The cells of the corpuscles of Stannius appeared to be more active in eels acclimated to sea water than in eels acclimated to fresh water. In acclimated eels, however, total and ionic plasma calcium concentrations and stanniocalcin titers did not differ. This suggests that levels of stanniocalcin in freshwater-acclimated eels are sufficient to maintain normocalcemia in sea water. When freshwater-acclimated eels were transferred directly to sea water, total and ionic calcium concentrations in the plasma increased significantly within 24 h but there was no apparent effect on the corpuscles of Stannius within the same time. This suggests that changes in secretory activity of the corpuscles of Stannius do not occur rapidly when they are presented with a hypercalcemic challenge. Conversely, when seawater-acclimated eels were transferred to fresh or distilled water there appeared to be a very rapid reduction in secretory activity in the corpuscles of Stannius, as indicated by the rapid accumulation of secretory granules. These data suggest that stanniocalcin turnover is more rapid in seawater-acclimated eels and that the secretory activity of the corpuscles of Stannius is rapidly reduced when a hypercalcic challenge is removed.
Euryhaline fish, like the eel, face extensive changes in environmental calcium concentrations upon migration between fresh water and sea water. In fresh water the ionic calcium level can be as low as one-tenth of that of the blood, whereas in sea water the ionic calcium level may reach values 10 times higher than that of the blood. Since euryhaline fish maintain their blood calcium levels quite precisely in both fresh and sea water (Chan and Chester Jones, 1968), they must be able to regulate their blood calcium in accordance with the environmental calcium challenges associated with migration.
Of all the fish endocrine systems studied, the corpuscles of Stannius (CS) have been most consistently implicated in the control of plasma calcium metabolism (Wendelaar Bonga and Pang, 1991). Stanniocalcin (STC), the hormone secreted by the CS, lowers plasma calcium levels by reducing gill calcium uptake (Fenwick and So, 1974; So and Fenwick, 1979; Milet et al. 1979; Lafeber and Perry, 1988; Lafeber et al. 1988a). Histological examination of the CS has revealed that the glands are consistently more active in seawater-adapted fish than in freshwater-adapted fish (Olivereau, 1964; Hanke et al. 1967; Wendelaar Bonga et al. 1976; Meats et al. 1978). In addition, CS of fish kept in calcium-deficient sea water appear to be inactive (Cohen et al. 1975; Wendelaar Bonga et al. 1980), suggesting that glandular activity is directly related to environmental calcium concentrations. Such observations have led to the suggestion that the release of STC is controlled directly by changes in plasma calcium levels (Aida et al. 1980; Wagner et al. 1989). This hypothesis was supported by observations on the effects on STC secretion of exposure of cells from the corpuscles of Stannius in vitro to media of various calcium concentrations. However, the rise in the extracellular calcium level required to stimulate STC release in vitro is outside the physiological range ([Ca2+]>2mmoir1; Hanssen et al. 1991).
We have studied the regulation of STC release by examining CS from eels exposed to changes in the water calcium concentration. The effects of rapid transfers between fresh water and sea water on the ultrastructure of the CS and on the plasma calcium and STC concentrations were examined.
Materials and methods
Sexually immature freshwater European eels (Anguilla anguilla L.) weighing between 100 and 300 g were obtained from a commercial fish dealer in the Camarque, France. Fish were kept in well-aerated 10001 tanks at approximately 12°C. Eels were acclimated to circulating city of Nice tap water (1.63 mmol 1-1 Ca2+, 0.12 mmol 1-1 Na+, 0.09 mmol 1-1 Cl-) or Mediterranean sea water (12.50mmol 1-1 Ca2+, 530 mmol 1’1 Na+, 552 mmol 1-1 Cl-) for at least 3 weeks. The eels were not fed.
Freshwater-acclimated eels were netted and transferred directly to 2001 tanks containing circulating fresh water or sea water for 24 h. Seawater-acclimated eels were transferred for a 24 h period to seawater tanks, freshwater tanks or tanks containing running distilled water (Ca2+ not detectable with a commercial calcium kit, pH7.0). At the end of the experiment fish were anesthetized in ethylamino-benzoate (MS222, 2.5 g I-1, pH 7.8) and blood samples were taken by puncture of the caudal vessels. The CS were removed as described before (Hanssen et al. 1989), and prepared for electron microscopy.
Plasma ionic and total calcium concentrations were measured as previously described (Hanssen et al. 1989). Plasma ionic calcium was determined with an ionic calcium analyzer (ICA-1, Radiometer). Total plasma calcium was measured with a commercial calcium reagent kit (Sigma).
Enzyme-linked immunosorbent assay (ELISA)
A competitive ELISA technique, developed and validated by Mayer-Gostan et al. (1991), was used to quantitate STC levels in plasma samples. This technique is based on competition between free STC in standard or plasma samples and STC immobilized on microtiter plates for the STC antibodies. The trout STC antiserum used (RADH-1; Kaneko et al. 1988) has a high degree of crossreactivity with eel STC. Purified trout STC (Lafeber et al. 1988b) served as a standard.
The wells of microtiter plates were coated with trout STC (1.85nmoll-1) in 200μl of coating buffer (0.05 mol I-1 bicarbonate buffer, pH 9.6), except for the wells in the first column of the microtiter plate (blanks), which received 200 μ1 of a bovine serum albumin (BSA) solution of equivalent protein content in coating buffer. Coating lasted for 1 h at 37°C; coated plates were stored at 4°C. Between incubation steps, coated plates were rinsed with washing buffer (0.01 mol I-1 phosphate-buffered saline; PBS, pH7.4, with 0.05% Tween-20; Bio-Rad). For competition, antigen and RADH-1 antiserum were diluted in dilution buffer (i.e. washing buffer containing 2% porcine serum). Equal volumes of antigen and diluted antiserum were incubated in 4 ml Minisorp tubes (Nunc) overnight at room temperature (final antiserum dilution 1:80000). Tubes containing diluted anti-serum (1:80000) only were incubated under the same conditions. After washing, the wells of coated plates were blocked with 200 μ1 of dilution buffer (containing 2 % porcine serum) at 37°C for 1 h. The plates were washed and filled with 200 μ1 per well of the antigen/antiserum mixture except for the wells in columns 1 (blanks) and 2 (maximum absorbance) which received 200/d of the diluted antibody. Plates were incubated for 2h at 37°C. The wells were washed and incubated with 200μ1 of 1:5000 goat anti-rabbit peroxidase immunoconjugate (Nordic) in dilution buffer for lh at 37°C. For quantification of the peroxidase immunoconjugate bound to the wells, an enzymatic reaction was used with o-phenylenediamine (OPD; Sigma) as a substrate. After washing, 200μ1 of substrate (0.05% OPD in citrate/phosphate buffer; 0.2moll-1 Na2HPO4, 0.1 mol I-1 citric acid, pH5.0, with 0.025% hydrogen peroxide) was added at room temperature. The reaction was allowed to proceed for 20-30 min in the dark and was stopped by the addition of 50 /d of 2 mol I-1 H2SO4. Absorbance was measured at 492 nm in a microplate reader (Titertek).
Freshly dissected CS were prefixed for 10 min in 3% glutaraldehyde in 0.1 mol I-1 cacodylate buffer, pH7.4, at room temperature. The CS were then fixed for lh at 0°C in 0.1 mol I-1 cacodylate buffer containing 1% (v/v) glutaraldehyde, 0.66% (w/v) osmium tetroxide and 1.66% (w/v) potassium dichromate. Post-fixation was in 2% (w/v) uranyl acetate in water. After dehydration in graded ethanols, the tissues were embedded in Spurt’s resin. Ultrathin sections were post-stained with Reynold’s lead citrate and examined in a Jeol 100 CXII electron microscope.
Calculations and statistics
ELISA results were evaluated using a logit transformation (Mayer-Gostan et al. 1991). A least-squares linear regression of the transformed linear part of the standard curve was used to estimate the STC concentration of the samples. All data are presented as means±s.E.M. The Mann-Whitney U -test was used for statistical evaluation. Significance was accepted for P<0.05.
Ultrastructurally, two types of endocrine cells may be distinguished in the CS of eels: large rounded cells with large secretory granules (type 1 cells) and small cells with long cytoplasmic extensions and small secretory granules (type 2 cells; Wendelaar Bonga and Pang, 1986). STC was immunocytochemically localized in the secretory granules of both cell types, indicating that both either sequester or produce stanniocalcin (Kaneko et al. 1989; Wendelaar Bonga et al. 1989). Because no differences were observed in the structure of the type 2 cells when eels were transferred between fresh water and sea water our present analyses focused on the type 1 CS cells.
The ultrastructure of eel CS cells before and after transfer is shown in Fig. 1. The cells of fish acclimated to fresh water appear to be quite inactive (Fig. 1A), being characterized by an abundance of large secretory granules and relatively little granular endoplasmic reticulum (GER). In contrast, the cells of fish acclimated for 3 weeks to sea water contain a well-developed and extensive GER and few secretory granules (Fig. 1B). Twenty-four hours after transfer of freshwater-adapted eels to sea water, the CS still did not exhibit the secretory features typical of the cells of seawater-acclimated fish (Fig. 1C) and the cells were similar in appearance to those of the freshwater-acclimated fish, containing many secretory granules. Examination of the cells of seawater-adapted fish transferred to fresh water for 24 h (Fig. 1D) or distilled water (Fig. 1E) does not reveal noticeable differences in the extent of the GER compared to the CS cells of the seawater-acclimated fish. However, accumulation of secretory granules is ob-served in the cells of both groups of fish, particularly in those transferred to distilled water.
The ionic and total calcium levels of the plasma of freshwater-and seawater-adapted eels did not differ significantly (Fig. 2). Twenty-four hours after transfer of freshwater-adapted eels to sea water, total plasma calcium level was signifi-cantly increased and the change was, for the most part, accounted for by an increase in plasma ionic calcium. Seawater-adapted eels 24 h after transfer to fresh water had plasma ionic and total calcium concentrations that were not significantly different from those of seawater-acclimated eels. Transfer to distilled water, however, produced a significant fall in plasma ionic calcium, which was reflected in the decreased plasma total calcium level.
Plasma STC concentrations in freshwater-and seawater-adapted eels were not significantly different. Values ranged from 0.90 to 1.30 nmol I-1 (Fig. 3). Plasma STC levels of freshwater-adapted eels transferred to sea water were not signifi-cantly altered, but seawater-adapted eels transferred to fresh water or distilled water showed a significant drop in plasma STC levels to values around 0.50nmol I-1.
In this study we investigated the relationship between the calcium concentration of the water and the CS activity in eels. In the short term, a transfer to hypercalcic water causes hypercalcemia but does not stimulate CS activity. This observation does not favor a direct regulatory role for plasma Ca2+ in STC secretion. Transfer to hypocalcic environments, however, induced a decrease in CS activity within 24 h, suggesting that a hypocalcic challenge is more adequately counteracted than a hypercalcic challenge.
A relationship between the activity of the CS and the calcium concentration of the water has been suggested previously. Urasa and Wendelaar Bonga (1987) reported that in tilapia a long-term elevation of the calcium concentration of the water led to an increase in CS volume accompanied by an elevation of the plasma calcium concentration. Their results were in line with the assumption of Bailey and Fenwick (1975) that a negative feedback exists between the secretory activity of the CS and the ionic calcium concentration of the plasma. Further circumstantial evidence for such a relationship came from in vitro studies by Aida et al. (1980) and Wagner et al. (1989), who concluded that increased medium calcium concen-trations enhanced the depletion of CS secretory granules and the release of immunoreactive STC. However, recent in vitro studies by our group on eel CS showed that STC release was only altered by an increase in extracellular calcium concentration well above that shown to follow most hypercalcic challenges (Hanssen et al. 1991).
In CS cells of freshwater-adapted eels we observed a moderately developed GER and an abundance of secretory granules, indicative of a relatively inactive cell. CS cells of seawater-adapted eels were more active than those of freshwater-adapted eels judging from the absence of many secretory granules and the presence of an extensive GER. Although granule depletion has been interpreted as evidence of high secretory activity of CS cells (Cohen et al. 1975), this does not necessarily apply to the CS cells of all seawater-adapted euryhaline teleost fish. Granular depletion, as found in the European eel (this study), was also evident in the killifish (Cohen et al. 1975; Wendelaar Bonga et al. 1980) but not in the stickleback (Wendelaar Bonga et al. 1976) or trout (Meats et al. 1978). However, in these two species a high incidence of exocytotic phenomena, indicating high secretory activity, was observed. Ultrastructural studies on CS cells have been carried out for a number of euryhaline teleost species. In these studies it was shown that CS cells of killifish (Cohen et al. 1975; Wendelaar Bonga et al. 1980), sticklebacks (Wendelaar Bonga et al. 1976), trout (Meats et al. 1978) and tilapia (Urasa and Wendelaar Bonga, 1987) are activated by transfer of the fish from fresh water to sea water. This change in glandular activity appears to be correlated to the calcium concentration of the water (Cohen et al. 1975; Wendelaar Bonga et al. 1980). Our transfer experiments show that transfer of an eel from fresh water to sea water for 24 h was not accompanied by activation of the CS. Transfer of an eel from sea water to fresh water or distilled water for 24 h did not affect the amount of GER in the CS cells. CS cells showed an extensive GER typical of activated cells. The CS of seawater-acclimated eels transferred to distilled water, and to a lesser extent the CS of seawater-acclimated eels transferred to fresh water, showed an accumulation of secretory granules. Accumulation of secretory granules was also found after a 24 h seawater-freshwater transfer of sticklebacks (Wendelaar Bonga et al. 1976). We conclude from our results that in acclimated eels the activity and the degree of granulation of CS cells are related to the calcium concentration of the water. In the short term, however, transfer between water containing high and low calcium levels does not necessarily have immediate effects on CS cell activity. It may have small effects on granulation, in particular after a hypocalcic challenge. This indicates that CS cell activity in the eel is not rapidly changed by fluctuations in water calcium concentration.
Plasma calcium levels
In freshwater-acclimated eels, ionic and total calcium concentrations of the plasma did not differ from previously reported values (1.3 mmol 1-1 and 2.5 mmol 1-1, respectively). Plasma calcium concentrations in seawater-acclimated eels did not differ significantly from those measured in freshwater-acclimated eels. Similarly, in the striped mullet no differences in plasma calcium levels were observed between fish adapted to freshwater or seawater habitats (Johnson, 1972). However, plasma calcium concentrations in trout (Meats et al. 1978) and tilapia (Urasa and Wendelaar Bonga, 1987) differed significantly between fish adapted to high or low levels of calcium in the water. A 24 h transfer of our freshwater-acclimated eels to sea water induced a significant increase in ionic calcium in the plasma to 1.8 mmol 1-1 and in total calcium to 3.25 mmol 1-1. These elevated plasma calcium levels upon transfer to sea water corroborate the report of Ogasawara and Hirano (1984) on Japanese eels, who found a transient elevation of the total plasma calcium level 24 h after transfer. We found that the ionic and total calcium concentrations in the plasma of seawater-acclimated eels after a 24 h transfer to fresh water did not change significantly. Similar observations were made by Ogasawara and Hirano (1984), who did not find significant effects on total plasma calcium in the Japanese eel after these transfers. Our seawater-acclimated eels transferred for 24 h to distilled water showed a decrease in both ionic and total plasma calcium levels to values around 0.8 and 2.0 mmol 1-1, respectively. This indicates that the eels lost control over their branchial Ca2+ permeability and we predict an increased net Ca2+ efflux under these conditions. A transfer to distilled water may lead to a shift of the transepithelial potential (TEP) across the gills to more negative values (with the plasma compartment negative relative to the ambient medium). Indeed, McWilliams and Potts (1978) demonstrated a positive TEP across the gills of brown trout at ambient calcium levels of 2.0 mmol 1-1 and higher; the TEP shifted to negative values at ambient calcium levels of 1.0 mmol 1-1 and lower. As a presumed shift in the TEP to negative values would facilitate Ca2+ uptake from the water, the loss of calcium in our fish when challenged with low ambient Ca2+ levels indicates a severe disturbance of the integumental permeability to Ca2+ and, hence, a net loss of this ion. Contrary to the effects on CS cellular activity, plasma ionic and total calcium concentrations were not related to the calcium concentration of the water after long-term acclimation, and one may interpret this to indicate calcium homeostasis. Transfer of eels between high-and low-calcium waters as reported here provokes only transient changes in plasma calcium levels.
Plasma stanniocalcin concentrations
In freshwater-adapted eels the plasma immunoreactive STC concentration was 1.25 nmol I-1. The plasma STC concentration in seawater-adapted eels did not differ significantly from this value. These values are lower than the 2.33 nmol I-1 reported recently by Mayer-Gostan et al. (1991) for freshwater-adapted eels. The increased calcium concentration of the plasma after a 24 h transfer of freshwater-adapted eels to sea water was not accompanied by a significant increase of the plasma STC concentration and thus plasma Ca2+ is not the primary trigger for STC release under these conditions. A 24h transfer of seawater-adapted eels to fresh water or distilled water reduced the plasma STC concentration by 50%. In acclimated eels plasma STC concentrations appear to be very constant and were not related to the calcium concentration of the water. A short-term hypocalcic challenge, however, immediately affects the plasma STC concentration. This indicates that such a challenge inhibits the release of STC.
In this first report dealing with the relationship between environmental calcium, plasma calcium and plasma STC concentrations we conclude that the CS of freshwater-adapted eels exhibit a relatively low secretory activity and that an increase in the calcium concentration of the water activates the CS, but only in the long term. A short-term increase in the calcium concentration of the water does not stimulate the CS despite a 30 % increase in the total calcium concentration of the plasma and a 38% increase in ionic calcium. Increased plasma calcium concentrations of this magnitude may well be within the physiological range of eel plasma. Indeed, in vitro experiments have shown that a 60% increase in the calcium concentration of the incubation medium did not affect STC release (Hanssen et al. 1991). The present in vivo experiments confirm our conclusion, based on the in vitro experiments, that plasma calcium fluctuations within the physiological range do not regulate STC release. High secretory activity was observed in the CS of seawater-adapted eels. This activity is not reflected by an increased plasma STC level, suggesting that the metabolic clearance of STC is higher in seawater-adapted eels. When seawater-adapted eels are transferred to fresh water or distilled water, the high synthetic activity of the endocrine cells of the CS seems to be maintained for at least 24 h, but increased granulation indicates that the release of secretory material is inhibited. The drop in plasma STC levels after transfer to low-calcium water suggests that although STC clearance remains high there is a decreased release of STC.
An anti-hypercalcemic role was attributed to STC by Fenwick and Forster (1972) on the basis of the ability of CS homogenates to prevent hypercalcemia after removal of the CS. Our present finding, that a hypercalcemic state evoked by transferring eels for 24 h from fresh water to sea water is not accompanied by a stimulated release of STC, indicates that a physiologically induced hypercalcemia is not a trigger for STC release. Rapid antihypercalcemic responses of the CS cells have so far only been reported after experimentally induced high plasma calcium levels well above the physiological range. Because a direct role of calcitonin in the immediate control of plasma calcium level is very unlikely in fish (Wendelaar Bonga and Pang, 1991), we conclude that regulatory mechanisms for the minute-to-minute control of plasma calcium are absent in eels, and possibly in fish in general. Calcium levels in plasma appear to be less tightly controlled in fish than they are in terrestrial vertebrates.
We thank Mr C. van der Meij for preparing the electron micrographs and Professor Dr B. Lahlou for his interest during our study. We thank Professor Dr P. Pang for providing the RADH-1 antiserum. R. Hanssen received a CNRS (Centre National de la Recherche Scientifique) grant through the Netherlands Organiz-ation for Scientific Research (NWO).