Multiplicity, Structures, and Endocrine and Exocrine Natures of Eel Fucose-binding Lectins*

Lectins, a group of proteins that bind to cell surface carbohydrates and play important roles in innate immunity, are widely used experimentally to distinguish cell types and to induce cell proliferation. Eel serum lectins have been useful as anti-H hemagglutinins and also in lectin histochemistry as fucose-binding lectins (fucolectins), but their structures have not been determined. Here we report the primary structures and the sites of synthesis of eel fucolectins. Eel serum fucolectins were separated by two-dimensional gel electrophoresis and sequenced. cDNA cloning, based on the amino acid sequence information, and Northern blot analysis indicated that 1) the fucose-binding lectins are secretory proteins and have unique structures among the lectins, exhibiting only weak similarities to frog pentraxin, horseshoe crab tachylectin-4, and fly fw protein; 2) there are at least seven closely related members; and 3) their messages are abundantly expressed in the liver and in significant levels in the gill and intestine. The lectin-producing hepatic cells were identified by immunostaining; in the gill, exocrine mucous cells were stained, suggesting that serum fucolectins derive from the liver. Using primary culture of eel hepatocytes, the message levels were shown to be increased by lipopolysaccharide, suggesting a role for fucolectins in host defense. SDS-polyacrylamide gel electrophoresis analysis showed that eel fucolectins have a SDS-resistant tetrameric structure consisting of two disulfide-linked dimers.

Lectins, apart from their biological roles in host defense (1), have long been useful research tools that recognize specific sugar side chains of glycoconjugates including bacterial lipopolysaccharides and cell surface glycoproteins (2,3). For example, anti-H hemagglutinins or fucose-specific lectins (fucolectins) have been found in eel serum and extracts of certain plant seeds and widely used for blood typing (4,5) and lectin histochemistry (6,7) aimed at discriminating cells based on their expression of sugar residues that differ depending on their developmental stages, differentiation, and malignancy. Reflecting these needs, a variety of lectins with different specificities are now commercially available including the above mentioned fucolectins (i.e. eel serum lectins, asparagus seed lectins, and gorsethree seed lectins). Concerning the eel serum lectins, however, their structures have not been determined. Determination of their cellular origin is also crucial for understanding their physiological significance.
While attempting to identify proteins specifically expressed in either freshwater or seawater eels, we cloned a group of 20-kDa proteins from seawater eels and named them osmoregulins, but later they turned out to be fucose-binding lectins present in both freshwater and seawater eels and considered to play important roles against microbial invasion. A literature survey indicated that, although eel fucolectins have a long history of research and practical applications (4, 8 -16), their primary structures have not been determined yet. We therefore decided to isolate corresponding cDNA clones. Characterization of the clones indicated that the fucolectins are a group of proteins with at least seven members. This heterogeneity or multiplicity may have evolved to provide a means for effective defense against a wide variety of pathogenic bacteria and appears to be consistent with their primary functions in innate immunity (1). Northern blot analysis and immunohistochemistry suggested that 1) eel serum fucolectins are of hepatic origin and 2) gill mucus cells also produce fucolectins in an exocrine fashion. A data base search revealed the presence of an interesting molecule in Drosophila, a chimeric protein of fucolectin and selectin.

EXPERIMENTAL PROCEDURES
Materials-Japanese eels (Anguilla japonica) were obtained from a dealer and blood was collected from the ventral aorta or caudal vein into tubes containing 0.5 M EDTA (final concentration, 10 mM), which was then centrifuged to obtain plasma. Immobilized pH gradient (IPG) 1 polyacrylamide gel or Immobiline DryStrip (13 cm) and IPG buffer (pH 4 -7) were obtained from Amersham Pharmacia Biotech; polyvinylidene difluoride (PVDF) membrane was from Millipore; lysyl endopeptidase Achromobacter protease I was from Takara, Kyoto, Japan; L-fucose agarose was from Seikagaku Corporation, Tokyo, Japan; Radiochemicals were from Amersham Pharmacia Biotech.
Two-dimensional Electrophoresis-Eel plasma proteins were precipitated with three volumes of ice-cold 10% trichloroacetic acid in acetone containing 0.07% 2-mercaptoethanol. After incubation at Ϫ20°C for 1 h, pellets were collected by centrifugation and washed with cold acetone containing 0.07% 2-mercaptoethanol. Air-dried pellets were dissolved in rehydration solution containing 8 M urea, 2% CHAPS, 2% IPG buffer, bromphenol blue, and 65 mM dithiothreitol. An IPG strip was rehydrated overnight at room temperature in 250 l of rehydration solution containing plasma proteins extracted from 100 l of eel plasma. It was then subjected to isoelectric focusing in a Multiphor II system with an EPS 3500 XL electrophoresis power supply (Amersham Pharmacia Biotech). The gel was run with the following voltage program: pre-focusing at 300 V for 1 min and focusing at 3500 V for 16 h. After soaking in equilibration buffer (1% SDS, 35% glycerol, 1% dithiothreitol, and 50 mM Tris-HCl, pH 6.8) for 10 min, the gel strip was loaded on a second-dimension 12.5% (w/v) SDS-polyacrylamide gel.
Amino Acid Sequencing-The amino acid sequences of proteasedigested fragments of the protein were determined by the method of Iwamatsu et al. (17). Following two-dimensional electrophoresis on SDS-polyacrylamide gel, separated proteins were transferred to a PVDF membrane and stained with Coomassie Brilliant Blue. The spots corresponding to the proteins of interest were cut, reduced, and digested with the lysyl endopeptidase Achromobacter protease I (Takara, Kyoto, Japan), and the fragments released (AP fragments) were collected. The peptides remaining on the membrane were further incubated with trypsin (Takara), and the digests (TP fragments) were collected. The proteolytic fragments generated were separated by reversed phase high performance liquid chromatography on a Bondasphere column (5-m C18; 100 Å, 3.9 ϫ 150 mm, Waters) using a linear gradient generated between solvents A (0.05% trifluoroacetic acid in water) and B (0.02% trifluoroacetic acid in 2-propanol/acetonitrile (7:3, v/v)) in 30 min at a flow rate of 0.5 ml/min and applied to a gas-phase peptide sequencer (PPSQ-21, Shimadzu) for their amino acid sequences. For the determination of the N-terminal sequences, proteins were sequenced directly from the PVDF blots.
Isolation of a Probe for Library Screening-To isolate a partial cDNA probe, PCR amplification was performed using eel gill and liver cDNA as templates using ExTaq DNA polymerase (Takara) and the following oligonucleotide primers: 5Ј-CCNAAYMGNTAYATHCARGARAAYG-T-3Ј (S1, sense primer corresponding to the N-terminal sequence), and 5Ј-TTNGGNARRTANACNGTNAC-3Ј (AS2, antisense probe corresponding to tryptic fragment 2 (TP-2) in Fig. 3). The PCR was run for 35 cycles by repeating denaturation at 94°C for 30 s, annealing at 45°C for 1 min, and polymerization at 72°C for 2 min. A specific PCR product of 360 bp was isolated by agarose gel electrophoresis, purified, and subcloned into the EcoRV site of pBluescript II (Stratagene) and sequenced.
Construction and Screening of cDNA Library-The eel gill cDNA library in ZAP II (Stratagene) was prepared as described (18). For construction of a liver cDNA library, total RNA was prepared from eel livers by the acid guanidinium thiocyanate/phenol/chloroform method and mRNA was isolated using an oligo(dT)-cellulose column (Amersham Pharmacia Biotech). An eel liver cDNA library was constructed in ZAP II phage vector (Stratagene) using an oligo(dT) primer and other reagents from the SuperScript Choice System (Life Technologies, Inc.). Approximately 2 ϫ 10 5 recombinant phage were obtained. The eel gill and liver cDNA libraries were plated out at a density of 3 ϫ 10 4 plaque-forming units/150-mm plate. Plaques were lifted onto nitrocellulose filters (Shleicher & Schuell), and the filters were prehybridized in 5ϫ SSPE (SSPE: 0.15 M NaCl, 1 mM EDTA, and 10 mM phosphate, pH 7.4), 50% formaldehyde, 5ϫ Denhardt's solution (Denhardt's: 0.1% each Ficoll, polyvinylpyrrolidone, and bovine serum albumin), and 0.5% SDS for 2 h at 42°C. The above cDNA probe was labeled with [ 32 P]CTP (3000 Ci/mmol) using random primers. Hybridization was performed for 16 h at 42°C in the same solution. Ten and five positive clones were isolated from the gill and liver cDNA libraries, respectively.
Northern Blot Analysis-Total RNA was isolated from various tissues of eel by the acid guanidinium thiocyanate/phenol/chloroform method. Total RNA (5 g or 20 g) was electrophoresed in 1% agaroseformaldehyde gel and transferred to Magna MT nylon membrane (Micron Separations) by capillary blotting overnight. Eel fucolectin and ␤-actin cDNAs were labeled with [␣-32 P]CTP using random primers and hybridized in 6ϫ SSPE, 50% formaldehyde, 5ϫ Denhardt's solution, and 1% SDS for 16 h at 42°C. After hybridization, membrane was washed twice with 2ϫ SSC (SSC: 0.15 M NaCl, 15 mM sodium citrate, pH 7.0) and 0.1% SDS at room temperature, washed with 1ϫ SSC and 0.1% SDS for 1 h at 55°C and twice with 0.1ϫ SSC and 0.1% SDS for 1 h at 55°C, and exposed to imaging plate (Fuji Film) in a cassette for 1 week. The result was analyzed using a Fujix BAS2000 bio-imaging analyzer (Fuji Film).
Total Southern Analysis-Five micrograms of eel genomic DNA were digested with restriction enzymes (BamHI, BglII, EcoRI, EcoRV, Hin-dIII, KpnI, and PstI), separated in a 1.0% agarose gel, and transferred to the Hybond N ϩ membrane (Amersham Pharmacia Biotech). The membrane was prehybridized in PerfectHyb hybridization solution (Toyobo) at 68°C for 1 h. The full-length eel fucolectin cDNA was labeled with [␣-32 P]CTP using random primers and hybridized in the same solution at 68°C for 16 h. The membrane was washed twice at 68°C for 5 min with 2ϫ SSC and 0.1% SDS and once at 68°C for 1 h with 0.5ϫ SSC and 0.1% SDS. After washing, the membrane was exposed to imaging plate in a cassette for 1 day. The result was analyzed using a Fujix BAS2000 bio-imaging analyzer.
Purification of Eel Serum Fucolectin-Eel plasma (10 ml) was diluted 10 times with TBS (0.15 M NaCl, and 10 mM Tris-HCl, pH 7.5), and filtered through a 0.22-m filter (Millex-GV, Millipore). The diluted plasma was applied to a column containing L-fucose agarose (5 ml), and washed with 10 volumes of TBS. The bound eel serum fucolectin was eluted with 50 mM L-fucose in TBS and dialyzed against three changes of 150 mM NaCl.
Preparation of Antiserum-Antiserum to eel fucolectin was prepared in a Japanese White rabbit. The purified eel serum fucolectin was emulsified with TiterMax Gold adjuvant (CytRx Corp.) and injected intramuscularly three times.
Western Blot Analysis-Eel gill was homogenized in five volumes of TBS and aliquots were dissolved in Laemmli sample buffer (2% SDS, 10% glycerol, 0.1% bromphenol blue, and 50 mM Tris-HCl, pH 6.8); eel plasma was diluted 1:100 with Laemmli sample buffer. The samples were separated by SDS-PAGE using 12.5% acrylamide gel and electroblotted to a PVDF membrane. After blocking in 150 mM NaCl, 0.05% Tween 20, and 10 mM Tris-HCl, pH 8.0, containing 5% nonfat milk for 30 min at room temperature, the membranes were incubated with anti-fucolectin antiserum, preimmune serum, or preabsorbed anti-  ) from various eel tissues were electrophoresed in a 1.0% agarose-formaldehyde gel, transferred to a nylon membrane, and hybridized with eel fucolectin and ␤-actin cDNA probes labeled with 32 P. Eel ␤-actin transcripts were present as ϳ1.6-kb (ubiquitous) and ϳ1.3-kb (specific for muscle; atrium and ventricle) species. eFL, eel fucolectin mRNA; nt, nucleotides; int., intestine. serum at 1:1000 dilution overnight at 4°C. Preabsorption was carried out by incubating 10 l of the antiserum with 40 g of purified fucolectin in 100 l of phosphate-buffered saline overnight at 4°C. The immune complexes on the filter were then reacted with alkaline phosphatase-conjugated goat anti-rabbit IgG at 1:3000 dilution for 1 h at room temperature. The bound secondary antibody was detected with 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium chloride as chromogenic substrate.
LPS Treatment and Blood Sampling-Eels were acclimated in a 1-ton freshwater tank for more than 1 week before use. Seawater eels were acclimated in a 0.5-ton seawater tank for more than 2 weeks. Water in the tank was continuously filtered, aerated, and thermoregulated at 18 Ϯ 0.5°C. Eels were not fed after purchase. They weighed 190 Ϯ 20 g (n ϭ 5) at the time of experiments. Eels were cannulated into the ventral aorta for injections and blood sampling. After anesthesia in 0.1% tricaine methanesulfonate for 10 min, a polyethylene tube (outer diameter, 0.8 mm) was inserted into the ventral aorta as reported previously (19). The catheters in the aorta were connected to plastic syringes filled with 0.9% NaCl solution. After more than 18 h post-operatively, eels were injected with LPS (20 g/eel/0.2 ml of saline), and the catheter was flushed with 50 l of saline. A small volume of blood (150 l) was collected before and 1, 3, 6, and 10 h after injection into a chilled syringe containing 10 l/ml blood of 2Na-EDTA.
Primary Culture of Hepatocytes-Isolation and culture of eel hepatocytes were carried out according to Hayashi and Ooshiro (20) and Hayashi and Komatsu (21). Liver was excised from an anesthetized eel and perfused with Ca 2ϩ -free Ringer solution (120 mM NaCl, 4.7 mM KCl, 2.4 mM CaCl 2 , 1.25 mM MgSO 4 , and 23 mM NaHCO 3 , pH 7.4), Ringer solution containing collagenase (15 mg/50 ml) at room temperature for 30 min, and then Ca 2ϩ -and Mg 2ϩ -free Ringer solution con-taining 2 mM EDTA. The liver was minced with scissors and filtrated through a 200-mesh grid. Filtrated liver cells were collected by centrifugation at 50 ϫ g for 1.5 min, resuspended with Ringer solution, and centrifuged again. This washing procedure was repeated three times. Isolated cells were seeded on a poly L-lysine-coated dish (55 cm 2 , Iwaki) at a density of 3 ϫ 10 5 cells/cm 2 and incubated at 28°C in a humidified atmosphere containing 5% CO 2 . Medium consisted of William's E medium (Life Technologies, Inc.) containing 5% fetal bovine serum, 0.16 M insulin from bovine pancreas, 100 units/ml penicillin, and 100 mg/ml streptomycin. The medium was changed to a fresh one after 2 days and the culture was further continued for 3 days before LPS treatment.
In Vitro LPS Treatment of Hepatocytes-After the 3-day culture mentioned above, LPS (final concentration, 1 g/ml) was added to the medium, and 20 l of medium was collected every 3 h and analyzed for secreted fucolectin levels by Western blotting. At 6-h intervals over a 24-h period, cells were harvested and analyzed for fucolectin mRNA levels by Northern blotting.
Immunohistochemistry-Eel gill and liver were fixed in 0.1 M phosphate-buffered saline, pH 7.4, containing 4% (w/v) paraformaldehyde, and frozen in Tissue-Tek OCT compound. Frozen sections were cut in a cryostat, washed in TBS containing 1 mM CaCl 2 , and fixed in methanol containing 3% H 2 O 2 to inactivate endogenous peroxidase for 30 min at room temperature and washed again in TBS containing 1 mM CaCl 2 . The sections were incubated with 2% (v/v) normal goat serum for 30 min at room temperature, then with anti-eel fucolectin antiserum, preimmune serum, or preabsorbed antiserum at 1:1000 (liver) or 1:5000 (gill) dilution overnight at room temperature, and treated with biotinylated goat anti-rabbit IgG and the avidin-biotin-conjugated peroxidase (Vectastain Elite ABC kit; Vector Laboratories) for 1 h each at room tem-

Separation by Two-dimensional Gel Electrophoresis and Partial Amino Acid Sequencing of Eel Fucolectins-When plasma
samples of a freshwater eel and a seawater eel were analyzed by SDS-PAGE, a band of Ϫ20 kDa was found to vary in its intensity between the samples (Fig. 1A). Although later analysis using an increased number of samples indicated that this variance is due to individual difference rather than the osmo-larity difference, we were interested in its nature and decided to determine the N-terminal amino acid sequence. Sequencing of the 20-kDa band, after transferring it to PVDF membrane, gave multiple peaks of phenylthiohydantoin-amino acids in each Edman degradation cycle, indicating that it represents a mixture of proteins of the same size. We therefore performed two-dimensional gel electrophoresis to separate the constituents of the 20-kDa band according to their isoelectric points (Fig. 1B) and sequenced three major spots of 20 kDa. Two of them (spots a and b) gave the same N-terminal sequence (AD- Gaps are introduced to obtain maximal sequence alignment. Numbers to the left and right refer to the first and last amino acid residues on the lines. eFL-1-3 are isolated from the liver and eFL-4 -7, from the gill. A data base search also identified a partial sequence of a rainbow trout homologue with a 63% similarity (T23111) in the GenBank Expressed Sequence Tag data base. The BLAST and MOTIF programs were used to search the GenBank data base at the NCBI web site. VPNRYIQENVAVRGKATQ, Fig. 3) and the third one (spot c), a slightly different sequence (ADIPEGYIQENVALRGRATQ). We further obtained internal amino acid sequences by digesting spot a with trypsin or lysyl-endospecific Achromobacter protease I, isolating the proteolytic fragments by high performance liquid chromatography, and sequencing them (Fig. 3,

AP-1, TP-1, and TP-2).
PCR Amplification of cDNA Probe and Northern Blot Analysis-Several combinations of PCR primers were designed based on the above amino acid sequence information and used for amplification of cDNA encoding the 20-kDa protein. One combination (primers AS2 and S1 described under "Experimental Procedures") yielded a prominent PCR band of Ϫ360 bp from gill and liver cDNA, whose identity was confirmed by nucleotide sequencing, namely the amino acid sequence deduced contained the sequence corresponding to that of tryptic fragment 1. Fig. 2 shows Northern blots of eel total RNA prepared from various eel tissues and probed with the PCR product labeled with 32 P. A very strong signal was seen in the liver sample and weak signals, in the gill and intestine preparations. The gill message appeared to be longer (Ϫ850 nucleotides) than that of the liver (700 -850 nucleotides). We therefore decided to screen, as described below, both liver and gill cDNA libraries for full-length cDNA clones.
cDNA Cloning and Sequence Analysis-cDNA libraries were constructed from liver and gill poly(A) ϩ RNA and screened with the same probe used for the Northern analysis. Fifteen positive clones were isolated and sequenced (Figs. 3 and 4). Sequence analysis revealed the presence of seven types of clones that code for similar but distinct proteins of Ϫ180 amino acid residues (Fig. 4): three clones from the liver (eFL-1-3) and four from the gill (eFL-4 -7). Their nucleotide and deduced amino acid sequences were deposited into the DDBJ/EMBL/Gen-Bank data base (accession nos. AB037867-AB037873), one of which is shown in Fig. 3 as a typical example. The cDNA clones from the eel liver and gill were mutually exclusive, suggesting that they are encoded by a separated group of genes whose promoters are active in either liver cells or gill cells. Most of the cDNA clones were almost identical in their length of the 5Јnoncoding region but differed in that of the 3Ј-noncoding region, providing an explanation for the difference in the mRNA size seen in the Northern blot analysis (Fig. 2). The overall sequence similarities among the seven proteins are about 70% but they shared a highly conserved core sequence of about 11 amino acids (residues 98 -108, Fig. 4).
A data base search indicated that the nucleotide and amino acid sequences determined here are weakly but significantly similar to those of tachylectin-4 (22), a horseshoe crab lectin with binding specificity to O-antigen of bacterial lipopolysaccharides, to the N-terminal domain of the Xenopus homologue of pentraxin 1, XL-PXN1, which consists of the N-terminal domain of unknown function and the C-terminal pentraxin domain (23), and to the N-terminal domain of Drosophila fw protein whose C-terminal domain is similar to selectin (Fig. 4, gray boxes in lower four lines; Fig. 10). This similarity raised the possibility that the Ϫ20-kDa proteins cloned here are the eel serum lectins known as anti-H hemagglutinins (4,8,10) or fucolectins (11)(12)(13) that are now commercially available and widely used in lectin histochemistry (6, 7, 14 -16) but whose primary structures have not been determined yet. We therefore characterized the Ϫ20-kDa proteins and established their identities as eel fucolectins as described below.
Properties and Subunit Structure-To confirm the identity of the Ϫ20-kDa proteins as fucolectins, we first performed affinity chromatography of eel serum on fucose-agarose (Fig. 5A). When eel serum was applied to a fucose-agarose column, the Ϫ20-kDa proteins were specifically bound to and eluted from the column. The molecular size of the eluate was estimated to be Ϫ80 kDa by gel filtration on Superdex 75 (data not shown). On mild SDS-PAGE (i.e. without reduction and heat treatment), it migrated as a band of Ϫ80 kDa (Fig. 5B, lane 6); on heat treatment, it gave a band of Ϫ40 kDa (lane 5); and on reduction, its size was further reduced to Ϫ20 kDa (lane 4). These results are essentially identical to those reported by previous investigators (11,13) in the binding specificity to fucose and the tetrameric subunit structure consisting of two noncovalently associated dimers whose constituent monomers are linked by disulfide bonds ((20 kDa)-SS-(20 kDa)) 2 . Furthermore, the partial N-terminal sequence of eel lectin determined by Horejsi et al. (11) and Kelly (13) is very similar to one of our sequences (Fig. 4). We therefore named the Ϫ20-kDa proteins "eel fucolectin-1 to Ϫ7." A unique property we found is that the tetrameric structure of eel fucolectins is very stable and per- sists even in the presence of SDS if not heated (Fig. 5B, lane 6).
Total Southern Analysis-Total Southern blot showed multiple bands even under highly stringent conditions using fucolectin-5 cDNA as a probe (Fig. 6). The multiple bands seen in Fig. 6 are consistent with the fact that there are at least seven closely related members in the eel fucolectin family.
LPS Treatment-Serum and glandular lectins play important roles in innate immunity by directly binding to pathogenic bacteria and usually do not undergo large changes in their levels. We were, however, interested in whether the plasma levels of eel fucolectins increase during bacterial infections since we observed, at the initial stage of this study, large variations in their levels in individual eels. To determine the changes, catheters were inserted into the ventral aorta for blood sampling and LPS injection. Blood samples were collected at 0, 1, 3, 6, and 10 h after LPS treatment, and the fucolectin levels were determined by measuring the density of the 20-kDa band on SDS-PAGE. Results were, however, negative; in four eels having low or high levels of fucolectins, no significant changes were observed (data not shown).
We next determined the effects of LPS on the expression of the fucolectin genes using primary culture of eel hepatocytes since the lack of response mentioned above might be due to effective inactivation of the infused LPS by circulating fucolectins. Hepatocytes were prepared from eel livers by collagenase digestion, mechanical dispersion, and passage through a mesh. Time course of their response to LPS was followed by measuring the amounts of fucolectins accumulated in the culture medium by Western blot analysis (Fig. 7). On exposure to LPS, the production of fucolectins increased significantly (Fig. 7A). Similarly, the fucolectin message levels were also found to be increased following stimulation of hepatocytes with LPS (Fig. 7B).
Immunohistochemistry on Liver and Gill Sections-An anti-serum was raised against eel serum fucolectins and characterized by Western blotting (Fig. 8). The antiserum, termed antie-fucolectins (e for eel), recognized well not only the serum fucolectins but also the gill fucolectins probably because of the presence of highly conserved regions among the serum and gill fucolectins (Fig. 4). Fig. 9 shows the results of immunohistochemistry performed on eel liver and gill sections. In the liver, certain groups of parenchymal hepatocytes were moderately stained (Fig. 9A). In the gill, mucous cells were strongly stained (Fig. 9C). Relatively weak staining of the liver compared with the gill may suggest that a large fraction of fucolectins synthesized in the liver are constitutively secreted into the circulation. The gill cells immunostained were identified as mucous cells by the carbohydrate-specific periodic acid-Schiff staining of serial sections (Fig. 9F). DISCUSSION There are currently three major sources of fucolectins: eel serum and seeds of asparagus (Lotus tetragonolobus) and gorsethree (Ulex europaeus), among which only the gorsethree seed lectin, UEA-1, has fully been characterized in terms of the primary structure (24). In the present study, we determined the primary and subunit structures of the eel fucolectins by cDNA cloning and SDS-PAGE, and demonstrated that they are a family of proteins with a tetrameric structure. Apparently no significant sequence similarity was seen between the eel and gorsethree lectins, suggesting a multiplicity of the lectin motif for fucose. Weak but significant similarities were found to horseshoe crab tachylectin-4 (22), to the N-terminal domain of Xenopus homologue of pentraxin 1, XL-PXN1, which comprises two domains: the N-terminal domain of unknown function and C-terminal pentraxin domain (23), and to the N-terminal domain of Drosophila melanogaster furrowed (fw) protein (accession no. AE003487-48) (25) whose C-terminal domain consists of a selectin domain, 10 complement binding repeats, and a transmembrane span (26) (Fig. 10). Since pentraxin and selectins are, by themselves, carbohydrate-binding proteins, XL-PXN1 and fw protein may represent a unique family of lectins with two separate sugar-binding sites within a single polypeptide chain that might have risen by fusion of the ancestral fucolectin and pentraxin or selectin genes. The presence of related proteins in the frog, horseshoe crab, and fruit fly indicates that the lectin motif in the eel fucolectin is useful for host defense and successfully employed in certain animal species as a defense strategy against microbial invasion. Among them, eels are unique in having multiple species of fucolectins generated probably by a series of gene duplications.
LPS-stimulated nature and endocrine and exocrine natures of eel fucolectins also suggest that they serve as powerful defense agents. Northern blot analysis and immunohistochemistry indicated that eel serum fucolectins are of hepatic origin and constitutively secreted proteins, whereas the gill fucolectins are secreted from the mucous cells into the extracellular space in a regulated manner. In this context, it is quite reasonable that the serum and gill fucolectins are encoded by separate genes allowing independent control of expression, namely constitutive or regulated expression.