Regulation of lipoprotein lipase activity in rainbow trout (Oncorhynchus mykiss) tissues

Lipoprotein lipase (LPL) is considered as a key enzyme in the lipid deposition and metabolism of many tissues. Information on LPL activity and its regulation in fish remains very scarce. In the present study, we have examined the nutritional regulation of LPL activity by conducting post-feeding and fasting experiments in rainbow trout (Oncorhynchus mykiss). As insulin plays an important role in the nutritional regulation of LPL activity in mammals, the effects of this hormone were tested in vivo by intraperitoneal administration. Moreover, we conducted in vitro studies using fat pads of rainbow trout to better clarify the direct role of insulin and tumor necrosis factor-alpha (TNFalpha) as possible regulators of LPL activity in rainbow trout. LPL activity in adipose tissue increased in response to feeding, 4h after ingestion of food, then decreasing to basal levels at 6h. No clear response was found in either red or white muscles, where LPL values were lower. Moreover, fasting produced a down-regulation of LPL activity in adipose tissue, concomitant with low levels of plasma insulin. While insulin administration stimulated LPL activity of adipose tissue 3h after injection, no response was observed in red or white muscles. Finally, in vitro studies using fat pads revealed that insulin significantly stimulated the proportion of LPL in active conformation at the extracellular level. On the other hand, TNFalpha did not greatly affect LPL activity using this in vitro model. These data indicate that LPL activity is regulated in a tissue-specific manner following food intake, and suggest that insulin is an important regulator of LPL activity in the adipose tissue of rainbow trout.     

Osmoregulatory actions of insulin-like growth factor-I in rainbow trout (Oncorhynchus mykiss)

The ability of insulin-like growth factor-I (IGF-I), insulin and GH to promote hypoosmoregulatory ability was examined in juvenile rainbow trout (Oncorhynchus mykiss). Following adaptation to 12 parts per thousand (p.p.t.) seawater for 5 days, fish were given a single injection of hormone or vehicle, then exposed to 29 p.p.t. for 24 h and examined for changes in plasma osmolarity, ions and glucose. Ovine GH (oGH; 0.2 micrograms/g) significantly improved the ability of rainbow trout to maintain plasma osmolarity and sodium levels following transfer to 29 p.p.t. seawater. Recombinant bovine IGF-I (0.01, 0.05 and 0.02 micrograms/g) also improved the hypoosmoregulatory ability of trout; the effect being dose-dependent and greater than that of oGH. Bovine insulin (0.01, 0.05 and 0.2 micrograms/g) had no statistically significant effect on plasma ions. The results indicate that IGF-I is a potential mediator of the action of GH in seawater adaptation of salmonids.     

Glucagon and insulin response to dietary carbohydrate in rainbow trout (Oncorhynchus mykiss)

The effects of high carbohydrate (CH) diets on circulating levels of insulin and glucagon were studied in rainbow trout (Oncorhynchus mykiss). Fish (76.87 +/- 8.11 g) were fed for 60 days with three isocaloric diets: control (C) (12% CH), W (28% CH mainly from wheat), and S (28% CH from wheat plus gelatinized starch). After the diet treatment, trout fed the enriched CH diets showed a higher hepatosomatic index and liver glycogen content than controls. In addition, plasma glucose levels were also higher but, on the contrary, circulating insulin levels were greater in the control group. After the period of diet adaptation, fish were sampled 3, 6, 9, and 24 h after food administration. Glycemia levels correlated with the dietary carbohydrate content, and were minimum in controls (from 96.02 +/- 3.77 to 118.97 +/- 6.08 mg/100 ml), followed by W group (from 124.60 +/- 7.46 to 172.19 +/- 11.36 mg/100 ml) and maximum in the S group (from 133.51 +/- 9.36 to 217.88 +/- 13.36 mg/100 ml). Postprandial glucagon profiles showed an inverse relationship with glycemia, indicating that the ingestion of glucose inhibits glucagon secretion. There were no significant differences in postprandial insulin levels between groups except for W group, which presented lower levels of circulating insulin 9 h after feeding. These results indicate that in trout fed a CH diet glucose affects the secretion of glucagon more than insulin during the postprandial period.     

Effect of somatostatin on prolactin in rainbow trout (Oncorhynchus mykiss) pituitary cells in primary culture.

To study the control of prolactin secretion in fish, an in-vitro technique using a monolayer cell culture system of rainbow trout pituitary glands was developed. Such secretion was characterized by measurement of both prolactin release and prolactin mRNA content using a trout prolactin cDNA as a probe. This cell culture technique, already used to study the regulation of gonadotrophin secretion in rainbow trout, was further validated by measuring total DNA and protein content. Both parameters appeared to be stable after 2 days of culture. Studying the effect of somatostatin (SRIF) on prolactin cells indicated that a maximal inhibitory effect (62%) was observed after 24 h of treatment. Significant inhibition of prolactin release was obtained for SRIF doses ranging from 50 nM to 1 microM. However, in the same experiment, SRIF was much more potent as an inhibitor of growth hormone release. Short-term (< 12 h) incubation with SRIF did not induce a significant change in prolactin release, whereas growth hormone release was reduced at as early as 1 h after SRIF exposure. SRIF did not have a significant effect on total prolactin content or prolactin mRNA levels, suggesting the absence of an effect on prolactin synthesis. No increase in the magnitude of the inhibitory effect of SRIF was observed when using pituitary cells from immature, mature male or mature female trout. When comparing effects on primary cultures containing cells from the whole pituitary with a prolactin cell-enriched population, SRIF appeared to have the same inhibitory effect on prolactin release, supporting a direct action of SRIF on prolactin cells. These results provide further support for SRIF being a prolactin-inhibiting factor in rainbow trout and acting as a modulator of a dominant stimulatory control of prolactin release.     

Disposition of genistein in rainbow trout (Oncorhynchus mykiss) and siberian sturgeon (Acipenser baeri)

Genistein (G) is a xenoestrogen from soy present in fish diet. In vivo, a 50-fold difference in sensitivity to genistein on vitellogenin (VTG) synthesis was found when comparing trout and sturgeon. This difference was not linked to the estrogen receptor affinity nor to the sensitivity of induction of the VTG pathway. The study was performed to check if differences in the G disposition in the two species could explain their difference of sensitivity to G. A pharmacokinetic analysis of radiolabeled G was performed to determine its bioavailability and metabolism in both species. G was used at levels corresponding to fish farm exposure. G plasma levels after chronic ingestion were found to be 15.6 times higher in sturgeon than in trout. Sturgeon primarily produces sulfate conjugates after G ingestion whereas trout mainly produces glucuronides. Sturgeon was able to excrete orobol glucuronide in bile. An important first pass effect was suggested in both species. No accumulation of G or its metabolites was observed in the two species. Trout muscles accounted only for 0.14 of radioactivity 48 h post-ingestion similarly to sturgeon. Trout viscera accounted for 15% of the radioactivity 48 h post-ingestion. In sturgeon, 48 h post-ingestion, viscera accounted for 21.5% of the radioactivity. These rates decreased rapidly thereafter. The study partly explains the difference in sensitivity to G, previously recorded between the two species. In addition, it shows that human exposure to G through farmed fish consumption is negligible.     

Actions of growth hormone on carbohydrate metabolism and osmoregulation of rainbow trout (Oncorhynchus mykiss)

Rainbow trout Oncorhynchus mykiss were injected intraperitoneally with slow-release implants of vegetable oil alone or containing ovine growth hormone (oGH) (2 and 5 microgg(-1) body weight), and sampled after 5 days to assess the simultaneous effects of GH on both osmoregulation and carbohydrate metabolism. An enhanced hypoosmoregulatory capacity of oGH-implanted fish is suggested by the increase observed in gill Na+,K+-ATPase activity, and the decrease observed in plasma ion concentration (Na+ and Cl-) and osmolality. GH treatment also elicited increased plasma glucose levels and metabolic changes in liver, gills, kidney, and brain. Major metabolic changes elicited by GH treatment included (1) decreased glycolytic potential and capacity for exporting glucose in liver, (2) enhanced glycogenolytic potential and capacity for use of exogenous glucose in gills and kidney, as well as increased glycolytic capacity in the later tissue, and (3) enhanced glycogenolytic and glycolytic capacities in brain. These metabolic changes elicited by GH treatment support a role for GH in the control of carbohydrate metabolism in salmonids that could be related either to the metabolic changes occurring during osmotic acclimation in nature (a process in which changes in GH levels and carbohydrate metabolism have both been reported) or to metabolic changes associated with growth.     

CO2 transport and excretion in rainbow trout (Oncorhynchus mykiss) during graded sustained exercise

A quantitative analysis of CO2 transport and excretion was conducted in seawater acclimated rainbow trout (Oncorhynchus mykiss) swimming at different sustained swimming velocities. CO2 excretion increased linearly with cardiac output during exercise but arterial P(CO2) (Pa(CO2)) and total CO2 levels also increased indicating a diffusion limitation to CO2 excretion. The elevated Pa(CO2) was not accompanied by a decrease in pH, indicating that the acid-base compensation was rapid. Mixed-venous P(CO2) increased to a greater extent than Pa(CO2) resulting in a large increase in the venous arterial difference in P(CO2) (Pv(CO2) - Pa(CO2)). The Pv(CO2) - Pa(CO2) difference was used to calculate the proportion of total CO2 excreted comprised of dissolved CO2 which accounted for less than 1% of total CO2 excreted in fish swimming at 11 cm sec(-1) but increased to about 9% at the greatest swimming velocity indicating that the pattern of CO2 excretion changes during exercise. There was no effect of exercise on the proportion of CO2 excreted which was dependent upon HCO3-/Cl- exchange (54%) or that which was dependent upon the dehydration of HCO3- that resided within the red cell prior to gill blood entry (42%). The large proportion of total CO2 excreted that was dependent upon HCO3-/Cl- exchange is significant because this is thought to be the rate limiting step in CO2 excretion.     

Effects of cortisol on aspects of 3,5,3'-triiodo-L-thyronine metabolism in rainbow trout (Oncorhynchus mykiss)

Aspects of 3,5,3'-triiodo-L-thyronine (T3) metabolism were studied in fed rainbow trout (Oncorhynchus mykiss) held at 11.5-14 degrees and intraperitoneally implanted with hydrogenated corn oil (controls) or oil containing cortisol. Cortisol implants caused dose-related plasma cortisol elevations within the physiological range for 2-3 weeks, loss in body weight, and depression in plasma T3 and free T3 index with no consistent change in plasma thyroxine (T4) or free T4 index. Plasma T3 clearance rate and plasma T3 appearance rate were both increased by cortisol, with no change in hepatic microsomal T4 5'-monodeiodinase activity (Km or Vmax), but with a significant decrease in muscle T3 concentration. It is concluded that chronic physiologic cortisol treatment enhances plasma T3 clearance without change in hepatic T4 to T3 conversion, resulting in a decline in T3 concentration in both plasma and tissue (muscle) compartments.     

Differential expression of two somatostatin receptor subtype 1 mRNAs in rainbow trout (Oncorhynchus mykiss)

Somatostatins (SSs) play important roles in the growth, development and metabolism of vertebrates. In this study, cDNAs for two unique somatostatin receptor variants were cloned and sequenced from rainbow trout. The two cDNAs, one consisting of 1755 bp and the other of 1743 bp, share 63.6% identity in nucleotide sequence and 94.1% identity in deduced amino acid sequence and presumably arose through gene duplication. Each cDNA encodes for a putative 371-amino acid somatostatin receptor (one designated sst1A and the other sst1B) containing seven transmembrane domains. Rainbow trout sst1A and sst1B have 64.4 and 65.5% similarity respectively with human sst1 and only 43-60% similarity with other subtypes. Trout sst1 mRNAs are differentially expressed, both in terms of distribution among tissues as well as in terms of abundance within selected tissues. Both sst1A and sst1B mRNAs were present in brain, stomach, liver, pancreas, upper and lower intestine, pyloric cecum, kidney and muscle, whereas only sst1B mRNA was present in the esophagus. sst1A mRNA was more abundant than sst1B in the optic tectum, whereas sst1B mRNA was more abundant than sst1A in liver. sst1A and sst1B mRNAs were equally abundant in pancreas. These findings contribute to the understanding of the evolution of the SS signaling system and provide insight into the mechanisms that regulate the expression of SS receptors.     

Cloning and characterization of teneurin C-terminus associated peptide (TCAP)-3 from the hypothalamus of an adult rainbow trout (Oncorhynchus mykiss)

Neuropeptides that evolved early in metazoan evolution may possess much larger networks of paralogous genes than later evolving peptides due to the increased exposure to gene and genomic duplication events. The corticotropin-releasing factor family of peptides, which also include invertebrate CRF-like peptides, are a candidate group that appear to have an early origin. We have attempted to find additional paralogous genes to the CRF family by doing a low-stringency screen of a rainbow trout hypothalamic cDNA library using a hamster urocortin probe. A clone was identified that represented the rainbow trout ortholog of teneurin-3. The C-terminal region of the last exon teneurin transmembrane protein gene possesses a neuropeptide-like sequence with a primary structure similarity to the corticotropin-releasing factor family of peptides. We have called this sequence teneurin C-terminal associated peptide (TCAP). The predicted peptide is 40 residues long and possesses an expected pyroglutamyl residue in the first position and an amidated carboxy terminus. A synthetic version of the rainbow trout (rt) TCAP-3 is potent at increasing the concentration of cAMP and stimulating proliferation in a neuronal cell line. The synthetic peptide can also either increase or decrease the expression of the teneurin-1 gene, depending upon its concentration. The teneurin/TCAP system may represent a novel and highly conserved regulatory signalling system in the vertebrate brain.     

Definitive role for natriuretic peptide receptor-C in mediating the vasorelaxant activity of C-type natriuretic peptide and endothelium-derived hyperpolarising factor

OBJECTIVE: C-type natriuretic peptide (CNP) has recently been suggested to represent an endothelium-derived hyperpolarising factor (EDHF) in the mammalian resistance vasculature and, as such, important in the regulation of local blood flow and systemic blood pressure. Additionally, this peptide has been shown to protect against ischaemia-reperfusion injury and inhibits leukocyte and platelet activation. Herein, we use a novel, selective natriuretic peptide receptor-C (NPR-C) antagonist (M372049) to highlight the pivotal contribution of CNP/NPR-C signalling in the EDHF-dependent regulation of vascular tone and investigate the mechanism(s) underlying the release and biological activity of CNP. METHODS: In vitro pharmacological investigation was conducted in rat (Sprague-Dawley) aorta and mesenteric resistance arteries. Relaxant responses to CNP, atrial natriuretic peptide (ANP), the nitric oxide donor spermine-NONOate (SPER-NO) and the endothelium-dependent vasodilator, acetylcholine (ACh) were examined in the absence and presence of M372049 or inhibitor cocktails shown previously to block endothelium-dependent dilatation in the resistance vasculature. RT-PCR was employed to characterize the expression of NPR subtypes in the vessels studied. RESULTS: M372049 produced concentration-dependent inhibition of the vasorelaxant activity of CNP in rat isolated mesenteric resistance arteries but not aorta; in contrast, M372049 did not affect relaxations to ANP or SPER-NO in either vessel. M372049 or ouabain alone produced small, significant inhibition of EDHF-dependent relaxations in mesenteric arteries and in combination acted synergistically to abolish such responses. A combination of M372049 with established inhibitors of EDHF-dependent relaxation revealed that multiple, distinct pathways coordinate the bioactivity of EDHF in the resistance vasculature, and that CNP/NPR-C signalling represents a major component. CONCLUSIONS: These data substantiate CNP/NPR-C signalling as a fundamental pathway underlying EDHF-dependent regulation of vascular tone in the rat mesenteric resistance vasculature. An increased understanding of the physiological roles of CNP/NPR-C signalling in the vasculature (now facilitated by the identification of a selective NPR-C antagonist) should aid determination of the (patho)physiological importance of EDHF and might provide the rationale for the design of novel therapeutics.     

Recombinant prolactin receptor extracellular domain of rainbow trout (Oncorhynchus mykiss): subcloning, preparation, and characterization

The cDNA of the extracellular domain of rainbow trout (Oncorhynchus mykiss) prolactin receptor (trPRLR-ECD) was cloned in the prokaryotic expression vector pMON to enable its expression in Escherichia coli after induction with nalidixic acid. The bacterially expressed trPRLR-ECD protein, contained within the refractile body pellet, was solubilized in 4.5 M urea, refolded, and purified on a Q-Sepharose column, pH 8, by stepwise elution with NaCl. The bioactive monomeric 26-kDa fraction was eluted in 0.2 M NaCl, yielding 20 mg/2.5 L of induced culture. The purified protein was over 98% homogeneous, as shown by SDS-PAGE in the presence or absence of reducing agent and by chromatography on a Superdex column. Binding experiments using [125I]ovine placental lactogen (oPL) as a ligand revealed that human growth hormone (hGH), oPL, and ovine prolactin (oPRL) were the most effective competitors, with respective IC50 values of 1.32, 2.27, and 2.70 nM. Chicken (ch) PRL did not compete at all, and homologous trPRL was much less effective, with a corresponding IC50 value of 1826 nM. Gel-filtration was used to determine the stoichiometry of trPRLR-ECD's interaction with oPL, hGH, and oPRL. Only oPL yielded a 2:1 complex, whereas hGH and oPRL formed only 1:1 complexes, with excess trPRLR-ECD being seen at the initial 2:1 trPRLR-ECD:hGH or trPRLR-ECD:oPRL ratios. No studies were performed with chPRL because of its inability to compete with [125I]oPL or with trPRL because of its low affinity toward trPRLR-ECD. The present results agree with previous findings indicating, as in mammals, that homologous PRL interacts transiently with its receptor and suggest that transient homologous PRL-induced homodimerization of the receptor is sufficient to initiate a biological signal, despite the fact that, in classical binding experiments, only low specific binding can be detected.     

Stress elevates corticotropin-releasing factor (CRF) and CRF-binding protein mRNA levels in rainbow trout (Oncorhynchus mykiss)

The objectives of this study were to characterize rainbow trout (Oncorhynchus mykiss) corticotropin-releasing factor (CRF)-binding protein (CRF-BP) cDNA and to examine the variations in CRF-BP and CRF mRNA levels in response to different intensities of stress. Trout were physically disturbed by a single or three consecutive periods of chasing until exhaustion followed by 2 h of recovery. The pituitary CRF-BP and preoptic area CRF1 mRNA contents were significantly increased only after repeated chasing events. Physical disturbance increased plasma cortisol levels with the largest change occurring in the group of trout that were exposed to repeated chasing events. Trout were also individually isolated in 120 l tanks or confined to 1.5 l boxes for 4, 24 or 72 h. CRF-BP mRNA levels in confined fish were greater than those of isolated fish at 72 h although there were no differences compared with the control group. CRF1 mRNA levels in the preoptic area were greater and remained elevated for a longer period in confined compared with isolated trout. Isolation led to a transient increase in plasma cortisol levels, but the higher cortisol values developed in the confined fish suggest that this treatment was more stressful than isolation. These results demonstrate that the intensity and duration of stress are important factors regulating CRF and CRF-BP mRNA levels in rainbow trout. We hypothesize that pituitary CRF-BP is involved in regulating the activity of the stress axis, possibly by reducing access to CRF1 receptors in the corticotropes.     

Participation of the pineal gland in sexual maturation of female rainbow trout (Oncorhynchus mykiss Walbaum)

In the study rainbow trout, pinealectomized at two different periods of their sexual cycle, i.e., at either 1 or 5 months before spawning, were studied. It was found that the lack of the pineal gland in the period directly preceding the spawning had no statistical effect on either spawning or the number of females that produced eggs. Pinealectomy performed during vitellogenesis delayed spawning by about 2 weeks and resulted in the absence of ovulation in 20% of the females. The results suggest that the pineal gland may influence the hypothalamo-pituitary-gonadal axis by altering the maturation period and controlling spawning in the rainbow trout.     

Control of adipose tissue lipid metabolism by tumor necrosis factor-alpha in rainbow trout (Oncorhynchus mykiss)

Tumor necrosis factor-alpha (TNF alpha) is a cytokine with multiple biological functions which, in mammals, has been shown to modulate muscle and adipose tissue metabolism. In fish, TNF alpha has been identified in several species. However, few studies have examined the role of TNF alpha in fish outside the immune system. In this study, we assessed the effects of human recombinant TNF alpha and conditioned media from rainbow trout lipopolysaccharide (LPS)-stimulated macrophages (LPS-MCM) on lipolysis in isolated rainbow trout adipocytes. Furthermore, we studied the effects of an LPS injection in vivo on lipid metabolism. In our study, human recombinant TNF alpha stimulated lipolysis in trout adipocytes in a time- and dose-dependent manner. Similarly, LPS-MCM stimulated lipolysis in trout adipocytes when compared with control conditioned medium. Experiments using specific inhibitors of the MAP kinase pathway showed that p44/42 and p38 are partially involved in the lipolytic effects of TNF alpha. On the other hand, adipocytes from LPS-injected rainbow trout showed higher basal lipolysis than adipocytes from control fish after 24 h, while this effect was not seen at 72 h. Furthermore, lipoprotein lipase (LPL) activity in adipose tissue of LPS-injected fish was lower than in the controls at 24 h. These data suggest that TNF alpha plays an important role in the control of lipid metabolism in rainbow trout by stimulating lipolysis in vitro and in vivo and by down-regulating LPL activity of adipose tissue in vivo.     

Characterization of epidermal growth factor binding to hepatic plasma membranes of rainbow trout (Oncorhynchus mykiss)

Epidermal growth factor (EGF) is a potent mitogen which exerts its effects through a transmembrane receptor located on target cells. Since little is known about EGF in nonmammalian animals, experiments were conducted to characterize the EGF receptor in rainbow trout hepatic cell membranes. The binding of mouse EGF with rainbow trout receptors was peptide-specific, saturable, reversible, and of high affinity. Optimal binding was observed at pH 7.2. Both monovalent and divalent cations augmented the specific binding of EGF, by a factor of two- to threefold over control values, but were not needed for binding to occur. Scatchard plot of the saturation data was curvilinear. Analysis of the data by kinetic methods indicated the curvilinear nature of the Scatchard plot was due to multiple receptor sites of differing affinity and not to site-to-site interactions. Rate constants for association (K11) were 9.38 X 10(8) M-1 * min-1 and 2.28 X 10(7) M-1 * min-1 for the high and low affinity sites, respectively. Rate constants for dissociation (K-1) were 2.03 X 10(-3) min-1 and 2.07 X 10(-1) min-1 for high and low affinity sites, respectively. The apparent dissociation constants determined from the rate constants for the high affinity (KD = 2.3 X 10(-11) M) and low affinity (KD = 9.1 X 10(-9) M) were in agreement with constants estimated by equilibrium methods. Maximum binding capacities were 13 fmol EGF bound/mg protein of protein and 270 fmol EGF bound/mg of protein for the high and low affinity receptor sites, respectively.(ABSTRACT TRUNCATED AT 250 WORDS)     

Environmental temperature increases plasma GH levels independently of nutritional status in rainbow trout (Oncorhynchus mykiss)

Like many poecilotherms, salmonids exhibit seasonal variations of growth rate in relation with seasonal temperatures and plasma GH level. However, temperature alters other parameters like food intake, which may directly modify the level of plasma GH. In order to determine whether temperature regulates plasma GH levels independently of nutritional status, fish were reared at 8, 12, or 16 degrees C and either fed ad libitum (fish with different food intake) to determine the global effect of temperature, or with the same ration (1.2%/body weight) to observe the temperature effect in fish with the same growth rate. Plasma insulin level was inversely proportional to the temperature (8, 12, and 16 degrees C) in fish fed ad libitum (12.1+/-0.3 ng/ml, 10.9+/-0.3 ng/ml, 9.5+/-0.4 ng/ml; P<0.001) and in restricted fish (14.0+/-0.3 ng/ml, 11.3+/-0.3 ng/ml, 10.0+/-0.2 ng/ml; P<0.0001), probably due to a prolonged nutrient absorption, and delayed recovery of basal insulin level at low temperature. Conversely, temperature did not affect plasma T3 level of fish fed ad libitum (2.5+/-0.2 ng/ml, 2.4+/-0.1 ng/ml, 2.5+/-0.1 ng/ml at 8, 12, and 16 degrees C) while fish fed with the same ration present less T3 at 16 degrees C than at 8 degrees C (1.83+/-0.1 ng/ml versus 1.2+/-0.1 ng/ml; P<0.001) throughout the experiment; these observations indicate that different plasma T3 levels reflect the different nutritional status of the fish. The levels of GH1 and GH2 mRNA, and GH1/GH2 ratio were not different for whatever the temperature or the nutritional status. Pituitary GH content, of fish fed ad libitum did not exhibit obvious differences at 8, 12, or 16 degrees C (254+/-9 ng/g bw, 237+/-18 ng/g bw, 236+/-18 ng/g bw), while fish fed with the same ration have higher pituitary GH contents at 16 degrees C than at 8 degrees C (401+/-30 ng/g bw versus 285+/-25 ng/g bw; P<0.0001). Interestingly, high temperature strongly increases plasma GH levels (2.5+/-0.3 ng/ml at 8 degrees C versus 4.8+/-0.6 ng/ml at 16 degrees C; P<0.0001) to the same extent in both experiments, since at a given temperature average plasma GH was similar between fish fed ad libitum or a restricted diet. Our results, demonstrate that temperature regulates plasma GH levels specifically but not pituitary GH content, nor the levels of GH1 and GH2 mRNA. In addition no differential regulation of both GH genes was evidenced whatever the temperature.