A method to culture tissue explants of the intestine from freshwater-adapted sockeye salmon (Oncorhynchus nerka) was developed to assess possible direct effects of cortisol on Na+-K+-ATPase activity. As judged by several criteria, explants from pyloric ceca and the posterior region of the intestine remained viable during short-term (6-day) culture, although Na+-K+-ATPase activity declined and basolateral components of the enterocytes were observed to be partially degraded. Addition of cortisol to the culture medium maintained Na+-K+-ATPase activity (over 2–12 days) above that of control explants and, in some cases, was similar to levels before culture. The response to cortisol was dose dependent (0.001–10 μg/ml). Within the physiological range, the response was specific for cortisol and showed the following hierarchy: dexamethasone ≥ cortisol > 11-deoxycortisol > cortisone. Insulin maintained Na+-K+-ATPase activity over controls in explants of ceca but not posterior intestine. To compare in vivo and in vitro responses, slow-release implants of cortisol (50 μg/g) were administered to salmon for 7 days. This treatment elevated plasma cortisol levels and stimulated Na+-K+-ATPase activity in both intestinal regions. The results demonstrate that the teleost intestine is a direct target of cortisol, this corticosteroid protects in vitro functionality of Na+-K+-ATPase, and explants retain cortisol responsiveness during short-term culture.
- Oncorhynchus nerka
- pyloric ceca
a primary mechanism of ion transport across intestinal epithelium is the basolateral Na+-K+-ATPase. This ATP-requiring enzyme provides an electrochemical gradient that drives solute and water transport (24, 48). The vertebrate intestine is a major osmoregulatory organ, and, in fish, intestinal salt and water uptake is essential for maintaining internal water balance, especially in seawater where osmotically lost water must constantly be replaced (44).
Electrolyte transport across the vertebrate intestine is modulated by steroids of the adrenal cortex, the glucocorticoids, and mineralocorticoids (5, 6, 22, 40). In teleost fishes, cortisol is the primary corticoid secreted by the diffuse interrenal tissue (adrenal homologue), and, because aldosterone is generally not found in plasma, cortisol exhibits a wide spectrum of action, including both ionoregulatory and metabolic functions (32).
In vivo, intestinal ion and water transport is regulated by cortisol in teleosts. In several species of the genera Anguilla (eels) and in Carassius auratus auratus (goldfish) and Fundulus heteroclitus (cyprinodont), adrenocorticotropic hormone or cortisol stimulates intestinal water absorption and Na+-K+-ATPase activity in intact fish and restores these mechanisms that are otherwise decreased with interrenalectomy or hypophysectomy (12–14, 16, 19, 20, 34). The augmentation of water absorption is specific for cortisol, since other C21 steroids are without effect (20, 36).
In juvenile migratory salmonids, intestinal Na+-K+-ATPase activity and fluid uptake become elevated during the seasonal event termed parr-smolt transformation that takes place in fresh water and is preparatory for seawater survival (4, 33, 50, 53). These changes in the intestine are part of the complex of physiological mechanisms that lead to increased salinity tolerance (21, 31). Cortisol is thought to mediate osmoregulatory adaptations of the salmonid intestine (46). The cortisol-synthesizing capacity of the coho salmon's interrenal tissue and circulating concentrations of cortisol in sockeye salmon are elevated during this period of development (15, 54), and cortisol stimulates intestinal Na+-K+-ATPase activity and fluid uptake in anadromous rainbow trout and brown trout (25, 39), Atlantic salmon (8), and chinook salmon (51). Finally, intestinal fluid absorption in Atlantic salmon (Salmo salar) is inhibited by the glucocorticoid receptor antagonist RU-486, indicating a cortisol-specific regulation (49).
Despite this body of evidence, the regulation of osmoregulatory mechanisms by cortisol in fish intestine have been limited to in vivo studies. A culture system would be advantageous to discriminate between direct and indirect actions of regulatory signals, particularly in salmon, since a host of endocrine factors are elevated during parr-smolt transformation that likely interact with cortisol and contribute to increased hypoosmoregulatory ability (28). These include growth hormone, insulin-like growth factor-I (IGF-I), and thyroid hormones (1, 37, 55).
In the present paper, a short-term culture system for intestinal explants from freshwater-adapted sockeye salmon (Oncorhynchus nerka) was developed. Two regions of the intestine were examined due to their prominent absorptive functions: the numerous pyloric ceca, which are attached and open to the anterior intestine, and the posterior region of the intestine. Although ceca are the primary site for nutrient uptake (3), both ceca and posterior intestine are major sites of salt and water balance (4, 50, 51). Cytological and physiological changes during culture were assessed. Na+-K+-ATPase activity was the endpoint for analyses of functional responses to cortisol. Part of these data have appeared in abstract form (52).
MATERIALS AND METHODS
Juvenile sockeye salmon, between 1 and 2 yr of age, were maintained in recirculating, freshwater aquaria at 12°C under a natural photoperiod. Before dissections, salmon were lethally anesthetized in 3-aminobenzoic acid ethyl ester (MS-222, 200 mg/l) buffered with sodium bicarbonate (400 mg/l). All maintenance and experimental manipulations of salmon and tissues were approved by the University of Otago Committee on Ethics in the Care and Use of Laboratory Animals and adhered to the APS's Guiding Principles in the Care and Use of Animals.
Salmon were anesthetized in 0.2% 2-phenoxyethanol and given intraperitoneal injections of cortisol (50 μg/g body wt) suspended in a lightly heated mixture (1:1) of vegetable oil and vegetable shortening as described in detail elsewhere (45). Controls were injected with vehicle.
The technique used was adapted from a previously published method for culturing gill filaments from salmon (29). Several pyloric ceca and the posterior region of the intestine were cut from the gastrointestinal tract, gently sliced open along the long axis, and laid flat (serosal side down) on an ice-cold glass sheet. Each intestinal region was then cut into pieces of ∼1 × 1.5 mm and carefully placed in 24-well culture plates (in duplicate) containing preincubation medium (MEM with Hanks' salts, 5 mg/ml BSA, 250 U/ml penicillin G, and 250 μg/ml streptomycin sulfate, adjusted to pH 7.8 at 14°C). After several hours, the medium was replaced with MEM containing Earle's salts (pH 7.8), 4 mg/ml BSA, 292 μg/ml l-glutamine, 50 U/ml penicillin G, and 50 μg/ml streptomycin sulfate. Explants were incubated at 14°C in an air-tight humidified chamber and gassed daily with 95% O2-5% CO2 (29). Every third day, the culture medium was replaced with freshly prepared MEM. Steroids were dissolved in ethanol and bovine insulin in 0.01 N HCl, before addition to culture media, so that the final solvent concentration in all plate wells was 1 μl/ml (including wells with control tissue). Although explants were occasionally found to adhere to the bottom of the plate well, they typically remained unattached during culture.
For analyses of tissues by light microscopy, intestinal explants were fixed in aqueous Bouin's fixative overnight at room temperature, dehydrated through an ethanol series, and infiltrated and embedded in paraffin. Sections were cut at 6 μm and stained with hematoxylin and eosin, all according to standard histological procedures. Transmission electron microscopy was carried out on tissues that were prepared as follows: explants were fixed in 3% glutaraldehyde in 0.1 M cacodylate buffer at 4°C overnight. Postfixation was in 1% osmium tetroxide-1.5% potassium ferrocyanide in 0.1 M cacodylate buffer for 1 h. Tissue was then dehydrated and embedded in Agar 100 resin. Ultrathin sections were cut at 80 nm, stained with uranyl acetate and lead citrate, and viewed on a Philips CM100 transmission electron microscope.
Measurement of tissue protein and sodium.
Before protein and sodium analyses, explants were dried overnight at 65°C, weighed (for normalizing protein and sodium concentrations), and then dissolved in 0.1 N HNO3 overnight. Methods for protein determinations followed the procedures of McCormick and Bern (29). Tissue was homogenized in 0.1% sodium deoxycholate and assayed for protein using bicinchoninic acid (BCA; Sigma) as chromogen. Sodium concentrations were measured by inductively coupled plasma spectroscopy at Chemsearch Laboratories (Department of Chemistry, University of Otago).
Measurement of plasma cortisol and Na+-K+-ATPase.
Concentrations of cortisol were measured by radioimmunoassay in unextracted plasma (54). The assay for determining Na+-K+-ATPase activity has been reported by Veillette and Young (53). Briefly, crude homogenates of intestine were prepared, and the hydrolysis of ATP was enzymatically coupled to the conversion of NADH to NAD+. The utilization of NADH was measured over 8 min in the presence and absence of ouabain. The difference in activity, normalized to amount of protein (BCA assay), yielded ouabain-sensitive Na+-K+-ATPase activity.
For all experiments, each gut region was analyzed separately. In experiments with a control and a single treatment, significant differences (P < 0.05) were assessed by one-way ANOVA. In all experiments involving results for multiple days, there was a significant interaction between treatment and day (two-way ANOVA). Therefore, for these results, each day was analyzed separately to determine differences between treatments. Two-way ANOVA was used to assess the simultaneous effects of 1) fetal bovine serum and cortisol and 2) insulin and cortisol. When appropriate, Tukey's honestly significant difference procedure was applied for post hoc comparisons between means. We used one-way ANOVA to analyze dose-response curves, followed by Dunnett's post hoc test to determine significant differences from controls (0 μg/ml). Analyses of half-maximal responses (effective concentrations = EC50) to cortisol and dexamethasone were performed with the use of four-parameter curve fitting on mean Na+-K+-ATPase activity for each dose.
Cytological changes in culture.
Light microscopy analyses showed that, after 3 or 6 days in culture, shrinkage of the tissue occurred in general, including a reduction in size of the mucosal folds and cells of the epithelium (Fig. 1). A few parts of the tissue exhibited signs of necrosis, and there was some edema of the tissue underlying the epithelium, the lamina propria. One of the more pronounced changes during culture was partial degradation at the base of the single layer of epithelial cells, resulting in a separation of the epithelium from the underlying tissue in some sections. However, at the ultrastructural level, the cells exhibited functionality on the basis of intact nuclei and other organelles, intact microvilli of the brush border, and intact cell-to-cell borders and junctions (Fig. 2). Cortisol-treated explants were not examined cytologically.
Tissue protein and sodium concentrations in culture.
Total protein and sodium content of intestinal explants remained relatively unchanged in both ceca and posterior intestine after 6 days in culture, although sodium content of ceca rose moderately (Table 1).
Effect of cortisol on tissue protein in culture.
Potential effects of cortisol on tissue protein during culture were examined because measurements of Na+-K+-ATPase activity were normalized to total protein and changes in protein content could affect estimations of enzyme activity. Table 2 shows the effects of cortisol exposure (1 μg/ml) on protein content of intestinal explants in culture for 6 days. The amount of protein in explants of pyloric ceca was unaffected by cortisol, but this was significantly lowered in the posterior intestine compared with controls.
Time course of cortisol action on Na+-K+-ATPase in culture.
Explants were cultured from 2 to 6 days in the presence and absence of 1 μg/ml cortisol and/or 10% fetal bovine serum. This concentration of cortisol was chosen because it has been shown to stimulate Na+-K+-ATPase activity in the cultured gill of several salmonids (29, 30). Although BSA was a constituent of the culture medium, addition of fetal bovine serum was explored to determine whether it affected Na+-K+-ATPase activity or could potentiate a response to cortisol, perhaps through growth factors (or other factors) present in the serum.
In the present and subsequent experiments, Na+-K+-ATPase activity generally declined after several days in culture in the absence of cortisol. Treatment with cortisol significantly maintained Na+-K+-ATPase activity over that shown in control tissue in pyloric ceca within 2 days of culture, and this response continued through 6 days of culture (Fig. 3). In the posterior intestine, cortisol exposure maintained Na+-K+-ATPase activity over controls after 4 and 6 days of culture. At no time did activity exceed initial, preculture levels in any of the treatment groups. Addition of 10% fetal bovine serum to the culture medium did not enhance the maintenance of in vitro Na+-K+-ATPase activity or change the response to cortisol (Fig. 3). Fetal bovine serum contained <2 ng/ml cortisol, as assessed by radioimmunoassay.
Addition of cortisol (1 μg/ml) to the culture medium 3 days after the start of explant culture resulted in significant elevations of Na+-K+-ATPase activity over controls in both intestinal regions within 2 days. This response continued for 4 days after addition of cortisol to the medium (Fig. 4).
Effect of insulin and cortisol on Na+-K+-ATPase in culture.
The in vitro effect of insulin (10 μg/ml), with or without cortisol (10 μg/ml), was explored. After 6 days of culture, cortisol significantly maintained Na+-K+-ATPase activity over controls in pyloric ceca and posterior intestine (Fig. 5). Insulin did not enhance the action of cortisol on either gut region, although the presence of insulin changed Na+-K+-ATPase activity, resulting in significantly elevated Na+-K+-ATPase activity in pyloric ceca and a small but significant decrease (main effect of insulin) in the posterior intestine. Cortisol maintained Na+-K+-ATPase activity over that of controls through 12 days of culture in the pyloric ceca and 9 days in the posterior intestine (Fig. 5), indicating a continued response of explants to hormone beyond 6 days of culture.
Dose-response relationship between several corticosteroids and Na+-K+-ATPase in culture.
The in vitro response of Na+-K+-ATPase activity to cortisol and dexamethasone was dose dependent in both regions of the intestine (Fig. 6). Significant responses (compared with controls, 0 μg/ml) to both cortisol and dexamethasone were seen at 0.1 μg/ml for pyloric ceca and at 1 μg/ml for posterior intestine. To compare the relative potency of these two steroids, four-parameter curve fitting was used to calculate the dose required for a half-maximal response (EC50). The EC50 for dexamethasone was lower than that for cortisol in both regions of the intestine (Table 3).
To further evaluate whether the response of Na+-K+-ATPase activity was specific to cortisol, the effects of cortisol, 11-deoxycortisol, and cortisone were compared at doses of 0–10 μg/ml (Fig. 7). Cortisol was more potent than the other natural corticosteroids. The only corticoid to induce a response near a physiological concentration (0.1 μg/ml) was cortisol, although responses were elicited by 11-deoxycortisol and cortisone at higher doses (1–10 μg/ml). EC50 results were not calculated for these steroids because a maximal response was not achieved with 11-deoxycortisol or cortisone. These results, together with those of Fig. 6, suggest the following hierarchy of potency for these steroids: dexamethasone ≥ cortisol > 11-deoxycortisol ≥ cortisone.
In vivo effect of cortisol implants.
Peritoneal implants of cortisol for 7 days elevated plasma cortisol concentrations over controls and significantly, although moderately, stimulated Na+-K+-ATPase activity in both pyloric ceca and posterior intestine (Table 4).
The most important finding of this study is that the intestine of sockeye salmon is responsive to cortisol in short-term culture, as evidenced by the maintenance of Na+-K+-ATPase activity. This is the first demonstration of a direct action of cortisol on the teleost intestine. The response to cortisol was dose dependent and specific for cortisol in a physiological range. The present results implicate the intestinal epithelium as a target for the osmoregulatory actions of cortisol in salmonids.
The specificity of the cortisol response reported here for salmon intestine is supported by corticoid-binding studies of the eel intestinal mucosa and responses of intestinal osmoregulatory functions to cortisol in vivo. DiBattista et al. (9, 10) demonstrated high-affinity and low-capacity cortisol binding sites in eel (Anguilla rostrata) intestinal mucosa, with relative binding affinities for various corticoids that matched our functional responses in culture. In addition, the in vitro potency of these corticosteroids in salmon intestine is corroborated by the relative ability of these steroids to stimulate in vivo intestinal sodium and water absorption in eel and goldfish (20, 36).
More recently, two glucocorticoid-like receptors have been cloned from rainbow trout (designated rtGR1 and rtGR2); these are highly expressed (mRNA) in intestine and share cortisol as the major ligand (2, 11). Interestingly, high concentrations of 11-deoxycortisol and corticosterone induce transcriptional activity in COS-7 cells transfected with rtGR2 (but not rtGR1), similar to the response of Na+-K+-ATPase activity in cultured salmon intestine. In a cichlid fish, four subtypes of corticoid receptors are present that all show high ligand selectivity for cortisol, although the one mineralocorticoid-like receptor has a higher affinity for cortisol than the glucocorticoid-like receptors (18). In contrast, only a single class of receptors is suggested by radioreceptor assay for gills of coho and Atlantic salmon (41, 43). Finally, a rainbow trout mineralocorticoid-like receptor exhibits higher affinity for cortisol than for dexamethasone, based on characterization of the steroid-binding domain or transactivation properties (7, 47), whereas, in the present study, the potency of dexamethasone was equal to or greater than the potency of cortisol. From this evidence and those from Colombe et al. (7) and Sturm et al. (47), changes in Na+-K+-ATPase activity in the cultured salmon intestine in response to cortisol are mediated by what would be operationally defined as a glucocorticoid-like receptor(s). At this time, specific functions are not known for each of the corticoid receptors identified in teleost fish.
The concentration of cortisol necessary to induce a response by the intestine in vitro compares favorably with those determined for other osmoregulatory epithelia in culture. Cortisol either maintains or stimulates Na+-K+-ATPase activity in the cultured gill of coho salmon at 10 μg/ml and opercular membrane of tilapia at 0.1 μg/ml (27, 29). In addition, the intestine and coho salmon gill share a similar specificity for cortisol and, along with the tilapia opercular membrane, a similar time-frame of action on Na+-K+-ATPase activity. For example, the response to cortisol is seen within 2 days in both gill and intestine. These data indicate that the intestine shares a similar sensitivity to cortisol with other osmoregulatory tissues.
Cortisol treatment enabled explants to retain Na+-K+-ATPase activity but did not increase activity over levels found in tissue before the start of culture. This is often the case for salmon gill in culture (30). In the present study, it is not clear whether the action of cortisol was due to direct effects on Na+-K+-ATPase or due to the potential for cortisol to retain explant integrity. Both are likely however. With regard to architecture of explants, glucocorticoids have direct actions on the morphology of mammalian intestine in organ culture, including the preservation of mucosal structure in rabbit ileum and increased mucosal maturation in fetal mouse (17, 38).
A possible explanation for the in vitro decrease in Na+-K+-ATPase activity is the partial degradation of basal components of the intestinal epithelium. Because Na+-K+-ATPase is localized in the basolateral membrane of the epithelium (39), any degradation in this region would presumably result in decreased activity. The absence of factors that are important for basement-membrane attachment, or the possibility of dedifferentiation occurring, might also account for some of the decreased enzyme activity in culture. Decreasing Na+-K+-ATPase activity is common during culture of the salmon gill (26, 30), although in the case of the gill Na+-K+-ATPase is concentrated in chloride cells.
The possibility that factors absent in culture might potentiate a cortisol response or interact positively with cortisol warrants further attention. Shrimpton and McCormick (42) have shown that growth hormone and triiodothyronine increase the abundance of corticosteroid receptors in Atlantic salmon gill and therefore may increase cortisol responsiveness. Also, IGF-I protects against a decline in Na+-K+-ATPase activity during culture of gill from coho salmon (26).
Protein content of the cultured posterior intestine declined by ∼20% over 6 days of cortisol exposure. An increase in catabolism may explain the decrease of protein concentrations, although it is unclear why no change occurred in ceca.
In pyloric ceca, Na+-K+-ATPase activity was higher in the presence of insulin, whereas insulin slightly decreased Na+-K+-ATPase activity in the posterior intestine. Although the reason for this is not clear, it may be that sensitivity of ceca to insulin is associated with a high capacity for nutrient uptake in this region of the intestine (3). Alternatively, the response to insulin in ceca is perhaps due to occupancy of IGF receptors. We used a dose in culture that is supraphysiological (23), and, at high concentrations, insulin and IGFs can exhibit cross talk between their respective receptors (35). Compared with ceca, insulin has a less pronounced effect on Na+-K+-ATPase activity of the salmon gill in culture (29).
Exogenous cortisol stimulated intestinal Na+-K+-ATPase activity in sockeye salmon, consistent with reports for other salmonids (25, 51). Comparing this result with our findings from culture, cortisol induced a response in vivo at plasma cortisol levels of 200 ng/ml, a value falling within the range of exposure concentrations (0.1–10.0 μg/ml) that induced a response in vitro. Resting levels of plasma cortisol for sockeye salmon in the smolt stage are ∼50 ng/ml (15), although cortisol concentrations increase in the short term after transfer of smolts to seawater (55). Although our implants exceeded this value, cortisol concentrations necessary for a half-maximal response in culture (13–30 ng/ml) were well within the physiological range. The magnitude of the response of Na+-K+-ATPase activity (to cortisol) in vitro is similar to changes that our group (51, 53) documented during seawater adaptation and parr-smolt transformation of chinook salmon: Na+-K+-ATPase activity in ceca and posterior intestine increase in the range of 50–100%.
In summary, we have shown that the salmon pyloric ceca and posterior segment of intestine are direct targets of the major corticosteroid product of the interrenal tissue, cortisol. The responsiveness of explants, as measured by Na+-K+-ATPase activity, was indicative of receptor-mediated mechanisms. With cortisol as a supplement, the present culture system may be useful for examining effects of a variety of neuroendocrine factors on the intestine. The retention of cortisol responsiveness in culture suggests that this system can be used to assess whether there are differences in sensitivity to this corticosteroid during parr-smolt transformation that are perhaps part of the developmental changes leading to improved hypoosmoregulatory ability in salmon (30, 53).
The expert histological teachings of, and assistance from, the late Gerald Stokes are gratefully acknowledged. We thank the staff at Chemsearch Laboratories (Otago) for ion analyses and Matthew Downes for invaluable contributions in preparing tissues for electron microscopy. Discussions with Dr. Jennifer Specker (University of Rhode Island) were particularly helpful while preparing the manuscript.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2005 the American Physiological Society