Interaction of the Hemolytic Lectin CEL-III from the Marine Invertebrate Cucumaria echinata with the Erythrocyte Membrane
1995; Elsevier BV; Volume: 270; Issue: 8 Linguagem: Inglês
10.1074/jbc.270.8.3560
ISSN1083-351X
AutoresTomomitsu Hatakeyama, Haruna Nagatomo, Nobuyuki Yamasaki,
Tópico(s)Silk-based biomaterials and applications
ResumoCEL-III is one of four Ca2+-dependent galactose/N-acetylgalactosamine (GalNAc)-binding lectins from the marine invertebrate Cucumaria echinata which exhibits hemolytic activity, especially toward rabbit and human erythrocytes. The hemolytic activity of CEL-III was also Ca2+-dependent and was found to be inhibited by galactose or GalNAc-containing carbohydrates, suggesting that the hemolysis was caused by CEL-III binding to specific carbohydrates on the erythrocyte membrane by Ca2+-dependent lectin activity, followed by partial destruction of the membrane. The activity of CEL-III was highest at 10°C and decreased markedly with increasing temperature, unlike usual enzymatic reactions. The hemolytic activity of CEL-III increased with increasing pH from neutral to 10, but almost no hemolysis was observed below pH 6.5. Immunoblotting analysis of proteins from the erythrocyte membrane after treatment with CEL-III indicated that CEL-III aggregates were irreversibly bound to the membrane. When erythrocytes were incubated with CEL-III in the presence of dextran with molecular masses greater than 4 kDa, lysis was impeded considerably, while a concomitant release of ATP was detected from these osmotically protected cells. It was found that CEL-III released carboxyfluorescein from artificial globoside-containing lipid vesicles, and it is suggested that CEL-III is a novel pore-forming protein with the characteristics of a Ca2+-dependent lectin, which may act as a toxic protein to foreign microorganisms. CEL-III is one of four Ca2+-dependent galactose/N-acetylgalactosamine (GalNAc)-binding lectins from the marine invertebrate Cucumaria echinata which exhibits hemolytic activity, especially toward rabbit and human erythrocytes. The hemolytic activity of CEL-III was also Ca2+-dependent and was found to be inhibited by galactose or GalNAc-containing carbohydrates, suggesting that the hemolysis was caused by CEL-III binding to specific carbohydrates on the erythrocyte membrane by Ca2+-dependent lectin activity, followed by partial destruction of the membrane. The activity of CEL-III was highest at 10°C and decreased markedly with increasing temperature, unlike usual enzymatic reactions. The hemolytic activity of CEL-III increased with increasing pH from neutral to 10, but almost no hemolysis was observed below pH 6.5. Immunoblotting analysis of proteins from the erythrocyte membrane after treatment with CEL-III indicated that CEL-III aggregates were irreversibly bound to the membrane. When erythrocytes were incubated with CEL-III in the presence of dextran with molecular masses greater than 4 kDa, lysis was impeded considerably, while a concomitant release of ATP was detected from these osmotically protected cells. It was found that CEL-III released carboxyfluorescein from artificial globoside-containing lipid vesicles, and it is suggested that CEL-III is a novel pore-forming protein with the characteristics of a Ca2+-dependent lectin, which may act as a toxic protein to foreign microorganisms. INTRODUCTIONAnimal lectins have been of great interest recently, as their various important functions in organisms are being suggested. Most animal lectins have been classified into one of two groups, i.e. Ca2+-dependent (C-type) and independent (S-type or galectin)(1Drickamer K. J. Biol. Chem. 1988; 263: 9557-9560Abstract Full Text PDF PubMed Google Scholar, 2Barondes S.H. Castronovo V. Cooper D.N.W. Cummings R.D. Drickamer K. Feizi T. Gitt M.A. Hirabayashi J. Hughes C. Kasai K. Leffler H. Liu F.-T. Lotan R. Mercurio A.M. Monsigny M. Pillai S. Poirer F. Raz A. Rigby P.W.J. Rini J.M. Wang J.L. Cell. 1994; 76: 597-598Abstract Full Text PDF PubMed Scopus (1085) Google Scholar). In vertebrates, C-type lectins are classified into six groups, which include proteoglycan core protein, hepatic lectin, mannose-binding protein, and selectin(3Drickamer K. Biochem. Soc. Trans. 1993; 21: 456-459Crossref PubMed Scopus (43) Google Scholar). These proteins are composed of C-type CRDs, 1The abbreviations used are: CRDcarbohydrate recognition domainPAGEpolyacrylamide gel electrophoresisTBSTris-buffered saline. which exhibit some degree of homology with each other, and additional domains or regions with the characteristic functions of individual proteins. C-type lectins purified from invertebrates, on the other hand, are generally smaller than those from vertebrates, and many of them consist of only a single CRD. One of the most probable roles of invertebrate C-type lectins is to act as humoral factors in the defense mechanism, as do immunoglobulins in vertebrates; enhanced fly lectin production after injury to the body wall and its activation of phagocytes in vitro have been demonstrated(4Komano H. Mizuno D. Natori S. J. Biol. Chem. 1980; 255: 2919-2924Abstract Full Text PDF PubMed Google Scholar, 5Nakajima H. Komano H. Esumi-Kurisu M. Abe S. Yamazaki M. Natori S. Mizuno D. Gann. 1982; 73: 627-632PubMed Google Scholar, 6Yamazaki M. Ikenami M. Komano H. Tsunawaki S. Kamiya H. Natori S. Mizuno D. Gann. 1983; 74: 576-583PubMed Google Scholar), and it has also been suggested that this lectin plays an important role during developmental stages.Some C-type lectins have been isolated from marine invertebrates, such as a sea urchin (Anthocidaris crassispina)(7Giga Y. Ikai A. Takahashi K. J. Biol. Chem. 1987; 262: 6197-6203Abstract Full Text PDF PubMed Google Scholar), an acorn barnacle (Megabalanus rosa)(8Muramoto K. Kamiya H. Biochim. Biophys. Acta. 1990; 1039: 42-51Crossref PubMed Scopus (66) Google Scholar), a tunicate (Polyandrocarpa misakiensis)(9Suzuki T. Takagi T. Furukohri T. Kawamura K. Nakauchi M. J. Biol. Chem. 1990; 265: 1274-1281Abstract Full Text PDF PubMed Google Scholar), and a sea cucumber (Stichopus japonicus)(10Himeshima T. Hatakeyama T. Yamasaki N. J. Biochem. (Tokyo). 1994; 115: 689-692Crossref PubMed Scopus (43) Google Scholar). We have recently purified four galactose/GalNAc-specific Ca2+-dependent lectins (CEL-I-CEL-IV) from the marine invertebrate Cucumaria echinata(11Hatakeyama T. Kohzaki H. Nagatomo H. Yamasaki N. J. Biochem. (Tokyo). 1994; 116: 209-214Crossref PubMed Scopus (95) Google Scholar), and we have determined the complete and partial amino acid sequences of CEL-IV and CEL-I, respectively. 2T. Hatakeyama, K. Ohuchi, M. Kuroki, S. Nishinohara, and N. Yamasaki, manuscript in preparation. It was found from the determined sequences that these two lectins have apparent structural homology with other C-type lectins, especially with those from marine invertebrates. Interestingly, CEL-III exhibited strong hemolytic activity when incubated with human and rabbit erythrocytes, while only weak agglutination was observed with chicken and horse erythrocytes. The hemolytic activity of CEL-III was inhibited by several galactose or GalNAc-containing carbohydrates and was prevented in the presence of EDTA, suggesting that the hemolysis was caused by CEL-III binding to galactose or GalNAc-containing carbohydrates on the erythrocyte membrane by Ca2+-dependent lectin activity, followed by partial destruction of the membrane.Hemolytic and cytolytic proteins have been isolated from various origins(12Ryoyama K. Biochim. Biophys. Acta. 1973; 320: 157-165Crossref PubMed Scopus (30) Google Scholar, 13Bertheussen K. Dev. Comp. Immunol. 1983; 7: 21-31Crossref PubMed Scopus (74) Google Scholar, 14Parrinello N. Rindone D. Dev. Comp. Immunol. 1981; 5: 33-42Crossref PubMed Scopus (25) Google Scholar, 15Parrinello N. Rindone D. Canicatti C. Dev. Comp. Immunol. 1979; 3: 45-54Crossref PubMed Scopus (42) Google Scholar, 16Canicatti C. Ciulla D. Dev. Comp. Immunol. 1988; 12: 55-64Crossref PubMed Scopus (20) Google Scholar, 17Canicatti C. Experientia (Basel). 1987; 43: 611-614Crossref PubMed Scopus (20) Google Scholar, 18Mangel A. Leitão J.M. Batel R. Zimmermann H. Müller W.E.G. Schröder H.C. Eur. J. Biochem. 1992; 210: 499-507Crossref PubMed Scopus (35) Google Scholar, 19Shinoda S. Ishida K. Oh E.-G. Sasahara K. Miyoshi S. Chowdhury M.A.R. Yasuda T. Microbiol. Immunol. 1993; 37: 405-409Crossref PubMed Scopus (38) Google Scholar, 20Füssle R. Bhakdi S. Sziegoleit A. Tranum-Jensen J. Krantz T. Wellensiek H.J. J. Cell Biol. 1981; 91: 83-94Crossref PubMed Scopus (258) Google Scholar, 21Lazarovici P. Primor N. Loew L.M. J. Biol. Chem. 1986; 261: 16704-16713Abstract Full Text PDF PubMed Google Scholar, 22Young J.D.-E. Leong L.G. Liu C.-C. Damiano A. Cohn Z.A. Proc. Natl. Acad. Sci. U. S. A. 1986; 83: 5668-5672Crossref PubMed Scopus (51) Google Scholar, 23Kaiser E.T. Kézdy F.J. Annu. Rev. Biophys. Biophys. Chem. 1987; 16: 561-581Crossref PubMed Scopus (159) Google Scholar, 24Park N.G. Lee S. Oishi O. Aoyagi H. Iwanaga S. Yamasita S. Ohno M. Biochemistry. 1992; 31: 12241-12247Crossref PubMed Scopus (45) Google Scholar). Lytic actions of some proteins have been ascribed to enzymatic activity(25Doery H.M. Magnusson B.J. Gulasekharam J. Pearson J.E. J. Gen. Microbiol. 1965; 40: 283-296Crossref PubMed Scopus (96) Google Scholar, 26Macfarlane M.G. Knight B.C.J.G. Biochem. J. 1941; 35: 884-902Crossref PubMed Google Scholar), perturbation of the activities of membrane-associated enzymes(27Hirayama T. Kato I. FEBS Lett. 1983; 157: 46-50Crossref PubMed Scopus (13) Google Scholar), or pore formation in the membranes (28Bhakdi S. Weller U. Walev I. Martin E. Jonas D. Palmer M. Med. Microbiol. Immunol. 1993; 182: 167-175Crossref PubMed Scopus (207) Google Scholar). In this study, we examined the interaction of CEL-III with the erythrocyte membrane as well as with artificial lipid vesicles, in order to elucidate the mechanism for hemolysis by CEL-III.EXPERIMENTAL PROCEDURESMaterialsThe C. echinata samples were a generous gift from N. Ikeda (Fukuoka Fisheries and Marine Technology Research Center). The samples were stored at −30°C until use. Chicken, horse, and rabbit blood samples were obtained from Nippon Bio-Test Laboratories (Tokyo). Inulin and egg yolk phosphatidylcholine were obtained from Nacalai Tesque (Kyoto). Dextran 4 and dextran 8 were from Serva. Luciferase-luciferin and human globoside were products of Sigma. Cellulofine GCL-2000-c was from Seikagaku Kogyo (Tokyo). Peroxidase-conjugated goat anti-mouse IgG was from Organon Teknika (West Chester, PA).Preparation of Lactosyl-Cellulofine and GalNAc-Cellulofine ColumnsThe columns were prepared by cross-linking lactose or GalNAc with Cellulofine GCL-2000-c using divinyl sulfone as described by Teichberg et al.(29Teichberg V.I. Aberdam D. Erez U. Pinelli E. J. Biol. Chem. 1988; 263: 14086-14092Abstract Full Text PDF PubMed Google Scholar).Purification of CEL-IIIIn a previous paper we reported the purification of the four C. echinata lectins by column chromatography using lactosyl-Sepharose 4B, Sephacryl S-200, and Q-Sepharose(11Hatakeyama T. Kohzaki H. Nagatomo H. Yamasaki N. J. Biochem. (Tokyo). 1994; 116: 209-214Crossref PubMed Scopus (95) Google Scholar). In the present study we also used a newly developed procedure, described below, which involved two affinity columns, i.e. lactosyl- and GalNAc-Cellulofine columns, to facilitate the separation of CEL-III.C. echinata body fluid (350 ml) was taken, and CaCl2 was added to 10 mM. This solution was applied to a lactosyl-Cellulofine column (3.6 × 9 cm) equilibrated with TBS (10 mM Tris-HCl buffer, pH 7.5, containing 0.15 M NaCl) and 10 mM CaCl2. After washing the column with the same buffer, adsorbed proteins were eluted with 20 mM EDTA followed by 0.1 M lactose in TBS. CEL-I and CEL-III were eluted from the column with 20 mM EDTA, and CEL-IV with 0.1 M lactose, whereas CEL-II was not retained in this column. The fraction containing a mixture of CEL-I and CEL-III was further separated using the GalNAc-Cellulofine column (1.3 × 2 cm) with the same buffer after dialysis against TBS and addition of CaCl2 to 10 mM. Since CEL-I has an extremely high affinity for GalNAc in addition to a moderate affinity for galactose-containing carbohydrates(11Hatakeyama T. Kohzaki H. Nagatomo H. Yamasaki N. J. Biochem. (Tokyo). 1994; 116: 209-214Crossref PubMed Scopus (95) Google Scholar), CEL-I was bound to the column more strongly than CEL-III, so the former was eluted with 0.1 M lactose, while the latter had to be eluted with 20 mM EDTA. A small amount of CEL-III aggregate was sometimes observed with this preparation, which could be removed by gel filtration on Sephacryl S-200. The CEL-III purified by this method exhibited the same specific activity as that purified by the previous method(11Hatakeyama T. Kohzaki H. Nagatomo H. Yamasaki N. J. Biochem. (Tokyo). 1994; 116: 209-214Crossref PubMed Scopus (95) Google Scholar). Chromatography was performed at 7°C. The proteins thus purified were dialyzed against TBS and stored frozen at −30°C.Protein DeterminationProtein concentrations were determined with bicinchoninic acid by the method given by Smith et al.(30Smith P.K. Krohn R.I. Hermanson G.T. Mallia A.K. Gartner F.H. Provenzano M.D. Fujimoto E.K. Goeke N.M. Olson B.J. Klenke D.C. Anal. Biochem. 1985; 150: 76-85Crossref PubMed Scopus (18445) Google Scholar), using bovine serum albumin as a standard.Determination of Hemolytic ActivityHemolytic activity of CEL-III was determined by the absorbance at 540 nm due to hemoglobin released from the erythrocytes. CEL-III in TBS containing 10 mM CaCl2 (100 μl) was mixed with the same volume of erythrocyte suspension (10%, v/v) in the same buffer. After incubation, the suspension was centrifuged, and lysis was determined by the supernatant's absorbance at 540 nm.AntiserumAnti-CEL-III antiserum was produced in mice using CEL-III purified by the previous method (11Hatakeyama T. Kohzaki H. Nagatomo H. Yamasaki N. J. Biochem. (Tokyo). 1994; 116: 209-214Crossref PubMed Scopus (95) Google Scholar) as the antigen. Initially, 20 μg of protein in complete Freund's adjuvant was injected intraperitoneally. Booster injections of the same amount of the protein in incomplete Freund's adjuvant were administered twice, at 3-week intervals. Blood was taken 1 week after the final injection, and the antiserum was prepared.ImmunoblottingA 5% (v/v) erythrocyte suspension was incubated with CEL-III (10 μg/ml) in 1 ml of TBS containing 10 mM CaCl2 for 15 min at 20°C. After lysis had been completed, the membrane was pelleted by centrifugation at 13,000 × g for 5 min and serially washed twice with 10 mM Tris-HCl, pH 8.0, once with the same buffer containing 0.1 M lactose, and once with the same buffer containing 10 mM EDTA. One-fifth of the membrane pellet was solubilized with the sample buffer containing 1% SDS and then subjected to 12.5% SDS-PAGE. The proteins separated in the polyacrylamide gel were transferred to a nitrocellulose membrane in the transfer buffer (20% ethanol, 25 mM Tris, 192 mM glycine, and 0.1% SDS) for 3 h at 180 mA. The membrane was blocked with 5% nonfat dry milk in TBS (blocking buffer) for 60 min at room temperature, then incubated with mouse anti-CEL-III antiserum (1000-fold dilution), followed by peroxidase-conjugated goat anti-mouse IgG (2000-fold dilution), in the blocking buffer at room temperature for 60 min. Proteins were detected by incubating the membrane with 0.1 M Tris-HCl, pH 7.6, containing 0.08% 3,3′-diaminobenzidine tetrahydrochloride and 0.05% H2O2.Measurement of ATP Release from ErythrocytesATP released from rabbit erythrocytes was measured by the firefly assay(31DeLuca M. McElroy W.D. Methods Enzymol. 1978; 57: 3-15Crossref Scopus (316) Google Scholar). A 10% (v/v) suspension of rabbit erythrocytes was incubated with CEL-III (16 μg/ml) in 0.8 ml of TBS containing 10 mM CaCl2 and 15 mM dextran 8 at 20°C. After centrifugation, 0.2 ml of supernatant was mixed with 0.1 ml of luciferase-luciferin solution (10 mg/ml), and its luminescence was measured at 560 nm (band path 10 nm) using a Hitachi 650-10SC fluorescence spectrophotometer without an excitation light.Preparation of Liposomes Containing CarboxyfluoresceinEgg yolk phosphatidylcholine (20 mg) and human globoside (0.4 mg) were dissolved in 0.5 ml of chloroform and dried under reduced pressure in a conical glass tube. After addition of 1 ml of TBS containing 0.1 M carboxyfluorescein, the lipids were hydrated by vortex mixing for 15 min at room temperature. The suspension was then sonicated for 10 min at room temperature using a Tomy Seiko UR-200P ultrasonic disruptor, and subjected to gel filtration on a column of Cellulofine GCL-2000-c in TBS. The fractions containing small unilamellar vesicles were collected.Measurement of the Release of Carboxyfluorescein from the LiposomesThe liposome trapping 0.1 M carboxyfluorescein in 0.5 ml of TBS containing 10 mM CaCl2 was mixed with CEL-III (0.45 μg/ml) in 0.1 ml of the same buffer at 20°C. The fluorescence of carboxyfluorescein at 523 nm excited at 470 nm was measured at appropriate intervals using a Hitachi 650-10SC fluorescence spectrophotometer.RESULTSTemperature and pH Dependence of the Hemolytic Activity of CEL-IIIOur previous results suggested that hemolysis by CEL-III was caused by the interaction of CEL-III with the erythrocyte membrane after it has bound to the Gal/GalNAc-containing carbohydrates on the erythrocyte surface(11Hatakeyama T. Kohzaki H. Nagatomo H. Yamasaki N. J. Biochem. (Tokyo). 1994; 116: 209-214Crossref PubMed Scopus (95) Google Scholar). To obtain information on the nature of the interaction between CEL-III and the erythrocyte membrane, we examined the hemolytic activity under different conditions. As shown in Fig. 1A, when the hemolytic activity of CEL-III was measured at various temperatures using rabbit erythrocytes, the highest activity was observed at 10°C, but activity decreased with increasing temperature. At 30°C, the hemolysis had decreased to 25% of that at 10°C. It is evident that such a decrease in hemolytic activity at higher temperatures is not due to the instability of the protein, since the hemolytic activity of CEL-III was retained after incubation at up to 50°C for 30 min (Fig. 1B). This suggests that the hemolysis is not an enzymatic reaction and might depend largely on the binding of CEL-III to the specific carbohydrates on the surface of the erythrocyte. On the other hand, the remarkable decrease in the activity below 10°C might be related to reduced membrane fluidity at low temperatures. Fig. 2 shows the pH dependence profile of the hemolytic activity and stability of CEL-III. The hemolytic activity of CEL-III increased with increasing pH up to pH 10, while almost no hemolysis was observed below pH 6.5 (Fig. 2A). However, after pretreatment at different pH values for 15 h, a decrease in hemolytic activity was observed only in the acidic region below pH 5 (Fig. 2B), indicating that the low activity in the acidic region shown in Fig. 2A was not due to irreversible inactivation of the protein. This might reflect the presence of some ionizable group of amino acid residue with a pKa around 9-10, which is involved in the hemolytic activity.Figure 2:pH dependence of the hemolytic activity (A) and pH stability (B) of CEL-III. The following buffers were used for both experiments. pH 2, 0.01 M HCl; pH 3-5.5, 10 mM sodium acetate buffer; pH 6-7, 10 mM Bis-Tris-HCl buffer; pH 7.5-8.5, 10 mM Tris-HCl buffer; pH 9-10, 10 mM sodium borate buffer. A, activity measurements were performed, in duplicate, with 0.5 μg/ml (○) or 5 μg/ml (•) CEL-III and 5% rabbit erythrocytes in the presence of 0.15 M NaCl and 10 mM CaCl2. B, pH stability was measured after treating CEL-III in the buffers at the pH values indicated for 15 h at room temperature, followed by dialysis against TBS containing 10 mM CaCl2. The protein concentration was adjusted to 5 μg/ml. The value for the erythrocytes lysed with 0.1% Triton X-100 (A) or the highest value indicated (B) was taken as 100%.View Large Image Figure ViewerDownload (PPT)Immunoblotting AnalysisSince there was a possibility that the erythrocyte membrane had been damaged by the irreversible binding of CEL-III, the interaction between CEL-III and the erythrocytes was examined by immunoblotting analysis of the proteins solubilized from the erythrocyte membranes treated with CEL-III (Fig. 3). In this experiment, the erythrocyte membranes treated with CEL-III were washed with 10 mM Tris-HCl buffer, pH 8.0, containing 0.1 M lactose, followed by the same buffer containing 10 mM EDTA, to remove any CEL-III bound to the membrane by its carbohydrate binding ability. The proteins solubilized from the membranes were then subjected to SDS-PAGE and immunoblotting analysis. The results clearly showed high molecular mass protein reacting with anti-CEL-III antiserum in the membranes from rabbit and human erythrocytes (Fig. 3A, lanes2 and 3) but not in those from horse and chicken erythrocytes. This suggests that CEL-III was tightly bound to the membranes of rabbit and human erythrocytes in an aggregated form, in accordance with the fact that these erythrocytes are susceptible to the hemolytic action of CEL-III. It is apparent that these aggregates were derived from the CEL-III monomer, since the antiserum used in this experiment was raised against the monomer protein purified by gel filtration on Sephacryl S-200(11Hatakeyama T. Kohzaki H. Nagatomo H. Yamasaki N. J. Biochem. (Tokyo). 1994; 116: 209-214Crossref PubMed Scopus (95) Google Scholar). On the other hand, when native CEL-III was subjected to SDS-PAGE, high molecular mass bands were often observed in addition to the monomer band. These high molecular mass species were also confirmed to be aggregates of CEL-III by the reaction with the antiserum (Fig. 3B), indicating that even native CEL-III has a tendency to form aggregates under the conditions for SDS-PAGE. The aggregates formed during SDS-PAGE are apparently different from those induced by interaction with the erythrocyte membranes, since the former always appear as multiple bands of smaller molecular size than the latter.Figure 3:Immunoblotting of the erythrocyte membranes treated with CEL-III (A) and aggregates of CEL-III formed on SDS-PAGE (B). A, the membranes prepared from the CEL-III-treated erythrocytes were solubilized with sample buffer containing 1% SDS and subjected to 12.5% SDS-PAGE under nonreducing conditions. After electrophoretic transfer to a nitrocellulose membrane, CEL-III was detected with mouse anti-CEL-III antiserum. Lane 1, chicken erythrocytes; lane 2, rabbit erythrocytes; lane 3, human erythrocytes; lane 4, horse erythrocytes. B, SDS-PAGE (lane 1) and its immunoblotting (lane 2) of native CEL-III.View Large Image Figure ViewerDownload (PPT)Osmotic Protection of Erythrocytes by CarbohydratesSince CEL-III was irreversibly bound to the membranes of the erythrocytes susceptible to its hemolytic action, it was postulated that the hemolytic activity of CEL-III was due to the formation of ion-permeable pores in the membrane, as with some cytolytic proteins(17Canicatti C. Experientia (Basel). 1987; 43: 611-614Crossref PubMed Scopus (20) Google Scholar, 18Mangel A. Leitão J.M. Batel R. Zimmermann H. Müller W.E.G. Schröder H.C. Eur. J. Biochem. 1992; 210: 499-507Crossref PubMed Scopus (35) Google Scholar, 19Shinoda S. Ishida K. Oh E.-G. Sasahara K. Miyoshi S. Chowdhury M.A.R. Yasuda T. Microbiol. Immunol. 1993; 37: 405-409Crossref PubMed Scopus (38) Google Scholar, 20Füssle R. Bhakdi S. Sziegoleit A. Tranum-Jensen J. Krantz T. Wellensiek H.J. J. Cell Biol. 1981; 91: 83-94Crossref PubMed Scopus (258) Google Scholar, 21Lazarovici P. Primor N. Loew L.M. J. Biol. Chem. 1986; 261: 16704-16713Abstract Full Text PDF PubMed Google Scholar, 22Young J.D.-E. Leong L.G. Liu C.-C. Damiano A. Cohn Z.A. Proc. Natl. Acad. Sci. U. S. A. 1986; 83: 5668-5672Crossref PubMed Scopus (51) Google Scholar). Therefore, we examined this possibility by performing the osmotic protection experiment using several carbohydrates as protectants. As shown in Fig. 4, when rabbit erythrocytes were incubated with CEL-III in the presence of several carbohydrates with varying molecular masses, lysis was inhibited increasingly as the size of the carbohydrates increased; sucrose and melezitose gave only slight protection against lysis of rabbit erythrocytes, whereas inulin, dextran 4 (4-6 kDa), and dextran 8 (8-12 kDa) gave 62%, 98%, and 99% protection against lysis, respectively. Moreover, washing the erythrocytes treated with CEL-III in the presence of 15 mM dextran 8 with TBS resulted in immediate lysis of the cells. These results suggest that the erythrocytes were ruptured by colloid-osmotic shock after CEL-III had formed ion-permeable pores in the membrane.Figure 4:Osmotic protection against hemolysis by various carbohydrates. The hemolysis of rabbit erythrocytes was measured in the presence of carbohydrates (15 mM) as indicated. Measurements were performed in duplicate with 5% rabbit erythrocytes and 5 μg/ml CEL-III in TBS containing 10 mM CaCl2.View Large Image Figure ViewerDownload (PPT)ATP Release from Erythrocytes Treated with CEL-IIIIn order to confirm the formation of the transmembrane pores by CEL-III, the release of ATP from the erythrocytes treated with CEL-III in the presence of dextran 8 was examined by the firefly luciferase assay (31DeLuca M. McElroy W.D. Methods Enzymol. 1978; 57: 3-15Crossref Scopus (316) Google Scholar). As shown in Fig. 5, ATP was released from rabbit erythrocytes protected with 15 mM dextran 8 when the cells were incubated with CEL-III, supporting the above idea. The increase in ATP release paralleled the hemolysis measured in the absence of dextran 8 with a slight delay, showing that the formation of ion-permeable pores preceded the hemolysis. In addition to rabbit erythrocytes, ATP release was also detected with human erythrocytes, but not with chicken and horse cells, in agreement with the susceptibility of these cells to the hemolytic action of CEL-III.Figure 5:CEL-III-induced release of ATP from rabbit erythrocytes osmotically protected by dextran 8, and hemolysis in the absence of dextran 8. Rabbit erythrocytes (5%) were incubated with CEL-III (5 μg/ml) in TBS containing 10 mM CaCl2 in the presence (○) or absence (•) of 15 mM dextran 8. The ATP released from the erythrocytes was measured by the firefly luciferase assay(31DeLuca M. McElroy W.D. Methods Enzymol. 1978; 57: 3-15Crossref Scopus (316) Google Scholar). The highest values were taken as 100%.View Large Image Figure ViewerDownload (PPT)Interaction of CEL-III with LiposomesThe interaction of CEL-III with the membrane was further examined by using artificial lipid vesicles as shown in Fig. 6. When CEL-III was added to a solution of carboxyfluorescein-trapping liposomes prepared with egg yolk phosphatidylcholine and a small amount of human globoside (2%, w/w) as a receptor, the fluorescence intensity at 523 nm was enhanced, due to the carboxyfluorescein released from the liposomes. At the same time, a slight turbidity was seen in the solution, due to cross-linking of the liposomes by CEL-III. In contrast, the increase in fluorescence intensity was much smaller both in the solution without the protein and in the presence of 0.1 M lactose. The release of carboxyfluorescein from the liposomes treated with CEL-III was quite slow compared with the hemolysis and ATP release form osmotically protected erythrocytes (Fig. 5).Figure 6:Interaction of CEL-III with the artificial lipid vesicles. CEL-III was incubated with liposomes trapping carboxyfluorescein in TBS containing 10 mM CaCl2 at 20°C in the absence (○) and presence (•) of 0.1 M lactose, and the increase in fluorescence at 523 nm was recorded with excitation at 470 nm. □, blank solution without CEL-III. The fluorescence intensity of the solution after treatment with 0.1% Triton X-100 was taken as 100%.View Large Image Figure ViewerDownload (PPT)DISCUSSIONSince galactose- or GalNAc-containing carbohydrates effectively inhibited the hemolysis, it was suggested that CEL-III exerted lytic action by damaging the erythrocyte membrane, after binding to the specific carbohydrates on the cell surface(11Hatakeyama T. Kohzaki H. Nagatomo H. Yamasaki N. J. Biochem. (Tokyo). 1994; 116: 209-214Crossref PubMed Scopus (95) Google Scholar). In order to elucidate this mechanism, the interactions of CEL-III with the erythrocyte membrane and with artificial lipid vesicles were investigated. When lysis of rabbit erythrocytes by CEL-III was measured under different conditions, one of the characteristic features was temperature dependence. The highest activity was exhibited at a low temperature (10°C), suggesting that hemolysis by CEL-III may be a nonenzymatic reaction. Furthermore, the composition of the major lipids in the rabbit erythrocyte membrane, e.g. phosphatidylcholine, phosphatidylethanolamine, and sphingomyelin, did not change after treatment with CEL-III, as confirmed by thin-layer chromatography (data not shown). Therefore, we postulated that hemolysis by CEL-III was caused by formation of transmembrane pores, as with some bacterial and invertebrate hemolysins(12Ryoyama K. Biochim. Biophys. Acta. 1973; 320: 157-165Crossref PubMed Scopus (30) Google Scholar, 13Bertheussen K. Dev. Comp. Immunol. 1983; 7: 21-31Crossref PubMed Scopus (74) Google Scholar, 14Parrinello N. Rindone D. Dev. Comp. Immunol. 1981; 5: 33-42Crossref PubMed Scopus (25) Google Scholar, 15Parrinello N. Rindone D. Canicatti C. Dev. Comp. Immunol. 1979; 3: 45-54Crossref PubMed Scopus (42) Google Scholar, 16Canicatti C. Ciulla D. Dev. Comp. Immunol. 1988; 12: 55-64Crossref PubMed Scopus (20) Google Scholar, 17Canicatti C. Experientia (Basel). 1987; 43: 611-614Crossref PubMed Scopus (20) Google Scholar, 18Mangel A. Leitão J.M. Batel R. Zimmermann H. Müller W.E.G. Schröder H.C. Eur. J. Biochem. 1992; 210: 499-507Crossref PubMed Scopus (35) Google Scholar, 19Shinoda S. Ishida K. Oh E.-G. Sasahara K. Miyoshi S. Chowdhury M.A.R. Yasuda T. Microbiol. Immunol. 1993; 37: 405-409Crossref PubMed Scopus (38) Google Scholar, 20Füssle R. Bhakdi S. Sziegoleit A. Tranum-Jensen J. Krantz T. Wellensiek H.J. J. Cell Biol. 1981; 91: 83-94Crossref PubMed Scopus (258) Google Scholar). To test this postulate, the binding of CEL-III to the erythrocyte membrane was analyzed by immunoblotting the proteins bound to the CEL-III-treated membranes (Fig. 3). Irreversibly bound proteins that reacted with anti-CEL-III antiserum were detected in high molecular mass form only in those erythrocytes susceptible to hemolysis by CEL-III, such as human and rabbit cells. This result suggested that CEL-III aggregated in the membrane after binding to the carbohydrates on the erythrocyte surface by lectin activity and then formed ion-permeable transmembrane pores so that the erythrocytes were ruptured by colloid-osmotic shock. Since dextran 4 (4-6 kDa) and dextran 8 (8-12 kDa) showed remarkable osmotic protection against lysis, the pores formed by the aggregates of CEL-III may have a functional radius smaller than 1.75 nm(32Scherrer R. Gerhardt P. J. Bacteriol. 1971; 107: 718-735Crossref PubMed Google Scholar). In this case the carbohydrate binding activity of CEL-III is obviously not directly responsible for lysis, since agglutination was observed instead of lysis when the erythrocytes were protected by the dextrans. This result also indicates that CEL-III has more than one carbohydrate-binding site per single polypeptide chain, since CEL-III is a monomeric protein in native form as confirmed by gel-permeation high performance liquid chromatography (data not shown).Native CEL-III exhibited multiple high molecular mass bands on SDS-PAGE (Fig. 3B), although no aggregates were observed in this CEL-III preparation when analyzed by gel-permeation high performance liquid chromatography (data not shown). These aggregates seemed to be induced by interaction with SDS and were different from those induced by interaction with the erythrocyte membrane. In fact, aggregate formation was also observed when CEL-III was incubated with other detergents, e.g. Triton X-100, Tween 20, and sodium deoxycholate. 3T. Hatakeyama, M. Furukawa, H. Nagatomo, and N. Yamasaki, unpublished results. This is similar to the case of Staphylococcus aureus α-toxin, which forms hexamers on treatment with deoxycholate. After binding to the erythrocyte membrane as a monomer, S. aureus α-toxin forms transmembrane pores composed of toxin hexamers induced by interaction with the membrane (33Walker B. Krishnasastry M. Zorn L. Bayley H. J. Biol. Chem. 1992; 267: 21782-21786Abstract Full Text PDF PubMed Google Scholar). Likewise, detergents such as SDS and deoxycholate might also induce a conformational change in CEL-III, promoting aggregation. This appears to be closely related to the fact that CEL-III forms aggregates irreversibly bound to the membrane when incubated with erythrocytes susceptible to lysis by CEL-III. It is probable that CEL-III forms aggregates in partially hydrophobic environments like detergent solutions. CEL-III pore formation may also be triggered by its conformational change in the partially hydrophobic environment at the surface of the membrane following binding to the galactose or GalNAc-containing carbohydrates of the erythrocytes.The experiment shown in Fig. 6 demonstrated that CEL-III induced the release of carboxyfluorescein trapped in liposomes consisting of egg yolk phosphatidylcholine and human globoside. Release of carboxyfluorescein increased with time following a rapid agglutination of the liposomes, which indicates that CEL-III binds to the vesicles via the carbohydrate moiety of globoside containing GalNAc at its nonreducing end. Although the release of carboxyfluorescein was much slower than the hemolysis, this result demonstrated that CEL-III could also form transmembrane pores in the artificial lipid vesicles. The difference in the apparent rate of reactions between liposomes and erythrocytes might be related to the composition of their lipid or carbohydrate receptors.Of the C. echinata lectins, CEL-I and CEL-IV were found to have apparent homology with C-type CRDs, like the other invertebrate Ca2+-dependent lectins whose primary structures have been determined. Therefore, CEL-III can also be expected to belong to the C-type lectin family because of the Ca2+-dependent nature of its hemolytic and hemagglutinating activity. Since C-type CRDs usually consist of 120-130 amino acid residues, the extra polypeptide portion of CEL-III (45 kDa) might be responsible for other function(s), e.g. for membrane binding or aggregate formation. Alternatively, it also seems possible that CEL-III has a structure consisting of multiple CRDs like the structure of the macrophage mannose receptor(34Taylor M.E. Drickamer K. J. Biol. Chem. 1993; 268: 399-404Abstract Full Text PDF PubMed Google Scholar).There have been few reports of lectins with cytolytic activity, whereas recently a hemolytic lectin from the mushroom Laetiporus sulfureus and egg vitelline coat lysins from Mytilus edulis sperm, the latter with homology with C-type lectins, have been reported(35Konska G. Guillot J. Dusser M. Damez M. Botton B. J. Biochem. (Tokyo). 1994; 116: 519-523Crossref PubMed Scopus (50) Google Scholar, 36Takagi T. Nakamura A. Deguchi R. Kyozuka K. J. Biochem. (Tokyo). 1994; 116: 598-605Crossref PubMed Scopus (40) Google Scholar). Although their lytic mechanisms are not clear, it would be interesting to investigate their relationship with CEL-III. Many invertebrate lectins are thought to be involved in the defense mechanisms of organisms, not only neutralizing foreign substances by binding to their carbohydrate moieties, but also by activating phagocytes(5Nakajima H. Komano H. Esumi-Kurisu M. Abe S. Yamazaki M. Natori S. Mizuno D. Gann. 1982; 73: 627-632PubMed Google Scholar, 6Yamazaki M. Ikenami M. Komano H. Tsunawaki S. Kamiya H. Natori S. Mizuno D. Gann. 1983; 74: 576-583PubMed Google Scholar). CEL-III may be a novel type of invertebrate Ca2+-dependent lectin that can act directly as a toxic protein to foreign microorganisms. In fact, CEL-III was found to exhibit cytotoxicity to some cultured cells. 4T. Hatakeyama, H. Nagatomo, N. Yamasaki, and T. Oda, unpublished results. Pore-forming proteins have been purified from some invertebrates (12Ryoyama K. Biochim. Biophys. Acta. 1973; 320: 157-165Crossref PubMed Scopus (30) Google Scholar, 13Bertheussen K. Dev. Comp. Immunol. 1983; 7: 21-31Crossref PubMed Scopus (74) Google Scholar, 14Parrinello N. Rindone D. Dev. Comp. Immunol. 1981; 5: 33-42Crossref PubMed Scopus (25) Google Scholar, 15Parrinello N. Rindone D. Canicatti C. Dev. Comp. Immunol. 1979; 3: 45-54Crossref PubMed Scopus (42) Google Scholar, 16Canicatti C. Ciulla D. Dev. Comp. Immunol. 1988; 12: 55-64Crossref PubMed Scopus (20) Google Scholar, 17Canicatti C. Experientia (Basel). 1987; 43: 611-614Crossref PubMed Scopus (20) Google Scholar, 18Mangel A. Leitão J.M. Batel R. Zimmermann H. Müller W.E.G. Schröder H.C. Eur. J. Biochem. 1992; 210: 499-507Crossref PubMed Scopus (35) Google Scholar), although they are not lectins. It is probable that the binding ability of CEL-III to galactose or GalNAc-containing carbohydrates makes this protein more specific to its target organisms. INTRODUCTIONAnimal lectins have been of great interest recently, as their various important functions in organisms are being suggested. Most animal lectins have been classified into one of two groups, i.e. Ca2+-dependent (C-type) and independent (S-type or galectin)(1Drickamer K. J. Biol. Chem. 1988; 263: 9557-9560Abstract Full Text PDF PubMed Google Scholar, 2Barondes S.H. Castronovo V. Cooper D.N.W. Cummings R.D. Drickamer K. Feizi T. Gitt M.A. Hirabayashi J. Hughes C. Kasai K. Leffler H. Liu F.-T. Lotan R. Mercurio A.M. Monsigny M. Pillai S. Poirer F. Raz A. Rigby P.W.J. Rini J.M. Wang J.L. Cell. 1994; 76: 597-598Abstract Full Text PDF PubMed Scopus (1085) Google Scholar). In vertebrates, C-type lectins are classified into six groups, which include proteoglycan core protein, hepatic lectin, mannose-binding protein, and selectin(3Drickamer K. Biochem. Soc. Trans. 1993; 21: 456-459Crossref PubMed Scopus (43) Google Scholar). These proteins are composed of C-type CRDs, 1The abbreviations used are: CRDcarbohydrate recognition domainPAGEpolyacrylamide gel electrophoresisTBSTris-buffered saline. which exhibit some degree of homology with each other, and additional domains or regions with the characteristic functions of individual proteins. C-type lectins purified from invertebrates, on the other hand, are generally smaller than those from vertebrates, and many of them consist of only a single CRD. One of the most probable roles of invertebrate C-type lectins is to act as humoral factors in the defense mechanism, as do immunoglobulins in vertebrates; enhanced fly lectin production after injury to the body wall and its activation of phagocytes in vitro have been demonstrated(4Komano H. Mizuno D. Natori S. J. Biol. Chem. 1980; 255: 2919-2924Abstract Full Text PDF PubMed Google Scholar, 5Nakajima H. Komano H. Esumi-Kurisu M. Abe S. Yamazaki M. Natori S. Mizuno D. Gann. 1982; 73: 627-632PubMed Google Scholar, 6Yamazaki M. Ikenami M. Komano H. Tsunawaki S. Kamiya H. Natori S. Mizuno D. Gann. 1983; 74: 576-583PubMed Google Scholar), and it has also been suggested that this lectin plays an important role during developmental stages.Some C-type lectins have been isolated from marine invertebrates, such as a sea urchin (Anthocidaris crassispina)(7Giga Y. Ikai A. Takahashi K. J. Biol. Chem. 1987; 262: 6197-6203Abstract Full Text PDF PubMed Google Scholar), an acorn barnacle (Megabalanus rosa)(8Muramoto K. Kamiya H. Biochim. Biophys. Acta. 1990; 1039: 42-51Crossref PubMed Scopus (66) Google Scholar), a tunicate (Polyandrocarpa misakiensis)(9Suzuki T. Takagi T. Furukohri T. Kawamura K. Nakauchi M. J. Biol. Chem. 1990; 265: 1274-1281Abstract Full Text PDF PubMed Google Scholar), and a sea cucumber (Stichopus japonicus)(10Himeshima T. Hatakeyama T. Yamasaki N. J. Biochem. (Tokyo). 1994; 115: 689-692Crossref PubMed Scopus (43) Google Scholar). We have recently purified four galactose/GalNAc-specific Ca2+-dependent lectins (CEL-I-CEL-IV) from the marine invertebrate Cucumaria echinata(11Hatakeyama T. Kohzaki H. Nagatomo H. Yamasaki N. J. Biochem. (Tokyo). 1994; 116: 209-214Crossref PubMed Scopus (95) Google Scholar), and we have determined the complete and partial amino acid sequences of CEL-IV and CEL-I, respectively. 2T. Hatakeyama, K. Ohuchi, M. Kuroki, S. Nishinohara, and N. Yamasaki, manuscript in preparation. It was found from the determined sequences that these two lectins have apparent structural homology with other C-type lectins, especially with those from marine invertebrates. Interestingly, CEL-III exhibited strong hemolytic activity when incubated with human and rabbit erythrocytes, while only weak agglutination was observed with chicken and horse erythrocytes. The hemolytic activity of CEL-III was inhibited by several galactose or GalNAc-containing carbohydrates and was prevented in the presence of EDTA, suggesting that the hemolysis was caused by CEL-III binding to galactose or GalNAc-containing carbohydrates on the erythrocyte membrane by Ca2+-dependent lectin activity, followed by partial destruction of the membrane.Hemolytic and cytolytic proteins have been isolated from various origins(12Ryoyama K. Biochim. Biophys. Acta. 1973; 320: 157-165Crossref PubMed Scopus (30) Google Scholar, 13Bertheussen K. Dev. Comp. Immunol. 1983; 7: 21-31Crossref PubMed Scopus (74) Google Scholar, 14Parrinello N. Rindone D. Dev. Comp. Immunol. 1981; 5: 33-42Crossref PubMed Scopus (25) Google Scholar, 15Parrinello N. Rindone D. Canicatti C. Dev. Comp. Immunol. 1979; 3: 45-54Crossref PubMed Scopus (42) Google Scholar, 16Canicatti C. Ciulla D. Dev. Comp. Immunol. 1988; 12: 55-64Crossref PubMed Scopus (20) Google Scholar, 17Canicatti C. Experientia (Basel). 1987; 43: 611-614Crossref PubMed Scopus (20) Google Scholar, 18Mangel A. Leitão J.M. Batel R. Zimmermann H. Müller W.E.G. Schröder H.C. Eur. J. Biochem. 1992; 210: 499-507Crossref PubMed Scopus (35) Google Scholar, 19Shinoda S. Ishida K. Oh E.-G. Sasahara K. Miyoshi S. Chowdhury M.A.R. Yasuda T. Microbiol. Immunol. 1993; 37: 405-409Crossref PubMed Scopus (38) Google Scholar, 20Füssle R. Bhakdi S. Sziegoleit A. Tranum-Jensen J. Krantz T. Wellensiek H.J. J. Cell Biol. 1981; 91: 83-94Crossref PubMed Scopus (258) Google Scholar, 21Lazarovici P. Primor N. Loew L.M. J. Biol. Chem. 1986; 261: 16704-16713Abstract Full Text PDF PubMed Google Scholar, 22Young J.D.-E. Leong L.G. Liu C.-C. Damiano A. Cohn Z.A. Proc. Natl. Acad. Sci. U. S. A. 1986; 83: 5668-5672Crossref PubMed Scopus (51) Google Scholar, 23Kaiser E.T. Kézdy F.J. Annu. Rev. Biophys. Biophys. Chem. 1987; 16: 561-581Crossref PubMed Scopus (159) Google Scholar, 24Park N.G. Lee S. Oishi O. Aoyagi H. Iwanaga S. Yamasita S. Ohno M. Biochemistry. 1992; 31: 12241-12247Crossref PubMed Scopus (45) Google Scholar). Lytic actions of some proteins have been ascribed to enzymatic activity(25Doery H.M. Magnusson B.J. Gulasekharam J. Pearson J.E. J. Gen. Microbiol. 1965; 40: 283-296Crossref PubMed Scopus (96) Google Scholar, 26Macfarlane M.G. Knight B.C.J.G. Biochem. J. 1941; 35: 884-902Crossref PubMed Google Scholar), perturbation of the activities of membrane-associated enzymes(27Hirayama T. Kato I. FEBS Lett. 1983; 157: 46-50Crossref PubMed Scopus (13) Google Scholar), or pore formation in the membranes (28Bhakdi S. Weller U. Walev I. Martin E. Jonas D. Palmer M. Med. Microbiol. Immunol. 1993; 182: 167-175Crossref PubMed Scopus (207) Google Scholar). In this study, we examined the interaction of CEL-III with the erythrocyte membrane as well as with artificial lipid vesicles, in order to elucidate the mechanism for hemolysis by CEL-III.
Referência(s)