Role of the Invertebrate Electrogenic 2Na⁺/1H⁺ Antiporter in Monovalent and Divalent Cation Transport

A wide variety of epithelial and nonepithelial vertebrate cell types possess a Na⁺/H⁺ exchange protein in their plasma membranes that catalyzes the net uptake of extracellular sodium coupled to the net extrusion of cytoplasmic protons (Grinstein, 1988). Biological functions of this antiport mechanism include the regulation of intracellular pH and cell volume as well as the transcellular transport of sodium and bicarbonate (Murer et al. 1976; Knickelbein et al. 1983; Boron and Boulpaep, 1983; Rothstein, 1989). Distinguishing characteristics of the vertebrate electroneutral Na⁺/H⁺ exchanger include a 1:1 cation exchange stoichiometry for several monovalent cations, an exclusion of divalent cations from binding sites, amiloride sensitivity at the external face of the transporter, an intracellular activator site for protons, which is distinct from another proton binding site on the same membrane surface that is involved in the cation exchange process, and an external sensitivity to quinidine. Aronson and Igarashi (1986) published a model for the vertebrate electroneutral Na⁺/H⁺ exchanger of renal epithelial brush-border membrane that displays many properties in common with similar exchangers in a large diversity of cell types. Recently, this model has been updated by Wakabayashi et al. (1992), who showed that the vertebrate exchanger is an oligomeric phosphorylated glycoprotein with 815 amino acids and 12 membrane-spanning elements.

During the last 5 years, two independent laboratories, using both fluorometric and radioisotopic techniques, have identified a Na⁺/H⁺ antiporter in brush-border membrane vesicles (BBMV) of gill, gut and renal epithelia from crustaceans and echinoderms. The physiology of the novel antiporter differs significantly from that of the mammalian antiporter in that it exhibits an electrogenic exchange with a transport stoichiometry of 2Na⁺/1H⁺ (Ahearn and Clay, 1989; Ahearn and Franco, 1990, 1991; Ahearn et al. 1990; Shetlar and Towle, 1989). In crustacean gut cells, two external cation binding sites with dissimilar binding properties were confirmed kinetically by external proton and amiloride inhibition of sodium transport; electrogenicity was demonstrated by the use of an imposed transmembrane electrical potential difference as the only transport driving force for cation exchange (Ahearn and Clay, 1989; Ahearn et al. 1990).

A monoclonal antibody has recently been developed to crustacean hepatopancreatic brush-border protein elements that may be the electrogenic 2Na⁺/1H⁺ antiporter (de Couet et al. 1993). This antibody inhibited 2Na⁺/1H⁺ exchange activity by lobster (Homarus americanus) hepatopancreatic BBMV, but was without effect on Na⁺-dependent D-glucose transport in the same preparation. Immunoreactivity was localized at the hepatopancreatic brush border, in antennal gland coelomosac podocytes and in posterior gill epithelial cells – all locations where unique 2Na⁺/1H⁺ exchange kinetics have been identified (Kimura et al. 1994). The antibody reacted with the digestive enzyme vacuoles of the hepatopancreatic B-cells that have a role in extracellular digestion, with the brush border of absorptive R-cells and with the calcium storage vacuoles of absorptive R-cells (Fig. 1). On Western blots of hepatopancreatic, antennal gland and gill proteins, this antibody recognized a single 185 kDa protein band with approximately twice the reported size of the mammalian Na⁺/H⁺ antiporter. The antigen recognized by this antibody may be widely distributed among crustacean epithelial cells, is considerably larger than the vertebrate Na⁺/H⁺ exchanger and appears to occur in several subcellular locations in the hepatopancreas, where it may perform a variety of different functions as a result of its unique electrogenic cation exchange properties.

This paper will present previously unpublished data concerning the role of the invertebrate electrogenic 2Na⁺/1H⁺ antiporter in divalent cation transport by the crustacean hepatopancreatic epithelium and suggest ways by which this transporter may function in the processes of exoskeletal calcification and heavy metal sequestration and detoxification.

The crustacean hepatopancreas is a diverticulum of the gut which consists of a single-layer epithelium that functions in digestive, absorptive and secretory processes. We have isolated, purified and characterized, in membrane vesicles from hepatopancreatic epithelial brush-border and basolateral membranes, transport proteins for a variety of organic and inorganic substances (Ahearn, 1988; Ahearn and Clay, 1988, 1989; Ahearn et al. 1990, 1992). The hepatopancreatic brush-border membrane of the Atlantic lobster Homarus americanus possesses three transport processes for the uptake of luminal Ca²⁺: (1) an amiloride-sensitive porter; (2) an amiloride-insensitive porter; and (3) a verapamil-inhibited, non-saturable, diffusion-like process responding to an induced transmembrane electrical potential (Fig. 2; Z. Zhuang and G. A. Ahearn, in preparation). Both porter processes were mediated by antiporters that were capable of exchanging external Ca²⁺ for either intravesicular Na⁺ or H⁺. The amiloride-sensitive system was electrogenic and exhibited the following Michaelis–Menten kinetic values: (1) K_t=0.58±0.02 mmol l⁻¹ (mean ± S.E.M., N=3); (2) J_max=1.67±0.04 pmol μg⁻¹ protein 3 s⁻¹. The amiloride-insensitive system was electroneutral and possessed the following values: (1) K_t=0.52±0.02 mmol l⁻¹; (2) J_max=0.55±0.01 pmol μg⁻¹ protein 3 s⁻¹ (N=3). External amiloride was a competitive inhibitor of ⁴⁵Ca²⁺ influx, inhibiting entry of the divalent cation at a single binding site with a K_i of 920 μmol l⁻¹. A significant fraction of Ca²⁺ entry into hepatopancreatic epithelial cells was postulated to occur by a previously reported electrogenic 2Na⁺/1H⁺ antiporter (Ahearn and Clay, 1989; Ahearn et al. 1990), by an electroneutral 2Na⁺/1Ca²⁺ antiporter (Ahearn and Franco, 1990, 1993) and by a verapamil-sensitive putative Ca²⁺ channel. This array of hepatopancreatic brush-border Ca²⁺ transport processes qualitatively resembled that previously reported for the luminal membrane of lobster antennal glands (Ahearn and Franco, 1993) and suggested that crustacean epithelial cells from different organs may handle this divalent cation by remarkably similar physiological means.

In 1991, the electrogenic 2Na⁺/1H⁺ antiporter that had been originally identified only in epithelia of protostomes (e.g. crustaceans) was reported in cells of deuterostomes [e.g. asteriod echinoderm (Pycnopodia helianthoides) pyloric ceca epithelia] as well (Ahearn and Franco, 1991), suggesting that this invertebrate exchanger may be widely distributed. As in crustacean cells, ²²Na⁺/H⁺ exchange by BBMV of the echinoderm epithelium was electrogenic and was inhibited by external amiloride at two external monovalent cation binding sites with markedly dissimilar apparent amiloride affinities (K_i=28 μmol l⁻¹ and 1650 μmol l⁻¹). A static head flux ratio analysis revealed a 2Na⁺/1H⁺ exchange stoichiometry in which a balance of driving forces (i.e. no net Na⁺ flux) was attained with a combination of a 10:1 Na⁺ gradient and a 100:1 H⁺ gradient. ⁴⁵Ca²⁺ uptake by purified pyloric cecal BBMV of the starfish occurred by electrogenic, amiloride-sensitive and electroneutral, amiloride-insensitive transporters, as found in crustacean epithelia (Z. Zhuang and G. A. Ahearn, in preparation). The amiloride-sensitive antiporter displayed the following apparent kinetic values: K_t=0.31±0.04 mmol l⁻¹; J_max= 0.16±0.01 pmol μg⁻¹ protein 3 s⁻¹ (N=3), whereas values of the amiloride-insensitive system were: K_t=1.11±0.03 mmol l⁻¹; J_max=0.88 pmol μg⁻¹ 3 s⁻¹ (N=3).

⁴⁵Ca²⁺ uptake in the presence of an outwardly directed H⁺ gradient and induced potential difference (ΔΨ) (inside negative) by pyloric cecal brush-border membrane vesicles of the starfish Pycnopodia helianthoides was significantly reduced by the addition of Zn²⁺ to the external medium (Fig. 3; Z. Zhuang and G. A. Ahearn, in preparation). Similarly, exogenous Fe²⁺ and Cd²⁺ were also found to inhibit Ca²⁺/H⁺ exchange by this membrane preparation. As shown in Fig. 3, exogenous Zn²⁺ acted as a competitive inhibitor of Ca²⁺ uptake, exhibiting a K_i of 6.3 mmol l⁻¹. As discussed above, the apparent Ca²⁺ binding affinity (K_t) of the starfish amiloride-sensitive transporter in pyloric cecal epithelium was 0.31±0.04 mmol l⁻¹ (N=3), whereas the competitive inhibitor constant (K_i) for Zn²⁺ was 6.3 mmol l⁻¹ (Fig. 3). Because the shared carrier exhibited a 20-fold higher apparent binding affinity for Ca²⁺ than for Zn²⁺, and because environmental concentrations of Ca²⁺ far exceed those of Zn²⁺ (Forstner and Wittmann, 1979), even in the most highly polluted waters, inhibition of Ca²⁺ uptake by Zn²⁺ from sea water may be minimal. However, starfish are carnivorous animals, and the food they consume, in comparison with sea water, may contain much higher concentrations of Zn²⁺ or other potential heavy metal competitive inhibitors of Ca²⁺ transport as a result of bioaccumulation and sequestration mechanisms in the cells of the prey. It is possible that gastrointestinal heavy metal concentrations during prey digestion could be much higher than those found in the surrounding the sea water or sediments and, therefore, could pose a potentially far more serious physiological problem.

A similar experiment was conducted with lobster hepatopancreatic BBMV to ascertain whether heavy metal inhibition of Ca²⁺ uptake by gastrointestinal epithelial cells is a common phenomenon among invertebrates. In this instance, ⁴⁵Ca²⁺ influx (0.1 and 0.5 mmoll⁻¹ Ca²⁺) was observed in proton-loaded vesicles in the presence and absence of external Cd²⁺ (0.01–5 mmoll⁻¹ Cd²⁺). Cadmium significantly reduced the uptake of Ca²⁺ at both concentrations, the metal acting as a competitive inhibitor of ⁴⁵Ca²⁺ influx into BBMV with a K_i of 66.5 μmol l⁻¹ (Fig. 4). This value was considerably lower than the K_i determined for Zn²⁺ inhibition of calcium uptake by starfish pyloric ceca and suggests that cadmium may be a more effective binding agent at the calcium transport site than is zinc. These data also suggest that cadmium may exert a significant inhibitory effect on calcium uptake at far lower environmental concentrations than would zinc. Future experiments with these alkaline metal ions will determine whether all agents are capable of moving across the hepatopancreatic brush-border membrane and into the cytoplasm.

A Percoll gradient centrifugation method originally developed for purifying epithelial basolateral membranes of mammalian cells (Davies et al. 1987) was adapted to lobster hepatopancreatic tissue. A vesicle band removed from near the top of this gradient was highly enriched in basolateral membrane marker enzymes (Na⁺/K⁺-ATPase purification factor=15.4-fold) with minimal contamination by brush-border membranes (alkaline phosphatase purification factor=1.2-fold) (Table 1). These putative basolateral membranes were used in flux studies to assess the nature of their cation transport processes.

The presence and characteristics of Na⁺/H⁺ exchange in purified lobster hepatopancreatic basolateral membrane vesicles (BLMV) were determined by preloading the preparation with buffer at pH 5.5 and incubating it in a medium at pH 8.5 containing 1 mmol l^{−1 22}Na⁺. Some of the vesicles preloaded in this manner were short-circuited (valinomycin; potassium gluconate both sides) or exhibited a transient inside-negative membrane potential (valinomycin; potassium gluconate inside, tetramethylammonium gluconate outside). In some cases, vesicles were exposed to incubation media containing 3 mmol l⁻¹ amiloride, 10 mmol l⁻¹ calcium or an inhibitory monoclonal antibody (de Couet et al. 1993).

A transmembrane pH gradient plus a transient potential difference (ΔΨ) caused a rapid uptake of ²²Na⁺ in exchange for protons (Fig. 5A). The fourfold overshoot phenomenon was maximal at approximately 1 min of incubation and gradually returned to equilibrium by 20 min. A transmembrane pH gradient alone stimulated the uptake of ²²Na⁺ about twofold over its equilibrium value. The presence of 3 mmol l⁻¹ amiloride completely blocked uptake of the isotope (Fig. 5A). Calcium was a partial inhibitor of the ²²Na⁺ uptake overshoot process, presumably interacting with Na⁺ on the Ca²⁺/Na⁺ antiporter described below (Fig. 5B). In contrast, the presence of a monoclonal antibody against protein elements of the brush-border electrogenic 2Na⁺/1H⁺ antiporter, which has previously been shown to inhibit the exchange process in BBMV (de Couet et al. 1993), was without effect on Na⁺/H⁺ exchange in BLMV.

In order to assess the nature of ²²Na⁺ influx by BLMV, vesicles were loaded at pH 5.5 with an induced inside-negative ΔΨ and were incubated for 2.5 s in medium containing ²²Na⁺ concentrations varying from 5 to 100 mmol l⁻¹. After correction for non-specific ²²Na⁺ binding to membranes and filters, the entry rate of sodium was a hyperbolic function of external sodium concentration (Fig. 6A). Apparent kinetic constants for basolateral sodium entry are displayed on the figure. In contrast with the hyperbolic kinetics of influx data for basolateral Na⁺/H⁺ exchange in lobster hepatopancreas, influx data for monovalent cation antiport in BBMV preparations from hepatopancreas from the marine lobster, the freshwater prawn (Macrobrachium rosenbergii) and the seawater starfish (Pycnopodia helianthoides) were all sigmoidal functions of exogenous sodium concentration (Fig. 6B; Ahearn et al. 1990; Ahearn and Franco, 1991). These comparisons between the Na⁺/H⁺ exchange kinetics of invertebrate brush-border and basolateral membranes suggest that markedly dissimilar transporters are present in the two locations and that this difference may be related to the apparent number of cation binding sites on each protein.

Further evidence for a difference in the apparent number of sodium binding sites on brush-border and basolateral Na⁺/H⁺ antiporters in lobster hepatopancreatic epithelium was obtained by examining the inhibitory effect of external amiloride on the exchange process in BLMV. Hepatopancreatic BLMV were loaded at pH 5.5 (inside-negative ΔΨ) and were exposed for 5 s to external medium containing ²²Na⁺ at either 1 or 5 mmol l⁻¹ and concentrations of amiloride ranging from 0.05 to 1 mmol l⁻¹. Amiloride inhibited Na⁺/H⁺ exchange in BLMV by acting as a competitive inhibitor at a single binding site (single slope at each [Na⁺]) on the antiporter (Fig. 7, Dixon plot). The single inhibitor constant (K_i=108.6 μmol l⁻¹) fell near the range for the effect of this drug on Na⁺/H⁺ exchange in vertebrate cells (i.e. 10–100 μmol l⁻¹). In contrast with the results from BLMV, amiloride was a competitive inhibitor of 2Na⁺/H⁺ exchange by invertebrate brush-border membranes at two external sites with markedly dissimilar apparent binding affinities (Ahearn and Clay, 1989; Ahearn et al. 1990; Ahearn and Franco, 1991).

These results with hepatopancreatic BLMV suggest that Na⁺/H⁺ exchange at the basolateral surface of the epithelium is significantly different from the analogous process at the brush border. Whereas influx of Na⁺ in BBMV occurs by an electrogenic exchanger with two external cation binding sites, entry of the cation at the basolateral border takes place by a transporter with a single sodium/amiloride binding site and, therefore, presumably displays an electroneutral 1Na⁺/1H⁺ exchange stoichiometry. Electrogenicity demonstrated in Fig. 5A under these conditions would result from the simultaneous operation of a membrane-potential-sensitive ion channel accommodating a diffusional flow of Na⁺ in response to an imposed membrane potential.

A conductive Na⁺ channel in the hepatopancreatic basolateral membrane is suggested by the results in Fig. 8. In this experiment, ΔpH was zero and ΔΨ was the only driving force affecting the movements of ²²Na⁺. The uptake of isotope by BLMV with an inside-negative ΔΨ was faster than that of vesicles possessing an electrically positive ΔΨ (Fig. 8). Exposure of vesicles under both conditions to a low concentration of amiloride (10 μmol l⁻¹) significantly reduced (P<0.05) their respective apparent initial uptake rates between 15 s and 1 min compared with that of vesicles lacking the drug. These results suggest that an amiloride-sensitive conductive ion channel is present in the basolateral membrane and support identification of the electroneutral Na⁺/H⁺ antiport discussed above.

In three different BLMV experiments, hepatopancreatic Ca²⁺ transport by basolateral membranes was compared with divalent cation uptake by the brush-border membrane. The first experiment was designed to assess whether a Ca²⁺/Na⁺ exchanger is present on the basolateral membrane. Short-circuited vesicles were loaded with 50 mmol l⁻¹ sodium at pH 7.5 and were incubated for 8 s in medium of the same pH containing ⁴⁵Ca²⁺ at Ca²⁺ concentrations from 0.01 to 1 mmol l⁻¹. There was a hyperbolic relationship between ⁴⁵Ca²⁺ influx, by Ca²⁺/Na⁺ exchange, and external calcium concentration (Fig. 9). These data provide evidence for an antiporter with the following apparent kinetic constants for Ca²⁺ influx: K_t=0.312±0.004 mmol l⁻¹; J_max=9.90±0.06 pmol μg⁻¹ protein 8 s⁻¹ (N=3). Detailed characterization of this antiporter and comparison of it with the brush-border Ca²⁺/H⁺ antiporter are planned.

The presence of a Ca²⁺-ATPase in lobster hepatopancreatic BLMV was confirmed in a second series of experiments with vesicles prepared without any transmembrane ion gradients or ΔΨ as potential driving forces. In these experiments ⁴⁵Ca²⁺ uptake by BLMV was measured in the presence of 5 mmol l⁻¹ exogenous ATP and in the presence and absence of the ATPase inhibitor vanadate. ATP significantly increased the rate of ⁴⁵Ca²⁺ accumulation and vanadate abolished it (Fig. 10). These results suggest the presence of a Ca²⁺-ATPase in this membrane and imply that intracellular calcium regulation in lobsters may involve the integration of secondary Ca²⁺/Na⁺ antiport and primary Ca²⁺-ATPase transport.

The possible presence of a calcium channel in the hepatopancreatic basolateral membrane was investigated in the third group of experiments by eliminating all possible driving forces for ⁴⁵Ca²⁺ uptake by BLMV except for an imposed transmembrane ΔΨ. In this instance, K⁺ and valinomycin were used to establish a transient inside-negative or inside-positive ΔΨ in BLMV and the Ca²⁺ channel blocker verapamil was added to the external medium to block diffusional flow of the divalent cation across the membrane in response to the established ΔΨ. In the presence of an inside-negative ΔΨ, ⁴⁵Ca²⁺ uptake was enhanced compared with that in the condition involving the application of an inside-positive ΔΨ (Fig. 11). In addition, external 100 μmol l⁻¹ verapamil significantly reduced ΔΨ -driven ⁴⁵Ca²⁺ flux. These data suggest that a verapamil-sensitive calcium channel may occur in the hepatopancreatic basolateral membrane and may play a significant role in calcium regulation.

Invertebrate gastrointestinal diverticula, such as the crustacean hepatopancreas or echinoderm pyloric ceca, perform a wide range of physiological activities that include digestion of food and absorption of amino acids, sugars, lipids and other organic solutes (Ahearn et al. 1992). In addition, these tissues secrete enyzmes, acids and other ionic substances such as the divalent anions (Cattey et al. 1994). These organs also appear to have a significant, but less studied, role in divalent cation regulation. The crustacean hepatopancreatic epithelium is known to be a cyclical storage site for exoskeletal calcium during the premolt phase of ecdysis. As an animal approaches molting, much of the calcium that is contained in the exoskeleton is transferred to the blood and either lost to the environment, across the gills and other permeable sites, or stored as calcium phosphate granules in hepatopancreatic epithelial cells and elsewhere (Greenaway, 1985). Following the loss of the old skeleton, hepatopancreatic calcium is mobilized and again transferred to the blood, where integumentary uptake mechanisms rapidly incorporate the divalent cation into the newly formed shell. Little is known about uptake or loss mechanisms for calcium by either hepatopancreatic brush-border or basolateral membranes during different phases of the crustacean molt cycle and even less is understood about the processes responsible for sequestration of the divalent cation in cytoplasmic calcium phosphate vacuoles.

It is known that hepatopancreatic calcium phosphate ‘concretions’ contain a variety of divalent and trivalent cationic heavy metals that are accumulated from the environment or diet and are concentrated in vacuoles by the epithelium as complexes with sulfate and phosphate (Viarengo, 1989; Viarengo and Nott, 1993). Concentrations of metals in vacuolar concretions can be several 100-fold higher than environmental concentrations of the individual metal. Heavy metal concretion formation is a detoxification mechanism used by numerous invertebrates to remove and isolate potentially lethal substances from the blood. Such concretions are eventually eliminated from the storage cells by exocytosis into the gastrointestinal lumen, where they are voided from the animal in the feces. The mechanisms by which divalent and trivalent heavy metal cations enter invertebrate epithelial cells, and the processes responsible for their sequestration, and therefore their detoxification, within epithelial vacuolar concretions are unknown. This study has begun to shed some light on the physiological mechanisms responsible for divalent cation transport across specific invertebrate gastrointestinal epithelial cell membranes that are closely involved in the processes of calcification and heavy metal detoxification.

Fig. 12 is a working model of monovalent and divalent cation transport mechanisms of the crustacean epithelial hepatopancreatic brush-border and basolateral membranes. Echinoderm pyloric cecal epithelial cells exhibit brush-border transporters similar to those described in this model, but the kinetic details of each mechanism are not currently as well clarified as those for crustaceans. Calcium and the heavy metal ions Zn²⁺ and Cd²⁺ are able to bind competitively to the previously described electrogenic 2Na⁺/H⁺ antiporters of hepatopancreatic and pyloric cecal brush-border membranes. The relative apparent affinities of the shared carrier for the cations were: Cd²⁺ > Ca²⁺ > Zn²⁺. Whether the bound metals eventually enter the epithelial cytoplasm and undergo subsequent intracellular physiological sequestration is unknown. Fig. 1 shows the subcellular localization of the antigen for the monoclonal antibody produced against protein elements of the electrogenic brush-border cation exchanger to hepatopancreatic calcium phosphate concretion vacuolar membranes and suggests that the same exchanger may be present in this location as occurs on the luminal membrane and that it may regulate Ca²⁺ and metal sequestration events at these vacuoles as well as cellular cation uptake across the brush border. Potential roles of the other brush-border cation transport mechanisms in heavy metal transport remain to be discovered, particularly in view of a recent study (Verbost et al. 1989) suggesting that cadmium may enter fish branchial epithelial cells through calcium channels.

Na⁺/H⁺ exchange at the hepatopancreatic basolateral membrane appears to use an electroneutral antiporter in combination with a membrane-potential-sensitive Na⁺ channel, rather than the electrogenic 2Na⁺/H⁺ antiporter (Fig. 12). The apparent electroneutral basolateral antiporter of hepatopancreas may be an isoform of the NHE gene family which has been thoroughly characterized in vertebrate cells (Tse et al. 1993) and which may also occur in the nematode Caenorhabditis elegans (Marra et al. 1993).

Three apparently distinct Ca²⁺ transport processes have been identified on the hepatopancreatic basolateral membrane: (1) a Ca²⁺/Na⁺ antiporter; (2) an ATP-dependent calcium ATPase, and (3) a verapamil-sensitive Ca²⁺ channel (Fig. 12). All of these proteins may be involved in intracellular Ca²⁺ regulation during intermolt, when an expected net absorption of Ca²⁺ from the luminal contents to the blood might occur in order to provide sufficient levels of the cation in the blood to supply organismic requirements. At premolt, when a large flux of Ca²⁺ into the blood from the old exoskeleton occurs, significant participation of the suggested basolateral verapamil-sensitive Ca²⁺ channel in uptake from the blood might occur prior to intracellular vacuolar sequestration.

Numerous invertebrate representatives of the Arthropoda, Mollusca, Echinodermata, Annelida and other phyla possess epithelial concretion vacuoles, which are apparently involved in detoxifying heavy metals. The mechanism(s) of metal sequestration and detoxification by these vacuoles has not been identified in any of these animals. Considerable work remains to substantiate the details of the model shown in Fig. 12 for crustacean hepatopancreatic epithelium and to link each transport mechanism to biological activities, such as calcification and heavy metal detoxification. In addition, separate studies of calcium phosphate concretion vacuolar membranes in crustaceans have yet to be undertaken to identify the processes responsible for metal and complexing anion accumulation. Characterization of the transport events associated with Ca²⁺ and heavy metal metabolism in this group of animals may provide an insight into universal cellular processes for regulating divalent cation levels in many invertebrate phyla.

Portions of this study were supported by a grant from the National Science Foundation (DCB89-03615). Experiments conducted on the starfish Pycnopodia helianthoides were performed at the University of Washington Friday Harbor Laboratories and the authors wish to thank the Director, Dr A. O. Dennis Willows, for providing the space and facilities to undertake these studies.