Biological activities of steroid glycosides from starfish

Comparative Biochemistry and Physiology Part B 134 (2003) 695–701

Biological activities of steroid glycosides from starfish Nina G. Prokof’eva*, Elena L. Chaikina, Alla A. Kicha, Natalya V. Ivanchina Pacific Institute of Bioorganic Chemistry, Far East Branch of the Russian Academy of Sciences, 690022 Vladivostok, Russia Received 1 August 2002; received in revised form 20 January 2003; accepted 21 January 2003

Abstract Glycosides of polyhydroxysteroids from starfish were compared with regard to hemolytic activities on mouse red blood cells and cytotoxicity on developing eggs of sea urchin Strongylocentrotus intermedius. Mediasteroside M1 from Mediaster murrayi with a 2-O-methyl group at a xylose residue exerted lower activity than mediasteroside M2 without this group. Ceramasteroside C2 from Ceramaster patagonicus having a 2,4-di-O-Me-Xylp residue had more activity in comparison with ceramasteroside C3 with 2-O-Me-Xylp. Activities of rathbuniosides R1 (four OH-groups) and R2 (five OH-groups) from Asterias rathbuni, and culcitosides C2 (six OH-groups) and C3 (five OH-groups) from Culcita novaeguineae were inversely proportional to the number of hydroxyl groups of the aglycon. There was a correlation between cytotoxic and hemolytic activities. The results indicated that salt concentration, temperature and pH values are important for the hemolytic activity of steroid glycosides. 䊚 2003 Elsevier Science Inc. All rights reserved. Keywords: Steroid glycosides; Starfish; Cytotoxic and hemolytic activities; pH; Temperature; Ionic strength

1. Introduction Structural studies on marine polyhydroxysteroids, glycosides of polyhydroxysteroids from the starfish are currently building up rapidly, largely exceeding the biological studies. Many of these compounds were reported to possess antifungal (D’Auria et al., 1990; Choi et al., 1999), antiviral, antineoplastic activities, anti-inflammatory and ichthyotoxic effects (Andersson et al., 1989), and cytotoxicity against developing eggs of sea urchin (Bruno et al., 1990; Levina et al., 2002). During studies of compounds from marine organisms we have isolated new steroid glycosides (Kicha et al., 1997, 1999; Ivanchina et al., 2000, 2001). It was shown the distribution of toxic steroid oligogly*Corresponding author. Fax: q7-4232-314050. E-mail addresses: [email protected] (N.G. Prokof’eva), [email protected] (N.G. Prokof’eva).

cosides in various body components of the starfish Patiria pectinifera and that sulfated cholestane monoglycoside ‘asterosaponin’ P1 is able to solubilize of the lipid suspension similarly to bile acids (Kicha et al., 2001). In this paper, we report the structure–activity interrelationships on hemolytic and cytotoxic effects of starfish polyhydroxysteroid glycosides and the influence of ionic strength, temperature and pH on their hemolytic activities. 2. Materials and methods 2.1. Compounds Mediasterosides M1 (1) and M2 (2) have been isolated from the deepwater starfish Mediaster murrayi, collected in the Philippines Sea (Kicha et al., 1999). Culcitosides C2 (3) and C3 (4) have been isolated from the starfish Culcita novaeguineae, collected at the Northwestern coast of the

1096-4959/03/$ - see front matter 䊚 2003 Elsevier Science Inc. All rights reserved. doi:10.1016/S1096-4959(03)00029-0

696

N.G. Prokof’eva et al. / Comparative Biochemistry and Physiology Part B 134 (2003) 695–701

Fig. 1. Structures of polyhydroxysteroid glycosides from starfish.

Madagascar (Kicha et al., 1986). Ceramasterosides C2 (5) and C3 (6) have been isolated from the starfish Ceramaster patagonicus, collected in the Okhotsk Sea (Kicha et al., 1997). Rathbuniosides R1 (7) and R2 (8) have been isolated from the starfish Asterias rathbuni, collected in the Bering Sea (Ivanchina et al., 2001) (Fig. 1). 2.2. Cytotoxicity Eggs and sperm from the sea urchin Strongylocentrotus intermedius gonads, collected at the Troitsa Bay of the Sea of Japan (Russia) were used. Eggs fertilization and cleavage were determined following the procedure described by Biyiti et al. (1990). Eggs were rinsed, filtered and diluted by natural seawater (NSW) to a concentration of

2000 eggs mly1. Sperm was collected ‘dry’ and shortly before use the semen was diluted (1:50) in NSW. Fertilization was obtained by adding 10 ml of sperm to 1 ml of eggs suspension at 20 8C. The percentage of divided cells was determined by light microscopy. The compounds were used in ethanol solution; the maximum concentration of 1% ethanol used in the experiments did not affect cell division. We estimated the minimum inhibiting concentration (MIC), defined as lowest concentration that totally inhibits cleavage of the eggs 30 min after the beginning of cell division in the control. 2.3. Hemolysis The method described by Kalinin et al. (1996) was employed to study hemolytic activity of gly-

N.G. Prokof’eva et al. / Comparative Biochemistry and Physiology Part B 134 (2003) 695–701

cosides. An erythrocyte suspension was prepared in 66 mM PBS (120 mM NaCl, 4 mM KCl, pH 7.4 or 5.5). A concentration of erythrocytes giving an optical density (OD) of 1.0 at 700 nm for nonhemolysed sample was used. Twenty microlitre of ethanolic solution of test substance were mixed with 2 ml of the erythrocyte suspension. (A) The samples were then incubated for 15, 30, 60 min at 37 or 45 8C and the OD of the red blood cells suspension was evaluated. Hemolytic activity of the compounds was expressed as molar concentration of glycoside causing 50% hemolysis (HC50). (B). Immediately following the addition of the test substance to the suspension the OD of each sample was monitored continuously at 700 nm in a spectrophotometer at room temperature for 5 min. Hemolysis was quantitated by determining the slope of the curves at 50% hemolysis and by the duration of the delay. 2.4. Statistics Data were analyzed by Student’s t-test. Differences from controls were considered significant at P-0.05. 3. Results and discussion The polyhydroxysteroid glycosides studied are amphiphilic molecules, consisting of different sugar groups linked by a glycosidic bond to a hydrophobic steroid nucleus. Glycosides from starfish have a broad spectrum of biological activity and membranotropic action plays an important role in their activity. It was shown that polyhydroxylated steroidal glycoside, asterosaponin P1, increased the ion permeability of bilayer lipid membranes, formatting of the ion channels (Likhatskaya and Kicha, 1997). It has been established that biological activity of starfish steroidal glycosides is dependent on the number of monosaccharide residues and type side-chain (Voogt and Van Rheenen, 1982; Komori, 1997). However, questions concerning the influence structures of a carbohydrate chain and the aglycone part on activity remained unresolved. In continuation of our investigation on starfish, we report herein the structure–activity interrelationships of polyhydroxysteroid glycosides (Fig. 1) on cytotoxic and hemolytic effects. 3.1. Cytotoxic and hemolytic potency Sea urchin gamete is an attractive bioassay since it is easy to obtain fertilized eggs, which give rise

697

Table 1 Cytotoxic and hemolytic activities of steroid glycosides 1–8 from starfish Compound

MICa

HC50b

1 2 3 4 5 6 7 8

68.0 40.0 120.0 50.0 62.0 150.0 74.0 10.0

132.0 66.0 100.0 50.0 70.0 120.0 45.0 97.0

a The concentration (in mM) that inhibits cleavage of the eggs 30 min after the beginning of cell division in the control. b Concentration of glycoside in (mM) causing 50% hemolysis at 37 8C in 30 min. The data shown are the mean values of three experiments. The S.E. values were less than 3% of the mean.

to cells dividing synchronously and have been recommended as model for bioassay and for investigation of allelopathic action. This model rapidly provides information on disruption of cell proliferation, calcium transport and enables targets of toxic substances to be investigated (Jacobs and Wilson, 1986; Biyiti et al., 1990; Pesando et al., 1991). In the present work we have shown that glycosides 1–8 blocked the first cell cleavage of sea urchin eggs, showing the MIC values ranging between 40.0 and 170.0 mM (Table 1). The dose response is remarkably steep as concentrations at least two times lower did not produce effect on the first cell cleavage. In this case, the eggs treated with glycosides induced polynucleated embryos and anomalies in the embryonic development. All saponins inhibited sea urchin embryos from developing further than the morula stage (data not shown). It is known that the action of steroid glycosides isolated from marine organisms is based on its ability to bind with membrane cholesterol (Elyakov and Stonik, 1988). Glycosides 1–8 inhibited the growth of sea urchin embryos, particularity those containing cholesterol in membrane lipids (Anisimov et al., 1983). Although the structural differences of compounds 1 and 2, 3 and 4, 5 and 6, 7 and 8 were minimal at first sight, the biological activity showed remarkable differences. The potency of compounds 4 and 7 was reduced by addition of a hydroxyl group to the aglycon. The addition of a methyl group at glycoside residue decreased the MIC value from 62.0 mM for 5 to 150 mM for 6. Lower striking was the difference between activities of mediasterosides 1

698

N.G. Prokof’eva et al. / Comparative Biochemistry and Physiology Part B 134 (2003) 695–701

and 2. Cytotoxic effects of the compounds were the result of cell membrane damage, since glycosides 1–8 (Table 1), in contrast with starfish sapogenins (Gorshkov et al., 1998) and disulfated polyhydroxysteroids (Aminin et al., 1995), possessed hemolytic activities. Table 1 showed that the glycosides displayed a moderate hemolytic activities: concentration of glycosides causing 50% hemolysis in 30 min incubation ranging between 45.0 and 132.0 mM. There was a correlation between their structure–activity relationships on cytotoxic and hemolytic activities. It confirmed a common mechanism of their action against membranes of erythrocytes and urchin eggs. Comparison of pairs: compounds 1 and 2, 5 and 6 showed that cytotoxic and hemolytic activities were depended on the presence of the O-methyl group and the number methyl groups at xylose residue. Thus, mediasteroside M1 (1) possessing a 2-Omethyl group at the xylose residue had lower activities than mediasteroside M2 (2) without this group. Ceramasteroside C2 (5) having 2,4-di-OMe-Xylp residue was more effective than ceramasteroside C3 (6) with 2-O-Me-Xylp . The variations in the activities of glycosides may be due to differences in structure as the hydrophilic head groups as steroidal subunits. Our data confirmed that the hemolytic activities of glycosides 3 and 4, 7 and 8, as their cytotoxic activity, were distinctly different based on the number hydroxyl groups in the aglycon. Culcitoside C3 (4) with five OHgroups and rathbunioside R1 (7) with four OHgroups in the aglycon were stronger than culcitoside C2 (3) with six OH-groups and rathbunioside R2 (8) with five OH-groups, respectively. Compounds with additional hydroxyl groups were less active in other investigations (Bergquist et al., 1993; Kuznetsova et al., 1999; Afiyatullov et al., 2000). It is known that certain mollusks exhibit escape responses to the presence of predatory starfish in their environment (Mackie and Turner, 1970) and that a biological function of oligoglycoside asterosaponins in starfish is to protect the animals against predators (Mackie et al., 1977). Thus, we showed that glycosides 1–8, as others polyhydroxysteroid glycosides (Komori, 1997; Levina et al., 2002), are cytotoxic compounds and their toxic effect is due to membranotropic action.

Table 2 The pH and temperature dependence on the hemolytic activities of culcitosides C2 and C3a Compound

3 4

pH 7.4

pH 5.5

37 8C

45 8C

37 8C

45 8C

125.0"5.0 84.0"3.2

40.0"2.7 25.0"1.8

80.0"4.5 48.0"3.0

25.0"1.6 12.5"2.6

a Concentration of glycoside (in mM) causing 50% hemolysis in 15 min (HC50 values). Values are means"S.E., ns4.

3.2. pH and temperature Since the pK of membranous oleic acid is close to the physiological pH (Kramer et al., 1997), small pH variations about the physiological pH may cause large differences in compound–lipid membrane interactions (Ilani and Granoth, 1990; Prokof’eva et al., 1990, 1999). Temperature is also a major factor for the biological activity of amphyphilic compounds (Kuznetsova et al., 1999; Prokof’eva et al., 1990, 1992, 1999). We have investigated culcitosides C2 (3) and C3 (4) membrane damaging effects on erythrocytes at 37 and 45 8C (Table 2 and Fig. 2). An increase of temperature produced a noticeable increase in the hemolytic activity of culcitosides, with compound 3 having less active at both temperatures than 4. The half-maximum hemolysis induced by 3 at 37 8C was seen at 60 min, but at 45 8C for 15 min. At 37 8C compound 4 induced 50% hemolysis of erythrocytes in 22 min; however, at 45 8C 50% erythrocytes were disrupted in 15 min (Fig. 3). In addition, solution temperature and pH affect hemolytic activities of culcitosides. The half-maximum hemolysis at 45 8C and pH 5.5 was seen for 25.0 and 12.5 mM compounds 3 and 4, respectively (Table 2). Hence, the general implication from the results that additional hydroxyl group in ring A of glycosides is due to decreased activity at all experimental conditions. Hemolytic activity of culcitosides is a stronger function of medium temperature than of pH value. 3.3. The role of ionic strength Membranotropic capacity of polyhydroxysteroid glycosides is complex, because it can be affected by the structure of compounds, temperature, pH and also salt concentration. The data showed that compounds 5 and 6 (100 mM) at 0.2 M NaCl did

N.G. Prokof’eva et al. / Comparative Biochemistry and Physiology Part B 134 (2003) 695–701

699

Fig. 2. Hemolytic activity of culcitoside C2 and C3 (50 mM) as a function of temperature. In controls hemolysis was not observed up to 60 min. Values are mean"S.E. (ns3).

not cause hemolysis in 5 min, white adding sodium chloride notably increases the activity of these glycosides. The effect of salt concentration was significantly larger for compound 5 than for 6. Thus, at 0.4 M NaCl ceramasterosides C2 and C3

induced hemolysis with lag times of 16 and 74 s, respectively, with a rate of hemolysis of 8.5% and 0.5% per s, respectively (Table 3). The timecourse of hemolysis induced by 50 mM rathbunioside R1 (7) at different salt concentrations is shown

Fig. 3. Time-course of hemolysis of erythrocytes induced by rathbunioside R1 (50 mM) as a function of NaCl concentration. In controls at 0.2–0.8 M NaCl, the value of the OD remained constant for 30 min. Values are mean"S.E. (ns3).

700

N.G. Prokof’eva et al. / Comparative Biochemistry and Physiology Part B 134 (2003) 695–701

Table 3 The rate and the delay time of hemolysis induced by ceramasteroside C2 and C3 (The concentration of compounds was 100 mM) Compound

5 6

Delay time (s)a

Slope (sy1)b

0.4 M NaCl

0.2 M NaCl

0.4 M NaCl

0.2 M NaCl

16.0"3.0 74.0"5.0

(–) (–)

0.085"0.015 0.015"0.005

(–) (–)

(–), hemolysis was not observed up to 5 min at 20 8C. The data represent the mean"S.E., ns3. a Delay time was defined as the point of intersection of the slope to the curve at a 50% hemolysis level with the line representing the zero percent hemolysis. b The slope—the rate of hemolysis, i.e. the fraction of cells rupturing per unit time at the 50% hemolysis level.

in Fig. 3. An increase in NaCl concentration is positively correlated with growth in the hemolytic activity of the glycoside. Thus, the delay time of the hemolysis induced by 7 in 0.2 M NaCl was 70 s and the rate of hemolysis was 0.44% per second. The same quantity of glycoside in 0.8 M NaCl show that the increased activity of the glycoside 7 was reflected in a reduced lag time (;15 s) and a higher lysis rate (5% per second) after the lag time. Control experiments indicated that OD of the red blood cells suspension at 0.2– 0.8 M NaCl was constant for 10 min. At high ionic strength, the kinetics of the rathbunioside R1 induced hemolysis exhibited a transient rapid phase (t-1 min). The rapid phase was attributed to the initial rapid association of the glycoside with membranes as was shown for sea cucumber triterpene glycosides (Kalinin et al., 1996). The dependent of steroid glycosides damaging potency on temperature, pH and salt concentration could result from changes of their aqueous solution properties as shown for quillaja saponin (Mitra and Dungan, 1997). In conclusion, we showed cytotoxic, hemolytic activities and the structure–activity interrelations of some polyhydroxysteroid glycosides. It was shown that the membrane damaging effects of glycosides depended on the structure, concentration of compounds and temperature, pH, ionic strength. The knowledge observed in these findings will help us to understand the functioning of these biological surfactants in starfish and the role in their environment. Acknowledgments This work was partially supported by RFFR grant No. 02-04-49491.

References Afiyatullov, Sh.Sh., Kuznetsova, T.A., Isakov, V.V., Pivkin, M.V., Prokof’eva, N.G., Elyakov, G.B., 2000. New diterpenic altrosides of the fungus Acremonium striatisporum isolated from a sea cucumber. J. Nat. Prod. 63, 848–850. Aminin, D.L., Agafonova, I.G., Fedorov, S.N., 1995. Biological activity of disulfated polyhydroxysteroids from the pacific brittle star Ophiopholis aculeata. Comp. Biochem. Physiol., Part C 112(2), 201–204. Andersson, L., Bohlin, L., Iorizzi, M., Riccio, R., Minale, L., Moreno Lopez, W., 1989. Biological activity of saponins and saponin-like compounds from starfish and brittle-stars. Toxicon 27(2), 179–188. Anisimov, M.M., Aminin, D.L., Rovin, Y.G., et al., 1983. On the resistance of the cells of the sea cucumber Stichopus japonicus to the action of endotoxin—stichoposide A. Dokl. AN SSSR 270, 991–993. Bergquist, J., Strandberg, C., Andersson, M., Sterner, O., Pesando, D., Girard, J.P., 1993. Structure–activity relationship for unsaturated dialdehydes. Comparative effects of 10 sesquiterpenoids on the sea urchin gamete fertilization. Toxic. In vitro 7(3), 205–212. Biyiti, L., Pesando, D., Puiseux-Da, O.S., Girard, J.P., Payan, P., 1990. Effect of two antibacterial plant flavonones on the intracellular calcium compartment involved in the first cleavage of sea urchin eggs. Toxicon 28, 275–283. Bruno, I., Minale, L., Ruccio, R., 1990. Starfish saponins, part 43. Structures of two new sulfated steroidal fucofuranosides (imbricatosides A and B) and six new polyhydroxysteroids from the starfish Dermasterias imbricata. J. Nat. Prod. 53, 366–374. Choi, D.H., Shin, S., Park, I.K., 1999. Characterization of antimicrobial agents extracted from Asterina pectinifera. Int. J. Antimicrob. Agents 11(1), 65–68. D’Auria, M.V., Fontana, A., Minale, L., Riccio, R., 1990. Starfish saponins. Part XLII. Isolation of twelve steroidal glycosides from the Pacific ocean starfish Henricia laeviuscola. Gazz. Chim. Ital. 120, 155–162. Elyakov, G.B., Stonik, V.A., 1988. Marine Organisms Steroids. Nauka, Moskow. Gorshkov, B.A., Kapustina, I.I., Kicha, A.A., Aminin, D.L., Gorshkova, I.A., 1998. Stimulatory effects of starfish sapogenins (Asterias amurensis and Lethasterias nanimensis chelifera) on molluscan heart (Spisula sachalinensis). Comp. Biochem. Physiol., Part C 120, 235–239.

N.G. Prokof’eva et al. / Comparative Biochemistry and Physiology Part B 134 (2003) 695–701 Ilani, A., Granoth, R., 1990. The pH dependence of the hemolytic potency of bile salts. Biochim. Biophys. Acta 1027, 199–204. Ivanchina, N.V., Kicha, A.A., Kalinovsky, A.I., et al., 2000. Hemolytic polar steroidal constituents of the starfish Aphelasterias japonica. J. Nat. Prod. 63, 1178–1181. Ivanchina, N.V., Kicha, A.A., Kalinovsky, A.I., Dmitrenok, P.S., Prokof’eva, N.G., Stonik, V.A., 2001. New steroid glycosides from the starfish Asterias rathbuni. J. Nat. Prod. 64, 945–947. Jacobs, R.S., Wilson, L., 1986. Sea urchin eggs as a model for detecting cell division inhibitors. In: Adorjan, A. (Ed.), Modern Analysis of Antibiotics. Marcel Dekker, New York, pp. 481–493. Kalinin, V.I., Prokof’eva, N.G., Likhatskaya, G.N., et al., 1996. Hemolytic activities of triterpene glycosides from the holothurian order Dendrochirotida: some trends in the evolution of this group of toxins. Toxicon 34, 475–483. Kicha, A.A., Kalinovsky, A.I., Levina, E.V., Andrijashenko, P.V., 1986. Culcitosides C2 and C3 from the starfish Culcita novaeguineae. Khim. Prir. Soedin. 592–596. Kicha, A.A., Kalinovsky, A.I., Stonik, V.A., 1997. New polyhydroxysteroids and steroid glycosides from the Far East starfish Ceramaster patagonicus. Izv. Acad. Nauk. Ser. Khim. 190–195. Kicha, A.A., Kalinovsky, A.I., Ivanchina, N.V., Stonik, V.A., 1999. New steroid glycosides from the deep-water starfish Mediaster murrayi. J. Nat. Prod. 62, 279–282. Kicha, A.A., Ivanchina, N.V., Gorshkova, I.A., Ponomarenko, L.P., Likhatskaya, G.N., Stonik, V.A., 2001. The distribution of free sterols, polyhydroxysteroids and steroid glycosides in various body components of the starfish Patiria (s Asterina) pectinifera. Comp. Biochem. Physiol., Part B 128, 43–52. Komori, T., 1997. Toxins from the starfish Acanthaster planci and Asterina pectinifera. Toxicon 35(10), 1537–1548. Kramer, S.D., Jakitsdeiser, C., Wunderlillenspach, H., 1997. Free fatty acids cause pH-dependent changes in drug-lipid membrane interactions around physiological pH. Pharm. Res. 14, 827–832. Kuznetsova, T.A., Prokof’eva, N.G., Chaikina, E.L., Afiatullov, S.S., 1999. Biological activity of diterpenoid altrosides from

701

marine isolate of the fungus Acremonium striatisporum. In: Chou, C.H., Waller, G.R., Reinhardt, C. (Eds.), Biodiversity and Allelopathy: From Organisms to Ecosystems in the Pacific. Academia Sinica, Taipei, pp. 193–197. Levina, E.V., Andriyashcenko, P.V., Kalinovsky, A.I., Dmitrenok, P.S., Stonik, V.A., Prokof’eva, N.G., 2002. Steroid compounds from the starfish Lysastrosoma anthosticta collected in the Sea of Japan. Izv. Akad. Nauk. Ser. Khim. 497–500 Russ. Chem. Bull., Int. Ed., 51, 3. 535–539. Likhatskaya, G.N., Kicha, A.A., 1997. Channel-forming activity of asterosaponin P1 in the model lipid membranes. In: First Euro Conference on Marine Natural Products, Abstracts. Greece, 135. Mackie, A.M., Turner, A.B., 1970. Partial characterization of a biologically active steroid glycoside isolated from the starfish Marthasterias glacialis. Biochem. J. 117, 543–550. Mackie, A.M., Singh, H.T., Owen, J.M., 1977. Studies on the distribution, biosynthesis and function of steroidal saponins in echinoderm. Comp. Biochem. Physiol., Part B 56, 9–14. Mitra, S., Dungan, S.R., 1997. Micellar properties of quillaja saponin. 1. Effect of temperature, salt, and pH on solution properties. J. Agric. Food Chem. 45, 1587–1595. Pesando, D., Graillet, C., Braekman, J.C., Dubreuil, A., Girard, J.P., Puiseux-Dao, S., 1991. The use of sea urchin eggs as a model to investigate the effects of crassolide, a diterpene isolated from a soft coral. Toxicol. In vitro 5(5y6), 395– 401. Prokof’eva, N.G., Likhatskaya, G.N., Aminin, D.L., et al., 1990. On the mechanism of interaction of the pH dependent cauloside C cytostatic agent with membranes of tumor cells and liposomes. Izv. AN. SSSR Ser. Biol. 3, 338–342. Prokof’eva, N.G., Likhatskaya, G.N., Volkova, O.V., 1992. Effect of betulafolienetetraol on the erythrocyte and model membranes. Biol. Membr. 9(9), 954–960. Prokof’eva, N.G., Shentsova, E.B., Stekhova, S.I., Anisimov, M.M., Pokhilo, N.D., Uvarova, N.I., 1999. A comparative study of the hemolytic potency of dammarane-type triterpene polyoles. Izv. Acad. Nauk, Ser. Biol. 493–496. Voogt, P.A., Van Rheenen, J.W.A., 1982. Carbohydrate content and composition of asterosaponins from different organs of the sea star Asterias rubens: relation to their haemolytic activity, and implications from their biosynthesis. Comp. Biochem. Physiol., Part B 72, 683–688.