A cell adhesion protein from the crayfish Pacifastacus leniusculus, a serine proteinase homologue similar to Drosophila masquerade.

A cDNA encoding a protein resembling masquerade, a serine proteinase homologue expressed during embryogenesis, larval, and pupal development in Drosophila melanogaster, was identified in hemocytes of the adult freshwater crayfish, Pacifastacus leniusculus. The crayfish protein is similar to Drosophila masquerade in the following aspects: (a) overall sequence of the serine proteinase domain, such as the position of three putative disulfide bridges, glycine in the place of the catalytic serine residue, and the presence of a substrate-lining pocket typical for trypsins; (b) the presence of several copies of a disulfide-knotted motif in the putative propeptide. This masquerade-like protein is cleaved into a 27-kDa fragment, which could be detected by immunoblot analysis using an affinity-purified antibody against a synthetic peptide in the C-terminal domain of the protein. The 27-kDa protein could be immunoaffinity-purified from hemocyte lysate supernatant and exhibited cell adhesion activity in vitro, indicating that the C-terminal domain of the crayfish masquerade-like protein mediates cell adhesion.

Serine proteinases, the most studied class of proteinases, belong to a diverse multigene family that shares a common catalytic mechanism and structural characteristics such as the presence of three conserved amino acid residues, His, Asp, and Ser, within the active site. These enzymes are involved in several biological processes in higher animals, including digestion, proenzyme, prohormone, and complement activation, as well as participate in defense mechanisms (1). All eukaryotic serine proteinases, most of which are digestive enzymes, are suggested to have originated from a single ancestral gene, and variants associated with other functions are thought to have arisen by gene duplication and mutations through evolution (2). Serine proteinases are typically synthesized as zymogens or inactive proenzymes, which are then activated by a specific and limited proteo-lytic cleavage at a specific peptide bond (3). Upon activation of many serine proteinases, the noncatalytic N terminus remains linked to the catalytic C terminus via a disulfide bond. The N-terminal domain has been shown to be important in the activation of the protein and may play an essential role for the normal regulation of enzymatic activity (4).
Several serine proteinase-inactive homologues have already been identified in vertebrates and invertebrates, and they have been suggested to have different biological functions: such as antimicrobial activities, e.g. human azurocidin (5) and horseshoe crab factor D (6), as a growth factor, e.g. human hepatocyte growth factor (7), an adhesion molecule, e.g. fruit fly glutactin (8), neurotactin (9), and masquerade (mas) 1 (10), or as an immune molecule, e.g. mosquito infection-responsive serine protease-like protein (ispl5) (11). These molecules show homology to serine proteinases except for the substitution(s) of the catalytic residues. The prophenoloxidase-activating system (proPO system), which carries out recognition and defense responses in invertebrates, is composed of an enzyme cascade consisting of several serine proteinases and prophenoloxidase, which is converted to an active enzyme following proteolytic cleavage (12). Several insect serine proteinases have been found to be involved in the activation of the proPO system (12), as a trypsin-like serine proteinase in crayfish (13). To obtain information about genes encoding crayfish serine proteinases, we have isolated and analyzed putative serine proteinase genes from crayfish hemocytes. Among these we found several trypsin-like enzymes and a mas-like protein. We here describe the cloning, purification, and cell adhesion activity of this mas-like protein from hemocytes of the freshwater crayfish Pacifastacus leniusculus.

EXPERIMENTAL PROCEDURES
Animals-Freshwater crayfish, P. leniusculus, were purchased from Berga Krä ftodling, Södermanland, Sweden and were maintained in tanks with aerated running water at 10°C. Only intermolt crayfish were used in this study.
PCR Amplification and cDNA Cloning of Crayfish mas-like Protein-Two degenerate oligonucleotides, 5Ј-TGGGTIGTIACIGCIGCICAYT-G-3Ј and 5Ј-ANIGGICCICCI(G/C)(T/A)NTCICC-3Ј (where N is any nucleoside), were designed from the consensus amino acid sequences WVVTAAHC and GDSGGPL in serine proteinases. Either a hemocyte cDNA library (14), 1 l of 10 7 plaque-forming units/l, or a hemocyte first-strand cDNA was used as DNA template for PCR in 50 l of total volume containing 10 mM Tris-HCl, pH 8 1 min at 72°C for 2 cycles; 1 min at 95°C, 1 min at 50°C, 1 min at 72°C for 28 cycles; and 7 min at 72°C. The PCR products were subcloned into the EcoRV site of pBluescript II KS(ϩ) (Stratagene). The cDNA library from crayfish hemocytes was screened using the initial PCR products, which had been labeled with 32 P by random priming using the Megaprime labeling kit (Amersham Pharmacia Biotech). Sequencing was performed in both directions by the dideoxy chain termination method using T7 sequencing mixes (Amersham Pharmacia Biotech). The cDNA sequence was analyzed with the MacVector 4.1.4 software (Eastman Kodak Co.). The deduced amino acid sequence was analyzed using the BLAST search program (National Center for Biotechnology International, Bethesda, MD).
Preparation of First-strand cDNA-Total RNA was isolated from the hemocytes by the acid guanidinium method (15). Poly(A ϩ ) mRNA was purified according to the protocol of the Poly(A)Ttract mRNA isolation system (Promega). First-strand cDNA was synthesized from the hemocyte poly(A ϩ ) mRNA using the first-strand cDNA synthesis kit (Amersham Pharmacia Biotech).
Northern Blot Analysis-Total RNA from hemocytes or hepatopancreas was run on a 1% agarose gel in the presence of formaldehyde (16) and transferred to a nylon membrane (Hybond-N, Amersham Pharmacia Biotech) by capillary blotting overnight. For hybridization, 10 Ci of a 32 P-labeled gene fragment was used in a hybridization solution containing 5ϫ SSPE (20ϫ SSPE is 3.6 M NaCl, 0.2 M sodium phosphate, and 0.02 M EDTA, pH 7.7), 0.1% (w/v) bovine serum albumin, 0.1% (w/v) Ficoll (Amersham Pharmacia Biotech), 0.1% (w/v) sodium dodecyl sulfate, and 100 g/ml salmon sperm DNA. The samples were hybridized overnight at 65°C and then washed three times for 20 min with 0.2ϫ SSPE and 0.1% SDS at 65°C. After drying, the filter was subjected to autoradiography.
Antibodies-A synthetic peptide CFTPQDLRVRWVSGRSTS corresponding to a part of the "catalytic" domain of crayfish mas-like protein was synthesized and then coupled to ovalbumin (Sigma) using sulfo-MBS (m-maleimidobenzoyl-N-hydroxysulfosuccinimide ester) (Calbiochem) as a coupling agent and used for production of a rabbit antiserum. The amount of peptide in each injection was 0.5 mg. Antibody was affinity-purified on a column containing the synthetic peptide coupled to CNBr-activated Sepharose 4B (Amersham Pharmacia Biotech). Affinity-purified rabbit antibodies against peroxinectin have been described (17). As a control, affinity-purified rabbit antibodies against a human ␤ 1 integrin cytoplasmic peptide (18) were used.
Protein Purification-Hemocyte lysate supernatant (HLS) was prepared by collecting crayfish hemolymph (blood) in 10 mM sodium cacodylate, 0.1 M CaCl 2 , 0.25 M sucrose, pH 7.0, followed by a centrifugation at 4°C and 800 ϫ g, and the cells were homogenized in 10 mM sodium cacodylate, 0.1 M CaCl 2 , pH 7.0; this preparation was called HLS1. Alternatively, anticoagulant (0.14 M NaCl, 0.1 M glucose, 26 mM citric acid, 30 mM trisodium citrate, 10 mM EDTA, pH 4.6 (19)) was used to collect the blood, and the cells were homogenized in 0.15 M NaCl and 2 mM EDTA to yield HLS2. HLS1 was added to an anti-mas-like protein antibody column (prepared by coupling 1.5 mg of affinity-purified antibodies to 0.5 ml of CNBr-Sepharose; the apparent coupling efficiency was 93%), previously equilibrated with TBS; the column was washed extensively with TBS, and the mas-like protein was eluted with 0.1 M glycine-HCl, pH 2.5, in 0.5-ml fractions, which were immediately neutralized with 0.1 volume of 1 M Tris-HCl, pH 8.0. Peroxinectin was isolated by cation exchange chromatography as described in Johansson and Söderhä ll (17).
SDS-PAGE and Immunoblotting-SDS-PAGE was performed in 10% polyacrylamide gels. Reduction was achieved by boiling the samples for 3 min in the presence of 5 mg/ml dithiothreitol. The gel was stained with Coomassie Brilliant Blue. For immunoblotting, the proteins were transferred to nitrocellulose at 0.21 A for 70 min in 25 mM Tris, 192 mM glycine. The filter was blocked for 1 h in TBS containing 3% BSA, incubated with 10 g/ml affinity-purified crayfish anti-mas-like protein antibodies in TBS-3% BSA for 1 h, washed 3 ϫ 10 min with TBS-0.1% Tween 20, then incubated with horseradish peroxidase-conjugated goat anti-rabbit IgG (Sigma) for 1 h, washed as before, washed for 3 ϫ 10 min with TBS, and finally developed in a mixture of 20 ml of 0.03% H 2 O 2 in TBS and 6 ml of 3 mg/ml 4-chloronaphthol in methanol.
Cell Adhesion Assay-Cell adhesion was assayed essentially as described (17). All glassware was rendered pyrogen-free by incubation at 180°C for 4 h. Glass coverslips (No 1.5, 22 ϫ 22 mm, Chance Proper, Ltd., Warley, UK) were placed in Falcon 6-well tissue culture plates (Becton Dickinson and Co., Franklin Lakes, NJ) and coated with 100 l of the sample to be tested at 20°C for 1 h. After coating, the coverslips were washed with filter-sterilized and autoclaved water three times, dried at 40°C, blocked with 100 l of 1% BSA for 5 min, and finally washed and dried again as before. Negative control coverslips received only BSA.
An isolated population of crayfish granular hemocytes was obtained by centrifugation in a density gradient of 70% Percoll in 0.15 M NaCl as described (19); this cell population consisted entirely of granular cells, and at least 90% of the cells were viable as judged by trypan blue exclusion. The cells were diluted 1:2 with 0.15 M NaCl, and the cell density of the diluted suspension was determined to generally be around 1.5 ϫ 10 4 cells/ml. The cells were added to the coverslips together with CaCl 2 (final concentration 10 mM). Each coverslip was overlaid with a total volume of 200 l.
After incubation for 1 h at 20°C, the coverslips were washed with crayfish saline (0.2 M NaCl, 5, 4 mM KCl, 10 mM CaCl 2 , 2.6 mM MgCl 2 , and 2 mM NaHCO 3 , pH 6.8) and fixed in 3.7% formaldehyde in crayfish saline, pH 6.8. The percentage of attached cells was assessed by counting the cells at 200ϫ magnification in an inverted microscope, covering at least 10% of the area initially covered by the cell suspension.
In one set of experiments, one volume of 6 g/ml mas-like protein or one volume of 2 g/ml peroxinectin was preincubated with one volume of 50 g/ml affinity-purified antibody at 20°C for 1 h, the mixture was clarified by centrifugation, and then the supernatant was used to coat the glass coverslips.

Cloning and Sequence Analysis of a mas-like Protein-
The catalytic domain of serine proteinases contains a highly conserved region that has been successfully amplified by the PCR method from a variety of species and tissues. During screening for hemocyte serine proteinases using degenerate primers designed from the signature sequence around the His and Ser active-site region, four independent clones were obtained. The presence of the characteristic catalytic His, Asp, Ser residues in three of these clones (data not shown) suggests that they are members of the serine proteinase superfamily, but in one of the clones derived from a hemocyte first-strand cDNA, the catalytic serine residue was found to be replaced by a glycine. Three of the serine proteinase clones are similar to digestive trypsins, whereas the fourth is similar to Drosophila melanogaster mas (10), a secreted serine proteinase-like protein in which the catalytic serine residue is replaced by a glycine.
To obtain a full-length clone, the cDNA clone, which was similar to D. melanogaster mas, was 32 P-labeled and used to screen the crayfish hemocyte cDNA library. From several overlapping positive clones, a cDNA encompassing 3277 base pairs, which if the second methionine in the putative open reading frame is assigned as start codon, will give rise to a protein of 978 amino acid residues with a predicted mass of 98.8 kDa. The crayfish mas-like protein has two domains, an N-terminal domain and a catalytic domain. The hydrophobic first 17 amino acids of the N-terminal end of the protein is probably a signal peptide sequence with a putative signal peptidase cleavage site between Ala-17 and Ala-18 (20) (Fig. 1a). Seven repeats of a putative disulfide-knotted motif and a region of seven repeats of a glycine-rich sequence are present in the N-terminal domain of the protein (Fig. 1a). The catalytic domain, Ile-714 to Lys-978, is characteristic to the corresponding region in other serine proteinases (Fig. 1b), and the presence of a glycine residue instead of a serine residue in the catalytic site was confirmed in two partially overlapping clones, indicating that this glycine residue is not a PCR artifact. In the 3Ј-untranslated region, a putative polyadenylation signal, AATAAA, occurs 13 base pairs upstream of the poly(A) tail. A putative cleavage site for proteolytic activation of this proenzyme occurs between Arg-713 and Ile-714, although this site does not contain the typical Ile-Val-Gly-Gly motif of serine proteinases. The Anopheles gambiae ispl5 (11) and Tachypleus tridentatus factor D (6) also lack such a motif. The predicted molecular mass of the catalytic domain is 29.8 kDa. The alignment (Fig. 1b) shows that this putative mas-like protein is similar to A. gambiae ispl5 (11) and factor D, factor B, and the proclotting enzyme of T. tridentatus (6, 21, 22) as well as D. melanogaster mas (10). If the N-terminal domain of the protein is included in the alignment, there is no significant homology seen. The residues lining the substrate binding pocket, which determines the substrate specificity of active serine proteinases, are present in crayfish mas-like protein and are typical for trypsin-like enzymes. The six cysteine residues, which may form three disulfide bridges in the crayfish mas-like protein, are conserved in most serine proteinases. A. gambiae ispl5, T. tridentatus factor D, and D. melanogaster mas also have amino acid substitutions in the active site of the molecules. A schematic comparison of the main structural features of crayfish mas-like protein with A. gambiae ispl5 and D. melanogaster mas (Fig. 1c) shows a modified serine proteinase domain, several disulfide-knotted motifs present in the N-terminal domain, and a repeated glycine-rich region only present in crayfish mas-like protein.
Expression of the mas-like Protein mRNA in Hemocytes-Northern blot analyses indicated that the crayfish mas-like protein mRNA was expressed in the hemocytes and not in the hepatopancreas, which is the major site for the production of digestive enzymes (data not shown). This mRNA is expressed in adult crayfish in contrast to the mRNA expression of D. melanogaster mas, which is only detected during embryonic, larval, and pupal development but not in the adult.
Immunoblotting and Purification of the mas-like Protein-Immunoblotting of a hemocyte lysate supernatant with the affinity-purified antibodies against a synthetic peptide made adjacent to a sequence in the C-terminal serine proteinase-like domain showed a single band of 27 kDa (Fig. 2), likely a processed form of the mas-like protein. However, if the blood was collected in a citrate-EDTA anticoagulant and the blood cells were homogenized in the presence of EDTA, only one band of 150 kDa reacted with the antibodies (Fig. 2). This mass is higher than predicted from the open reading frame, i.e. 98.8 kDa, and it is therefore possible that other attached groups such as carbohydrates may contribute to the size as estimated from SDS-PAGE. The open reading frame contains two putative N-glycosylation sites. No mas-like protein was detected in the plasma (data not shown). The 27-kDa protein could be purified by immunoaffinity chromatography (Fig. 3) using the anti-mas-like protein antibodies. It had approximately the same apparent molecular mass under both reducing and nonreducing conditions (data not shown). An estimated amount of 3 g was obtained from 150 crayfish (corresponding to about 0.3 g of HLS). The N-terminal of the 27-kDa protein was determined by amino acid sequencing to be Ile-Lys-Asn-Asn-Asp-Leu-Leu-Tyr-Tyr-Gln-Thr-His-Phe-Ala-Glu corresponding to the amino acids 714 -728 (see Fig. 1b) in the putative open reading frame, indicating that this cDNA clone is authentic and that the 27-kDa protein is the C-terminal part of the mas-like protein. Despite repeated attempts, we could not isolate the 150-kDa protein from HLS2 using the immunoaffinity column.
Cell Adhesion Activity of the 27 kDa mas-like Polypeptide- Granular blood cells from crayfish adhered to the purified 27-kDa mas-like protein at relatively low coating concentrations. Halfmaximal adhesion was obtained at about 2 g/ml (Fig. 4), and specific adhesion over control was achieved above 0.1 g/ml. Cell adhesion to the mas-like protein was specifically inhibited by affinity-purified anti-mas-like protein antibodies (Fig. 5). The 76-kDa protein peroxinectin is the only protein supporting adhesion described previously from these cells (17,23); in a parallel experiment with the same cell preparations, peroxinectin gave half-maximal adhesion at less than 0.3 g/ml (Fig. 4). Affinitypurified anti-peroxinectin antibodies did not affect cell adhesion to the mas-like protein (Fig. 5). Conversely, the anti-mas-like protein antibodies did not influence cell adhesion to peroxinectin (Fig. 6), whereas the anti-peroxinectin antibodies inhibited this adhesion ( Fig. 6 and Ref. 17). Taken together, these results show that the crayfish mas-like protein and peroxinectin are two distinct adhesive molecules from these cells (and that there was no cross-contamination of the purified proteins). In a crude blood cell homogenate (in the presence of CaCl 2 and after preincubation with ␤-1,3-glucans), both proteins were active, since the adhesion activity of this preparation could be partially inhibited by either the anti-mas-like protein or the anti-peroxinectin antibodies (data not shown). Binding of peroxinectin to cells can be detected by immunofluorescence. This binding was not affected by preincubation with the mas-like protein, indicating that the mas-like protein does not bind to the same cell membrane site as peroxinectin.

DISCUSSION
Crayfish mas-like protein is unlikely to possess enzyme activity because of a substitution of an essential active serine residue, but it shares common structural features with the catalytic domains of serine proteinases, suggesting that the protein can adopt a similar conformation as that of normal serine proteinases. Human haptoglobulin heavy chain (24), hepatocyte growth factor (7), bovine protein Z (25), fruit fly mas (10), horseshoe crab factor D (6), and mosquito ispl5 (11) are examples of serine proteinase homologues lacking proteolytic activity due to the absence of a critical residue(s) in the catalytic site. The modified proteinase domain lacking enzymatic activity has been suggested to mediate protein-protein interactions or to act as an antagonist molecule of serine proteinases to regulate and control their enzymatic activity (10).
As suggested to be the case in several arthropod serine proteinases, the disulfide-knotted motif within the N-terminal domain may play a role in regulating the processing of a proenzyme to the active enzyme (26). Thus the knot has been suggested as a recognition site for the activation of the proenzyme. Seven repeats of a putative disulfide-knotted motif are present in the N-terminal domain of the crayfish mas-like protein. In these motifs six cysteine residues assigned to form three intramolecular disulfide bonds in the T. tridentatus proclotting enzyme (21) are conserved in crayfish mas-like protein and in other arthropod serine proteinase proproteins including T. tridentatus factor B (22), D. melanogaster easter, stubbled-stubbloid gene, mas (10,27,28), and A. gambiae ispl5 (11) (Fig. 7a). The sequence of the disulfide-knotted motif also shows similarity to that of T. tridentatus big defensin (29) (Fig. 7a), an antibacterial protein, suggesting that this motif may act as an antimicrobial substance perhaps after being released upon zymogen activation (30). The biological significance of the presence of several copies of this motif in D. melanogaster mas (5 . Specific inhibition of cell adhesion to the crayfish maslike protein by affinity-purified anti-mas-like protein antibodies. Isolated crayfish granular blood cells were added to glass coverslips previously coated with a mixture of the mas-like protein (final concentration 3 g/ml) and affinity-purified anti-mas-like protein antibodies (ab) (anti-mas), anti-peroxinectin antibodies (anti-pxn), or control antibodies (ctrl ab) (final antibody concentration 25 g/ml) as described under "Experimental Procedures." The experiment was performed three times. copies), A. gambiae ispl5 (2 copies), and crayfish mas-like protein (7 copies) is, however, unknown, but they may have an additional role beyond participating in the activation of a zymogen.
The repeated region has 7 repeats of a 31-amino acid sequence that contains a high number of glycine residues (Fig.  7b). No significant similarity to these repeats was found; however, it is worth noticing that repeats of several amino acids, principally serine and threonine, occur in the N-terminal domain of T. tridentatus proclotting protein as well as that of Drosophila mas, easter, and stubbled-stubbloid gene. The role of the repeated glycine-rich region of mas-like protein clearly needs to be studied.
The affinity-purified antibodies recognize only one protein in the hemocyte lysate supernatant, indicating that the cloned cDNA corresponds to the protein recognized by antibodies. The presence of a 27-kDa band in HLS1 and a 150-kDa band in HLS2 by immunoblot analyses suggests that the proenzyme is cleaved into a detectable polypeptide of 27 kDa and, as many other serine proteinases, is activated by processing. The proprotein may be protected from proteolytic cleavage in the HLS2 preparation, since components of the proPO system, e.g. proteinases, are active in HLS1 in contrast to those in HLS2 that contains EDTA. The 27-kDa protein, recognized by the affinitypurified antibodies against a synthetic peptide positioned in the C-terminal domain, is similar to the estimated mass of the C-terminal domain, 29.8 kDa, indicating that the serine proteinase domain is released upon cleavage of the proprotein, possibly by a trypsin-like activity produced upon activation of the proPO system. The affinity-purified protein from HLS1 also has a mass of 27 kDa. The purified 27-kDa protein could support adhesion of crayfish granular hemocytes in a dose-dependent manner. This cell adhesion could be specifically inhibited by affinity-purified anti-mas-like protein antibodies but not by anti-peroxinectin antibodies. This shows that these two crayfish cell adhesion molecules, the mas-like protein and peroxinectin, are distinct.
Using immunofluorescence, no binding of the 27-kDa maslike protein to fixed suspended cells could be detected. This may, however, not be entirely surprising, since binding of soluble adhesive ligands to suspended cells via adhesion receptors, is usually of low affinity and difficult to detect, despite the fact that these receptors, by clustering, mediate high avidity adhesion to immobilized substrata (31)(32)(33). In contrast, the other studied cell adhesion protein from crayfish blood cells, peroxinectin (17,23) can be detected to bind to suspended cells by the use of immunofluorescence (34). The peroxinectin binds to the suspended cells through a cell-surface superoxide dismutase (34). It is unlikely that that the mas-like protein interacts with the dismutase since preincubation of the suspended cells with mas-like protein had no effect on the binding of peroxinectin to them. The peroxinectin-mediated adhesion of the blood cells was suggested to involve an integrin receptor that may bind the peroxinectin directly through its putative integrin binding sequence, KGD (23). Indirect support for this idea comes from the recent finding that human myeloperoxidase, a homologue of peroxinectin, mediates cell adhesion via the ␣ M ␤ 2 integrin (35). Alternatively, another KGD site present in the dismutase may mediate binding of the peroxinectindismutase complex to an integrin receptor of the adhesive blood cell. Recently, an integrin, which is one candidate receptor for binding peroxinectin or the peroxinectin-dismutase complex, was isolated from crayfish blood cells, and its ␤ subunit was identified and cloned (36).
Catalytically inactive serine proteinase-like domains may function as integrin ligands in both invertebrates and vertebrates. In humans, azurocidin and haptoglobin, in which the catalytic serine is replaced with other amino acids, have been reported to bind the ␣ M ␤ 2 integrin (37,38). However, these proteins have not yet been shown to promote cell adhesion. The serine proteinase-like domain in the C-terminal part of masquerade, which has been demonstrated to be present at muscle attachment sites (10), may directly mediate cell adhesion by binding a muscle cell receptor. The mechanism for the interaction between the crayfish mas-like protein and blood cells is unknown; the findings reported here suggest, however, that the modified catalytic domain of this protein mediates cell adhesion.
Acknowledgment-We thank Anbar Khodabandeh for excellent technical assistance.
FIG. 6. Specific inhibition of cell adhesion to peroxinectin by affinity-purified anti-peroxinectin antibodies. Isolated crayfish granular blood cells were added to glass coverslips previously coated with a mixture of peroxinectin (final concentration 1 g/ml) and affinity-purified anti-mas-like protein antibodies (ab) (anti-mas), anti-peroxinectin antibodies (anti-pxn), or control antibodies (ctrl ab) (final concentration 25 g/ml) as described under "Experimental Procedures." The experiment was performed three times.