ABSTRACT
Ca2+ transport by hepatopancreatic basolateral mem-brane vesicles of Atlantic lobster (Homarus americanus) occurred by at least two independent processes: (1) an ATP-dependent carrier transport system, and (2) a Na+-gradient-dependent carrier mechanism. The sensitivity of ATP-dependent Ca2+ transport to vanadate indicated that it was probably due to a P-type ATPase. This system exhibited an extremely high apparent affinity for Ca2+ (Kt=65.28±14.39 nmol l−1; Jmax=1.07±0.06 pmol µg−1 pro-tein 8 s−1). The Na+-gradient-dependent carrier transport system exhibited the properties of a Ca2+/Na+ antiporter capable of exchanging external Ca2+ with intravesicular Na+ or Li+. Kinetic analysis of the Na+-dependence of the antiport indicated that at least three Na+ were exchanged with each Ca2+ (n=2.91±0.22). When Li+ replaced Na+ in exchange for 45Ca2+, the apparent affinity for Ca2+ influx was not significantly affected (with Na+, Kt=14.57±5.02 µmol l−1; with Li+, Kt=20.17±6.99 µmol l−1), but the maximal Ca2+ transport velocity was reduced by a factor of three (with Na+, Jmax=2.72±0.23 pmol µg−1 pro-tein 8 s−1; with Li+, Jmax=1.03±0.10 pmol µg−1 protein 8 s−1). It is concluded that Ca2+ leaves hepatopancreatic epithelial cells across the basolateral membrane by way of a high-affinity, vanadate-sensitive Ca2+-ATPase and by way of a low-affinity Ca2+/Na+ antiporter with an apparent 3:1 exchange stoichiometry. The roles of these transporters in Ca2+ balance during the molt cycle are discussed.
INTRODUCTION
Animal studies, largely with vertebrate species, suggest that epithelial Ca2+ absorption occurs via both an active and a passive pathway (Bronner et al. 1986; Favus, 1985; Murer and Hildmann, 1981). The passive process occurs exclusively through the paracellular pathway, while the active step in Ca2+ transport is located at the basolateral domain of the enterocytes (Hildmann et al. 1982; Nellans and Popovitch, 1981).
The presence of a high-affinity Ca2+-ATPase isolated from basolateral membranes of rat duodenum (Ghijsen et al. 1986), jejunum (Nellans and Popovitch, 1981) and kidney (Tsukamoto et al. 1986), as well as from plasma membranes of gill epithelia of the crab (Carcinus maenas) (Flik et al. 1994) and a fish (Oreochromis mossambicus) (Flik et al. 1993), has been reported. These Ca2+ pumps are thought to be involved in Ca2+ translocation across the plasma membrane against an unfavorable electrochemical gradient. In addition to the occurrence of this primary active transport system, the presence of a basolateral Ca2+/Na+ exchange process, which may be responsible for a significant portion of Ca2+ transport at this cell pole, has been suggested for gill epithelia from crustaceans (Towle, 1993; Flik et al. 1994), and it may also be present on the basolateral membrane of rat small intestine and kidney (Hildmann et al. 1982; Jayakumar et al. 1984; Tsukamoto et al. 1986; Van Heeswyk et al. 1984).
Our recent studies with lobster (Homarus americanus) hepatopancreatic and starfish (Pycnopodia helianthoides) pyloric cecal brush-border membrane vesicles indicated that a considerable fraction of the luminal Ca2+ content is transported into epithelial cells through a combination of three transport processes: (1) an amiloride-sensitive carrier system; (2) an amiloride-insensitive carrier system; and (3) a potential-sensitive, verapamil-or nifedipine-inhibitable, Ca2+ channel (Ahearn and Franco, 1990, 1991, 1993; Zhuang et al. 1995; Zhuang and Ahearn, 1996). Little is known about Ca2+ transport into the hemolymph across the lobster hepatopancreatic basolateral membrane, with the exception of our preliminary studies suggesting the occurrence of a Ca2+-ATPase and a Ca2+/Na+ antiporter in the basolateral membranes of this crustacean organ (Ahearn and Zhuang, 1996). In the present study, the basolateral Ca2+ transport processes of the lobster hepatopancreas are investigated in detail and are discussed in relation to the interplay of basolateral and brush-border Ca2+ transport systems during the molt cycle.
MATERIALS AND METHODS
Live Atlantic lobsters (Homarus americanus Milne-Edwards; 0.5 kg each) were purchased from commercial dealers in Hawaii and maintained unfed at 10°C for up to 1 week in filtered sea water. Most animals were in either intermolt or early premolt as assessed by the molt stage classification scheme introduced by Aiken (1973).
Hepatopancreatic basolateral membrane vesicles (BLMVs) were prepared from fresh organs of individual lobsters. BLMVs were prepared freshly for each experiment utilizing a self-orienting Percoll gradient centrifugation technique adapted from a procedure developed for mammals (Davies et al. 1987). Hepatopancreatic tissue was quickly excised and placed into a hypotonic (relative to blood) buffered sucrose medium containing 250 mmol l−1 sucrose, 20 mmol l−1 Tris/HCl, 300 µmol l−1 phenylmethylsulfonylfluoride (PMSF), pH 7.4, and homogenized with a Kinematica GmbH polytron. An initial centrifugation at 2500 g removed large cellular debris. A crude separation of basolateral membranes was obtained from a second centrifugation at 20 400 g. The crude basolateral membrane pellet was resuspended in the hypotonic sucrose buffer, combined with a premixed dilution of Percoll, and centrifuged at 47 800 g for 1 h. The resulting Percoll-gradient bands were enzymatically assayed for enrichment of the basolateral marker enzyme Na+/K+-ATPase and the brush-border membrane marker enzyme alkaline phosphatase. High levels of Na+/K+-ATPase enrichment (15.2-fold) and minimal alkaline phosphatase enrichment (1.2-fold) occurred in a band containing a high concentration of BLMVs with minimal contamination by brush-border membrane vesicles (BBMVs; Duerr and Ahearn, 1996).
Vesicle orientation and leakiness were determined by the Na+/K+-ATPase latency technique described for this basolateral membrane preparation using a sucrose gradient purification method described previously (Ahearn et al. 1987). This method consists of using the asymmetry of ligand binding sites of Na+/K+-ATPase in the presence and absence of a detergent to estimate the percentages of right-side-out, inside-out and leaky-sheet membranes in a hepatopancreatic basolateral preparation. Preparing hepatopancreatic basolateral vesicles by a sucrose gradient method resulted in a vesicle preparation that possessed 62.9% leaky sheets, 31.1% right-side-out vesicles and 6.0% inside-out vesicles (Ahearn et al. 1987). The same method of assessing vesicle sidedness was applied to the present preparation isolated using Percoll density gradient centifugation. Table 1 indicates that very similar results were obtained for the orientation of vesicles isolated by Percoll. In this instance, 65.24% of the membranes were leaky sheets,
28.72% were right-side-out vesicles and 6.04% were inside-out vesicles. Where capacities between different transport systems are discussed in this paper, appropriate corrections for the orientation of functional vesicles have been considered.
The basolateral membrane fraction was homogenized with inside buffer (compositions varying with experiment, see figure legends for details) and then centrifuged for 40 min at 47 800 g. The resulting pellet was resuspended and washed in fresh intravesicular medium and centrifuged a final time for 40 min at 47 800 g to collect the pellet.
Long-term and short-term 45Ca2+ uptake time-course experiments were conducted in a manner similar to that reported previously (Ahearn et al. 1995; Ahearn and Zhuang, 1996; Zhuang and Ahearn, 1996). At the beginning of a transport experiment, a volume (e.g. 20 µl) of basolateral membrane vesicles was added to a volume (e.g. 160 µl) of radiolabelled medium containing 45Ca2+ (ICN Radiochemicals). Following incubation times ranging from 8 s to 60 min, a known volume (e.g. 20 µl) of this reaction mixture was withdrawn and plunged into 2 ml of ice-cold stop solution (composition differing with different experiments; see figure legends). The resulting suspensions were rapidly filtered through Millipore filters (0.65 µm pore diameter) to retain the vesicles and washed with another 5 ml of stop solution. Filters were then added to ICN Ecolume scintillation cocktail and counted for radioactivity in a Beckman LS-8100 scintillation counter. Incubation times and intravesicular media varied for different experiments as indicated in the figure legends. Ca2+ uptake values were expressed as pmol per microgram of protein (Bio-Rad protein assay) per filter using the specific activity of 45Ca2+ in the incubation medium.
Valinomycin (50 µmol l−1) and bilaterally equal K+ concentrations across the vesicular wall were present to short-circuit the membranes, and 5 mmol l−1 MgCl2 was used in all ATPase-related experiments. Preliminary experiments have shown that this concentration of ionophore is sufficient to short-circuit the membrane or impose a membrane potential without resulting in significant non-specific transport effects. Each experiment was usually repeated three times using membranes prepared from different animals. Within a given experiment, each point was determined from 3–5 replicate samples. Data are presented as means ± S.E.M. of these replicates in a single representative experiment. Similar qualitative findings were obtained in the repetition of an experiment. Data were analyzed using the computer program SigmaPlot (Jandel), which provides an iterative best fit to experimental values.
Ca2+ and Mg2+ activity values in external buffers were calculated in all kinetic experiments using the computer program Chelator (Schoenmakers et al. 1992) and the chelators EGTA, N-hydroxyethyl-EDTA (HEEDTA) and nitrilotriacetic acid (NTA) (0.5 mmol l−1 each) taking into account the appropriate pH, temperature and ionic strength of the external media.
45Ca2+ was obtained from New England Nuclear Corp., Boston, MA, USA, while reagent grade chemicals valinomycin, tetramethylammonium hydroxide (TMA-OH) and D-gluconic acid lactone were purchased from Sigma Chemical Co., St Louis, USA.
RESULTS
Osmotic reactivity and Ca2+ binding properties of lobster hepatopancreatic BLMVs
To determine whether 45Ca2+ uptake was into an osmotically active space or represented membrane binding, BLMVs were incubated in outside solutions with various osmolalities provided by using different sucrose concentrations, and 45Ca2+ uptake was determined at 15 min (Kikuchi et al. 1988). Fig. 1 indicates that a significant (P<0.01) linear relationship existed between vesicular 45Ca2+ content at 15 min and the reciprocal of the incubation medium osmolality for all membrane preparations. Extrapolation of the curve to the y-axis provided an index of non-specific surface 45Ca2+ binding to vesicles at 15 min and amounted to approximately 60% of total 45Ca2+uptake under control osmotic conditions (0 mmol l−1 sucrose) after 15 min of incubation. These results suggest that a significant proportion of the hepatopancreatic BLM preparations were sealed, osmotically reactive and displayed a binding component that had to be taken into account during subsequent influx assessments. A high osmotic sensitivity, with little or no binding of 22Na+ to membrane preparations, has already been reported in a study of 22Na+ uptake by hepatopancreatic BLMVs under the same osmotic conditions as described in the present study (Duerr and Ahearn, 1996). The present results show that considerably more 45Ca2+ than 22Na+ tends to bind to hepatopancreatic BLMVs. To correct for non-specific 45Ca2+ binding, blanks were prepared during all uptake experiments by exposing membrane preparations and isotope simultaneously to ice-cold stop solution and filtering immediately to collect vesicles for counting of radioactivity. The resulting bound Ca2+ was subtracted from total Ca2+ uptake at selected exposure intervals to provide estimates of transported Ca2+ alone.
Effects of the Ca2+ ionophore A23187 and chelator EGTA on Ca2+ efflux
Fig. 2 depicts 45Ca2+ efflux from preloaded vesicles in the presence of either EGTA alone or EGTA plus Ca2+ ionophore A23187. BLMVs were preloaded with 45Ca2+ and, after a designated preloading period, a volume of loaded vesicles was removed and diluted in an incubation buffer containing 100 mmol l−1 mannitol, 50 mmol l−1 TMA gluconate, 50 mmol l−1 KCl, 1 mmol l−1 EGTA and 20 mmol l−1 Hepes/Tris (pH 7.4) with or without 20 µmol l−1 (final concentration) A23187. As illustrated in Fig. 2, 45Ca2+ efflux was significantly accelerated in the presence of A23187 compared with the treatment lacking the ionophore, indicating the presence of 45Ca2+ within an osmotically reactive intravesicular space.
Occurrence of Ca2+-ATPase in lobster hepatopancreatic BLMVs
To investigate whether Ca2+-ATPase occurred in the hepatopancreatic basolateral membrane preparation, a series of time-course experiments was undertaken. Fig. 3 shows 45Ca2+ uptake in the presence and absence of 5 mmol l−1 ATP, indicating that 45Ca2+ uptake was significantly greater (P<0.01) in the presence of ATP than in its absence at each time point. The presence of 0.1 mmol l−1 vanadate in one group of vesicles with ATP significantly reduced the 45Ca2+ uptake, providing further evidence that a P-type Ca2+-ATPase occurred in the basolateral membrane of lobster hepatopancreas. It is noteworthy that 45Ca2+ uptake in the presence of ATP was linear during the first 2 min of incubation and therefore approximated the initial transport rate. A 1 min incubation time was used in subsequent kinetic experiments on Ca2+-ATPase. To characterize the optimum pH for Ca2+-ATPase in lobster hepatopancreatic BLMVs, buffers at three different pH values were used in one time-course experiment. The results displayed in Fig. 4 suggest that pH 7.4 was near the optimum pH in terms of 45Ca2+ transport, and this value was therefore was chosen for all later experiments concerning Ca2+-ATPase.
Kinetic properties of the Ca2+-ATPase in lobster hepatopancreatic BLMVs
where J is the total 45Ca2+ influx in pmolµg−1 protein 8 s−1, Jmax is the apparent maximal carrier-mediated influx, Kt (in nmol l−1) is the apparent Ca2+ activity resulting in half-maximal influx, and [Ca2+]o is the external Ca2+ activity determined for each total Ca2+ concentration using Chelator.
A nonlinear, iterative, best-fit computer program using equation 1 was employed to analyze the data in Fig. 5. Apparent kinetic parameters for the ATP-dependent carrier process calculated in this manner were Jmax= 1.07±0.06 pmol µg−1 protein 8 s−1; Kt=65.28±4.39 nmol l−1.
Occurrence of Ca2+/Na+exchange and the effects of membrane potential on Ca2+/Na+ exchange in hepatopancreatic BLMVs
The possible presence of a Ca2+/Na+ antiporter in hepatopancreatic BLMVs was investigated in a time-course experiment using short-circuited vesicles that had been preloaded with Na+ or H+, providing outwardly directed Na+ and H+ gradients. The data shown in Fig. 6 indicate that 45Ca2+ uptake in vesicles with such a Na+ gradient was significantly greater than in vesicles without a Na+ gradient or in vesicles with an outwardly directed H+ gradient. This ‘overshoot phenomenon’ under these conditions suggests the occurrence of a carrier system that can transiently transport Ca2+ against a concentration gradient using energy inherent in the outwardly directed Na+ gradient.
Fig. 7 illustrates an experiment to clarify the specificity of the Na+-gradient-dependent 45Ca2+ uptake process. In this experiment, the uptake of 45Ca2+ by vesicles with outwardly directed Na+ or Li+ gradients was determined at 15 s and at 6 min intervals. The results of this experiment show that the Ca2+/Na+ antiporter clearly demonstrated a preference for Na+ as an internal exchangeable substrate over Li+. Diltiazem, a known inhibitor of Ca2+/Na+ exchange in mitochondrial membranes (Gunter et al. 1994), was ineffective against the antiporter in hepatopancreatic BLMVs (Fig. 7), suggesting a possible fundamental difference between the cation exchangers in the two cellular locations.
Fig. 8 describes the effects of an imposed transmembrane potential (inside negative or positive; K+/valinomycin) on Na+-gradient-dependent 45Ca2+ uptake in hepatopancreatic BLMVs. Fig. 8 indicates that an inside-negative vesicular membrane potential reduced the stimulatory effects of a Na+ gradient in terms of 45Ca2+ uptake, while an inside-positive membrane potential enhanced the effect of the Na+ gradient. These results strongly suggest that an electrogenic process is responsible for the exchange of 45Ca2+ with Na+.
Kinetic properties of Ca2+/Na+ exchange in lobster hepatopancreatic BLMVs
Preliminary 45Ca2+ uptake experiments with hepatopancreatic BLMVs using exposure intervals of 1–15 s (data not shown) suggested that initial rates of 45Ca2+/Na+ exchange could be approximated using an exposure interval of 8 s. 45Ca2+ influx in BLMVs was measured as a function of external 45Ca2+ activity (12.5 nmol l−1 to 180 µmol l−1) in the presence of an outwardly directed Na+ or Li+ gradient, and the results are displayed in Fig. 9. All influx values were corrected for non-specific binding, as reported previously.
As shown in Fig. 9, 45Ca2+ influxes into vesicles preloaded with either intravesicular Na+ or Li+ were both hyperbolic functions of external Ca2+ activity, possessing rates that could be described by equation 1. Michaelis–Menten constants for vesicles preloaded with Na+ were Kt=14.57±5.02 µmol l−1 and Jmax=2.72±0.23 pmol µg−1 protein 8 s−1. Constants determined for vesicles preloaded with Li+ were Kt=20.17±6.99 µmol l−1 and Jmax=1.03±0.10 pmol µg−1 protein 8 s−1. These results suggest that both internal cations could support the exchange process with external Ca2+, but that the maximal transport velocity was three times greater with Na+ than with Li+. There was no apparent difference between the effects of the two preloaded cations on 45Ca2+ binding to the external membrane surface (e.g. Kt).
Apparent stoichiometry of 45Ca2+/Na+ exchange
where Jmax is maximal 45Ca2+ influx, (Kt)n is an affinity constant modified to accommodate multisite interactions (interaction coefficient) and the Hill coefficient n is an estimate of the number of reactive Na+ binding sites on the internal vesicular surface. The positive vertical axis intercept of Fig. 10 at 0 mmol l−1 Na+ suggests the occurrence of a Na+-independent Ca2+ uptake process such as diffusion under these conditions. The SigmaPlot curve-fitting program was used to obtain estimates for the three kinetic parameters using equation 2. The best-fitting curve provided the following values for these constants: Jmax=1.40±0.02 pmol µg−1 protein 8 s−1; (Kt)n=10.34±0.29 mmol l−1; and n=2.91±0.22. These results, and those of the electrogenic experiment shown in Fig. 8, suggest that this cation antiporter has an exchange stoichiometry of at least 1Ca2+/3Na+.
DISCUSSION
The present study focuses upon the characterization of Ca2+ transport mechanisms of the lobster hepatopancreatic epithelial basolateral membrane. This membrane was prepared using a Percoll gradient centrifugation technique that produced sealed, osmotically reactive vesicles (Figs 1, 2; Table 1). These results, together with the occurrence of an ‘overshoot phenomenon’ in time-course Ca2+ uptake experiments (Fig. 6), suggest that the Ca2+ uptake values reported in this study represent transport processes rather than non-specific binding.
Data from basolateral vesicle orientiation experiments resulting from a Percoll isolation procedure (Table 1) confirmed the previously reported distribution of leaky sheets, right-side-out vesicles and inside-out vesicles generated using a sucrose gradient (Ahearn et al. 1987) and suggested that five times more sealed vesicles were oriented right-side-out than in the other direction.
Sea water contains approximately 10 mmol l−1 Ca2+, and it has been reported that lobsters (Homarus americanus) drink a considerable volume of sea water. In addition, they may consume much or all of their old exoskeleton during the process of molting (Mykles, 1980). A significant amount of Ca2+ is therefore available to crustacean hepatopancreatic epithelial cells during ecdysis, and this ion is also probably a component of the stomach contents and the hepatopancreatic ducts during normal feeding activities in intermolt. Our previously reported results (Ahearn and Zhuang, 1996; Zhuang and Ahearn, 1996) suggest that Ca2+ influx into lobster hepatopancreatic brush-border membrane vesicles occurs by a combination of three transport processes: (1) an amiloride-sensitive carrier system; (2) an amiloride-insensitive carrier system; and (3) a verapamil-inhibited, potential-dependent ion channel. Similar findings were also reported for Ca2+ transport by apical membrane vesicles of the kidneys in the same animal, suggesting a common physiological theme for regulating the transmembrane flow of this divalent cation by marine crustacean epithelia (Ahearn and Franco, 1993). It is critical for a crustacean to harden its exoskeleton rapidly during the post-molt stage. Therefore, efficient hepatopancreatic basolateral mechanisms are essential to translocate Ca2+ from their sequestration sites, such as foregut gastroliths and hepatopancreatic R-cell organelle concretions, to the blood where it is needed for exoskeletal hardening.
Our previously reported preliminary results and the data from the present investigation suggest that Ca2+ efflux from lobster hepatopancreatic basolateral membrane vesicles occurred by a combination of two carrier-mediated transport processes: (1) a high-affinity, vanadate-sensitive, ATP-dependent Ca2+-ATPase (Kt=65.28 nmol l−1); and (2) a low-affinity, electrogenic 1Ca2+/3Na+ antiporter system (Kt=14.57 µmol l−1). These mechanisms generally resemble those previously described for crab gill basolateral membranes (Towle, 1993; Flik et al. 1994). A quantitative comparison between the carrier-mediated Ca2+ efflux kinetic constants for the lobster hepatopancreatic basolateral membrane and those from crab gill, fish gill, fish intestine and rat kidney are displayed in Table 2. Comparison of the values from the five types of animals shown in this table suggests that, while the apparent affinity constant for the lobster Ca2+-ATPase is well within the range described for other tissues, the affinity constant for the electrogenic 1Ca2+/3Na+ antiporter is considerably greater than the other values reported for vertebrates and invertebrates. The markedly low apparent affinity of the hepatopancreatic basolateral membrane may relate to the role of this tissue in Ca2+ sequestration during the molt cycle. Hepatopancreatic R-cells are known to sequester significant concentrations of Ca2+ during premolt when this ion is released from the old exoskeleton and enters the blood. Ca2+ is also stored in stomach gastroliths and must pass through the hepatopancreatic epithelium to reach this latter stomach storage depot. After the molt, the new exoskeleton must be recalcified by the Ca2+ kept in the sequestration sites, and transfer of Ca2+ occurs through the hepatopancreatic epithelium to the blood. The low apparent affinity of the hepatopancreatic basolateral electrogenic antiporter may be an adaptation of the cell to its role in molting, enabling it to regulate larger swings in intracellular Ca2+ activity than those that occur in similar epithelial cells from crustacean gills or vertebrate gut and renal tissues.
Fig. 11 is a comparison of the transporting roles of the lobster hepatopancreatic basolateral Ca2+-ATPase and the electrogenic 1Ca2+/3Na+ antiporter at intracellular Ca2+ activities between 1 and 100 000 nmol l−1 calculated using apparent Kt and Jmax values obtained from this study (Table 2), with the maximal transport rates being modified under the assumption that only the protein associated with inside-out vesicles was responsible for Ca2+ transport by the Ca2+-ATPase and the 1Ca2+/3Na+ antiporter. This assumption is justified for the ATPase because it is assumed that ATP can only gain access to its binding site from the cytoplasmic membrane surface. In the absence of other driving forces, the external addition of ATP would only stimulate transport by inside-out vesicles. The asymmetric binding properties of the antiporter similarly justify the hypothesis that only inside-out vesicles are responsible for the exchange properties of Na+-loaded vesicles. Similar reasoning has been applied to Ca2+ transport by the ATPase and antiporter of the intestinal epithelium of the fish Oreochromis mossambicus by Schoenmakers and Flik (1992). While crustacean and fish epithelia may differ quantitatively in the relative amounts of Ca2+ transferred by each protein, the qualitative roles of inside-out and right-side-out vesicles in Ca2+ movements in these two species are probably similar.
Given the above assumptions and caveats concerning vesicle orientation, Fig. 11 suggests that, at intracellular Ca2+ activities that might occur in typical epithelial cells (e.g. 100–500 nmol l−1), more than 90% of Ca2+ efflux takes place by way of the Ca2+-ATPase. Between 1000 and 10 000 nmol l−1, the range of intracellular activities that might occur during temporary Ca2+ storage or transcellular Ca2+ movements taking place at certain times in the molt cycle, the electrogenic 1Ca2+/3Na+ exchanger assumes a greater role in moving the divalent cation out of the cell. In Fig. 11, the crossover point (where efflux by both processes is approximately equal) is 10 000 nmol l−1 (a value near the apparent Kt for the exchanger). Flick et al. (1994) described Ca2+ efflux by these two transporters in crab gill epithelia and found a crossover point at 500 nmol l−1, considerably below that of the hepatopancreas. These results suggest that this crossover point may have biological relevance to the roles the different cell types play in Ca2+ balance in the two animals. Because the gills are unlikely to store a large amount of Ca2+ or to move massive quantities of this cation to and from sites of sequestration, and because their maximal transport capacities are lower than those of the hepatopancreas (Table 2), it can be assumed that the gill cells may experience smaller fluctuations in intracellular Ca2+ activity than those of the hepatopancreas and therefore do not biologically require as high a crossover point as do the cells of the hepatopancreas. From this reasoning, it is likely that in hepatopancreas the Ca2+-ATPase probably serves a housekeeping role by maintaining relatively low intracellular Ca2+ levels during intermolt when the animal feeds. However, during molting, when massive Ca2+ movements occur into and through this epithelium, the low-affinity, electrogenic 1Ca2+/3Na+ exchanger may assume a more significant role in accommodating the increased intracellular Ca2+ activity that results.
Data from the present investigation with intermolt lobsters provide tentative clues about how the membrane proteins of the hepatopancreatic brush-border and basolateral membranes may be involved in transcellular Ca2+ movements during different molt stages. During the intermolt or feeding stage, the previously reported (Ahearn, 1996; Ahearn and Zhuang, 1996) brush-border amiloride-sensitive carrier system (i.e. an electrogenic 2Na+/1H+ antiporter) may play a major role in Ca2+ absorption from the diet into epithelial cells, followed by Ca2+ efflux into the blood through the Ca2+-ATPase and 1Ca2+/3 Na+ antiporter in the basolateral membranes. During the premolt stage, some physiological processes are highly activated to store Ca2+ solubilized from the old exoskeleton. Hepatopancreatic epithelial cells probably participate in these processes by upregulating presently undescribed basolateral Ca2+ uptake mechanisms from the blood (e.g. putative Ca2+ channels or carrier processes) so that Ca2+ can then flow efficiently down its electrochemical gradient into the epithelial cells. Some of this Ca2+ may be sequestered into mitochondria or other organelles, and the rest is transported out of the cell through the brush border by the amiloride-insensitive carrier system for sequestration by the gastroliths in the stomach. Following ecdysis, an increase in Ca2+ concentration in the stomach due to the solubilization of Ca2+ from the gastroliths may trigger the opening of the previously described Ca2+ channels in the brush-border membrane, which combine with Ca2+ uptake by the amiloride-sensitive carrier system to accelerate Ca2+ translocation from lumen to cell. Within the cell, the mitochondria and other organelles may release stored Ca2+, which joins with entering gastrolith Ca2+ to be transported to the blood by way of the Ca2+-ATPase and 1Ca2+/3Na+ antiporter in order to harden the new exoskeleton. Clearly, considerably more work needs to be undertaken to complete this tentative picture of the role of the hepatopancreas in Ca2+ balance, but the present study represents a significant step towards our understanding of the cellular mechanisms regulating the massive transcellular Ca2+ movements that accompany the growth processes of crustaceans.
ACKNOWLEDGEMENTS
This investigation was supported by National Science Foundation Grant IBN-9317230 and by Environmental Protection Agency Grant R-823068-01-0.