3 domains, which are the binding sites for other basement membrane proteins; these were highly conserved between entactin and entactin-2. Furthermore, two potential calcium-binding sites and an RGD integrin recognition sequence, which are involved in cell adhesion, were conserved on the rod-like domain as for entactin. These observations therefore strongly suggest that entactin-2 is a new gene member of the entactin family. Furthermore, as the functional region in entactin is conserved in entactin2* we hypothesized that entactin-2 might have a role similar to that of entactin. Tissue distribution of both entactin and entactin-2 gene resembled each other. The expression of two homologous proteins in the same tissue suggests that entactin-2 might have a distinct function from entactin. Interestingly, short mRNA species were detected in both genes. Most of the carboxyl terminal half of entactin gene, rod domain and G3 domain, is encoded by 12 exons [38]. These together with the cDNA fragment corresponding to N-terminal region were used as probes; these short transcripts might be alternatively spliced variants with some deleted C-terminal exons, although the genomic region of entactin2 has not been characterized. As entactin-2 (and also human osteonidgen) was cloned from osteoblasts, we compared the expression level of both entactin-2 and entactin genes during osteoblastic differentiation, employing two kinds of osteoblast cell lines, KUSA and MC3T3-E1. Our KUSA cells displayed high ALP activity (2.18 { 0.04 mmol/min/ mg) and high osteocalcin gene expression (Fig. 6A), which are characteristic markers of osteoblastic differentiation. In contrast, the levels of either ALP activity (9.7 { 0.4 1 1003 mmol/min/mg) or osteocalcin gene expression (Fig. 6A) were very low in MC3T3-E1 cells. These observations suggest that KUSA cells are at a late stage in osteoblastic differentiation, whereas MC3T3-E1 cells appear as progenitors of osteoblasts at an early stage. The expression of both entactin-2 and

ecal

ENTACTIN-2, A NEW BASEMENT MEMBRANE PROTEIN

43

FIG. 6. Analysis of expression of entactin-2 and entactin. (A) A 1/50 vol of single strand cDNA derived from 1 mg of total RNAs of KUSA cells, MC3T3-E1 cells, and NIH3T3 cells was amplified in 50 ml of PCR reaction mixture containing specific primers for entactin-2, entactin, osteocalcin, and mouse G3PDH cDNA. After 25 cycles (entactin-2 and G3PDH), 30 cycles (osteocalcin), or 40 cycles (entactin) of PCR reaction, the DNA samples (10 ml) were electrophoresed on 1% agarose gel and stained with EtBr. (B) KUSA cells were cultured for 8 days after confluency. Total RNA was extracted at 1 and 8 days of culture and used for amplification with specific primers for entactin-2 (35 cycles) and G3PDH (25 cycles) cDNA and subjected as above. (C) MC3T3-E1 cells were maintained in the absence of b-glycerophosphate for 1, 6, and 9 days and in the presence of b-glycerophosphate for 18 and 39 days. Total RNA was extracted from each cells and used for amplification of entactin-2 (30 cycles), entactin (25 cycles), osteocalcin (30 cycles), and G3PDH (25 cycles) cDNA and subjected as above.

entactin gene was scarcely detectable in KUSA cells, but high in both MC3T3-E1 and NIH3T3 cells. However, when the differentiation was induced in MC3T3E1 cells, expression of mRNA for both of these genes was markedly suppressed. These results indicate that expression of both proteins, entactin-2 and entactin, is completely downregulated transcriptionally during osteoblastic differentiation. Although entactin/nidogen is known to be a major component in basement membrane, entactin-2’s proximity is as yet undetermined; the proximity of entactin is of interest in terms of elucidating on whether or not entactin-2 is involved in the regulation of osteoblastic differentiation through its association with extracellu-

AID

ECR 4016

/

6i31$$$321

04-24-98 11:09:44

lar matrix. Recent studies have also demonstrated that the components in basement membrane substrate (matrigel) influence the phenotypes of cells [39]. In our present study, downregulation of entactin and entactin-2 gene and sequential expression of osteocalcin gene during progressive osteoblast differentiation suggest that changes in the type of ECM components synthesized by osteoblast-like cells, from those of basement membrane matrix to bone-type ECM, might be involved in the process of bone cell differentiation. Dong et al. [24] had shown that both G2 and rod domain in entactin had cell attachment activity using a panel of GST-fusion proteins that encompasses the four major structural domains of entactin, G1, G2, rod, and

ecal

44

KIMURA ET AL.

G3. According to their strategy, we examined the cell attachment activity of the GST-fusion protein of rod domain in recombinant entactin-2. This fusion protein was found to promote cell attachment within several cell lines. This result allowed us speculate that cell attachment occurs through the RGD sequence in this domain, as was the the case with entactin [24]. Having said the above, it remains to be seen whether or not native entactin-2 folds with similar conformation compared to our experimental recombinant form, as the tertiary structure of these may differ; this could affect the nature of their respective cell attachment activities. Both mouse entactin-2 and human osteonidgen were isolated from osteoblasts of their respective species. As for entactin, expression of entactin-2 was markedly detected in normal fibroblast NIH3T3 cells, and widely in mouse tissues. In addition, recombinant entactin-2 had cell-binding activity not only with osteoblastic cells and KUSA cells but also in other types of cell lines (NIH3T3 cells, fibroblast cell line, and P5 cells, stromal cell line). Therefore, we concluded that entactin-2 might be found in several tissues. Based on several similarities to entactin/nidogen, its decreased expression during differentiation, and implications for cellbinding activity, it is of importance to see whether or not entactin-2 is a new constituent of basement membrane and is involved in the process of differentiation.

8.

9.

10.

11.

12.

13. 14.

15. We are extremely grateful to Dr. Akihiro Umezawa who established and provided us with KUSA cells. We thank also Dr. Hitoshi Nomura and Dr. Peter Kowalski-Saunders for useful comments and discussion concerning the manuscript.

16.

REFERENCES 1. Umezawa, A., Maruyama, T., Segawa, K., K. Shadduck, R., Waheed, A., and Hata, J. (1992). Multipotent marrow stromal cell line is able to induce hematopoiesis in vivo. J. Cell. Physiol. 151, 197–205. 2. Tashiro, K., Tada, H., Heilker, R., Shirozu, M., Nakano, T. and Honjo, T. (1993). Signal sequence trap: A cloning strategy for secreted proteins and type I membrane. Science 261, 600–603. 3. Shirozu, M., Tada, H., Tashiro, K., Nakamura, T., Lopez, N. D., Nazarea, M., Hamada, T., Sato, T., Nakano, T., and Honjo, T. (1996). Characterization of novel secreted and membrane proteins isolated by the signal sequence trap method. Genomics 37, 273–280. 4. Chakravarti, S., Tam, M. F., and Chung, A. E. (1990). The basement membrane glycoprotein entactin promotes cell attachment and binds calcium ions. J. Biol. Chem. 265, 10597–10603. 5. Senior, R. M., Gresham, H. D., Griffin, G. L., Brown, E. J., and Chung, A. E. (1992). Entactin stimulates neutrophil adhesion and chemotaxis through interactions between its Arg-Gly-Asp (RGD) domain and the leukocyte response integrin. J. Clin. Invest. 90, 2251–2257. 6. Yelian, F. D., Edgeworth, N. A., Dong, L. J., Chung, A. E., and Armant, D. R. (1993). Recombinant entactin promotes mouse primary trophoblast cell adhesion and migration through the Arg-Gly-Asp (RGD) recognition sequence. J. Cell Biol. 121, 923–929. 7. Nicosia, R. F., Bonanno, E., Smith, M., and Yurchenco, P.

AID

ECR 4016

/

6i31$$$321

04-24-98 11:09:44

17.

18.

19. 20.

21.

22.

23.

24.

(1994). Modulation of angiogenesis in vitro by laminin-entactin complex. Dev. Biol. 164, 197–206. Chung, A. E., Freeman, I. L., and Braginski, J. E. (1977). A novel extracellular membrane elaborated by a mouse embryonal carcinoma-derived cell line. Biochem. Biophys. Res. Commun. 79, 859–868. Fox, J. W., Mayer, U., Nischt, R., Aumailley, M., Reinhardt, D., Wiedemann, H., Mann, K., Timpl, R., Krieg, T., and Engel, J. (1991). Recombinant nidogen consists of three globular domains and mediates binding of laminin to collagen type IV. EMBO J. 10, 3137–3146. Nagayoshi, T., Sanborn, D. J., Hickok, N., Olsen, D. R., Fazio, M. J., Chu, M. L., Knowlton, R., Mann, K., Deutzmann, R., and Timpl, R. (1989). Human nidogen: Complete amino acid sequence and structural domains deduced from cDNAs, and evidence for polymorphism of the gene. DNA 8, 581–594. Durkin, M. E., Chakravarti, S., B. Bartos, B., Liu, S. H., Friedman, R. L., and Chung, A. E. (1988) Amino acid sequence and domain structure of entactin. Homology with epidermal growth factor precursor and low density lipoprotein receptor. J. Cell Biol. 107, 2749–2756. Mann, K., Deutzmann, R., Aumailley, M., Timpl, R., Raimondi, L., Yamada, Y., Pan, T. C., Conway, D., and Chu, M. L. (1989). Amino acid sequence of mouse nidogen, a multidomain basement membrane protein with binding activity for laminin, collagen IV and cells. EMBO J. 8, 65–72. Carpenter, G., and Cohen, S. (1979). Epidermal growth factor. Annu. Rev. Biochem. 48, 193–216. Gray, A., Dull, T. J., and Ullrich, A. (1983). Nucleotide sequence of epidermal growth factor cDNA predicts a 128,000-molecular weight protein precursor. Nature 303, 722–725. Mercken, L., Simons, M. J., Swillens, S., Massaer, M., and Vassart, G. (1985). Primary structure of bovine thyroglobulin deduced from the sequence of its 8,431-base complementary DNA. Nature 316, 647–651. Yamamoto, T., Davis, C. G., Brown, M. S., Schneider, W. J., Casey, M. L., Goldstein, J. L., and Russell, D. W. (1984). The human LDL receptor: A cysteine-rich protein with multiple Alu sequences in its Mrna. Cell 39, 27–38. Mann, K., Deutzmann, R., and Timpl, R. (1988). Characterization of proteolytic fragments of the laminin-nidogen complex and their activity in ligand-binding assays. Eur. J. Biochem. 178, 71–80. Reinhardt, D., Mann, K., Nischt, R., Fox, J. W., Chu, M. L., Krieg, T., and Timpl, R. (1993). Mapping of nidogen binding sites for collagen type IV, heparan sulfate proteoglycan, and zinc. J. Biol. Chem. 268, 10881–10887. Timpl, R. (1989). Structure and biological activity of basement membrane proteins. Eur. J. Biochem. 180, 487–502. Lin, C. Q., and Bissell, M. J., (1993). Multi-faceted regulation of cell differentiation by extracellular matrix. FASEB J. 7, 737– 743. Ruoslahti, E., and Pierschbacher, M. D. (1987). New perspectives in cell adhesion: RGD and integrins. Science 238, 491– 497. Pierschbacher, M. D., and Ruoslahti, E. J. (1987). Influence of stereochemistry of the sequence Arg-Gly-Asp-Xaa on binding specificity in cell adhesion. Biol. Chem. 262, 17294–17298. Dedhar, S., Jewell, K., Ojiani, M., and Gray, V. (1992). The receptor for the basement membrane glycoprotein entactin is the integrin alpha 3/beta 1. J. Biol. Chem. 267, 18908 – 18914. Dong, L. J., Hsieh, J. C., and Chung, A. E. (1995). Two distinct cell attachment sites in entactin are revealed by amino acid substitutions and deletion of the RGD sequence in the cysteine-

ecal

ENTACTIN-2, A NEW BASEMENT MEMBRANE PROTEIN rich epidermal growth factor repeat 2. J. Biol. Chem. 270, 15838–15843. 25. Takebe, Y., Seiki, M., Fujisawa, J., Hoy, P., Yokota, K., Arai, K., Yoshida, M., and Arai, N. (1988). SR alpha promoter: An efficient and versatile mammalian cDNA expression system composed of the simian virus 40 early promoter and the R-U5 segment of human T-cell leukemia virus type 1 long terminal repeat. Mol. Cell. Biol. 8, 466–472. 26. Chirgwin, J. M., Przybyla, A. E., MacDonald, R. J., and Rutter, W. J. (1979). Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 18, 5294–5299. 27. Sang, J., Lim, Y. P., Panzica, M., Finch, P., and Thompson, N. L. (1995). TA1, a highly conserved oncofetal complementary DNA from rat hepatoma, encodes an integral membrane protein associated with liver development, carcinogenesis, and cell activation. Cancer Res. 55, 1152–1159. 28. Cao, Q. P., and Crain, W. R. (1995). Expression of SGP-1 mRNA in preimplantation mouse embryos. Dev. Genet. 17, 263–271. 29. Steele, F. R., Chader, G. J., Johnson, L. V., and Tombran-Tink, J. (1993). Pigment epithelium-derived factor: Neurotrophic activity and identification as a member of the serine protease inhibitor gene family. Proc. Natl. Acad. Sci. U.S.A. 90, 1526– 1530. 30. Swaroop, A., Hogan, B. L., and Francke, U. (1988). Molecular analysis of the cDNA for human SPARC/osteonectin/BM-40: Sequence expression, and localization of the gene to chromosome 5q31-q33. Genomics 2, 37–47. 31. French, B. T., Lee, W. H., and Maul, G. G. (1985). Nucleotide sequence of a cDNA clone for mouse pro alpha 1(I) collagen protein. Gene 39, 311–312. 32. Scholzen, T., Solursh, M., Suzuki, S., Reiter, R., Morgan, J. L., Buchberg, A. M., Siracusa, L. D., and Iozzo, R. V. (1994). The murine decorin. Complete cDNA cloning, genomic organization, chromosomal assignment, and expression during organogenesis and tissue differentiation. J. Biol. Chem. 269, 28270–28281. 33. Solem, M., Rawson, C., Lindburg, K., and Barnes, D. (1990). Transforming growth factor beta regulates cystatin C in serum-

34.

35.

36.

37.

38.

39.

40.

41.

42.

free mouse embryo (SFME) cells. Biochem. Biophys. Res. Commun. 172, 945–951. Kodama, H. A., Amagai, Y., Koyama, H., and Kasai, S. (1982). A new preadipose cell line derived from newborn mouse calvaria can promote the proliferation of pluripotent hemopoietic stem cells in vitro. J. Cell. Physiol. 112, 89–95. Sudo, H., Kodama, H. A., Amagai, Y., Yamamoto, S., and Kasai, S. (1983). In vitro differentiation and calcification in a new clonal osteogenic cell line derived from newborn mouse calvaria. J. Cell Biol. 96, 191–198. Hauschka, P. V., Lian, J. B., Cole, D. E., and Gundberg, C. M. (1989). Osteocalcin and matrix Gla protein: Vitamin K-dependent proteins in bone. Physiol Rev. 69, 990–1047. Lian, J. B., and Gundberg, C. M. (1988). Osteocalcin. Biochemical considerations and clinical applications. Clin. Orthop. 226, 267–291. Durkin, M. E., Wewer, U. M., and Chung, A. E. (1995). Exon organization of the mouse entactin gene corresponds to the structural domains of the polypeptide and has regional homology to the low-density lipoprotein receptor gene. Genomics 26, 219–228. Stadler, E., and Dziadek, M. (1996). Extracellular matrix penetration by epithelial cells is influenced by quantitative changes in basement membrane components and growth factors. M. Exp. Cell. Res. 229, 360–369. Rees, D. J., Jones, I. M., Handford, P. A., Walte, S. J., Esnouf, M. P., Smith, K. J., and Brownlee, G. G. (1988). The role of betahydroxyaspartate and adjacent carboxylate residues in the first EGF domain of human factor IX. EMBO. J. 7, 2053–2061. Handford, P. A., Mayhew, M., Baron, M., Winship, P. R., Campbell, I. D., and Brownlee, G. G. (1991). Key residues involved in calcium-binding motifs in EGF-like domains. Nature 351, 164– 167. Selander-Sunnerhagen, M., Ullner, M., Persson, E., Teleman, O., Stenflo, J., and Drakenberg, T. (1992). How an epidermal growth factor (EGF)-like domain binds calcium. High resolution NMR structure of the calcium form of the NH2-terminal EGFlike domain in coagulation factor X. J. Biol. Chem. 267, 19642– 19649.

Received July 23, 1997 Revised version received December 15, 1997

AID

ECR 4016

/

6i31$$$321

04-24-98 11:09:44

45

ecal