Search

Gastrointestinal function in vertebrates is influenced by stressors, such as fasting and refeeding, different types of diet and hormonal factors. The aim of this paper was to analyze the effect of a Spirulina (Arthrospira platensis) diet, a microalga known for its nutraceutical properties, on the gastrointestinal tract of zebrafish (Danio rerio) regarding expression of oligopeptide transporter 1 (PepT1) and ghrelin (GHR). Food deprivation and refeeding was investigated to elucidate expression of PepT1 and GHR at a gastrointestinal level and the zebrafish compensatory mechanism. PepT1 is responsible for absorbing di- and tripeptides through a brush border membrane of intestinal mucosa. GHR is a brain-gut peptide in fish and mammals, stimulating growth hormone secretion and regulating appetite. Samples were taken after 2 and 5 days of specimen fasting, and 2 and 5 days of refeeding with Sera Spirulina tabs, in which the major constituent is Spirulina sp. (50.2% protein). Morphological and immunohistochemical analysis of PepT1 and GHR were carried out. Control specimen intestinal tract showed normal morphology of the digestive tract. Fasting caused fold structural changes and intestinal lumen constriction. Immunohistochemical analysis showed a PepT1 level reduction after fasting and an increase after refeeding, reaching very high levels after 5 days, compared to controls. GHR levels increased after food deprivation and gradually decreased after refeeding. Increased expression of PepT1 in refeeding fish suggests a compensatory physiological mechanism, as does the increase in GHR levels in fasting fish followed by a reduction after refeeding. A compensatory mechanism may be induced by fasting and refeeding and by a higher protein Spirulina diet. The microalga, for its nutraceutical properties, is an excellent candidate for animal breeding and human diet.

In the present work, morphological and molecular effects of short-term feed deprivation and refeeding with Spirulina (Arthrospira platensis) on zebrafish digestive tract were determined. Once elucidated the proximate composition of Spirulina feed, immunohistochemical and western blot analyses of peptide transporter (PepT1) and cholecystokinin (CCK8) were carried out in the gastrointestinal tract of zebrafish, previously morphologically investigated. Two and five fasting days caused not only morphostructural alterations, but also the downregulation of PepT1 and CCK8 proteins. Conversely, the recovery of normal morphological conditions, along with an increased PepT1 and CCK8 expression, were observed after refeeding with Spirulina. The increase of PepT1 expression in zebrafish may be responsible for the enhanced CCK8 secretion, so that both proteins may contribute to an improved digestion process during refeeding. These observations could be supported not only by compensatory mechanisms induced by fasting and refeeding but also by an higher protein quality of Spirulina-based diet.

A short fasting-refeeding experience was applied to specimens of red porgy, Pagrus pagrus (Teleostei, Sparidae) to assess its effects on some physiological parameters. Haematological (haematocrit), biochemical (serum cortisol and glucose) and immunological (lysozyme, haemolytic and haemagglutinating activities) parameters were measured. For this study, two fish groups were considered: one was fasted for 14 days and then refed to satiation during further 7 and 15 days (indicated as fasted/refed group), the other was fed throughout the study and was taken as a control group. Significantly lower values were recorded for the condition index, the hepato-somatic index and viscero-somatic index in the fasted/refed group compared to the fed one. Fasting did not affect significantly the examined parameters, except for cortisol; refeeding for 7 days induced a significant increase in the haemoagglutinating titre and the spontaneous haemolytic activity, but when refeeding was extended to 14 days haemagglutinating and haemolytic values remained lower than those measured in fed fish.

Growth, haematological (haematocrit), biochemical (serum cortisol and glucose), and non-specific immune (lysozyme, serum haemolytic and haemagglutinating activities, extracellular respiratory burst activity) parameters, were monitored in European sea bass Dicentrarchus labrax and blackspot sea bream Pagellus bogaraveo subjected to a 31 days starvation compared to fed fish, to assess the responses to feed deprivation of these health status indicators. While haematocrit, serum cortisol, glucose and haemolytic activity of both species did not undergo significant variation following starvation, probably due to the short period applied, some non-specific immune parameters were affected significantly. In the starved sea bass, mucus lysozyme content doubled (1.8 U/mL) compared to the initial value. Haemagglutinating activity was significantly lower in starved sea bass than in fed fish after 31 days. In blackspot sea bream, a slight, not significant, reduction in haemagglutinating activity occurred 11 days after starvation. Respiratory burst activity decreased significantly in the starved fish. In spite of the limited number of examined parameters, the opportunity to use a panel of several indicators to obtain a more complete picture of health status in fish was underlined.

The physiological effects of short-term starvation on some haematological, biochemical and non-specific immune response parameters together with the histological structure of the skin, were investigated in the European eel, Anguilla anguilla. Blood haemoglobin and haematocrit, serum glucose and cortisol, hemolysins, haemagglutinins, and lysozyme in the plasma, kidney and epidermal extract, were measured in fish after 31, 42 and 58 days of starvation, and compared to those of fed fish. Starvation did not affect haemoglobin and haematocrit values, while an increase in glucose and cortisol levels was found in starved eels by day 42. Haemolytic and haemagglutinating activities decreased in starved eels. On the other hand, starvation caused an increase in the lysozyme content in the epidermal extracts, while no significant variations were observed in kidney and plasma. On the whole, no major changes in metabolic, haematological and non-specific immune parameters were observed when short-term (less than 2 months) starvation was applied to the European eel, suggesting an adaptive response to starvation, rather than a typical alarm-stress response, allowing this species to withstand food deprivation.

Ion transport in the intestine of Gobius niger, a euryhaline teleost, was studied in both isotonic and hypotonic conditions. Isolated tissues, mounted in Ussing chambers and bilaterally perfused with isotonic Ringer solution, developed a serosa negative transepithelial voltage and a short circuit current indicating a net negative current in absorptive direction. Bilateral removal of Cl- and Na+ from the bathing solutions as well as the luminal removal of K+in the presence of Ba2+(10(-3) M) almost abolished both Vt and Isc. Similar results were obtained by adding bumetanide (10(-5)M) to the luminal bath while other inhibitors of Cl- transport mechanisms were ineffective. These observations suggest that salt absorption begins with a coupled entry of Na+, Cl-, and K+ across the apical membrane; a Ba2+inhibitable K+ conductance, demonstrated also by micropuncture experiments, recycles the ion into the lumen. Salt entry into the cell is driven by the operation of the basolateral Na+/K(+)-ATPase since serosal ouabain (10(-4)M) completely abolished both Vt and Isc; this pump also completes the Na(+) absorption. The inhibitory effect of both serosal bumetanide (10(-4)M) and SITS (5 x 10(-4)M) suggests that Cl- would leave the cell via the KCl cotransport, the Cl/HCO3- antiport and/or conductive pathways. Bilateral exposure of tissues to hypotonic media produced a reduction of both the transepithelial voltage and the short circuit current probably due to the activation of homeostatic ionic fluxes involved in cell volume regulation. The results of experiments with both isolated enterocytes and intestine exposed to hypotonic solution suggested that the recovery of cell volume, after the initial cell swelling, involves a parallel opening of K+ and Cl- channels to facilitate net solute and water effluxes from the cell. J. Exp. Zool. 301A:49-62, 2004.

Sea bass (Dicentrarchus labrax, Moronidae), European eel (Anguilla anguilla, Anguillidae) and blackspot seabream (Pagellus bogaraveo, Sparidae) were studied to check their mucus, blood sera and tissue samples for antibacterial and agglutinating activity against a variety of Gram negative and positive bacteria. Samples were also examined for their haemolytic properties against sheep red blood cells. The highest antibacterial activity was detected in the blood sera of blackspot seabream and European eel (against Vibrio alginolyticus) and in the kidneys of sea bass (against Photobacterium damselae subsp. piscicida). Haemolytic properties against sheep red blood cells were observed in the mucus of sea bass and blackspot seabream, as well as in the sera of eel and sea bass. The sera of sea bass and eel showed also agglutinating activity against Pseudomonas aeruginosa and Staphylococcus aureus; the mucus of sea bass was able to agglutinate isolates of Vibrio fluvialis, V. alginolyticus and A. hydrophila. The study suggested that the mucus secretions, biological fluids and organs of the examined fish species can be regarded as an interesting source of bioactive compounds with antibacterial and haemolytic properties.

An experiment was carried out in red porgy, Pagrus pagrus (Teleostei, Sparidae), to assess the effects of a 14-day fasting period, followed by refeeding to apparent satiation, on the contents of digestive enzymes (total proteases, and particularly pepsin, trypsin, chymotrypsin, carboxypeptidases A and B; amylase and lipase). Two fish groups were considered: one (indicated as fasted/refed group) was fasted for 14 days and then refed during further 7 and 15 days, and the other was fed throughout the study and was taken as a control group. The measured enzymatic values showed that fasting resulted in a generalized, not significant decrease, of the activity of digestive enzymes. Refeeding caused a significant increase for most of the assayed enzymes: total proteases both in the middle and distal intestine, pepsin in the stomach, trypsin in the middle intestine, and amylase and lipase in the proximal intestine. Nevertheless, the detection in the fasted/refed fish of enzymatic values still lower than those measured in the control fish suggested that fish experiencing short-term fasting were partially impaired in their digestive capacity.

The effects of acetylcholine analogues, serotonin and catecholamines on ion transport were studied in both the middle and the posterior intestine of Anguilla anguilla, mounted in an Ussing chamber, with the aim of understanding whether these regulators affect different mechanisms in the different tracts. In the middle intestine, acetylcholine analogues and serotonin decreased the serosa negative transepithelial potential and short-circuit current without altering the transepithelial resistance; catecholamines reversed the inhibitory effects of both regulators. Similar opposite effects were produced by both the acetylcholine analogues and noradrenalin in the posterior intestine. However, the lowering of the short-circuit current elicited by serotonin was paralleled by the decrease of the transepithelial resistance, whilst noradrenalin had the opposite effects on both parameters. These observations, together with the results of experiments performed by measuring the dilution potential in the control condition and in the presence of either serotonin or serotonin plus noradrenalin, led us to hypothesize that serotonin increases the anion conductance of the paracellular pathway while noradrenalin decreases it. In both the middle and posterior intestine, these regulators probably affect transcellular transport mechanisms by acting on the Na-K-Cl transporter; both acetycholine and serotonin decrease its activity while noradrenalin increases it.

The fundus of an eel stomach was mounted in an Ussing chamber and bathed with control Ringer on the serosal side and with unbuffered solution on the mucosal side. The gastric mucosa exhibited a mucosa negative transepithelial voltage (V t), a “short circuit” current (I SC) and a small spontaneous acid secretion rate (J H). All these parameters were abolished by cimetidine treatment. Bilateral ion substitution experiments in tissues lacking spontaneous acid secretion suggested that a net Cl− transport from serosa to mucosa was responsible for the genesis of the I SC in the absence of H+ secretion. Serosal application of histamine (10−4 mol · l−1) or carbachol (10−4 mol · l−1) stimulated both I SC and J H. The action of carbachol was independent of histamine. The control as well as the histamine-stimulated I SC was sensitive to both serosal bumetanide (10−5 mol · l−1), inhibitor of the Na+-K+-2Cl− cotransport, and 4,4-diisothiocyano-stilbene-2,2-disulphonic acid (DIDS, 5 · 10−4 mol · l−1), inhibitor of the Cl−-HCO− 3 exchange, while the I SC stimulated by carbachol was nullified by serosal DIDS. These data suggested that the non-acidic Cl− uptake across the serosal membrane was linked to the activity of both Na+-K+-2Cl− cotransport and Cl−-HCO− 3 antiporter; histamine stimulated both transporters while carbachol was limited to the anion exchanger. The finding that the acid secretion was strictly dependent on serosal Cl− and was completely blocked by serosal DIDS suggested that the Cl− accompanying H+ secretion entered the cell through the serosal membrane by the Cl−-HCO− 3 exchange. In addition, the acid secretion stimulated by carbachol was also dependent on serosal Na+ and sensitive to the application of 5-N-N-dimethyl-amiloride in the serosal bath, suggesting that the increased activity of the Cl−-HCO− 3 during carbachol treatment was linked to the activation of serosal Na+-H+ exchange. The inhibitory effect of luminal omeprazole (10−4 mol · l−1) on acid secretion suggested the presence of the H+-K+ pump on the luminal membrane.

In intensive aquaculture, basic studies on fish digestive capacity and metabolism play a key role for diet formu- lation. In particular, intensive rearing of a new species is strictly related to the development of appropriate nutritional pro- tocols and therefore knowledge of specific features concerning physiology and nutrition is needed. Nutrient utilization by fish is a direct function of the availability of suitable digestive enzymes along the gastro-intestinal tract. Therefore the de- termination of specific enzymatic activities (proteases, carbohydrases and lipases) may provide qualitative information about the digestive capacity and the efficiency of each reared species to use feeding components. Recent observations on digestive enzymatic profiles and on digestive ability of some Teleost species with large aquaculture potential and recog- nized as new candidates for product diversification have been reviewed in this paper. Attention has been given to the dif- ferent features (dietary composition, digestive phases, life stages, contribution of bacterial .enzymes) that may affect en- zymatic patterns.

The involvement of Ca2+ in the regulatory volume decrease (RVD) mechanism was studied in both isolated enterocytes and intestine of the eel, Anguilla anguilla. Videometric methods and electrophysiological techniques were respectively employed. The isolated enterocytes rapidly swelled following a change from isotonic (315 mOsm/kg) to hypotonic (180 mOsm/kg) saline solutions. Afterwards, they tended to recover their original size. This homeostatic response was inhibited both in the absence of extracellular Ca2+ and in the presence of TMB8, an inhibitor of Ca2+ release from intracellular stores. It is likely that Ca2+ entry through verapamil-sensitive Ca2+ channels is responsible for RVD since the blocker impaired the ability of the cell to recover its volume after the hypotonic shock. The observation that a 10-fold increase of K+ concentration as well as the presence of quinine in the hypotonic solution completely abolished RVD indicated the involvement of K+ in this response. Experiments performed with the isolated intestine suggested that the opening of basolateral K+ channels facilitates K+ loss (and hence water efflux) from the cell during RVD and that this opening is probably due to Ca2+ entry into the cell through both the mucosal and the serosal membranes.

The role of Ca++ on the regulation of the paracellular pathway permeability of the middle intestine of Anguilla anguilla was studied by measuring the transepithelial resistance and the dilution potential, generated when one half of NaCl in the mucosal solution was substituted iso-osmotically with mannitol, in various experimental conditions altering extracellular and/or intracellular calcium levels. We found that removal of Ca++ in the presence of ethylene glycol-bis(beta-aminoethyl ether) (EGTA) from both the mucosal and the serosal side, but not from one side only, reduced both the transepithelial resistance and the magnitude of the dilution potential. The irreversibility of this effect suggests a destruction of the organization of the junction in the nominal absence of Ca++. However a modulatory role of extracellular Ca++ cannot be excluded. The decrease of the intracellular Ca++ activity, produced by using verapamil to block the Ca++ entry into the cell, or by adding 3,4,5-trimethoxybenzoic acid 8-(diethylamino) octyl ester (hydrochloride) (TMB-8), an inhibitor of Ca++ release from the intracellular stores, reduced both the transepithelial resistance and the magnitude of the dilution potential, indicating a role of cytosolic Ca++ in the modulation of the paracellular permeability. However the rise of calcium activity produced by the Ca++ ionophore calcimycin (A23187) evoked an identical effect, suggesting that any change in physiological intracellular Ca++ activity alters the paracellular permeability.

The sensitivity to external pH of Cl- absorption was studied in isolated stripped intestinal mucosa of the eel, Anguilla anguilla, mounted in Ussing chambers. Short-circuit current, transepithelial potential difference and conductance were measured in bathing solutions containing various combinations of HCO3--concentration (0–25 mmol·l-1), partial pressure of CO2 (0–76 mm Hg) and pH (6.9–7.9). A linear relationship was found between pH and short-circuit current in the range of pH studied both in HCO3-/CO2 Ringer and in Hepes Ringer. The pH effect was almost completely reversible. It was not affected by the presence of mucosal Ba2+ (10-3 mol·l-1) but it was inhibited by the presence of luminal (10-5 mol·l-1) or serosal (10-4 mol·l-1) bumetanide. The results obtained suggest that the Cl- absorption in the European eel intestine is pH sensitive. The data do not indicate whether the pH affects directly the Na+−K+−Cl- cotransport and/or the basolateral Cl- conductance or other mechanisms indirectly linked to Cl- absorption.

The role of HCO3−/CO2 buffer in Cl− absorption was examined in the in vitro perfused eel intestine adapted to seawater. Cl− absorption, expressed as short/circuit current (Isc), was measured in either 20 mM HCO3−/1% CO2 Ringer or HEPES Ringer, pH 8.0. Unilateral (mucosal or serosal) substitution of HCO3−/CO2 with HEPES/O2 was without effect on Isc and transepithelial voltage (Vt), whereas bilateral removal of HCO3−/CO2 reduced Isc and Vt by 50%, indicating that the presence of HCO3−/CO2 buffer at one side of the epithelium is sufficient to keep Cl− absorption at the maximum rate. We examined in further detail the individual components of the HCO3−/CO2 system that stimulates Cl− absorption. We found that, in tissues bathed with HEPES Ringer, addition of 1% CO2 to the luminal or serosal solution (final pH=7.6 in the chamber) had no effect on Isc and Vt, while both electrical parameters could be restored to control values by unilateral (luminal or serosal) substitution of HEPES Ringer with 20 mM HCO3−/1% CO2 Ringer or 20 mM HCO3−alone. Stimulation of Isc induced by unilateral (luminal or serosal) HCO3−/CO2 was inhibited by luminal or serosal 4-acetamido-4′-isothiocyanostilbene-2,2′-disulphonic acid (SITS) (0,25 mM) or by serosal Na+ removal, whereas amiloride (1 mM), luminal or serosal, had no effect. Acetazolamide (0.1 mM, both sides) inhibited stimulation of Isc induced by luminal addition of HCO3−/CO2, whereas it was without effect when HCO3−/CO2 was added serosally or bilaterally. We reached the following conclusions, (a) Cl− absorption is stimulated by HCO3−/CO2 buffer via an increase in intracellular HCO3−concentration and/or pHi changes consequent to the HCO3−uptake mediated by HCO3−transport systems operating on both cell membranes, (b) A Na+-dependent SITS-inhibitable HCO3−transport mechanism operates at the basolateral membrane, (c) The transfer of HCO3−through the luminal membrane is mediated by the carbonic anhydrase enzyme located on the brush-border membranes of the enterocyte: the movement of HCO3−, via a SITS-sensitive transport system, occurs most likely in form of OH−, which originates from the dehydration reaction of HCO3−catalysed by the carbonic anhydrase. (d) There is no apparent amiloride-sensitive Na+/H+ antiporter on either cell membrane.

The possible mechanisms of Cl−-exit across the basolateral membrane of the enterocyte in sea water (SW) and freshwater (FW) eel intestine were investigated by observing the response of the short-circuit current, Isc, equivalent to net Cl−-absorption, and transepithelial voltage, Vt, to specific or relatively specific inhibitors of different Cl−-transport pathways. Serosal NPPB [5-nitro-2(3 phenylpropylamino)-benzoate] (10−4 M), the potent blocker of Cl−-channels in different epithelia, produced a negligible effect on Isc and Vt either in presence or in absence of HCO3− in the bathing solutions. In the presence of HCO3− serosal bumetanide, better than furosemide and torasemide, at a concentration of 10−4 M inhibited Isc and Vt by ∼90% within 15–30 min without changing the transepithelial resistance, Rt. In HCO3−-free solutions the drug completely collapsed Isc and Vt. The results were similar in both SW and FW eels. In SW fishes serosal SITS (2.5 × 10−4 M) reduced Isc and Vt by ∼50% when HCO3− was present in the solutions. In absence of HCO3− SITS inhibited Isc and Vt by ∼35%, a finding consistent with the presence of a Cl−-conductance in the basolateral membrane. Under these conditions SITS inhibition disappeared after bumetanide pretreatment. There is little evidence for the presence of a basolateral K+-conductance since serosal Ba2+ (1 mM) elicited variable effects on Isc and Vt. The results are interpreted as evidence for the presence of bumetanide- and SITS-sensitive Cl−- conductance channels in the basolateral membrane of SW and FW eel intestine which would contribute at least 35% of the net Cl−-absorption. An HCO3−-dependent Cl−-transport was also identified since Isc and Vt were reduced by ∼50% when tissues were perfused with HCO3−-free salines. © 1992 Wiley-Liss, Inc.

The mechanism(s) of absorption of chloride ions (Cl-absorption), as measured by the short-circuit current (Isc, μA·cm−2), was investigated in the intestinal middle part of the seawater-(SW) and the freshwater-(FW) adapted eel. Tissues, mounted in an Ussing chamber and continuously perfused by gravity flow on both sides with 20 mM HCO3− - 1% CO2 Ringer's solution at pH 8.0, developed a serosa-negative transepithelial voltage (Vt, mV) and an Isc that were both lower in FW-adapted than in SW-adapted eels. Bilateral removal of Na+ or Cl− from the bathing solutions reduced both electrical variables to zero. Removal of K+ from or addition of Ba2+ (1 mM) to the luminal bath markedly decreased Isc and Vt. When a K+-free plus Ba2+-containing solution perfused the lumen, Isc and Vt were almost completely abolished. Luminal bumetanide (10−5 M), an inhibitor of Na+-K+-2Cl− cotransport, reduced Isc and Vt by about 90% in a few minutes. Amiloride (10−3 M) and DIDS (4,4′-diisothiocyanostilbene-2,2-disulphonic acid; 10−4 M), inhibitors of Na+/H+ and Cl−/HCO3− exchangers, and hydrochlorothiazide (10−4 M), an inhibitor of the Na-Cl symport, had no effect on Isc and Vt when added luminally. Serosal ouabain (10−3 M) abolished Isc and Vt in about 30 min. All the above experimental manoeuvres led to similar results in both SW-adapted and FW-adapted eels. The data suggest that in the intestine of both SW and FW eels, Cl−-absorption is mediated by a luminally located neutral Na+-K+-2Cl− cotransport system. The basolaterally located Na+, K+-ATPase, by generating an inwardly directed electrochemical gradient for Na+, provides the driving force for the active accumulation of Cl− by the electroneutral cotransporter. The K+ accumulated is refluxed by a Ba2+-inhibitable luminal K+ conductance. © 1992 Wiley-Liss, Inc.

The ultrastructural changes of the plantaris muscle of the rat have been investigated during long-term compensatory hypertrophy. The latter was induced by the removal of the synergistic gastrocnemius muscle. Transmission electron microscope (TEM) observations were performed on the 20th, 40th and 60th days after surgery on overloaded and control muscles. The plantaris muscle had hypertrophied 54.8% after 20 days, 74.4% after 40 days and 79.8% after 60 days. It was observed that activation of the satellite cell developed as shown by an increase in cytoplasmic volume, together with diffuse presence of polyribosomes and a rough endoplasmic reticulum. On the 40th day masses of swollen mitochondria grouped at the edge of muscle fibers were observed. On the 60th day structures completely covered with basal lamina, whose cytoplasm was filled with mitochondria and polyribosomes, were partially detached from the fibers. The above ultrastructural changes were absent in control contralateral muscles. These findings suggest that surgically induced overload produces a relevant ultrastructural re-arrangement in muscle tissue.

The effects of Hydroxyurea treatment on food intake and on the self-selection between a protein free diet and a complete one are different from those observed during Cyclophosphamide and Mustine Hydrochloride treatment and appear to be similar to those observed during Methotrexate treatment; these may be referred exclusively to a toxic action of the drug on the liver and on the mucous membranes of digestive apparatus. The urea, a product of Hydroxyurea metabolism, does not seem to induce modifications on food intake both as to the quantitative and qualitative aspect.

The protective effect of endogenous prostaglandins on the fish gastric mucosa was evaluated by studying the effect of indomethacin and aspirin, known cyclooxigenase inhibitors, on the mucosal ulceration in the isolated gastric sacs of Anguilla anguilla. Gastric sacs devoid of muscle layers were incubated in the presence of indomethacin (10(-4) mol x l(-1)) or aspirin (10(-4) mol x l(-1)) in different experimental conditions. Both the antiinflammatory drugs produced ulcers, but the effects were more severe in the presence of histamine and in the absence of HCO3- in the incubation bath. The effects of prostaglandin E2 (PGE2) on acid secretion rate (J(H)) and on alkaline secretion rate (J(OH)) were evaluated (with the aid of the pH stat method) in isolated gastric mucosa mounted in Ussing chambers. We found that PGE2 (10(-8)-10(-5) mol x l(-1)) increased JH in a dose-dependent manner. In tissues pretreated with luminal omeprazole (10(-4) mol x l(-1)), PGE2 stimulated gastric alkaline secretion. It was nullified by serosal removal of HCO3- or Na+ and by serosal ouabain (10(-4) mol x l(-1)). These results suggested that prostaglandins also exert their protective effects in fish gastric mucosa. This protection seems partially due to a stimulation of exogenous HCO3- transport from the serosal to the mucosal side. It is likely that this transport is an active transcellular mechanism coupled to Na+ transport.