Ca²⁺ Transport by the Sarcoplasmic Reticulum Ca²⁺-ATPase in sea Cucumber (Ludwigothurea Grisea) Muscle

Holothuroids are an abundant and diverse group of marine invertebrates. Most of the approximately 1400 species described, comprising 200 genera, occur in benthic environments (Smiley, 1994). The sea cucumber Ludwigothurea grisea (Echinodermata: Holothuroidea) is an ectothermic animal that buries itself in the sand at the bottom of the sea and moves at a very slow pace with accompanying marked changes in body length (Motokawa, 1982; Motokawa, 1984; Eylers, 1982). The body wall is rigid and firm and is composed of a thin pigmented outer layer, often with spicules, and a dense white inner layer of connective tissue. A layer of circular muscle and five pairs of longitudinal smooth muscle are attached to the body wall. Each longitudinal muscle is composed of two strips situated along the longitudinal median juncture (Prosser and Mackie, 1980; Chen, 1986). When the animal is stressed, the five longitudinal muscles contract in order to eject its viscera out of the body (evisceration). These muscles are also involved in gas exchange by circulating oxygenated water through the body. As in other muscles, contraction is triggered by an increase in cytosolic free Ca²⁺ concentration (Lehman et al., 1973; Hill et al., 1978; Devlin, 1993).

It is well established that, in vertebrate skeletal muscle, the contraction–relaxation cycle is regulated by the release and uptake of Ca²⁺ from the sarcoplasmic reticulum (SR) (de Meis, 1981; Inesi, 1985); however, for different types of smooth muscle, the sources of Ca²⁺ that activate the contractile mechanism remain unclear. Electron micrographs of the longitudinal muscle of the body wall (LMBW) of the sea cucumber Isostichopus badionotus reveal small myocytes connected in small bundles with extensive areas of tight surface-contact between the cells and projections at the centre of the bundles (Hill et al., 1978). The myocytes contain only a few small mitochondria and negligible amounts of SR (Fig. 1), suggesting that the contraction of the LMBW fibres is caused mainly by the inward movement of extracellular Ca²⁺. However, in the same study, it was shown that, to abolish muscle contractility, the isolated muscle must be incubated for up to 10 h in Ca²⁺-free solutions (containing chelating agents). Only after such treatment does the contraction become directly dependent on added external Ca²⁺. Additional evidence for the presence of an intracellular Ca²⁺ pool in sea cucumber LMBW was obtained from experiments using caffeine. This substance induces the release of Ca²⁺ from mammalian SR through the ryanodine-sensitive Ca²⁺ channel and, when applied to the sea cucumber LMBW, causes contraction even in Ca²⁺-free solutions (Hill et al., 1978). Suzuki (Suzuki, 1982) observed, using electron microscopy, the presence of Ca²⁺ as a pyroantimonate precipitate, localised in resting fibres, in intracellular Ca²⁺ stores at both the subsarcolemmal vesicles and at the inner surface of the plasma membrane (Fig. 2). In contrast, in fibres fixed during contraction, the precipitate concentration is greatly decreased at the subsarcolemma vesicles and at the plasma membrane, and is observed diffusely distributed in the myoplasm (Fig. 3). Electron-probe X-ray microanalysis showed that the precipitate contained significant amounts of Ca²⁺ (Suzuki, 1982). These observations clearly indicate that the contraction–relaxation cycle of the longitudinal muscles involves not only the movement of Ca²⁺ between the extracellular medium and the cytosol of the muscle cells but also the release and accumulation of Ca²⁺ from intracellular stores.

Ca²⁺-ATPases are proteins bound in the SR membrane that pump Ca²⁺ into this intracellular compartment against an electrochemical Ca²⁺ gradient using the energy derived from ATP hydrolysis. These pumps maintain a low cytoplasmic Ca²⁺ concentration (<10⁻⁷ mol l⁻¹) and promote muscle relaxation. The SR Ca²⁺-ATPase (SERCA) of rabbit skeletal muscle has been extensively studied (de Meis, 1981; Inesi, 1985; Mintz and Guillain, 1997; Wolosker et al., 1998). However, little is known about the SERCA in the muscle of marine invertebrates.

Echinoderms occupy an interesting phylogenetic position in animal evolution. Although they are invertebrates, the echinoderms are deuterostomes like the vertebrates, and some chordate-like structures present in this group indicate their close relationship to the chordates (Morris, 1993). Thus, comparative studies between invertebrate and vertebrate Ca²⁺-ATPase isoforms may lead us to a better understanding of the evolution of the SERCA molecules and of how this pump acts during muscle contraction. This review focuses mainly on the study of the mechanism of Ca²⁺ transport and energy transduction catalysed by the SERCA found in the LMBW from a sea cucumber (Ludwigothurea grisea). Functional comparisons between the echinoderm enzyme and the well-studied vertebrate SERCAs are also made throughout this review.

The various sarco/endoplasmic reticulum Ca²⁺-ATPase (SERCA) isoforms found in mammalian tissues are cation-transporting ATPases that belongs to the family of P-type or E1/E2 ATPases. During the catalytic cycle of these enzymes, aspartyl phosphate is formed, and the protein undergoes a transition between two major conformations designed E1 and E2. Other members of this family include the Na⁺/K⁺-, Ca²⁺- and H⁺/K⁺-ATPases of animal cells and the H⁺-ATPase found in the plasma membrane of fungi and plants (de Meis and Vianna, 1979; Pick, 1982; Pedersen and Carafoli, 1987a; Pedersen and Carafoli, 1987b). They are made up of a single polypeptide chain with a molecular mass of approximately 110 kDa. The SERCA isoforms are expressed in several tissues and are encoded at least by three different genes. The SERCA1 gene is expressed exclusively in fast skeletal muscle (MacLennan, 1985; Brandl et al., 1987). The SERCA2 gene gives rise to the SERCA2a and 2b isoforms by alternative splicing (Lytton and MacLennan, 1988). The SERCA2a isoform is muscle-specific, being expressed in cardiac and slow-twitch skeletal muscles, whereas SERCA2b is ubiquitous and is the isoform predominant in the cerebellum (Lytton et al., 1989). SERCA3 is expressed in non-muscle tissues such as blood platelets and lymphoid tissue (Burk et al., 1989). All the SERCA isoforms encode a cytoplasmic region that contains the catalytic site for ATP hydrolysis and a transmembrane domain that forms a channel-like structure that allows Ca²⁺ permeation through the membrane (Inesi, 1985). The protein contents of different preparations of microsome vesicles derived from sea cucumber smooth muscle, rabbit skeletal muscle and dog cardiac muscle were analysed by SDS–PAGE and silver staining and revealed a major band at 110 kDa (Landeira-Fernandez et al., 2000b). When western blots of these preparations were probed with either anti-SERCA1 or anti-SERCA2a monoclonal antibodies, only the 110 kDa band of sea cucumber SR vesicles showed cross-reactivity with both anti-SERCA1 and anti-SERCA2a antibodies, implying that sea cucumber vesicles express a SERCA-type ATPase (Landeira-Fernandez et al., 2000b). Vesicles derived from sea cucumber smooth muscle are able to accumulate Ca²⁺ at the expense of ATP hydrolysis (Landeira-Fernandez et al., 2000b). Ca²⁺ uptake into these vesicles was not inhibited by ouabain (a specific inhibitor of the Na⁺/K⁺-ATPase), sodium azide (a classical inhibitor of the mitochondrial ATP synthase) or the proton ionophor carbonyl cyanide p-trifluoromethoxyphenylhydrazone (FCCP) (Table 1). In contrast, the sea cucumber Ca²⁺-ATPase was inhibited by general inhibitors of the E1/E2-type ATPases, vanadate and fluorescein isothiocyanate (FITC), and by the specific inhibitors of the SERCA isoforms, thapsigargin and cyclopiazonic acid (Table 1), suggesting that the Ca²⁺-ATPase found in sea cucumber microsomes is related to the SERCA isoforms and not to the Ca²⁺-ATPase found in the plasma membrane of muscles. A comparison between the mammalian Ca²⁺-ATPase isoforms and the sea cucumber isoform shows that the kinetics and the sensitivity to inhibitors are very similar between the different isoforms (Table 2). The sea cucumber Ca²⁺-ATPase shows close similarities with both SERCA1 and SERCA2 isoforms, the exception being that the slightly alkaline optimum pH is similar to that of the SERCA3 isoform (Table 2).

Most of the E1/E2-type transport ATPases have two apparent K_m values for ATP, one of high affinity in the range 1–10 μmol l⁻¹ (K_m1) and other of low affinity in the range 0.05–0.4 mmol l⁻¹ (K_m2). In these enzymes, the binding of high ATP concentrations leads to a three-to sixfold increase in the V_max (de Meis, 1981; Engelender and de Meis, 1996). The high-affinity K_m1 reflects the binding of ATP to the catalytic site of the enzyme, while the low-affinity K_m2 reflects the binding of ATP to a regulatory site on the enzyme that accelerates the rate of conversion of the form E2 into the form E1. This conversion is the rate-limiting step in the catalytic cycle of the ATPase (de Meis, 1981; de Meis and Sorenson, 1989; Engelender and de Meis, 1996). In the presence of oxalate, a condition that leads to a lower intravesicular free Ca²⁺ concentration, and measuring the initial rate of Ca²⁺ uptake, the sea cucumber ATPase showed a high affinity for Mg.ATP with a K_m value similar to those of other E1/E2-type ATPases (Table 2). No second K_m for ATP was detected and, instead of activating the enzyme, high concentrations of ATP inhibit Ca²⁺ transport both in the presence and absence of K⁺ (Fig. 4A). In fact, high concentrations of ATP promote Ca²⁺ leakage from the vesicles (Fig. 4B). Curiously, the inhibition is limited to Ca²⁺ uptake and is not detected for the measurements of the Ca²⁺-dependent ATPase activity (Landeira-Fernandez et al., 2000b). This result suggests that high Mg.ATP concentrations uncouple Ca²⁺ transport from ATP hydrolysis by the sea cucumber Ca²⁺-ATPase (Fig. 4).

Rabbit skeletal muscle heavy SR contains the ryanodine Ca²⁺ release channel that possesses a low-affinity regulatory ATP binding site (Meissner, 1994; Xu et al., 1996). Because of the heterogeneous nature of this preparation, this channel could be present and could account for the inhibition of Ca²⁺ uptake observed at high ATP.Mg concentrations (Fig. 4B). Thus, the uptake measurements presented in Fig. 4B were made in the presence and absence of ryanodine and Ruthenium Red, which are known to block the Ca²⁺ release channel (Xu et al., 1999). Neither ryanodine nor Ruthenium Red was able to prevent the release of Ca²⁺ induced by the addition of 4 mmol l⁻¹ ATP.Mg, and they also failed to antagonise the inhibition of Ca²⁺ uptake observed in the presence of 4 mmol l⁻¹ ATP.Mg (data not shown). In a different set of experiments, 10 mmol l⁻¹ caffeine, which is known to open the Ca²⁺ release channel (Meissner, 1994; Ozawa, 1999), was added to medium containing 0.1 mmol l⁻¹ ATP.Mg after the vesicles had been loaded and a steady state had been reached. In this experiment, no efflux of Ca²⁺ was observed after the addition of caffeine (data not shown). These data indicate that the effect of high ATP concentrations is not related to opening of the Ca²⁺ release channel.

In addition, the sea cucumber Ca²⁺-ATPase can use also ITP, GTP and CTP as substrates, but with a lower affinity than ATP (Landeira-Fernandez et al., 2000b).

During catalysis, an aspartyl residue located in the catalytic site of the different E1/E2-type ATPases is phosphorylated by either ATP or inorganic phosphate (P_i). For the SERCA isoforms, Ca²⁺ transport and ATP hydrolysis are initiated by phosphorylation of the enzyme by ATP; in the reverse process, the synthesis of ATP is initiated by phosphorylation of the ATPase by P_i (de Meis and Vianna, 1979; de Meis, 1981). We were not able to measure phosphorylation of the sea cucumber ATPase by either ATP or P_i. High levels of enzyme phosphorylation by either ATP or P_i are only detected with the SERCA1 isoform. For all other isoforms, the steady-state level of phosphoenzyme is very low and difficult to measure (Engelender et al., 1995). The steady-state level of phosphoenzyme depends on the rate of phosphoenzyme formation and phosphoenzyme hydrolysis and, in SERCA1, the rate of formation has been shown to be at least one order of magnitude faster than the rate of cleavage (de Meis, 1981). The ability to transport Ca²⁺ and the inhibitors used indicate that the sea cucumber belongs to the the E1/E2 family of transport enzymes and should be phosphorylated by both ATP and P_i. Therefore, the fact that we were not able to measure the intermediate formation of phosphoenzyme suggests that, as for isoforms SERCA2 and 3, the rate of phosphoenzyme cleavage of the sea cucumber enzyme is faster than the rate of phosphoenzyme formation.

These kinetic data and immunological approaches indicate that the sea cucumber enzyme has some unique properties as well as some similarities with the various mammalian SERCA isoforms, suggesting that the sea cucumber Ca²⁺-ATPase consists of a distinct but related SERCA isoform.

In contrast to the vertebrate SERCA isoforms, Ca²⁺ transport and ATP hydrolysis catalysed by the sea cucumber Ca²⁺-ATPase are highly activated by monovalent cations (Fig. 5). As in microsomes derived from some vertebrate tissues (Wolosker et al., 1997; Mitidieri and de Meis, 1999), two different ATPase activities could be distinguished in the sea cucumber microsomes: (i) a Mg²⁺-dependent activity measured in the absence of Ca²⁺, and (ii) a total ATPase activity measured in the presence of both Mg²⁺ and Ca²⁺. The difference between the two activities is referred to as Ca²⁺-dependent ATPase and is responsible for the translocation of Ca²⁺ through the microsomal membrane. All three cations tested, K⁺, Na⁺ and Li⁺, inhibit the Mg²⁺-dependent activity of the sea cucumber microsomes, while only Na⁺ and K⁺ increase of the total activity (Landeira-Fernandez et al., 2000b). The activation of Ca²⁺ uptake by the sea cucumber vesicles promoted by K⁺ and Na⁺ varies depending on whether oxalate or P_i is used as the Ca²⁺-precipitating agent (Fig. 5). These two anions are known to increase the Ca²⁺-loading capacity of microsomes isolated from a variety of tissues, including skeletal muscle, blood platelets and brain (de Meis et al., 1974; de Meis, 1981; Wolosker et al., 1997). During transport, they diffuse through the membrane and form calcium phosphate and calcium oxalate crystals in the vesicle lumen (de Meis et al., 1974). The rate of Ca²⁺ uptake is faster (Landeira-Fernandez et al., 2000b) and the amount of Ca²⁺ retained by the vesicles greater during oxalate treatment than in the presence of P_i (Fig. 5). Also, K⁺ and Na⁺ activate both Ca²⁺-dependent ATP hydrolysis and Ca²⁺ transport to the same extent during oxalate treatment (Fig. 5A,B). However, when P_i is used, K⁺ activates Ca²⁺ transport more effectively than Na⁺ (Fig. 5C). This is not observed for ATP hydrolysis, where the degree of activation by Na⁺ and K⁺ is the same in the presence of 5 mmol l⁻¹ oxalate or 10 mmol l⁻¹ P_i (Fig. 5B,D). One explanation for the distinct effect of K⁺ and Na⁺ may be a variable free Ca²⁺ concentration inside the vesicles, which is determined by the solubility of the calcium oxalate (approximately 0.1 mmol l⁻¹) and calcium phosphate (approximately 5 mmol l⁻¹) precipitates formed inside the vesicles (de Meis et al., 1974). The degree of activation promoted by monovalent cations is greater for ATP hydrolysis than for Ca²⁺ uptake, being fourfold and twofold higher, respectively (Fig. 5). However, the concentrations needed for half-maximal activation of Ca²⁺ uptake and of ATP hydrolysis are almost the same for K⁺ and Na⁺, ranging in both cases from 30 to 40 mmol l⁻¹ (Fig. 5).

The effect of monovalent cations on vesicles derived from the SR of rabbit skeletal muscle has been extensively studied. K⁺ and Na⁺ modify different intermediate steps of the catalytic cycle of the enzyme (Chaloub and de Meis, 1980; Moutin and Dupont, 1991; Champeil et al., 1997). The overall effect, however, varies depending on the conditions used. For instance, in presence of a low concentration of ATP (2 μmol l⁻¹) and Ca²⁺ (less than 1 μmol l⁻¹), monovalent cations inhibit Ca²⁺ transport and ATP hydrolysis, with Na⁺ being more effective than K⁺ (de Meis and Hasselbach, 1971; de Meis, 1971; Shigekawa and Pearl, 1976; Duggan, 1977; Ribeiro and Vianna, 1978). In the presence of saturating Ca²⁺ concentrations (10–50 μmol l⁻¹), monovalent cations increase the rate of Ca²⁺ uptake, with an order of effectiveness of K⁺>Na⁺@Li⁺. Activation of Ca²⁺ uptake by K⁺ is also dependent on temperature. A two-to threefold activation of Ca²⁺ transport by 100 mmol l⁻¹ KCl in preparations at 10 °C has been reported (Duggan, 1977; Shigekawa and Pearl, 1976). However, in rabbit skeletal muscle, rat brain and human blood platelets, a much smaller activation of both Ca²⁺ uptake and ATP hydrolysis by K⁺ is observed in preparations at their physiological temperatures (de Meis and Suzano, 1994; Rocha et al., 1996; Engelender et al., 1995). The finding that the sea cucumber ATPase is activated by K⁺ under all experimental conditions tested (Landeira-Fernandez et al., 2000b) indicates that this enzyme has an absolute dependence on K⁺ to reach maximal transport efficiency, a feature different from the other SERCA so far studied. A similar activation by monovalent cations such as K⁺, and to a lesser extent Na⁺, on ATP hydrolysis has been described in microsomes derived from the striated adductor muscle of deep sea scallops (Mollusca) (Castellani et al., 1989; Kalabokis et al., 1991). Although the scallop microsomes were able to transport Ca²⁺, the possible effect of these monovalent cations on Ca²⁺ transport, as far as we know, was not measured.

We have previously reported that heparin and other glycosaminoglycans can, in vitro, inhibit the activity of a variety of ATPases of the E1/E2-type family, including the various SERCA isoforms and the plant plasma membrane H⁺-ATPase (de Meis and Suzano, 1994; Landeira-Fernandez et al., 1996; Rocha et al., 1996; Rocha et al., 1998). These results are summarised in Table 3, in which the values obtained with sea cucumber Ca²⁺-ATPase are compared with the values for several different mammalian SERCA isoforms. In control experiments, K⁺ has almost no effect or promotes minimal activation of Ca²⁺ uptake. However, in the presence of heparin, the activity is inhibited, and K⁺ is able to activate Ca²⁺ transport in preparations from a variety of animals. Thus, rabbit SERCA is not dependent on KCl, but in the presence of heparin the ATPase becomes highly sensitive to K⁺ in a manner similar to that of the sea cucumber Ca²⁺-ATPase (Fig. 6). The blockade of the inhibition does not appear to be due to neutralisation of the negative charges of the sulphate residues because Li⁺, the most electropositive ion among the alkali ions tested, has no effect. It is unlikely that binding of Ca²⁺ to the sulphate polysaccharides plays an important role in the effects of these compounds on the Ca²⁺-ATPase. The Ca²⁺ affinity constant for fucosylated chondroitin sulphate (Ruggiero et al., 1994) and heparin (Mattai and Kwak, 1981) is approximately 50 μmol l⁻¹, while the Ca²⁺ affinity of the ATPase is approximately 0.1 μmol l⁻¹ (de Meis, 1981). In addition, the concentration of sulphate polysaccharide used in these experiments was very low (approximately 10 μg ml⁻¹) to decrease the amount of free Ca²⁺ in the medium.

The possible presence of glycogen granules contaminating the SR sea cucumber vesicles was evaluated enzymatically (Arlington, 1990). It was found that the vesicles were contaminated with 1.3 mg glycogen mg⁻¹ vesicle protein. We then digested the contaminating glycogen using the enzyme amiloglucosidase for 3 h at room temperature (25 °C) and at pH 7.0, conditions that completely digested the contaminant glycogen. It was found that, similar to the vesicles incubated in the absence of amiloglucosidase, KCl activated the enzyme two-to threefold, indicating that the KCl-dependence of sea cucumber vesicles is not related to the presence of these glycogen granules. In addition, the effect of high concentrations of glycogen (up to 40 mg ml⁻¹) was measured using rabbit SR vesicles. In this experiment, glycogen did not modify the rate of Ca²⁺ uptake either in the presence or in the absence of heparin.

A unique fucosylated chondroitin sulphate isolated from sea cucumber connective tissue can mimic the effects of heparin (de Meis and Suzano, 1994; Landeira-Fernandez et al., 1996; Rocha et al., 1996). However, the possible physiological implication of these results can be questioned since these highly sulphated polysaccharides (such as heparin and fucosylated chondroitin sulphate) are not found in plants or mammalian tissues where these ATPases are present.

The activation by K⁺ of the native sea cucumber ATPase suggested that the smooth muscle may contain endogenous polysaccharides. To test this hypothesis, we attempted to isolate sulphated polysaccharides from the entire muscle of the sea cucumber and from the vesicles derived from its SR. The longitudinal muscles of the sea cucumber body wall revealed a high content of sulphated polysaccharides. Protease digestion and purification by anion exchange chromatography, followed by molecular mass and chemical composition determination, showed that the predominant polysaccharide present in the echinoderm tissue is a fucosylated chondroitin sulphate (Landeira-Fernandez et al., 2000a). However, when the sea cucumber muscle was homogenised and the vesicles derived from its SR purified, the fucosylated chondroitin sulphate was no longer detected. Instead, a sulphated polysaccharide was eluted from the anion-exchange column at higher salt concentrations; this polysaccharide had a pattern of molecular mass distribution unusual for sulphated polysaccharides. In contrast, vesicles obtained from the SR of rabbit muscles contained no detectable sulphated polysaccharide (Landeira-Fernandez et al., 2000a).

A nuclear magnetic resonance ¹H (NMR) spectrum of the polysaccharide isolated from the sea cucumber vesicles shows a broader and poorly resolved signal, indicating a heterogeneous chemical structure. This polysaccharide differs from mammalian glycosaminoglycans since it is resistant to chondroitin lyase digestion and to nitrous acid deamination. Determination of its structure with chemical methods has not been possible because of the limited amounts of material available, but NMR and methylation analyses did not show a simple repetitive structure. Partial acid hydrolysis (150 mmol l⁻¹ H₂SO₄, 100 °C for 30 min) removes most of the sulphate fucose from the polymer, but not galactose and hexosamines. Thus, the sulphate fucose residues are apparently branched units. Other attempts to prepare small fragments using partial hydrolysis procedures were unsuccessful.

The colocalization of this unique polysaccharide along with the sea cucumber Ca²⁺-ATPase led us to speculate about a possible physiological relationship. The effect of the sulphated polysaccharides extracted from sea cucumber vesicles was tested on rabbit Ca²⁺-ATPase. Surprisingly, the sulphated polysaccharide inhibits Ca²⁺ uptake by rabbit muscle-derived vesicles at an even lower concentration than heparin and restores the dependence on monovalent cations (Fig. 7). These results suggest that the sulphated polysaccharide present in the sea cucumber vesicles can act as an ‘endogenous’ Ca²⁺-ATPase inhibitor.

Purified vesicles from the sea cucumber muscle were incubated with papain or solutions containing either a high salt concentration (1.0 mol l⁻¹ KCl) or detergent (5 % C₁₂E₈). The extracted polysaccharides were analysed by anion-exchange chromatography and polyacrylamide gel electrophoresis. Most of the sulphated polysaccharide (>90 %) was extracted exclusively by protease. A small proportion (approximately 30 %) was also removed by concentrated salt solution, whereas detergent was ineffective. Electrophoresis on a polyacrylamide gel reveals some differences in the proportions of the various bands of sulphated polysaccharide extracted by papain and by high salt concentration. These procedures may preferentially remove different populations of the sulphated polysaccharides. Curiously, the detergent solution removes two different groups of sulphated polysaccharides; one is a high-molecular-mass component that remains at the origin of the gel, and the other is a low-molecular-mass component that migrates as a broad band. The sulphated polysaccharides extracted from the sea cucumber vesicles by papain inhibited rabbit Ca²⁺-ATPase, while the KCl-soluble sample had a less inhibitory effect and the detergent-extracted material was inactive (data not shown; Landeira-Fernandez et al., 2000a). The sulphated polysaccharide is preferentially extracted from the muscle-derived vesicles by protease, suggesting a direct interaction with proteins present in the vesicles and intracellular localisation of these molecules. Thus, it is conceivable that the polysaccharides may be located in the lumen of the SR; however, the nature of this interaction requires further investigations.

The sea cucumber muscle contains two different pools of sulphated polysaccharides. Fucosylated chondroitin sulphate is present at very high concentrations (equivalent to those found in the connective tissue of the invertebrate) and is located around the muscle fibres. Another unique sulphated polysaccharide (almost undetectable when extraction is performed using the entire muscle) is found in vesicles derived from the SR. The physiological action of these molecules remains unclear, but some hypotheses have been proposed. The body wall of the sea cucumber can rapidly and reversibly alter its mechanical properties. These alterations, which are thought to be neuronally controlled, allow the tissue to change its length by more than 200 % (Trotter and Koob, 1989; Trotter et al., 1995). The connective tissue of the sea cucumber Cucumaria frondosa contains parallel collagen fibrils that are able to slide past one another during length changes but are inhibited from sliding when the tissue is in ‘catch’. It has been suggested that sulphated polysaccharides (specifically the fucosylated chondroitin sulphate) are an important component of the stress-transfer matrix in echinoderms (Trotter et al., 1995; Motokawa, 1982; Motokawa, 1984). The function of this polysaccharide in the echinoderm muscle is still not known. One explanation is that it has a high negative charge density (Mourão et al., 1996), so its capacity to retain water in the extracellular matrix produces a ‘space’ among muscle fibres allowing the intense change in length, a phenomenon that may occur in echinoderm muscle but is not observed in mammalian muscles.

Most of the studies involving Ca²⁺ transport were made using the skeletal muscle Ca²⁺-ATPase as a model, but other invertebrate Ca²⁺-ATPases isoforms have been described in the literature. Cario et al. (Cario et al., 1996) demonstrated, using cytochemical analyses, the presence of Ca²⁺-ATPase activity both in SR and plasma membrane of the smooth muscle of the marine invertebrate Beroe ovata (ctenophore) and a Ca²⁺-loading capacity of the SR using oxalate as a precipitating agent. Coelenterates (cnidarians and ctenophores) are more primitive organisms than echinoderms, but evidence indicates that, like the sea cucumber, they are in the same evolutionary line leading to deuterostomes (Morris, 1993; Cario et al., 1996). The amino acid sequence of the Ca²⁺-ATPase present in the invertebrate Artemia franciscana (Crustacea) reveals a high homology with rabbit SR (71 %) and low similarities with either Na⁺/K⁺-ATPase and plasma membrane Ca²⁺-ATPase (24 % and 25 %, respectively) (Palmero and Sastre, 1989). After overexpression of a Ca²⁺-ATPase of the endoplasmic reticulum of the protozoon Trypanosoma brucei, kinetic data showed properties similar to the mammalian SERCA isoforms so far described, including a high affinity for Ca²⁺ and sensitivity to the general E1/E2-type ATPases inhibitor vanadate and to the specific SERCA inhibitor thapsigargin (Nolan et al., 1994). Thus, these data suggest that the mechanism of energy transduction catalysed by the Ca²⁺-ATPase is highly conserved through evolution. The main differences between the SERCA isoforms seem to be related to both tissue-specific expression and physiological regulation. For the Ca²⁺-ATPase found in sea cucumber vesicles, the high dependence on monovalent cations for maximal activity and the presence of unique sulphate polysaccharides seem to be specific features of this enzyme and are not found in other isoforms so far described. After extraction and purification of these sulphate polysaccharides from sea cucumber vesicles, their effect was tested on rabbit SR Ca²⁺-ATPase. In these vesicles, the activity of the Ca²⁺ pump is inhibited and K⁺ is able to activate the enzyme several-fold. These results suggest that it is possible that the K⁺-dependency of the sea cucumber Ca²⁺-ATPase is due to the presence of sulphate polysaccharides interacting with the enzyme.

I wish to express special thanks to Dr Leopoldo de Meis for introducing me into the scientific world and for his partnership throughout this work. I am also grateful to Dr Antonio Galina for opening the doors to marine invertebrate bioenergetics and to Dr Monica Montero-Lomeli for collaboration in the experiments and help in the preparation of this paper. I also thank Dr Paulo A. de Souza Mourão and Dr Luiz Claudio F. Silva for collaborating in the experiments with sulphate polysaccharides and for support during part of this work. I also thank Dr Vivian Rumjanek and Dr João Batista Rocha for reading a preliminary version of the manuscript and Mr Valdecir A. Suzano and Mr Antonio Carlos Miranda for the excellent technical assistance. This work was supported by grants from PRONEX – Financiadora de Estudos e Projetos (FINEP), Fundação de Amparo à Pesquisa do Rio de Janeiro (FAPERJ) and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq).