T w o Lectins from the Marine Sponge Halichondria okadai AN N-ACETYL-SUGAR-SPECIFIC LECTIN (HOL-I) AND AN N-ACE’IYLLACTOSAMINE-SPECIFIC LECTIN (HOL-II)*

Two lectins (HOL-I and HOL-11) were isolated from the marine sponge Halichondria okadai by affinity chromatography on a bovine submaxillary mucin (BSM)-Toyo- pearl and an acid-treated Sepharose 4B columns, respec-tively. In hemagglutination inhibition assays, GlcNAc, GalNAc, and their methyl glycosides were the most potent inhibitors among the monosaccharides tested against the HOL-I-mediated hemagglutination, suggesting that HOL-I can especially recognize the N-acetyl groups of the sugars. This N-acetyl specificity was supported by ‘H NMR analyses; the highest field-shifts of the signal of the N-acetyl group among all the signals in Me PGlcNAc were observed in ‘H NMR spectra of mix- tures of HOL-I and the sugar. Among the oligosacchari-des tested, GlcNAcpl+4(GlcNAc~l+2)Manc~1-O(CH~)~ CHs was the most potent inhibitor, and the inhibitory potency of the

Halichondria okadai is a toxic marine sponge, and potent toxins, okaidaic acid (11, and halichondrins (2, 3) were isolated from this sponge. However, there is no report about high molecular substances from the sponge. Herein the isolation and characterization of two unique lectins from the sponge are described.
* This work was supported by a Grant-in-Aid for Scientific Research on Priority Areas "Dynamic Aspects of Natural Product Chemistry" from the Ministry of Education, Science and Culture of Japan. The costs of publication of this article were defrayed in part by the payment of ment" in accordance with 18 U.S.C. Section 1734 solely to indicate this page charges. This article must therefore be hereby marked "aduertisefact.
$. To whom correspondence should be addressed.
All the other sugars for the tests were obtained from Nacalai Tesque (Japan). All the other chemicals were reagent grade.
Preparation of Afinity Adsorbent-Acid-treated Sepharose 4B was prepared by the method of Errson et al. (7). BSM' was conjugated to Amino-Toyopearl by following the instructions of the manufacturer. Isolation of HOLs-All the procedures were carried out at 4 "C. The frozen sponges (42.0 g) were defrosted at room temperature, immediately homogenized with saline (100 ml) in a blender, and extracted with stirring overnight. The resulting suspension was filtered with gauze and the filtrate was centrifuged (10,000 x g , 15 min) to remove insoluble residues. The supernatant was directly applied to a BSM-Toyopearl column (4 1.5 x 8.5 em) linked to an acid-treated Sepharose 4B (7) column (4 1.5 x 8.5 cm) equilibrated with PBS. After extensive washing, the two columns were disconnected. The fraction containing HOL-I was desorbed with 0.5 N acetic acid from the former column. After dialysis against PBS and concentration by ultrafiltration with PM-IO, the eluates were further applied to a Toyopearl 55F column (+ 1.6 x 80 cm). Purified HOL-I was obtained by dialysis against distilled water and then liophilization (1.86 mg). Purified HOL-I1 was obtained by elution with distilled water from the acid-treated Sepharose 4B affinity column and dialysis against distilled water followed by liophilization (1.98 mg).
Hemagglutination Test-Erythrocytes and Pronase-treated erythrocytes for agglutinating test were prepared as reported previously (8). Agglutination of erythrocytes and inhibition of the agglutination by sugars and glycoproteins were done in microtiter U-plates. The titer was defined as reciprocal of the end-point dilution causing hemagglutination. Inhibition was expressed as the minimum concentration of each sugar or glycoprotein required for inhibition of hemagglutination of titer 4 of the lectin.
SDS-Polyacrylamide Gel Electrophoresis-SDS-polyacrylamide gel electrophoresis was carried out by the method of Laemmli (9). Samples were heated in the presence or absence of 2-mercaptoethanol for 10 min at 100 "C. Gels were stained with Coomassie Brilliant Blue. The molecular weight standards (Pharmacia, Sweden) used were phosphoryl-' The abbreviations used are: BSM, bovine submaxillary mucin; PBS, phosphate-buffered saline. All the sugars used are of the D configuration unless otherwise stated.
Amino Acid Analysis-Amino acids were analyzed with a 470N130N 920A Derivatizer-Analyzer System (Applied Biosystems). During HCI hydrolysis, 1-dodecathiol was added as a scavenger for preventing Trp destruction.
N-terminal Amino Acid Analysis-The N-terminal amino acids of the proteins were analyzed on a model 477A pulsed liquid proteidpeptide sequencer (Applied Biosystems) equipped with a high performance liquid chromatography system (model 120A on-line phenylthiohydantoin analyzer, Applied Biosystems).
Sugar Content-Sugar was measured by the phenol-H2S04 method with reference to glucose (10).
Thermostability-Samples in PBS were heated for 30 min a t the temperatures indicated, cooled on ice, and titrated.
Efect of Metal Cations on the Lectin ActiuitpTo examine metal cation requirements of the hemagglutination by the lectins, the samples were demetalized by the method of Pandrolfino and Magnuso (11).

RESULTS AND DISCUSSION
The purification procedure was summarized in Table I. Since crude saline-extract ofH. okudui exhibited no remarkable binding specificity to any simple sugar and showed the specificity against asialo-BSM and BSM in the hemagglutination inhibition assay, BSM-Toyopearl was chosen as the affinity support. The lectin activity of the extract was largely adsorbed to the affinity column and eluted with 0.5 N acetic acid. The eluate from the column was further purified by gel filtration, and pure lectin which was named HOL-I was obtained. On the other hand, some lectin activity was recovered from the non-adsorbed fraction in the affinity chromatography in all experiments. The hemagglutination inhibition test of the non-adsorbed fraction demonstrated that the fraction did not bind to asialo-BSM and BSM but to galactose unlike the crude extract before the chromatography. Therefore, the fraction was applied to an acidtreated Sepharose 4B affinity column, and all the activity was adsorbed. Elution from the column was carried out with distilled water, and another purified lectin, HOL-11, was obtained. Gal p 1+4GlcNAc ( p 1-4GlcNA~)~ GalpljBGlcNAc Galpl+6GlcNAc Galpl4GlcNAc-pPNP Galpl-3GalNAc Galplh4GalNAc G a l p l 4 G a l N A c Galpl-6Gal-pPNP

Manp1-O(CHz)8COzCH3
N-AC-Glu N-AC-Gly N-Ac-Leu Each purified lectin was appeared as a single band in SDSpolyacrylamide gel electrophoresis (Fig. 1). HOL-I showed a single band corresponding to a molecular weight of M , = 21,000 regardless of the presence or absence of 2-mercaptoethanol. But without 2-mercaptoethanol the band was somewhat diffuse. The reason is not clear but the subunit might have some intramolecular S-S linkage(s1. The native molecular weight of the lectin was 84,000 on the basis of the result of gel filtration in PBS, suggesting that HOL-I consists of four identical subunits ofM, = 21,000. On the other hand, HOL-I1 showed a band of molecular weight 42,000 in SDS-polyacrylamide gel electrophoresis, and the estimation of the molecular weight of native HOL-I1 by gel filtration gave the same molecular weight. This result indicates that HOL-I1 exists as a monomeric molecule in its native state. A comparison of the amino acid compositions of HOL-I and HOL-I1 (Table 11) indicated that the two lectin forms are similar but differed mainly in their amount of Asx, Thr, Met, Leu, and Trp. Although 2 nmol of each protein was applied to the sequencer, N-terminal amino acid could not be detected. This result suggests that N termini of the proteins may be blocked.
In isoelectric focusing of the lectins, both lectins indicated families of bands in pH zone near 4.5 (data not shown). The carbohydrate contents of HOL-I and HOL-I1 amounted 2.0 and 7.5%, respectively. The hemagglutination of the lectins was not affected by demetalization, and addition of CaCl,, MgCl,, ZnC12, or MnC1, to the demetalized lectins did not cause any change of the activities. Thermostability of HOL-I and HOL-I1 was shown in Fig. 2. HOL-I was stable below 50 "C, but HOL-I1 was sensitive to temperature, and the activity was rapidly decreased over 40 "C. Both lectins were stable under wide range pH regions (Fig. 3).
The two lectins did not agglutinate any type of native human erythrocytes, but Pronase-treated blood cells caused agglutination by the lectins: the titer of HOL-I (100 pg/ml) was 1024 to the treated type A cells, 128 to the type B, and 512 to the type 0 ones. The titer of HOL-II(100 pg/ml) was 128 to the treated The results of hemagglutination inhibition assays of HOL-I are shown in Table 111. GlcNAc 1, GalNAc 2, and both anomers of their methyl glycosides 5-8 were the best inhibitors to the hemagglutination by HOL-I among monosaccharides tested 1-10. N-Acetylneuraminic acid 9 and N-acetylated amino acids such as N-acetylglutamine 31, N-acetylglycine 32, and N-acetylleucine 33 were also inhibitory at higher concentrations. This result suggests that HOL-I mainly recognizes Nacetyl groups. This suggestion was further supported by NMR analyses of mixtures of HOL-I and Me PGlcNAc (Figs. 4 and 5 and Table IV). In the binding equilibrium between a lectin and its specific binding sugar, resonance of the sugar may shift andlor broaden due to chemical changes between the free and bound environments (12,13). Furthermore, the binding site in the sugar to the lectin can be deduced by comparison of the chemical shift of each proton in the sugar bound to the lectin with those of free sugar (13). Fig. 4 shows 'H NMR data for Me P-GlcNAc, which is one of the best haptenic sugars of HOL-I, in the absence or presence of the lectin. The resonances of Nacetyl groups of the sugar in the presence of the lectin (Fig 4, A-F) caused the highest field shifts among all the signals in the sugar. The chemical shift changes are summarized in Table nT and Fig. 5. The shifts were dependent on the mole ratio of the sugar to the lectin; a lower mole ratio caused higher field shift.
The signals of H-2 in the sugar also showed similar but smaller changes than N-acetyl groups. The chemical shifts of the other protons, however, were almost the same as those in the absence of the protein. Almost the same results were obtained in the case of GalNAc (data not shown). These data indicate that HOL-I can specifically bind to the N-acetyl group in GlcNAc or GalNAc, and the binding to the other parts in the sugar was much weaker than that of the acetyl group. Among the oligosaccharides tested (1130) in the hemagglutination inhibition assays, oligosaccharide 28 was the best inhibitor, and oligosaccharides 24 and 30 were second best. All the saccharides having terminal GlcNAc residues, 24-30, were more inhibitory than the others (11-23). All the results of the hemagglutination inhibition tests and NMR analyses show that HOL-I can specifically recognize the N-acetyl group of the terminal GlcNAc residue. The detailed binding properties to GalNAc residues of HOL-I remains to be determined. The hemagglutination by HOL-I was also inhibited by BSM 34, asialo-BSM 35, fetuin 36, and asialofetuin 37. Among them, asialo-BSM exhibited the most inhibitory effect. Since asialo-BSM has mainly terminal P-galactosyl residues in the sugar chains, this result was inconsistent with those of monosaccharides and oligosaccharides. However, trypsinor Pronase-digested asialo-BSM showed much weaker inhibitory effect on HOL-I-mediated hemagglutination than native one (data not shown). This result suggests that the inhibitory effect of asialo-BSM is not due to only