Unique Catabolic Pathway of Glycosphingolipids in a Hydrozoan, Hydra magnipapillata , Involving Endoglycoceramidase*

Endoglycoceramidase (EGCase; EC 3.2.1.123) is an enzyme capable of cleaving the glycosidic linkage between oligosaccharides and ceramides of various glycosphingolipids. We detected strong EGCase activity in animals belonging to Cnidaria , Mollusca , and Annelida and cloned the enzyme from a hydra, Hydra magnipapillata . The hydra EGCase, consisting of 517 amino acid residues, showed 19.2% and 50.2% identity to the Rhodococcus and jellyfish EGCases,

components of the plasma membranes of vertebrates. The ceramide portion of the GSL molecule is embedded in the fluid phase of the plasma membrane, and a sugar chain faces the external environment. Some GSLs are considered to be receptors for microorganisms and their toxins as well as modulators of cell growth and differentiation (1). Recently, GSLs were found to be enriched with other sphingolipids and cholesterol to form microdomains on ectoplasmic membranes (2). These lipid microdomains, known as detergent-insoluble glycolipid-enriched domains, glycolipid-enriched membranes, or rafts, assemble receptors and signaling molecules coupled to Src family kinases and G-proteins on their inner surface and mediate cell adhesion, membrane trafficking, and signaling activities (3). In mammals, after recycling between the plasma membrane and intracellular organs, GSLs are eventually transported to lysosomes where they are all hydrolyzed sequentially from the non-reducing end by exo-type glycohydrolases (4). However, the catabolic pathway of GSLs in invertebrates has yet been elucidated.
In this study, we cloned a new acidic EGCase from the Hydra magnipapillata and analyzed the function of this animal EGCase. We found that the enzyme, present in endodermal cells, was transiently released into the gastric cavity during the feeding process. The EGCase was able to hydrolyze dietary GSLs to produce oligosaccharides and ceramides. When GM1a 2 was injected into the gastric cavity, it was incorporated and catabolized directly to ceramide and oligosaccharide. The for-* This work was supported in part by a Grant-in-aid for Scientific Research on Priority Areas 12140204 and Research (B) 15380073 from the Ministry of Education, Science and Culture of Japan, and Takeda Science Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the DDBJ/GenBank TM /EBI Data Bank with accession number(s) AB179748.
Hydra Culture-Strain 105, the standard wild-type strain of H. magnipapillata, was used in this study. Animals were cultured in hydra medium (1 mM Tris-HCl buffer, pH 7.6, 1 mM NaCl, 1 mM CaCl 2 , 0.1 mM KCl, and 0.1 mM MgSO 4 ) at 18°C using the standard method (18). Freshly hatched brine shrimp (Artemia salina) were fed to hydra three times a week, and the medium was changed daily. Before the experiment, animals were starved for 3 days.
EGCase Assay-Ten nanomoles of GM1a (containing 100 pmol of [ 14 C]stearyl-GM1a, 5 nCi) was incubated at 37°C for a specific period with an appropriate amount of the enzyme in 20 l of 25 mM sodium acetate buffer, pH 3.0, containing 0.2% (w/v) Triton X-100. The reaction was stopped by heating in a boiling water bath for 5 min. After drying, the sample was dissolved in 20 l of 50% methanol and applied to TLC plates, which were then developed with chloroform/methanol/0.02% CaCl 2 (5/4/1, v/v). Radioactive GM1a and ceramide separated by TLC were analyzed with an imaging analyzer (FLA-5000, Fujifilm, Tokyo, Japan) and quantified with Image Gauge version 3.0 (Fujifilm). One unit of the EGCase was defined as the amount of enzyme that catalyzes the release of 1 mol of GM1a/min under the conditions described above. A value of 10 Ϫ3 , 10 Ϫ6 , and 10 Ϫ9 units of enzyme was expressed as 1 milliunit, 1 microunit, and 1 nanounit, respectively.
For substrate specificity, non-radioactive GSLs (10 nmol of each) were employed. GSLs and oligosaccharides were visualized by spraying the TLC plates with orcinol-H 2 SO 4 regent and scanning them by using a Shimadzu CS-9300 chromatoscanner (Kyoto, Japan) with the reflectance mode set at 540 nm.
Purification of Hydra EGCase-H. magnipapillata (5.2 g) was homogenized in 20 ml of 20 mM sodium acetate buffer, pH 6.0, containing 0.1% Triton X-100 and Complete TM , a protease inhibitor mixture (Roche Diagnostics, Germany), and centrifuged at 10,500 ϫ g for 10 min. The supernatant (specific activity, 0.5 milliunit/mg) was then applied to a column of Hi-Trap phenyl-Sepharose (1 ml, Amersham Biosciences, Uppsala, Sweden) equilibrated with 20 mM sodium acetate buffer, pH 5.0. The enzyme was eluted from the column with the same buffer containing 1% Lubrol PX. The active fractions from the phenyl-Sepharose column were pooled (specific activity, 2.2 milliunits/mg) and subjected to chromatography using Hi-Trap SP-Sepharose (1 ml, Amersham Biosciences) previously equilibrated with 20 mM sodium acetate buffer, pH 5.0, containing 0.1% Triton X-100. The enzyme was eluted from the column with the same buffer containing 1.0 M NaCl. The active fractions were pooled (specific activity, 16.9 milliunits/mg), and the buffer was replaced with 20 mM sodium acetate buffer, pH 5.0, using a Centriprep-10 (Millipore). The enzyme was then applied to a TSK gel CM-5PW column (5 ϫ 50 mm, Tosoh Co., Tokyo, Japan) using the BioCAD SPRINT system (Applied Biosystems) that had been equilibrated with 20 mM sodium acetate buffer, pH 5.0, containing 0.1% Lubrol PX. The column was washed with the same buffer and eluted with a linear gradient of NaCl (0 to 0.1 M) in the same buffer at a flow rate of 1 ml/min. The fractions showing the enzyme activity were replaced with 20 mM sodium acetate buffer, pH 5.0, using a Centriprep-10. Characterization of the native hydra EGCase was performed using this final preparation (specific activity, 38.2 milliunits/mg). Throughout the course of purification, specific activity was increased by 76.4 times with 22.8% recovery. The preparation showed no detectable exoglycosidases except ␤-glucocerebrosidase.
Molecular Cloning and Sequencing-Nucleotide sequences were determined by the dideoxynucleotide chain termination method with a Bigdye Terminator Cycle Sequencing Ready Reaction kit (Applied Biosystems) and a DNA Sequencer (model 377A, Applied Biosystems). The initial search for the hydra EGCase cDNA was conducted using PCR. Based on the conserved sequence in microbial and jellyfish EGCases, degenerate primers of both sense and antisense strands were designed. A sense oligonucleotide primer, YLD (5Ј-CTNGAYATGCAYCARGA-3Ј), and an antisense primer, YNE (5Ј-GCRAANGGYTCRTTDAT-3Ј), were synthesized. PCR amplification was performed using these primers with first strand cDNA as a template in a GeneAmp PCR System 9700 (Applied Biosystems) using AmpliTaq Gold (Applied Biosystems). The PCR products were cloned into pGEM T-easy vector (Promega, Madison, WI, USA), and their DNA sequences were determined. The PCR products were also labeled with [␣-32 P]dCTP using a Ready-To-Go DNA labeling kit (Amersham Biosciences) and used as a probe to screen a hydra cDNA library constructed using Lambda ZAPII (Stratagene, La Jolla, CA). Finally, a full-length cDNA clone encoding hydra EGCase was obtained by plaque hybridization. The plasmid containing hydra EGCase was designated pSKHEGC.
In Situ Hybridization-Digoxygenin (DIG)-labeled RNA probes corresponding to the sense and antisense strands of the full-length cDNA of hydra EGCase were prepared using a DIG RNA Labeling Mix (Roche Diagnostics). In situ hybridization of whole mounts was carried out as described previously (20). Briefly, hydra were fixed with 4% paraformaldehyde after relaxation of the polyps with 2% urethane. Specimens were subsequently treated with ethanol and proteinase K. To stabilize the digested tissues, specimens were re-fixed with 4% paraformaldehyde and then prehybridized in hybridization solution (50% formamide, 5ϫ SSC, 1ϫ Denhardt's solution, 200 g/ml tRNA, 0.1% Tween 20, 0.1% CHAPS, and 100 mg/ml heparin) to block nonspecific hybridization sites. This was followed by hybridization for 16 h with the DIG-labeled RNA probe and subsequent washing in hybridization solution and then in 1ϫ SSC. The specimens were further washed in MAB (100 mM maleic acid, 150 mM NaCl, pH 7.5) and pre-blocked for 2 h in MAB with 2% blocking reagent (Roche Diagnostics). This was followed by a 16-h incubation at 4°C in the same solution with anti-DIG antibody (Roche Diagnostics). The specimens were washed eight times with MAB and then briefly in alkaline phosphatase buffer (100 mM Tris-HCl buffer, pH 9.5, containing 50 mM MgCl 2 , 100 mM NaCl, and 0.1% Tween 20). Specimens were then stained with BM-purple (Roche Diagnostics).
Expression of Hydra EGCase in CHOP Cells-A cDNA fragment encoding the hydra EGCase was prepared by PCR using a 5Ј primer containing a Kozak sequence (21) and EcoRI site (5Ј-GGAATTCACCAT-GGTAAGCGTCGCACTTAT-3Ј), a 3Ј primer containing an XhoI site (5Ј-CCCTCGAGTTCTGATAGATGTTTATCACCCA-3Ј), and pSKHEGC as a template with Pyrobest DNA polymerase (Takara Bio Inc., Otsu, Japan). The PCR product was digested with EcoRI and XhoI and introduced into these sites in the vector of pcDNA3.1/Myc-His(ϩ)A (Invitrogen, Carlsbad, CA). CHOP cells, Chinese hamster ovary cells expressing polyoma LT antigen to support the replication of eukaryotic expression vectors (22), were grown in ␣-minimal essential medium supplemented with 10% fetal bovine serum, 100 g/ml streptomycin, and 100 units/ml penicillin in a humidified incubator containing 5% CO 2 . Cells were seeded at 10 5 cells/dish. Transfection with vector alone or vector containing hydra EGCase was performed using Lipofect-AMINE TM Plus (Invitrogen) according to the instructions of the manu-facturer. After 4 h of incubation with the mixture, the cells were grown in fresh medium. After 24 h, cells were harvested and suspended in 100 l of 20 mM sodium acetate buffer, pH 6.0, containing 0.1% Triton X-100. The enzyme activity in the cell lysate was measured as described above.
Release of EGCase-One-hundred starved hydra were fed brine shrimp, washed, and then cultured in fresh hydra medium (1 ml/well on 6-well plates). After a period of time, 100 l of the medium was withdrawn and then centrifuged at 17,400 ϫ g for 5 min. The EGCase activity of the supernatant was assayed at 37°C for 2 h using [ 14 C]stearyl-GM1a as a substrate.
In Vivo Digestion of [ 14 C]GM1a in Hydra-50 pmol of [ 14 C]GM1a (2.5 nCi), dissolved in hydra medium (ϳ1 l), was injected through the mouth into the gastric cavity of the hydra using a micropipette. Just before the injection, the hydra were fed brine shrimp. After the animals were cultured for a period at 18°C, total lipids were extracted from 10 animals by homogenization with 400 l of chloroform/methanol (1/2, v/v). The extract was evaporated dry, redissolved in 20 l of the same solvent, and analyzed by TLC, using a solvent system of chloroform/ methanol/water (60/35/8, v/v). Radioactive lipids were analyzed by the method described above.
Extraction of Glycolipids from Brine Shrimp and Fast Atom Bombardment Mass Spectrometry-Total lipid was extracted from 13 g of freshly hatched brine shrimp by homogenization with 15 ml of chloroform/methanol (2/1, v/v). The extraction was repeated twice. Combined extracts were divided into upper and lower phases with Folch's partition (23). After being evaporated to dryness, the lower phase was incubated in 0.1 M methanolic KOH at 37°C for 1 h to eliminate glycerolipids. After the removal of salt using Folch's partition, the lipid was dissolved in chloroform/methanol (98/2, v/v) and applied to a column of Sep-Pak plus silica (Waters Co., Milford, MA). GSLs were eluted from the column with methanol. Each fraction was spotted onto a TLC plate, developed in a solvent system of chloroform/methanol/0.02% CaCl 2 (5/4/1, v/v), and visualized by spraying with orcinol-H 2 SO 4 reagent. For fast atom bombardment mass spectrometry, GSLs were purified by TLC. The portions corresponding to the orcnol-H 2 SO 4 -positive bands were scraped from the unstained duplicate TLC plate and then extracted with chloroform/methanol (1/1, v/v). After removal of the silica by centrifugation, the supernatant was dried under N 2 gas. Approximately 10 g of GSL was mixed with triethyleneglycol as a matrix and subjected to FAB mass spectrometric analysis, which was performed in the negative ion mode.
Fatty Acid and Sugar Composition Analysis-The fatty acid and sugar compositions of hydra GSL were analyzed by the methods described previously (24,25). Briefly, methyl esters of fatty acids and methyl glycosides were obtained by methanolysis of GSLs with 1.0 M anhydrous methanolic HCl at 100°C for 3 h. Fatty acid methyl esters were extracted with hexane, and the methyl glycosides remaining in the methanolic solution were concentrated, N-acetylated, and trimethylsilylated. The fatty acid and sugar compositions were determined separately by GLC on a Shimadzu GC-14A equipped with a fused silica capillary column (DB-1, 0.25 mm ϫ 30 m, J & W Scientific).
Preparation of Anti-hydra EGCase Antibody-To obtain the recombinant hydra EGCase, the ORF was subcloned into the bacterial expression vector pET32b (Novagen, Madison, WI), and introduced into AD494(DE3)pLysS, an Escherichia coli strain, for protein expression. The cells were cultured at 37°C in 1.5 liters of LB medium containing 50 g/ml ampicillin, 15 g/ml kanamycin, and 35 g/ml chloramphenicol. When the absorbance at 600 nm reached 0.4, isopropyl-1-thio-␤-Dgalactopyranoside was added to a final concentration of 0.5 mM, and the mixture was incubated for 5 h at 37°C. Cells were harvested by cen-trifugation at 7,000 ϫ g for 10 min at 4°C, and the pellet was lysed with 20 mM Tris-HCl buffer, pH 7.5, containing 1% Triton X-100. After centrifugation at 15,000 ϫ g for 10 min, the pellet (inclusion bodies) was lysed by sonication in 50 mM Tris-HCl buffer, pH 7.5, containing 8 M urea. The recombinant protein was purified using a Hi-Trap chelating column (Ni 2ϩ , Amersham Biosciences) according to the manufacturer's instructions. The purified protein was then subjected to 10% SDS-PAGE, and the protein band was extracted from the gel using 20 mM Tri-HCl buffer, pH 7.5, containing 1% SDS. Antiserum against recombinant hydra EGCase was obtained from a rabbit immunized with the purified protein.
SDS-PAGE and Western Blotting-SDS-PAGE was carried out according to the method of Laemmli (26). The transfer of protein to polyvinylidene difluoride membrane was performed using Trans-Blot SD (Bio-Rad) as described previously (27). After treatment with 5% skim milk in Tris-buffered saline containing 0.1% Tween 20 (T-TBS) for 1 h, the membrane was incubated with anti-hydra EGCase antiserum for 1 h at room temperature. After a wash with T-TBS, the membrane was incubated with horseradish peroxidase-conjugated anti-rabbit antibody for 1 h. After another wash with T-TBS, the ECL reaction was performed for 2-3 min as recommended by the manufacturer, and chemiluminescent signals were visualized with an ECL Mini-camera (Amersham Biosciences).
Immunohistochemistry and Fluorescence Microscopy-Hydra were fixed with 4% paraformaldehyde for 2 h after relaxation of the polyps with 2% urethane. The samples were rinsed with PBS and then infiltrated with 20% sucrose in PBS overnight at 4°C. They were embedded in Tissue-Tek OCT compound (Sakura Finetek Co., Tokyo, Japan), rapidly frozen using liquid nitrogen, and stored at Ϫ80°C. The frozen materials were cut into 9-to 10-m thick sections using a cryostat (Leica CM1850, Wetzlar, Germany) and mounted on poly-L-lysinecoated glass slides. After treatment with 5% skim milk in PBS (blocking buffer) for 20 min at room temperature, the samples were incubated with the anti-hydra EGCase antiserum diluted 1:500 with a blocking buffer for 2 h at room temperature followed by Cy3-labeled anti-rabbit IgG (Amersham Biosciences) at room temperature for 1 h. For controls, pre-immune serum was used as the primary antibody. For nuclear staining, SYTOX Green (Invitrogen) was used. To identify more precisely the EGCase-expressing cells, macerated samples were used. Hydra were placed in maceration fluid, acetic acid/glycerol/water (1/1/13, v/v), for 10 min, and then shaken to disperse the cells (28). The suspended cells were fixed using 4% paraformaldehyde and spread on a gelatin-coated glass slide. Immunostained samples were observed under a confocal laser-scanning microscope (Digital Eclipse C1, Nikon, Tokyo, Japan).
Other Methods-The protein content was measured by the bicinchoninic acid protein assay (Pierce) with bovine serum albumin as the standard. Glycopeptidase F treatment was conducted using the purified hydra EGCase as described previously (10). Exoglycosidases and sialidase were assayed using p-nitrophenyl glycosides (29) and 4-methylumbelliferyl-N-acethylneuraminic acid (12), respectively, as the substrate.

Distribution of EGCase in Animals-
The EGCase activity in various invertebrates was examined using [ 14 C]GM1a ([ 14 C]stearic acid, d18:1 sphingenine) as the substrate (Table I). The activity was found to be widely distributed in animals of the phyla Cnidaria, Mollusca, and Annelida. In addition to the animals listed in Table I, we detected EGCase activity in sea sponges, baby octopus, ear shells, etc., but not in mammalian or insect cell lines such as HEK293 cells (human embryonic kidney cells), COS-1 cells (African green monkey cells), and Schneider 2 cells (Drosophila melanogaster cells) (data not shown). Among the animals tested, we found a relatively high specific activity of EGCase in cnidarians and thus selected H. magnipapillata as a model to study the biological roles of animal EGCase. General Properties of the Hydra EGCase-First, EGCase was purified from extracts of H. magnipapillata by conventional column chromatography as described under "Experimental Procedures." The optimum pH of the purified enzyme was found to reside in an extremely acidic range, pH 3.0 -4.0, when [ 14 C]GM1a was used as a substrate (Fig. 1A). The enzyme required detergents for the hydrolysis of GSLs. The optimal concentrations of Triton X-100 and Lubrol PX were 0.2% (w/v) and 0.4% (w/v) at pH3.0, respectively, although taurodeoxycholate at 0.1-1% (w/v) strongly inhibited the enzyme activity. Cu 2ϩ , Zn 2ϩ , Mn 2ϩ , Ca 2ϩ , Mg 2ϩ , and EDTA had no significant effects on the activity at 5 mM. Fig. 1B shows the specificity of the purified enzyme for various GSLs. The extent of hydrolysis of GSLs by the enzyme was expressed under different conditions; one expressed the relative initial reaction velocity (30 min, open bar) and the other the degree of hydrolysis after exhaustive digestion (16 h, solid bar). Of all the substrates tested, GSLs having a gangliotetraose structure appeared to be the best substrates for the enzyme regardless of the presence or absence of sialic acid residues. GT1b was the most favored substrate under both conditions. Hydra EGCase hydrolyzed globoside very slowly compared with ganglio-series GSLs, indicating that the enzyme is a type II EGCase (5). In addition to GSLs from mammalian origins, we examined the specificity of the enzyme to invertebrate GSLs. The enzyme hydrolyzed L-5, GL-4, and Lipid IV, in which an oligosaccharide is linked to ceramide via ␤-glucosyl linkage, whereas it did not hydrolyze TGC in which an oligosaccharide is linked to ceramide via ␤-galactosyl linkage, indicating that hydra EGCase is an oligoglycosylglucosylceramide glycohydrolase, like the enzyme from other sources. It is noteworthy that the hydrolysis extent of invertebrate GSLs (L-5, GL-4, and Lipid IV) by the hydra enzyme was quite low compared with that of the gangliotetraose-series GSLs (Fig. 1B). Fig. 1C shows a Western blot of the hydra EGCase before and after treatment with glycopeptidase F. The apparent molecular mass of the hydra enzyme was reduced by glycopeptidase F treatment from 56.0 to 52.9 kDa on SDS-PAGE, indicating that hydra EGCase was glycosylated with N-glycans.
Cloning and Expression of a cDNA Encoding the Hydra EGCase-We reported that the Asn-Glu-Pro (NEP) sequence, the active site of the glycoside hydrolase (GH) family 5 endo-␤-1,4-glucanases, was conserved in both Rhodococcus (30,31) and jellyfish EGCases (10). In addition, we found that the region Leu 131 -Val 137 in the deduced amino acid sequence of jellyfish EGCase is highly conserved in two Rhodococcus EGCases (10,30,31). Thus, we designed the sense and antisense degenerate primers based on these two conserved regions and performed PCR as described under "Experimental Procedures." The 300-bp PCR product generated was labeled with 32 P and then used as the probe for plaque hybridization. After the screening of 5 ϫ 10 5 clones, a cDNA clone having a putative ORF of the hydra EGCase was isolated. Fig. 2A shows the nucleotide and deduced amino acid sequences of the hydra EGCase. The 1,551-bp ORF encodes 517 amino acid residues, and the molecular mass and pI of the EGCase were calculated to be 59,681 and 5.47, respectively. A hydrophobic motif composed of 19 amino acid residues starting with the initiation Met was found in the deduced amino acid sequence. The presence of  the hydrophobic motif was also clearly indicated by a hydrophobicity plot (Fig. 2B). The deduced amino acid sequence of the hydra EGCase contained four potential N-glycosylation sites, consistent with the notion that the hydra EGCase is glycosylated by N-glycans (Fig. 1C). The ORF showed 19.2% and 50.2% identity to that of the Rhodococcus M-777 and jellyfish EGCases, respectively, at the amino acid level (Fig. 3). In addition to the NEP sequence, seven amino acid residues, which are essential for the catalytic activity of GH5 endo-␤-1,4glucanases (32), were all conserved in these three EGCases; Glu 230 and Glu 349 are thought to be an acid/base catalyst and a nucleophile, respectively, and His 125 is thought to be important for interaction with glucose (Fig. 3).
To verify whether the putative ORF actually encodes the EGCase, CHOP cells were transfected with an expression vector containing the ORF. EGCase activity was detected in the   FIG. 2. Nucleotide and deduced amino acid sequences (A) and hydropathy plot (B) of the hydra EGCase. A, the deduced amino acid sequence is shown in one-letter code below the nucleotide sequence. Amino acid residues are numbered beginning with the first methionine, and the translation termination codon is denoted by an asterisk. The putative signal sequence is bold. Potential N-linked glycosylation sites are boxed, and the putative active site (NEP sequence) is double boxed, respectively. B, the deduced amino acid sequence of the EGCase was analyzed by the method of Kyte and Doolittle for hydropathy plotting (37). lysate of the CHOP cells transfected with the ORF-containing plasmid but not in that of the mock transfectant (Fig. 4A). The activity of hydra EGCase was not detected in the bacterial expression system using E. coli as a host cell (data not shown), suggesting that N-glycosylation is necessary for the enzyme activity. This speculation was supported by the finding that the hydra EGCase showed no activity when expressed in CHOP cells in the presence of tunicamycin, a specific inhibitor for N-glycosylation (data not shown).
The pH dependence of the recombinant hydra EGCase was quite similar to that of the purified native enzyme (Fig. 1A), the optimum pH being 3.0 -4.0 (Fig. 4B). This result indicates that the enzyme is a typical acid EGCase like the jellyfish enzyme.
Whole Mount in Situ Hybridization and Immunocytochemical Analysis of the Hydra EGCase-The expression of the EGCase mRNA in hydra was analyzed with whole mount in situ hybridization. The transcripts were detected specifically in the endodermal layer throughout the body. It was noticeable that signals were strong in the tentacles and peduncles (Fig.  5A, right panel). No signals were detected with the sense RNA probe (Fig. 5A, left panel). The expression of EGCase protein was also examined using anti-hydra EGCase antibody under a   FIG. 3. Alignment of hydra, jellyfish, and Rhodococcus EGCases. Hydra (H. magnipapillata), jellyfish (C. nozakii), and microbial (Rhodococcus sp. M-777) EGCases were aligned using the ClustalW algorithm (38). Identical amino acids are indicated by asterisks and chemically similar amino acids by dots. Gaps inserted into the sequences are indicated by dashed lines. Amino acid residues conserved in EGCases and endo-␤-1,4-glucanases are shown on a black background. Two glutamic acids, possibly functioning as an acid/base catalyst (Ⅺ) and a nucleophile (f), respectively, are indicated in the sequence. The arrows indicate the amino sequences used for the preparation of PCR primers. fluorescent microscope. As shown in Fig. 5B, strong signals for EGCase were detected in endodermal cells. This result indicates that the site of expression of the EGCase protein is well consistent with that of the mRNA. Endodermal cells are made up of digestive cells and gland cells. To identify which cells express the EGCase, macerated cells were subjected to immunostaining with anti-hydra EGCase antibody. As shown in Fig.  5C, the enzyme was specifically detected in digestive cells but not gland cells. To investigate the distribution of EGCase activity in the hydra, the animal was cut into three parts (tentacles, gastric cavity, and peduncle regions), from each of which the enzyme was extracted and the activity was measured. The specific activity in the three parts was almost the same (tentacles, 0.25 milliunit/mg; gastric cavity, 0.21 milliunit/mg; and peduncle regions, 0.23 milliunit/mg), suggesting that EGCase was expressed in the endoderm throughout the whole body.
Release of the Hydra EGCase in the Gastric Cavity-It is reported that digestive enzymes such as proteases are secreted into the gastric cavity when hydra ingest food (33). To clarify the possibility that hydra EGCase functions as an extracellular digestive enzyme, its release into the medium during the feeding process was examined. Interestingly, EGCase activity was released into the surrounding medium when brine shrimp was fed to the hydra (Fig. 6A). Notably, however, no EGCase activity was detected in the medium without prey (brine shrimp). It was confirmed that the EGCase activity was not derived from the prey, because the extract of brine shrimp contained no EGCase activity (data not shown). Next, we examined the total EGCase activity in the whole body of the hydra before and after release of the enzyme. EGCase activity in body was transiently reduced when the prey was given (Fig. 6B). This timing coincides with the increase of the enzyme activity in the medium. EGCase activity levels were restored after 24 h (Fig. 6B). These results strongly suggest that the hydra EGCase in the digestive cells is released during the digestive process into the gastric cavity and eventually into the surrounding medium. The enzyme seemed to be released from digestive cells but not from gland cells, because EGCase was exclusively detected in the digestive cells but not gland cells (Fig. 5C).
Digestion of Brine Shrimp GSLs by the Hydra EGCase-The results described above suggest that EGCase hydrolyzes dietderived GSLs. We thus investigated whether brine shrimp possesses GSLs, which are actually degraded by hydra EGCase. We detected several alkaline-resistant GSLs in brine shrimp (Fig. 7A, lane 1). These GSLs were not stained with either resorcinol or Dittmer reagent, indicating that these compounds contain no sialic acids or phosphate groups (data not shown). Brine shrimp GSLs were degraded by Rhodococcus EGCase II to generate oligosaccharides (Fig. 7A, lane 4) and ceramide (data not shown). As previously reported (5), the microbial EGCase did not hydrolyze glucosylceramide (GlcCer), which then remained intact after the enzyme treatment. On the other hand, the treatment of brine shrimp GSL with the " Arrows indicate the signals for mRNA expression of EGCase. B, frozen section of the gastric cavity was stained with anti-hydra EGCase antibody followed by staining with goat anti-rabbit IgG-Cy3 (red). Samples were counterstained with SYTOX Green to visualize nuclei (green). Arrowheads indicate the mesoglea. Details are described under "Experimental Procedures." Bars indicate 40 m. C, digestive cells and gland cells those consist of the endoderm were immunostained with antihydra EGCase antibody and goat anti-rabbit IgG-Cy3 (red) and counterstained with SYTOX Green (green). The same samples were also observed under a differential interference contrast microscope (each left panel). Bars indicate 20 m.
partially purified hydra EGCase diminished levels of GlcCer and other GSLs with the simultaneous production of glucose and GSL-derived oligosaccharides (Fig. 7A, lane 2). It is of note that the R F of GSL-derived oligosaccharides by either EGCase was the same. The addition of Conduritol B epoxide, a specific inhibitor for ␤-glucosylcerebrosidase (34), suppressed the production of glucose and decrease of GlcCer, suggesting that the hydra EGCase preparation contained a ␤-glucosylcerebrosidase. In fact, the hydra extract showed ␤-glucosidase activity when measured with BODIPY-GlcCer as a substrate. It is noteworthy that the activity was clearly separated from EGCase activity using CM-5PW chromatography, indicating that the major ␤-glucocerebrosidase activity in the hydra is not due to EGCase (data not shown). However, the possibility that the hydra EGCase per se hydrolyzed GlcCer can not be ruled out at present, because the leech ceramide glycanase (EGCase) hydrolyzes GlcCer very slowly (7). In any case, the dietary GSLs could be degraded by the combination actions of EGCase and ␤-glucocerebrosidase in the hydra.
The major EGCase-digestible GSL of brine shrimp (GSL-X) was subjected to negative ion FAB-MS analysis. The parental molecular ion [M-H] Ϫ was observed at m/z 1265. Fragment ions of GSL-X indicated the sequential elimination of N-acetylhexosamine, deoxyhexose, hexose, and hexose being detected at m/z 1062, 916, 754, and 592, respectively (Fig. 7B). The sugar composition of GSL-X, determined by GLC with trimethylsilyl derivatives, was GlcNAc, Fuc, Man, and Glc in a molar ratio of about 1:1:1:1. GSL-X seems to be composed of two bands on TLC, thus, each was extracted from the TLC and analyzed separately for sugar and fatty acid composition. The major fatty acids of the upper and lower bands were C22:0 and 2-hydroxy-C22:0, respectively, whereas the sugar compositions of both were the same. Conclusively, brine shrimp, a prey of the hydra, possesses an EGCase-digestible neutral GSL, although the precise structure of the GSL remains to be determined.

Digestion of [ 14 C]GSLs by EGCase in Vivo-
To examine whether hydra EGCase can hydrolyze exogenously added GSLs in vivo, [ 14 C]GM1a labeled at the fatty acyl chain or sugar chain was injected into the gastric cavity of the hydra just after the animal had captured brine shrimp, and then the catabolism of GM1a was monitored. After a period, radioactive metabolites were extracted from the hydra and analyzed by TLC as described under "Experimental Procedures." When [ 14 C]stearyl GM1a was fed to hydra, [ 14 C]ceramide first appeared after 0.5 h as a digestion product, after which [ 14 C]stearic acid and its putative metabolites were observed at 1-2 h (Fig. 8A). This FIG. 6. Release and restoration of the hydra EGCase. A, release of EGCase from hydra. One hundred hydra, fed with brine shrimp (Feed) or not (Non-feed), were cultured in 1 ml of the medium. EGCase activity in the medium was measured as described under "Experimental Procedures." B, restoration of EGCase activity in the hydra body. Fifteen non-budding hydra were fed brine shrimp. The animals were washed with fresh medium, proteins were extracted from the animal at the time indicated, and the EGCase activity was measured as described under "Experimental Procedures." Values are the means Ϯ S.D. for triplicate determinations (*, p Ͻ 0.05 versus control).

FIG. 7. Digestion of brine shrimp GSLs by EGCase (A) and FAB-MS spectra of the major GSL of brine shrimp (B). A, GSLs
were extracted from brine shrimp as described under "Experimental Procedures." GSLs (corresponding to about 10 g) were incubated with 1 milliunit of hydra or Rhodococcus EGCase in 20 l of 25 mM sodium acetate buffer, pH 3.0, for the hydra enzyme and pH 5.5 for the Rhodococcus enzyme, respectively. After incubation at 37°C for 16 h, the reaction mixture was analyzed by TLC with a solvent system of chloroform/methanol/0.02% CaCl 2 (5/4/1, v/v). The GSLs remaining and oligosaccharides released were stained with orcinol-H 2 SO 4 reagent. Lane 1, GSLs isolated from brine shrimp; lane 2, GSLs plus hydra EGCase; lane 3, GSLs plus hydra EGCase plus conduritol B epoxide (glucosylcerebrosidase inhibitor); lane 4, GSLs plus Rhodococcus EGCase II. Arrows indicate the R F values of standards (GlcCer, glucosylceramide; LacCer, lactosylceramide; Gb 3 Cer, globotryaosylceramide; GM1a and glucose) and oligosaccharides released from GSLs by the action of EGCases. B, FAB-MS spectra of the major GSL (GSL X) of brine shrimp in the negative ion mode. Hex, hexose; dHex, deoxyhexose; HexNAc, N-acetylhexosamine. Details are described under "Experimental Procedures." result indicated that the EGCase hydrolyzed the exogenously added GM1a to generate the ceramide relatively early during the digestive process. To clarify why [ 14 C]stearic acid was produced just after the generation of [ 14 C]ceramide, the hydra was examined for ceramidase activity using a fluorescent ceramide as a substrate at different pHs. Ceramidase is an enzyme capable of hydrolyzing the N-acyl linkage of ceramide to generate fatty acid and sphingosine (EC 3.5.1.23). As shown in Fig.  8B, an acid ceramidase having pH an optimum at 5.0 was actually found in the hydra. Metabolites (a-d) could be derived from the stearic acid, because they were also detected when [ 14 C]stearic acid was injected into the gastric cavity of hydra instead of [ 14 C]GM1a (Fig. 8C). Alkaline treatment of metabolites resulted in the disappearance of a-d but not the ceramide, stearic acid, e, and f. Metabolites a-d were suggested to be triacylglycerol (TAG), phosphatidylethanolamine, phosphatidylcholine, and phosphatidylserine, respectively, judging from the R F values for the TLC and alkaline sensitivity. It is noteworthy that metabolites c and f were also detected by metabolic labeling of hydra with [ 14 C]choline (data not shown). Metabolite e showed the same R F value of the standard GlcCer, and f was hydrolyzed by bacterial sphingomyelinase to generate ceramide, suggesting that alkaline-resistant e and f were cerebroside and sphingomyelin, respectively. However, alkaline-resistant EGCase-sensitive GSLs were not found in hydra in this study (data not shown). Fig. 8D shows the quantification of (2.5 nCi) was injected into the gastric cavity of 10 hydra. After incubation at 18°C for the periods indicated, radioactive lipids and oligosaccharides were extracted and analyzed as described under "Experimental Procedures." decreased during the course of the experiment, whereas levels of ceramide, stearic acid, and other metabolites (a ϩ b ϩ c ϩ d) increased. These results suggested that the dietary GSLs were degraded by EGCase and ceramidase to produce fatty acids, which were further metabolized to various glycerophospholipids and triacylglycerol.
The [ 14 C]sugar-GM1a was also fed to hydra, and its catabolism was monitored. As shown in Fig. 8E, GM1a-oligosaccharide was first appeared after 0.5-1 h by the action of the EGCase. The oligosaccharide was gradually converted to the shorter-chain oligosaccharides, apparently GM2-and GM3-oligosaccharides judged from the R F of the TLC. They were possibly generated by the action of ␤-galactosidase or combination actions of ␤-galactosidase and ␤-N-acetylgalactosaminidase, respectively. Both enzymes were actually detected in hydra, whereas sialidase was not detectable in hydra (data not shown). Interestingly, [ 14 C]GM1a was directly converted by EGCase to GM1a-oligosaccharide while hydra exoglycosidases did not attack GM1a directly under the conditions used because no GM2 or GM3 was found after prolong incubation (Fig.  8E). In summary, catabolism of GSLs in the hydra is initiated by the removal of sugar chains from GSLs by EGCase; subsequently, sugar chains and ceramides are degraded by exoglycosidases and ceramidase, respectively, and finally the fatty acids generated are metabolized to various glycerophospholipids and triacylglycerols. In conclusion, the hydra has a novel catabolic pathway for GSLs involving EGCase that is likely to be missing in mammals. DISCUSSION The catabolism of GSLs has been studied extensively for several decades mainly in mammals. In mammals, endogenous GSLs are finally decomposed by various exoglycosidases in lysosomes to generate monosaccharides and ceramide. The latter is further degraded by ceramidase to fatty acid and sphingosine. On the other hand, there have been few reports on the catabolism of dietary GSLs. Nillson (35) reported that exogenously added glucosylceramide was catabolized by the actions of glucosylcerebrosidase (glucosylceramidase, EC 3.2.1.45) and ceramidase to generate sphingosine and fatty acid, the latter of which was further converted to TAG and PC in the intestinal tract of rats. It should be emphasized that EGCase is not likely to be present in mammals. Thus, the new catabolic pathway for GSLs involving EGCase in hydra is completely different from that in mammals. This catalytic pathway seems to be unique but quite ubiquitous in invertebrates except insects, because EGCase activity is widely distributed in Cnidaria, Mollusca, Annelida, Porifera, Echinodermata, and others.
Hydra are the most primitive group of animals with a defined body plan and widely used for the study of pattern formation, cell differentiation, and morphogenesis (36). The animal has a simple tube-shape form consisting of a head, body column, and foot and is made up of two multicellular layers separated by an extracellular matrix, named the mesoglea. The outer layer (ectodermal cell layer) consists of epitheliomuscular cells and cells in the interstitial cell lineage (nematocysts, nerve, and germ cells), whereas the insider layer (endodermal cell layer) is made up of digestive cells and gland cells. During the feeding process, the prey are captured and paralyzed with exploded nematocysts and organelle in nematocytes, most of which are located on the tentacle ectodermal epithelial cells called battery cells. The captured prey are ingested into the gastric cavity and degraded into small particles by digestive enzymes secreted from gland cells. The digestive cells phagocytize the particles into phagosomes/lysosomes (food vacuoles) where the particles are further digested into organic compounds to be used as nutrients (33). Hydra EGCase seems to participate in the catabolic processing of GSLs in two different stages. First, EGCase functions as an extracellular digestive enzyme in the gastric cavity. This conclusion is supported by the finding that EGCase is transiently released into the cavity during feeding (Fig. 6, A and B) and actually hydrolyzes the diet-derived GSLs (Fig. 7A). However, the enzyme seems to be released from the digestive cells but not from gland cells, because the enzyme is exclusively found in the former cells in the endoderm (Fig. 5, B and C). Second, the enzyme could be involved in the catabolism of GSLs in the phagosomes/lysosomes of digestive cells. This was suggested by the extremely acidic pH optimum of the hydra EGCase (Figs. 1A and 4B).