Regulatory, Integrative and Comparative Physiology

Estradiol impairs hyposmoregulatory capacity in the euryhaline tilapia, Oreochromis mossambicus

Mathilakath M. Vijayan, Akihiro Takemura, Thomas P. Mommsen


Freshwater (FW)-adapted tilapia (Oreochromis mossambicus) were treated with estradiol (E2) for 4 days to stimulate protein synthesis and sampled at 0, 4, and 24 h after exposure to 50% seawater (SW). E2 increased circulating vitellogenin (VTG) levels in large amounts, indicative of unusually high rates of hepatic protein synthesis. E2 treatment prevented the recovery of plasma osmolality in 50% SW that was evident in the sham group. Plasma sodium concentration was significantly elevated with E2 in FW, but the levels did not change in 50% SW. Gill Na+-K+-ATPase activity was significantly lower in the E2 group compared with sham-injected tilapia in 50% SW. No significant differences were noted in plasma cortisol, thyroxine, triiodothyronine, or glucose concentration with E2 in 50% SW. E2 significantly lowered several key liver enzyme activities and also decreased gill lactate dehydrogenase and malate dehydrogenase activities over a 24-h period. Together, our results suggest that E2 impairs ion regulation in tilapia, partially mediated by a decreased metabolic capacity in liver and gill. The decreased tissue metabolic capacity is likely due to E2-induced energy repartitioning processes that are geared toward VTG synthesis at the expense of other energy-demanding pathways.

  • stress
  • intermediary metabolism
  • vitellogenin
  • ion regulation
  • gill sodium-potassium-adenosinetriphosphatase

the mozambique tilapia (Oreochromis mossambicus) tolerates wide ranges in salinity attributable to efficient ion-regulating mechanisms (6, 7). Transient changes in gill chloride cell dynamics including Na+-K+-ATPase activity play a key role in the ion regulation process after seawater (SW) exposure in this species (5, 9, 11). Changes seen in SW-exposed gills argue for a temporary increase in the epithelial cell energetics to maintain ion balance. Indeed, the short-term energetic cost associated with SW acclimation is as high as 20% of the total body metabolism in O. mossambicus (27). As protein synthesis accounts for the majority of the energetic costs (29), the higher oxygen consumption rate upon SW transfer may be due to the transient increases in chloride cell biogenesis and/or including the synthesis of ion pumps and ion channels (13).

The increased energy demand at the gills is predominantly supported by oxidation of glucose obtained from the circulation (23, 24,30). As liver is the main site for glucose production, the metabolic demand imposed on this tissue in SW would be greater, especially since gluconeogenesis is energetically expensive. Therefore, any alteration in the availability, mobilization, or oxidation of energy substrates will likely affect the SW acclimation process. To this end, studies with food-deprived O. mossambicusconfirmed that a decreased energy status did indeed compromise the ability of this euryhaline species to cope with SW transfer (34). Food-deprived fish were able to mount metabolic and hormonal responses crucial for SW acclimation, while being unable to regulate plasma ion levels in SW compared with fed fish (34). Together, the results suggest that ion transport mechanism(s) may be tightly linked to the prior energy status and/or the energetic demand imposed upon the animal.

In this study, we specifically tested the hypothesis that high rates of protein synthesis induced by estradiol (E2) will compromise the ability to cope with exposure to 50% SW in a euryhaline teleost. A similar approach had been taken earlier by Madsen and coworkers (20) who showed that extended exposure to E2or nonylphenol, a xenoestrogen, negatively affected smoltification in Atlantic salmon. The rationale for formulating this hypothesis stems from the fact that SW transfer is energetically expensive, and therefore increased energy demand (protein synthesis) as part of SW transfer will impart considerable strain on energy partitioning. We used exogenous E2 treatment to increase the protein synthetic requirements prior to 50% SW exposure. Previous studies had shown that this steroid induces a number of estrogen-responsive genes in the liver (26), elevates protein synthetic rates (14, 15) resulting in extraordinarily high circulating levels of vitellogenin (VTG), and alters metabolic performance of the liver (16, 37). In addition, E2 treatment also compromises the ability of male and/or juvenile salmon to acclimate to SW (18-20). Therefore, our objective was to address the impact of E2 treatment on short-term changes (4 and 24 h) in hyposmoregulatory abilities after exposure to 50% SW inO. mossambicus. Specifically, we examined changes in1) the concentration of plasma hormones (cortisol and thyroid hormone; 22) and metabolites (glucose; Refs. 28and 34) that are important in the SW acclimation process, 2) plasma osmolality and sodium concentration and gill Na+-K+-ATPase activity as indicators of osmoregulatory capacity, and 3) the activities of liver and gill enzymes involved in the intermediary metabolism to determine the metabolic capacity of these tissues in 50% SW. Plasma VTG concentration was assessed to confirm E2 stimulation of protein synthesis in tilapia.



O. mossambicus (∼70 g body mass) were collected from rivers in Okinawa in July 1999 and maintained in running freshwater (FW) at the Sesoko Station, Tropical Biosphere Research Center, University of the Ryukyus, Okinawa, Japan. The fish were maintained at ambient temperature (24°C) under a natural photoperiod (12:12-h light-dark) and fed once daily to satiety with a commercial tilapia feed (41M; Zen-no, Tokyo, Japan). The fish were acclimated at least for 3 wk.

Experimental protocol.

Groups of 13 male fish each were distributed into four tanks and maintained in static FW; the FW was replenished every second day. Fish were acclimated in these tanks and maintained as above for at least a week prior to the experiment. Fish were lightly anesthetized and injected intraperitoneally with either peanut oil alone (sham group) or peanut oil containing E2 at 5 mg/kg fish. Fish were not fed during the experimental period. Four days after injection, half of the water in all four tanks was drained, and two tanks each were refilled with either FW (control) or 100% SW (34‰). Salinities in the SW tanks were 50% SW (17‰) within 20 min. Fish were sampled either 4 h or 24 h after SW transfer. Sampling consisted of quickly netting six fish from each tank into a small bucket with tank water containing 2-phenoxyethanol (1:10,000; Kanto Pure Chemicals, Tokyo, Japan). Fish were bled from the caudal vessels using heparinized syringes, and blood was centrifuged at 13,000 g for 5 min. The resulting plasma was stored frozen at −70°C for later hormone and metabolite analyses. Fish were killed and weighed, and the liver wet weight was obtained to calculate the hepatosomatic index (liver mass/fish mass × 100). Pieces of liver and gill were quickly frozen in liquid nitrogen for protein and enzyme measurements (25, 34). Gill Na+-K+-ATPase activity was measured according to McCormick (21). Plasma cortisol (Medicorp, Montreal, Quebec, Canada) and triiodothyronine (T3) and thyroxine (T4) concentrations (Amerlex, Amersham, Montreal, Quebec) were quantified with commercial radioimmunoassay kits. Plasma glucose (Trinder method; Sigma, St. Louis, MO) and protein concentrations (bicinchoninic acid method, bovine serum albumin as standard; Sigma) were determined using commercially available kits. Plasma VTG concentration was measured with a sandwich enzyme-linked immunosorbent assay using antibodies raised against O. mossambicus egg homogenate protein and purified tilapia VTG as standard (33). Plasma osmolality was assessed with a vapor-pressure osmometer (model 5500, Wescor). A polarized Zeeman atomic absorption spectrophotometer (model Z-6100, Hitachi) was used to quantify plasma sodium concentrations.

Statistical analyses.

Data were analyzed using two-way ANOVA with treatment (sham and E2) and time (0, 4, and 24 h in 50% SW) as independent variables. As the 4- and 24-h time points for sham and E2 in FW showed no significant differences (unpairedt-test), these two time points were pooled as FW group (0 h in 50% SW) for two-way ANOVA. Significant interactions between treatment and time were compared using the Student-Newman-Keuls test. Significance level was set at 95% (P < 0.05). Regression analysis (slopes of the regression lines) for gill lactate dehydrogenase (LDH) and malate dehydrogenase (MDH) activities over time in 50% SW with and without E2 were compared using ANOVA. Log-transformed data were used wherever necessary to satisfy homogeneity of variance; however, nontransformed data are shown in Tables 1-3 and Figs. 1-4.

View this table:
Table 1.

Effect of estradiol treatment on total body mass, HSI, liver protein and plasma VTG concentration in tilapia

View this table:
Table 2.

Effect of estradiol treatment on plasma cortisol, T4, T3, and glucose concentration following exposure to 50% SW in tilapia

View this table:
Table 3.

Effect of estradiol treatment on hepatic enzyme activities following exposure to 50% SW in tilapia

Fig. 1.

Effect of estradiol (E2) treatment (5 mg/kg) on plasma osmolality following exposure to 50% seawater (SW) in tilapia (Oreochromis mossambicus). Values are means ± SE (n = 5–12; 12 animals at time 0). The numerals (“1”, “2”, and “3”) denote significant time effect, and values with different numbers are statistically significant (P < 0.05, two-way ANOVA); the letters (“a” and “b”) denote significant time effects within each group, and bars with different letters are significantly different (P< 0.05, two-way ANOVA).

Fig. 2.

Effect of E2 treatment (5 mg/kg) on plasma sodium concentration following exposure to 50% SW in tilapia. Values are means ± SE (n = 5–12; 12 animals at time 0). *Significantly different from the corresponding sham group (P < 0.05, two-way ANOVA); letters denote significant time effect in the sham group, and bars with different letters are significantly different (P < 0.05, two-way ANOVA).

Fig. 3.

Effect of E2 treatment (5 mg/kg) on gill Na+-K+-ATPase activity following exposure to 50% SW in tilapia. Values are means ± SE (n = 5–12; 12 animals at time 0). The numerals denote significant time effect, and values with different numerals are significantly different (P < 0.05, two-way ANOVA); the letters denote significant time effects within each group, and bars with different letters are significantly different (P < 0.05, two-way ANOVA). *Significantly different from the corresponding sham group (P < 0.05, two-way ANOVA).

Fig. 4.

Effect of E2 treatment (5 mg/kg) on gill lactate dehydrogenase (LDH, top) and malate dehydrogenase activity (MDH, bottom) following exposure to 50% SW in tilapia. Values are means ± SE (n = 5–12; 12 animals at time 0). *Significantly different from the corresponding sham group (P < 0.05, two-way ANOVA).+Significantly different from time 0(P < 0.05, two-way ANOVA). Regression lines are shown on the graph, and the insets give the slope andr 2 values for each group. Regression lines for LDH and MDH were significant (P < 0.005) for the E2 group, but not the sham group.


Body mass, hepatosomatic index, hepatic protein, and plasma VTG.

No differences in body mass, hepatosomatic index or liver protein concentration were noted between the sham and E2-treated fish (Table 1). Concentrations of plasma VTG were below the detection limit in the sham group, whereas E2 treatment resulted in elevated plasma VTG (>2 mg/ml) (Table 1). Since exposure to 50% SW exerted no obvious effect on the E2-dependent induction of plasma VTG, data from the E2 groups were pooled for subsequent comparisons with the sham group (Table 1).

Plasma hormone and metabolite concentration.

Plasma cortisol, T3, T4, and glucose concentration showed no significant differences after exposure to 50% SW either with or without E2 (Table2). In contrast, E2 treatment significantly elevated plasma T3, but not cortisol, T4, or glucose concentration, regardless of the duration of exposure to half-strength SW (Table 2).

Plasma osmolality and sodium concentration and gill Na+-K+-ATPase.

Exposure to 50% SW significantly increased plasma osmolality compared with FW at 4 h, but not at 24 h in the sham fish. E2 treatment resulted in significantly higher plasma osmolality at both 4 and 24 h after exposure to 50% SW (Fig.1). E2 treatment in FW led to significantly higher plasma sodium concentration compared with the sham fish (Fig. 2). Exposure to 50% SW significantly increased plasma sodium concentration at 4 and 24 h in the sham fish, but not in the E2-treated fish (Fig. 2). Gill Na+-K+-ATPase activity was markedly altered by exposure to 50% SW, including significant treatment (E2 < sham) and temporal effects (0 and 4 < 24 h), as well as a significant interaction between time and treatment (Fig. 3). In the sham group, gill Na+-K+-ATPase activity was not significantly different from the FW group at 4 h after 50% SW exposure. However, at 24 h, the activity was significantly higher than the FW control. In the E2 group, Na+-K+-ATPase activity significantly decreased at 4 h, but recovered at 24 h to that of the FW group (Fig.3). The Na+-K+-ATPase activity in the E2 group was significantly lower than the corresponding sham group at 4 and 24 h after 50% SW exposure (Fig. 3).

Liver and gill enzymes.

Exposure to 50% SW exerted no significant effect on any of the liver enzymes measured except for glutamine synthetase (GS) (24 > 0, 4 h) (Table 3). Regardless of SW exposure, E2 treatment significantly decreased the activities of several key enzymes, including GS, aspartate aminotransferase (AspAT), phosphoenolpyruvate carboxykinase (PEPCK), MDH, and 6-phosphogluconate dehydrogenase (Table 3). E2 treatment did not affect gill LDH or MDH activities compared with the sham group in FW (Fig.4). Exposure to 50% SW significantly decreased gill LDH and MDH activities in the E2 group, but not in the sham group (time × treatment interaction; two-way ANOVA) (Fig. 4). Also, regression analysis of LDH and MDH activities upon exposure to 50% SW showed a significant difference between the E2 and sham group's regression lines (ANOVA; Fig. 4).


Seawater exposure.

Plasma osmolality increased transiently with exposure to 50% SW, and this observation is in agreement with other studies showing similar increases in O. mossambicus (9, 11). The regulation of plasma osmolality to FW levels occurs within 12–24 h and coincides with the activation of salt extrusion mechanisms (7, 9). In 50% SW, sodium levels were significantly higher (Fig. 2) and concur with previous studies showing elevated plasma sodium and chloride concentrations for at least a day after exposure to 70–75% SW in O. mossambicus (27,28). The plasma sodium levels decline to FW levels at 3 days after transfer to 75% SW, and the transient activation of Na+-K+-ATPase in SW is thought to facilitate this hyposmoregulatory process (27). Na+-K+-ATPase activity did increase at 24 h after exposure to 50% SW in O. mossambicus (Fig. 3), which perhaps allows for efficient sodium and chloride regulation in SW (9, 11). This increase in the sodium pump activity is crucial for SW acclimation in this species (7, 10).

Hyposmoregulatory hormones, including cortisol and thyroid hormones, play key roles in the salt extrusion processes in SW-exposed fish, including activation of Na+-K+-ATPase activity (22). Our results did not show any changes in plasma concentration of cortisol, T3, and T4 with exposure to 50% SW in tilapia. A similar lack of significant changes in plasma cortisol levels 1–4 days after SW transfer was reported in O. mossambicus (27, 34). Cortisol stimulates sodium pump activity in fish gills (22), but the absence of any plasma cortisol changes in O. mossambicus, despite higher sodium pump activity, suggests either an increased clearance of the hormone from circulation (31) or the possibility that cortisol may not be playing a prominent role in the sodium pump activation in this species. The lack of change in plasma glucose concentration after SW transfer suggests a lack of cortisol stimulation in SW in the present study, because cortisol treatment increases plasma glucose levels in tilapia (27, 34, 35). One possibility for the lack of cortisol stimulation in SW may be due to the already elevated cortisol levels in FW. However, plasma levels of cortisol reported in this study are similar to those reported previously for resting tilapia (27, 34, 35). Also, plasma glucose levels that are under cortisol stimulation in these fish were similar to those seen in a resting fish and much lower than the stressed levels reported in tilapia (35). Taken together, these results argue against elevated cortisol in FW fish as a reason for the lack of cortisol stimulation in 50% SW. However, the cortisol response in SW becomes apparent in animals with a lower energy status such as in food-deprived tilapia (34). The higher glucose levels after SW transfer in O. mossambicus reported by Nakano and coworkers (28) may also be due to the stress (and as a result cortisol stimulation) associated with handling during fish transfer. These results imply that prior stress/energy status of the animal may be an important determinant in the tissue responses to cortisol stimulation in SW. It is, therefore, arguable that the higher sodium pump activity in SW tilapia is mediated by increases and/or decreases in other hormones, notably growth hormone and prolactin (22, 27, 28, 34, 39).

We noted no significant changes in the activities of liver or gill enzymes involved in the intermediary metabolism in SW-exposed fish. Liver GS activity was higher at 24 h after exposure to 50% SW, but the reason for the increase is not clear. Our results suggest that tilapia can efficiently regulate plasma osmolality in SW by activating the sodium pump. The increased sodium pump activation may be through several routes, including phosphorylation and/or membrane insertion of the protein and increased chloride cell biogenesis, as well as synthesis of new proteins. However, the changes in sodium pump activity were evident only at 24 h, but not at 4 h, suggesting that the changes are genomic and may be due to enhanced synthesis of Na+-K+-ATPase associated with chloride cell differentiation/turnover in this species upon SW exposure (7,13). It appears that the energy requirements for this salt extrusion process do not involve any major shift in the liver and/or gill metabolic capacity and this is reflected in the lack of changes in plasma cortisol and glucose concentration in 50% SW in our tilapia.

Effect of E2.

E2 treatment results in activation of energetically demanding protein synthetic pathways (2, 14, 15). Our results confirm this enhanced protein synthetic capacity, as the plasma VTG levels were significantly higher in the E2-treated fish compared with the sham fish (Table 1). The synthesis of VTG was in milligram levels within 4 days of E2 treatment, suggesting a very high protein synthetic rate (14, 15). The lack of change in liver total protein concentration suggests that synthesized VTG is not stored in the hepatocytes but was exported immediately into circulation. A similar lack of change in liver total protein concentration, despite significant increases in hepatic protein synthetic activity with E2, was seen in flounder (Platichthys flesus) (15) and eelpout (Zoarces viviparus) (14). However, in the longer term, usually hepatic hyperplasia and hypertrophy are noted, reflected in increases in the hepatosomatic index (16,20). That no such changes were seen in the tilapia may be due to the relatively short exposure to the estrogen. As protein synthesis accounts for a significant portion of total oxygen consumption and nearly 80% of hepatocyte oxygen demand (29), a repartitioning of the energy budget is likely with E2stimulation in fish. This is reflected in the observation that in hepatocytes isolated from E2-treated rainbow trout (Oncorhynchus mykiss), amino acids are increasingly channeled into oxidation relative to gluconeogenesis, resulting in the reduction of de novo glucose synthesis from alanine (16,37). Even in the present study, a marked reduction in enzyme activities was noted with E2 treatment, indicative of a general decrease in the liver intermediary metabolic capacity. Thus it appears likely that a shift in the energy balance associated with E2 stimulation in fish occurs at the expense of other energy-demanding pathways critical for allowing animals to cope with environmental challenges.

E2 treatment has been shown to decrease plasma glucose concentration and hepatocyte glucose production capacity in rainbow trout (16, 36), but E2 did not modify plasma glucose concentration in our tilapia (Table 2). We did not measure glucose turnover rates in these animals and therefore cannot make a strong case as to the production and utilization of glucose with E2 treatment in fish. Also, we cannot base our results on an earlier study with rainbow trout that showed that E2treatment did not affect the glucose turnover rate (36), because of species and experimental protocol differences and also because the rainbow trout were cannulated and confined to individual boxes, a protocol that is known to stress the fish and increase plasma cortisol concentration (8). We did, however, observe a significantly lower liver PEPCK activity, a key enzyme in gluconeogenesis from amino acids, in the E2-treated tilapia, suggesting a decreased liver gluconeogenic capacity as reported earlier for rainbow trout (16, 37). Also, the activity of liver AspAT, but not alanine aminotransferase (AlaAT), was significantly lower in the E2-treated tilapia, hinting at a curtailed carbon (C3 precursor) channeling for glucose production, at the expense of protein synthesis and oxidation. Similar decreases in transaminases with E2 have been reported previously in brook trout (Salvelinus fontinalis) (38) and flounder (15). The decrease in transaminases may be representative of the general metabolic status of the liver with E2, which favors protein synthesis (14, 15). Support for this lies in the observation that cortisol treatment, which increases liver gluconeogenic capacity, increases AspAT, but not AlaAT, activity (34). Thus, arguably, the hepatic capacity for glucose production is compromised perhaps to channel energy and precursors into protein synthesis. The mechanism behind this decrease is not clear, although effects of E2 in mammals are thought to be indirect and mediated by corticosteroids (12).

Our study did not show any obvious effect of E2 treatment on plasma cortisol or T4 concentration (Table 2). E2 treatment did increase the plasma concentration of T3 in the present study. This increase in T3may be associated with the increased anabolic activity evident in the E2-treated fish, as T3 is a known anabolic hormone in fish (2, 32). However, the increase in T3 with E2 in our tilapia is at odds with studies on rainbow trout showing that E2 significantly decreased plasma T3 concentration (4,17). The difference may be attributed to a species difference, and also to differences in sampling time, as both trout studies sampled fish 7–14 days after E2 treatment, whereas we sampled fish 4 days after E2 treatment. It is possible that the lower T3 levels with longer-term E2 treatment may reflect the lowering of the anabolic potential as the protein synthetic capacity decreases following the initial increase seen in E2-treated fish (15).

E2 did not significantly affect plasma osmolality or gill Na+-K+-ATPase activity in FW, but plasma sodium levels were significantly higher than the sham group (Figs. 1-3). A previous study using repeated E2 injections over a 15-day period showed sodium concentrations to be significantly lower in sea trout (Salmo trutta) (19). The difference between the two studies may be due to the additional stress factor involved in the sea trout study, due to the repeated injections of E2. Stress and cortisol treatment are known to decrease plasma E2 levels and hepatic E2 receptor binding capacity in trout (3) and to lower plasma sodium levels in FW (1). The reason for the higher sodium concentration in the E2-treated fish is not clear.

E2 impact on SW osmoregulation.

E2 treatment prevented the drop in plasma osmolality that was evident in the sham group after exposure to 50% SW (Fig. 1). The absence of a drop in osmolality with E2 suggests osmoregulatory disturbances that may impair SW acclimation in O. mossambicus. Further support for this assumption arises from the observation that plasma sodium concentration did not change with exposure to 50% SW in the E2-treated group. This lack of change in plasma sodium with SW exposure in the E2 fish may be due to the already significantly higher sodium levels in FW with levels similar to that seen after SW exposure in the sham group (Fig.1).

Unlike the stable plasma sodium levels in 50% SW, clear differences were apparent in gill Na+-K+-ATPase activity with E2, suggesting disturbances in the salt extrusion process. The absence of increased Na+-K+-ATPase activity in 50% SW in the E2-treated fish clearly implicates the sodium pump as a major site for E2-induced impairment of ion regulation, as previously shown for naturally smolting Atlantic salmon, where exposure to estrogen negatively affected the activity of gill Na+-K+-ATPase (20). The significant transient decrease in Na+-K+-ATPase activity 4 h after exposure to 50% SW in the E2 group is surprising, but perhaps represents an energy-repartitioning strategy to cope with the other energy-demanding metabolic processes upon exposure to hyperosmotic environment. This argument finds support from the significantly lower gill LDH and MDH activities with E2 in 50% SW (Fig. 4), suggesting an overall decrease in the gill metabolic potential. Although the gill sodium pump activity is recovered to the preexposure levels at 24 h after exposure, the activity is significantly lower than the sham fish and may account for the lack of a decrease in plasma osmolality in the E2-treated fish. In sea trout, the gill sodium pump activity was significantly lower in E2-treated fish in SW compared with the sham fish and perhaps the reason for the lack of SW acclimation and elevated mortality in E2-treated presmolts (19). We tentatively conclude that the lack of activation of gill Na+-K+-ATPase activity may be a mechanism for E2 impairment of transient SW acclimation process in O. mossambicus. Another possible route may be a direct effect of E2 on chloride cell turnover, thereby affecting the sodium pump activity, but this remains to be determined.

Our study did not show any obvious effect of E2 treatment on plasma cortisol, T3, or T4 concentration in 50% SW (Table 2), suggesting that impaired ion regulation in the E2-treated fish may be independent of these hormones. The lack of hyposmoregulatory ability in hyperosmotic environment with E2 may be mediated by other hormones, including growth hormone, prolactin, and/or insulin-like growth factor, independent of cortisol stimulation.

In conclusion, E2 treatment decreases the intermediary metabolic capacity of the liver in tilapia, pointing to an overall energy repartitioning perhaps to cope with the increased energy demand associated with synthesis of proteins encoded by estrogen-responsive genes. Concurrent decreases of activities of two housekeeping genes (LDH and MDH) in gill in the E2 fish in 50% SW imply curtailed gill metabolic potential. Additional support for decreased metabolic potential in the gill comes from the lower gill sodium pump activity in SW. Together our results suggest that the impairment of hyposmoregulatory ability in 50% SW may be directly related to E2-induced energy repartitioning processes, affecting both liver and gill.


In salmonid fish, vitellogenesis impairs hyposmoregulatory ability of smolts in SW to such a degree that it has been hypothesized that sexual maturation and smoltification are antagonistic. However, the processes compromising SW acclimation remain elusive. Based on our results, we propose that increased energy demands due to E2-induced metabolic reorganization in liver and gill are key determinants in the SW acclimation process. The partitioning of energy toward synthesis of proteins encoded by estrogen-responsive genes, including that coding for VTG, perhaps occurs at the expense of other energy-demanding pathways, among them gluconeogenesis, synthesis of maintenance proteins involved in intermediary metabolism, and Na+-K+-ATPase activation. This metabolic hypothesis would also explain the negative effect associated with xenoestrogens on smoltification and SW survival in salmonids. Considering that the actions of estrogen and of SW acclimation are by no means restricted to fish liver and gill, many other sites of common or antagonistic target can be envisaged and should be analyzed experimentally.


We thank M. S. Rahman immensely for assistance during the study.


  • This research in Okinawa was supported in part by an international collaboration program fund to A. Takemura from the Ministry of Education, Science, Sports and Culture of Japan. M. M. Vijayan and T. P. Mommsen also acknowledge research grant support from the Natural Sciences and Engineering Research Council of Canada.

  • Address for reprint requests and other correspondence: M. M. Vijayan, Dept. of Biology, Univ. of Waterloo, Waterloo, Ontario, Canada N2L 3G1 (E-mail: mvijayan{at}

  • 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.


View Abstract