Insulin and Igf-I Receptors in Trout Adipose Tissue are Physiologically Regulated by Circulating Hormone Levels

In fish, adipose tissue is one of the tissues that, together with liver and muscle, is able to store lipids, mostly as triacylglycerols. The maintenance of lipid stores in adipose tissue is determined by a balance between lipid mobilization and deposition. Since the primary role of adipose tissue is as an energy store, a number of influences, seasonal, developmental or physiological, can shift the balance to lipid mobilization or to lipid deposition, according to the particular circumstances of the animal. In mammals, adipose tissue is under the control of the sympathetic nervous system and the endocrine system (Shapiro, 1977). In particular, lipid mobilization is known to be stimulated by endocrine signals, such as catecholamines and glucagon, and inhibited primarily by insulin. However, lipid deposition is also hormonally regulated, and insulin is the main endocrine signal involved. A number of growth factors are also involved in the development and metabolism of adipose tissue. In fish, the enzymatic machinery responsible for mobilization and deposition of lipids in adipose tissue is similar to that found in mammalian adipose tissue (Sheridan, 1994). However, the endocrine factors directly involved in the regulation of adipose tissue function are not known, and there is conflicting information regarding the ability of fish adipose tissue to respond to hormonal stimulation (Christiansen et al., 1985; Harmon and Sheridan, 1992; Migliorini et al., 1992; Murat et al., 1985).

In addition to its primary role as a lipid store, adipose tissue may also play a role in other physiological processes in fish because a number of studies in mammals have suggested multiple functions for adipose tissue. To date, there is a considerable amount of evidence indicating that adipose tissue is an important target for hormones and is, itself, a source of factor(s), for example leptin, that affect other parts of the organism (Flier, 1995). It is possible that fish adipose tissue could also have this regulative capacity, as suggested from studies correlating levels of adiposity with food intake (Jobling and Miglavs, 1993) and with the onset of sexual maturation (Rowe et al., 1991; Silverstein et al., 1997).

In mammals, it is well known that adipose tissue is an important target for insulin and that insulin action is mediated through specific insulin receptors. In fish, however, no information is available regarding the direct effects of insulin on adipose tissue (Mommsen and Plisetskaya, 1991; Sheridan, 1994) or on the presence of specific insulin receptors, although insulin receptors, as well as insulin-like growth factor I (IGF-I) receptors, have been detected in a variety of fish tissues, including liver, skeletal and red muscle, heart and ovary (Baños et al., 1997; Gutiérrez et al., 1993, 1995; Párrizas et al.,1994b, 1995b). The piscine insulin and IGF-I receptors are physiologically regulated and appear to be structurally similar to their mammalian counterparts (Navarro et al., 1999).

To investigate further the hormonal regulation of adipose tissue function in fish, we initiated studies on the possible role of insulin and growth factors in the regulation of lipid metabolism in adipose tissue. As a first step, we set out to confirm the presence of receptors for insulin and IGF-I in adipose tissue of brown trout (Salmo trutta) and to determine the characteristics and physiological regulation of these receptors.

Two-year-old brown trout (Salmo trutta) were obtained from the fish hatchery Piscifactoria de Bagà (Departament de Medi Natural, Generalitat de Catalunya), Barcelona, Spain. Fish were maintained in tanks with open fresh water circuits under natural conditions of light and temperature. They were fed commercial trout pellets.

In each experiment, one group of fish (N=4–6) received one intraperitoneal injection of L-arginine (6.6 µmol g⁻¹ fish; Sigma, St Louis, MO, USA) after an overnight fast. Arginine is an amino acid known to cause an increase in plasma levels of insulin (Plisetskaya et al., 1991) and a subsequent upregulation of skeletal muscle insulin receptors in salmonid fish (Párrizas et al., 1994a). Another group of fish (control, N=4–6) received one injection of the vehicle (saline) under the same conditions as the arginine-injected group. After 4 h, blood and adipose tissue samples from the control and arginine-injected fish were taken. Four separate experiments were conducted between November 1996 and February 1997.

At the beginning of each experiment, fish were divided in two groups (10 fish per group): one received food daily (fed controls) and the other was deprived of food for the entire experiment (fasted). After 45 days of fasting, known in this species to be a non-life-threatening fast after which plasma insulin levels are significantly reduced (Navarro and Gutiérrez, 1995), fish from both groups were killed, and blood and adipose tissue samples were taken. Three separate experiments were conducted in April, June and November 1997.

Fish were anaesthetized in 3-aminobenzoic acid ethyl ester (0.1 g l⁻¹; Sigma, St Louis, MO, USA) dissolved in fresh water. Blood was drawn from the caudal vein and immediately centrifuged, and the plasma was stored at −80 °C until assayed for hormone content. After the anaesthetized fish had been killed by a blow to the head, the entire mesenteric adipose tissue was dissected out, immediately frozen in liquid N₂ and stored in liquid N₂ until assayed for receptor binding.

Adipose tissue from male and female brown trout was pooled since no significant differences in insulin and IGF-I binding characteristics were detected between sexes (data not shown). Adipose tissue (10–15 g, pooled from a minimum of three fish) was homogenized with a Polytron in Tris/HCl buffer (25 mmol l⁻¹ Tris/HCl, 5 mmol l⁻¹ CaCl₂, pH 7.6) and centrifuged at 600 g for 10 min at 4 °C. The supernatant was centrifuged at 40 000 g for 30 min at 4 °C, and the resulting pellet was resuspended in a buffer containing 25 mmol l⁻¹ Hepes, 4 mmol l⁻¹ EDTA, 4 mmol l⁻¹ EGTA, 2 mmol l⁻¹ Phenylmethylsulphonyl fluoride, 1 mmol l⁻¹ Bacitracin, 1 mmol l⁻¹ Leupeptin and 1 mmol l⁻¹ Pepstatin at pH 7.6. Membranes were solubilized by adding Triton X-100 to a final concentration of 2 % and stirring for 1 h at 4 °C. After centrifugation at 150 000 g for 90 min at 4 °C, the supernatant was recycled three times through an agarose-bound wheat germ agglutinin column. Receptors were eluted by washing the column with buffer containing 25 mmol l⁻¹ Hepes, 0.1 % Triton X-100 and 0.3 mol l⁻¹N-acetyl-D-glucosamine, and the protein concentration was determined using a commercial assay kit (BioRad, Richmond, CA, USA). For each experiment, between three and eight different purifications were performed.

Approximately 20 µg of partially purified receptors were incubated for 16 h at 4 °C in 30 mmol l⁻¹ Hepes, 0.1 % bovine serum albumin (BSA) and 100 units ml⁻¹ Bacitracin, with increasing concentrations of unlabelled porcine insulin (Lilly Co., Indianapolis, IN, USA) or recombinant human IGF-I (Chiron Corp., Emeryville, CA, USA) and the radiolabelled ligand (recombinant human Tyr A14-[¹²⁵I]-insulin or 3-[¹²⁵I]-IGF-I; specific activity 2000 Ci (74 TBq) mmol⁻¹; Amersham Life Sciences, Arlington Heights, IL, USA) at tracer concentrations (35 pmol l⁻¹). Partially purified receptors were precipitated with bovine gamma globulin (0.08 % w/v) and polyethylene glycol (10.4 % w/v) and centrifuged at 14 000 g for 7 min at 4 °C. Nonspecific binding was estimated as the percentage of ¹²⁵I-labelled ligand bound in the presence of 400 nmol l⁻¹ unlabelled hormone. The affinity and capacity (R_o, receptor number) of the receptors were calculated only from the high-affinity, low-capacity binding sites revealed in Scatchard plots. The use of mammalian hormones in our binding studies with partially purified receptors from fish tissues was justified by the finding that mammalian and fish insulin and IGF-I bind the corresponding piscine receptors equally well (Gutiérrez and Plisetskaya, 1991; Gutiérrez et al., 1995; Leibush et al., 1996).

Plasma insulin levels were measured by radioimmunoassay (RIA) that used bonito insulin as standard and rabbit anti-bonito insulin antibodies as antiserum (Gutiérrez et al., 1984) and that had been validated for brown trout plasma (Navarro et al., 1991, 1993). Plasma IGF-I levels in extracted plasma samples were measured by RIA that used human recombinant IGF-I as standard and rabbit anti-human IGF-I antibodies as antiserum (Pérez-Sánchez et al., 1994) and that had been validated for brown trout plasma (Baños et al., 1999).

Differences between groups were analysed for statistical significance using Student’s t-test. Differences were considered significant at P<0.05. Values are presented as means ± S.E.M.

Binding for both insulin and IGF-I was detected in adipose tissue of brown trout (Table 1). The binding for IGF-I was significantly (P<0.05) greater (approximately fourfold) than the binding for insulin, and the number of binding sites for IGF-I was also greater (approximately twofold) than for insulin. In addition, IGF-I receptors showed higher (approximately twofold) affinity for their ligand than did insulin receptors (Table 1; Fig. 1 inset). Displacement curves showed that unlabelled IGF-I was able to displace 50 % of ¹²⁵I-IGF-I at a lower concentration (approximately fourfold) than that needed by unlabelled insulin to displace 50 % of ¹²⁵I-labelled insulin bound to the partially purified receptor preparation (Fig. 1). Furthermore, cross-displacement experiments also demonstrated that IGF-I receptors are more specific than insulin receptors because unlabelled insulin was 10–30 times more potent than unlabelled IGF-I in displacing ¹²⁵I-labelled insulin, whereas unlabelled IGF-I was over 400 times more potent than unlabelled insulin in displacing ¹²⁵I-labelled IGF-I (data not shown).

Differences in specific binding for insulin and IGF-I in trout adipose tissue were found to be dependent on the time of the year when the samples were collected (Fig. 2). IGF-I binding in adipose tissue was high in the autumn (October–December) and gradually declined until spring (April–June). Insulin binding in adipose tissue did not change significantly between the autumn and winter (January–March) and was no longer detectable in the spring.

Treatment with arginine caused a significant (P<0.05) increase in the plasma levels of insulin and IGF-I (Fig. 3A). In addition, arginine treatment significantly (P<0.05) increased the specific binding for insulin and IGF-I (Fig. 3B). The 1.7-fold increase in specific binding for insulin and the 2.3-fold increase in specific binding for IGF-I, compared with the control group, were accompanied by increases in the number of receptors (Fig. 3B, inset). Thus, arginine treatment caused an upregulation of insulin and IGF-I receptors in adipose tissue in the autumn and winter months.

Food deprivation for a 45-day period (initiated in November 1997) caused a significant (P<0.05) decrease in total body mass (244.4±10.61 g in fed controls; 152.3±9.64 g in fasted animals, N=10) and mesenteric fat index (100×mesenteric fat mass/body mass; 1.9±0.23 % in fed controls; 0.74±0.09 % in fasted animals, N=10). In addition, fasting caused a significant (P<0.05) decrease in the plasma levels of insulin and IGF-I (Fig. 4A) and a significant (P<0.05) decrease in insulin and IGF-I binding in adipose tissue (Fig. 4B). Insulin binding in the fasted animals decreased by 3.6-fold compared with that in the fed controls, whereas IGF-I binding in the fasted animals decreased by 2.2-fold compared with that in the fed controls. In both cases, the decreases in specific binding were associated with decreases in receptor number, although the decrease was only significant for IGF-I receptors (Fig. 4B, inset). Therefore, insulin and IGF-I receptors in adipose tissue are downregulated by food deprivation.

A similar fasting experiment initiated in April 1997 also resulted in a significant (P<0.05) decrease in IGF-I-specific binding in adipose tissue (23.04±2.07 % in fed fish; 15.53±0.57 % in fasted fish; two receptor purifications), while insulin binding was not detectable in either the fed or fasted groups. In a third fasting experiment initiated in June 1997, the binding of insulin remained undetectable and IGF-I binding in adipose tissue decreased, showing no significant differences between fed and fasted fish (6.1±2.14 % for fed fish and 8.33±1.33 % for fasted fish; two receptor purifications).

In the present study, specific receptors for insulin and IGF-I were detected in trout adipose tissue. Insulin and IGF-I receptors in trout adipose tissue appeared to be physiologically regulated by the endocrine status of the fish, showing a direct correlation between plasma hormone levels and the number of receptors.

To our knowledge, this is the first report of the presence of specific insulin and IGF-I receptors in adipose tissue in an ectothermic vertebrate. In trout adipose tissue, IGF-I receptors were more abundant than insulin receptors and had a higher affinity for their natural ligand, which is characteristic of all fish tissues examined to date (Baños et al., 1997; Gutiérrez et al., 1993; Párrizas et al., 1994b, 1995b) and of amphibians and reptiles (Párrizas et al., 1995a). Furthermore, insulin and IGF-I receptors in adipose tissue showed a comparable or even higher binding capacity and affinity than those found in other tissues. We have previously detected differences in receptor characteristics among different fish tissues, with cardiac muscle and ovary also showing high-affinity, high-capacity binding sites for insulin and IGF-I (Gutiérrez et al., 1995; Maestro et al., 1997).

Our results on the physiological regulation of insulin and IGF-I receptors (upregulation in response to arginine treatment and downregulation in response to fasting) indicate the presence of adaptive mechanisms in fish adipose tissue similar to those described previously in trout skeletal muscle (Párrizas et al., 1994a, 1995b) and liver (Gutiérrez and Plisetskaya, 1991). These adaptive mechanisms would couple the presence of insulin and IGF-I in the circulation with the presence of specific receptors in adipose tissue, conferring the ability to respond to these hormones. Thus, similar binding characteristics and patterns of regulation of insulin and IGF-I receptors are found in trout tissues that serve as energy stores. This indirectly suggests a role for insulin and IGF-I in the maintenance of energy stores in fish, including adipose tissue.

It is worth noting that in red and cardiac muscle of trout (Baños et al., 1997; Moon et al., 1996), two tissues with high rates of aerobic metabolism, the number of insulin and IGF-I receptors is inversely related to the hormone levels and, thus, the situation is the opposite of that found in trout adipose tissue (the present study), skeletal muscle and liver (Gutiérrez and Plisetskaya, 1991; Párrizas et al., 1994a, 1995b). Thus, in the case of fish insulin and IGF-I, it appears that the relationship between hormone levels and hormone receptor number may reflect the metabolic characteristics of the target tissue. It is also interesting that the pattern of regulation of insulin and IGF-I receptors in trout adipose tissue differs from that in mammals, in which high insulin levels are associated with the downregulation (decreased numbers) of insulin receptors (Livingstone et al., 1978) and fasting is associated with the upregulation (increased numbers) of insulin receptors (Kasuga et al., 1977). It is not known whether these differences in receptor regulation between the adipose tissue of fish and mammals reflect differences in adipose tissue function between these two vertebrate groups.

The presence of specific insulin and IGF-I receptors in trout adipose tissue indicates that this tissue is a potential target for insulin and IGF-I in trout. In birds and mammals, insulin receptors are present in adipose tissue and mediate the known stimulatory effects of insulin on lipogenesis and glucose uptake in adipocytes (Bassas et al., 1988; Campbell and Scanes, 1988; Livingstone et al., 1978). IGF-I receptors are also present in mammalian and avian adipose tissue (Bassas et al., 1988; Smith et al., 1988) and mediate the proliferative and differentiative effects of IGF-I on adipocyte precursor cells (Butterwith and Goddard, 1991; Smith et al., 1988) and the insulin-like effects of IGF-I on mature adipocytes (Campbell and Scanes, 1988; Kern et al., 1989). In ectothermic vertebrates, no clear roles have yet been elucidated for insulin and IGF-I in the regulation of adipose tissue function (Sheridan, 1994). In fish adipose tissue, insulin has been reported to have anti-lipolytic effects (Harmon and Sheridan, 1992; Mommsen and Plisetskaya, 1991), but direct effects of insulin on lipogenesis in fish adipocytes have not been demonstrated (Mommsen and Plisetskaya, 1991). The reported absence of effect of insulin on glucose uptake by trout adipocytes (Christiansen et al., 1985), together with data on the predominant role of the liver over adipose tissue in lipogenesis (Christiansen and Klungsøyr, 1987; Sheridan, 1994), led to the suggestion that adipose tissue could be a storage organ for previously synthesized or absorbed fat. Nevertheless, the question of the direct role of insulin in the regulation of lipogenesis remains unanswered. While information on the effects of insulin on adipose tissue function in ectothermic vertebrates is regrettably scarce, there is a complete lack of information regarding the potential role of IGF-I. The high numbers of IGF-I receptors detected in trout adipose tissue suggest a role for IGF-I in adipose tissue function. Current efforts in our laboratory are directed towards determining the effects of insulin and IGF-I on adipose tissue function.

The observed variation in insulin and IGF-I binding in trout adipose tissue in the present study could be interpreted as underlying possible seasonal variations in insulin and IGF-I binding in trout adipose tissue, since both show a gradual decrease in binding between autumn and spring. Previous reports on seasonal variations in circulating levels of insulin in brown trout (the subject of this study) and Atlantic salmon (Salmo salar) indicate that insulin levels increase throughout winter and reach maximum values during spring (Dickhoff et al., 1989; Navarro et al., 1991). In the present study, plasma insulin levels were also found to be significantly (P<0.05) higher in trout sampled during late spring (June 1997; 21.36±0.17 ng ml⁻¹, N=8) than during autumn (November 1997; 14.48±0.73 ng ml⁻¹, N=10), despite the lack of detectable insulin binding in spring. These findings, coupled with the significant decrease in IGF-I binding detected between autumn and spring, despite similar IGF-I plasma levels, suggest that the postulated long-term (circannual) changes in insulin and IGF-I binding in adipose tissue may not be temporally (seasonally) associated with changes in plasma hormone levels. Since a direct relationship has been demonstrated between hormone levels and receptor binding for insulin and IGF-I in short-term (arginine treatment) and medium-term (fasting) experiments, it is possible that the postulated long-term changes in receptor binding could be initiated by several factors in addition to the ligands themselves. In the light of these results, it would be interesting to examine the endocrine interrelationships between insulin and IGF-I and their receptors in adipose tissue, as well as functional aspects of the tissue, throughout the entire annual cycle. Future studies in our laboratory will be directed towards answering these questions.

In conclusion, the results from this study on the presence and regulation of specific insulin and IGF-I receptors in fish adipose tissue indicate that this tissue can be a target for the actions of insulin and IGF-I. Despite the lack of information on the physiology and endocrine regulation of fish adipose tissue function, these results suggest that fish adipose tissue could indeed be regulated by insulin and IGF-I. Thus, this study is a first step towards our understanding of the role of insulin and IGF-I in regulating adipose tissue function in fish. Subsequent work should aim to determine the direct effects of insulin and IGF-I on lipid metabolism in fish adipose tissue.

We thank Mr Antonino Clemente at the Piscifactoria de Bagà (Departament de Medi Natural, Generalitat de Catalunya) for providing the trout and facilities to conduct some of the sampling and for his generous assistance. We also thank Dr Simon MacKenzie for critically reading the manuscript. We are very grateful to Chiron Corp. (Emeryville, CA, USA) for providing recombinant human IGF-I. This study was supported by grants from the European Union (FAIR CT95-0174), DGICY Spain (AGF98-0325) and CIRIT (1998SGR-00037) to J.G., and from DGICY Spain (PB97-0902) to I.N.