- Email: [email protected]
Social stress affects circulating melatonin levels in rainbow trout
GENERAL AND COMPARATIVE
ENDOCRINOLOGY General and Comparative Endocrinology 136 (2004) 322–327 www.elsevier.com/locate/ygcen
Social stress affects circulating melatonin levels in rainbow trout Earl T. Larson,a,b Svante Winberg,c Ian Mayer,d,e Olivier Lepage,c Cliff H. Summers,f and Øyvind Øverlif,g,* a Department of Neuroscience, Uppsala University, Box 593, SE-75124 Uppsala, Sweden Departments of Biology and Psychology, Northeastern University, Boston, MA 02115, USA Department of Comparative Physiology, Evolutionary Biology Centre, Norbyv€agen 18A, Uppsala University, SE-75236 Uppsala, Sweden d Department of Zoology, Stockholm University, SE-10691 Stockholm, Sweden e Department of Fisheries and Marine Biology, University of Bergen, Box 7800, N-5020 Bergen, Norway f Department of Biology and Neuroscience Group, University of South Dakota, 414 East Clark Street, Vermillion SD 57069, USA g Division of General Physiology, Department of Biology, University of Oslo, P.O. Box 1051, N-0316 Oslo, Norway b
c
Received 9 September 2003; revised 25 November 2003; accepted 13 January 2004
Abstract In salmonid fishes there are indications that socially subordinate individuals avoid competition with larger, dominant fish by adjusting daily feeding and activity cycles. As in other vertebrates, the pineal organ and its hormone melatonin act as synchronizers of daily rhythms to the external light/dark cycle in salmonids. Social defeat may act as a potent stressor; inducing elevated glucocorticoid secretion and a general behavioral inhibition. Here, we show that social stress also affects circulating melatonin levels in rainbow trout, a species known to display strong dominance hierarchies both in the wild and under captive rearing. Subordinate individuals had significantly higher nighttime melatonin levels than dominant fish or controls. There was no effect of social rank on the much lower melatonin levels observed in animals sampled during the day. Correlations between circulating glucocorticoids and melatonin depended on circadian cycles as well as social context. This study suggests that altered melatonin production contributes to the physiological and behavioral profile of subordinate animals. Social status, and other determinants of the stress level of experimental animals, therefore should be taken into consideration as potential factors influencing the results from in vivo research on this hormone. Ó 2004 Elsevier Inc. All rights reserved. Keywords: Aggression; Behavior; Hierarchy; Melatonin; Pineal function; Stress response
1. Introduction In a range of vertebrate species socially subordinate individuals often exhibit behavioral inhibition, such as suppressed aggressive behavior, reduced feeding, and decreased locomotor activity and exploration (Albonetti and Farabollini, 1994; Fuchs and Fl€ ugge, 2002; H€ oglund et al., 2001; Larson and Summers, 2001; Meerlo et al., 1997; Øverli et al., 1998; Summers and Greenberg, 1994; Winberg and Nilsson, 1993a). Salmonid fishes are widely used species in research and aquaculture, and in groups of these fish there are indications that unsuccessful com* Corresponding author. Fax: 47-22854664. E-mail address: [email protected] (Ø. Øverli).
0016-6480/$ - see front matter Ó 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.ygcen.2004.01.005
petitors might limit interactions with dominant individuals by adjusting their diurnal activity pattern, rather than by a general behavioral inhibition (Alan€ ar€ a et al., 2001; Chen et al., 2002; Kadri et al., 1997). The pineal organ and its hormone melatonin are regarded as synchronizers of daily rhythms to the external light/dark cycle in most vertebrates (Cassone, 1998; Falc on, 1999), including teleost fishes (Ekstr€ om and Meissl, 1997; Okimoto and Stetson, 1999). A limited number of reports indicate that social stress leads to altered melatonin release in mammals (Fuchs and Schumacher, 1990; Heinzeller et al., 1988). The current study was designed to investigate whether the formation of dominance–subordination relationships also affects circulating melatonin levels in rainbow trout, a species
E.T. Larson et al. / General and Comparative Endocrinology 136 (2004) 322–327
known to display strong dominance hierarchies both in the wild (Jenkins, 1969) and when reared in the laboratory (Øverli et al., 1999; Sloman et al., 2001; Winberg and Lepage, 1998).
2. Materials and methods Experimental fish were juvenile (1.5 year old) rainbow trout, Oncorhynchus mykiss, weighing between 125 and 160 g. Prior to the experiment fish were maintained inside in a 1 m3 holding tank at a density of approximately 0.03 kg L1 for more than 1 month. The holding tank was continuously supplied with aerated Uppsala tap water at 8–12 °C. The light/dark regime was adjusted to conditions at 51° N latitude. In the holding tanks and throughout the experiments, the fish were fed with commercial trout pellet food (EWOS ST40) at 1% body weight. All procedures were approved by the Uppsala University Animal Care Committee. The experiment was conducted in glass aquaria (100 50 50 cm) continuously supplied with aerated tap water (0.9 L min1 , 8–12 °C). Experiments were performed in April and May, and light/dark cycles were set at 12:12 with the light period beginning at 06:00 and ending at 18:00 h. Light was provided by 2 30 W Lumilux daylight fluorescent tubes placed 10 cm above the water surface. Each aquarium was divided into four compartments by removable PVC walls. Rainbow trout were transferred from the holding tank, weighed, and marked for individual recognition of pair members by small cuts (<3 mm) in the tail fin. The fish were then held in isolation in observation aquaria for 1 week. During this acclimation period, fish were fed pellet food once a day between 13:00 and 14:00 h. After 1 week of acclimation, paired interactions were started by removing walls separating isolated fish. As in previous experiments (Øverli et al., 1999), distinct dominance–subordination relationships were formed rapidly after pairing. After initial fights lasting between 1 and 45 min, aggressive behavior was unidirectional in all pairs with only one fish (the dominant individual) performing any aggressive acts. Fish were then allowed to interact in pairs for 30 h prior to sampling. This time span slightly exceeds the 24 h shown to be sufficient for a complete reversal of the stress response induced by fighting in dominant fish, while subordinate pair members were chronically stressed after 24 h of paired interactions in a previous experiment (Øverli et al., 1999). At either 12:00 or 24:00, dominant and subordinate fish were netted simultaneously, anesthetized in 0.5 g mL1 tricaine methanesulfonate (MS 222), and 1 mL blood was drawn from the caudal sinus within 3 min after netting. Capture of fish at night was conducted by briefly using a darkroom safe red light in order to minimize disturbance. Plasma was separated by centrifugation and kept at
323
)80 °C until analysis. In all, seven pairs of fish were sampled at night, and eight pairs were sampled during the day. In addition, nine isolated and undisturbed fish were sampled as controls at each time point. A small number of samples (no more than one in any given experimental group) were lost during analysis. Plasma levels of melatonin were measured by a direct radioimmunoassay (RIA) method as previously described by Mayer (2000), based on that of Fraser et al. (1983). Melatonin was measured in duplicate 250 lL plasma samples. The lower limit of detection was 8 pg melatonin mL1 plasma. An inhibition curve obtained from a serial dilution (1:2) of salmon plasma collected during darkness showed good parallelism with the standard curve. The intra- and interassay coefficient of variance of a pooled plasma sample containing ca. 200 pg mL1 was 6.9 and 10.4%, respectively. Cortisol analysis was performed directly on rainbow trout plasma (duplicate 20 lL plasma samples) without extraction, using a validated RIA modified from Olsen et al. (1992) as described by Winberg and Lepage (1998). The lower detection limit was approximately 0.5 ng mL1 and intra- and interassay coefficient of variance of a pooled plasma sample was 2.1 and 7.1%, respectively. Group differences in hormone concentrations were analyzed statistically by ANOVA followed by least significant difference (LSD) post hoc test. Day- and nighttime data were analyzed separately, for two reasons: 1. Circadian cycles in melatonin and cortisol levels have been published numerous times in the past. 2. The differences in hormone levels were so large that normality and homogeneity of variance criteria of parametric statistics were not fulfilled even after extensive transformations and removal of outliers in the complete data set. Data on cortisol concentrations were logtransformed prior to analysis, while melatonin concentrations fulfilled normality and homogeneity of variance criteria without transformation when day- and nighttime data were analyzed separately. Dallal and Wilkinson approximation to LillieforsÕs method were used as normality test and homogeneity of variance was checked by LeveneÕs test. Spearman correlation was used to test relationships between plasma cortisol and melatonin within experimental groups. Statistical software was GraphPad Instat (GraphPad Software, San Diego, CA) and Statistica for Windows (StatSoft, Tulsa, OK).
3. Results and discussion In agreement with previous data (Øverli et al., 1999), social interaction led to significantly increased plasma cortisol concentrations in subordinate pair members sampled during the day (ANOVA, Fð2;21Þ ¼ 13:33, p < 0:001), as compared to dominant fish as well as isolated controls (Fig. 1A). There was no significant effect of
324
E.T. Larson et al. / General and Comparative Endocrinology 136 (2004) 322–327
Fig. 1. Plasma cortisol and melatonin in control, dominant, and subordinate rainbow trout sampled during the day (A) and night (B) (means + SEM, group n between 7 and 9). Experimental groups assigned different letters displayed statistically significant differences (p < 0:05, ANOVA followed by least significant difference post hoc test).
social rank on plasma cortisol in fish sampled at night (ANOVA, Fð2;18Þ ¼ 1:43, p ¼ 0:24) (Fig. 1B). Most likely, fish did not behave aggressively in darkness, since neither dominant nor subordinate fish were observed to move or even try to evade dip-netting during sampling at night. In fish sampled during the day, plasma melatonin levels were similarly low in all experimental groups (ANOVA, Fð2;21Þ ¼ 0:17, p ¼ 0:85) (Fig. 1A). Subordinate individuals, however, showed significantly higher nighttime melatonin levels than dominant fish and controls (ANOVA, Fð2;19Þ ¼ 3:96, p ¼ 0:03) (Fig. 1B), while there was no difference in circulating melatonin levels between dominant individuals and controls at either of the two time points. Significant positive correlations between plasma cortisol and melatonin concentrations were seen in dominant and subordinate individuals sampled during the day (Fig. 2A), but not in fish sampled during the night (Fig. 2B). No relationship between melatonin and cortisol was seen in unstressed controls at any time point. These data demonstrate that social subordination is associated with increased nighttime levels of circulating melatonin in rainbow trout. The mechanism by which stressful experiences affect changes in circulating melatonin levels, however, remains unclear. Over a 10-day experimental period, Fuchs and Schumacher (1990) observed a sustained increase in levels of the melatonin metabolite 6-sulfatoxymelatonin (aMT6s) in morning urine of subordinate male tree shrews (Tupaia belangeri). Thus, it seems unlikely that the effect of social stress on
plasma melatonin levels is due to decreased melatonin metabolism. In an experiment with male gerbils (Meriones unguiculatus), Heinzeller et al. (1988) found that a single short-term (3 min) subjugation in a resident–intruder test was sufficient to induce an immediate doubling of pineal melatonin content, further supporting the suggestion that social interactions affect melatonin synthesis in the pineal gland. The rapidity of this response, and the fact that it was dependent on intact sympathetic innervation (Heinzeller et al., 1988), suggests that catecholamine modulation may have mediated this effect of social stress in gerbils. In mammals, pineal synthesis of melatonin is indirectly controlled by photoperiod via the hypothalamic circadian oscillator, the supraschiasmatic nucleus (Borjigin et al., 1999; Klein and Moore, 1979; LeSauter and Silver, 1998). In salmonid fish, on the other hand, melatonin synthesis in the pineal gland is directly controlled by ambient illumination (reviewed by Ekstr€ om and Meissl, 1997). Norepinephrine, the b-adrenergic receptor agonist isoproterenol, and dopamine were all found to be without effect on melatonin synthesis in the rainbow trout pineal (Meissl et al., 1996), indicating that catecholaminergic mechanisms are not of great importance for signal transduction in the pineal organ of salmonid fishes. Melatonin is synthesized from serotonin (5-hydroxytryptamine, 5-HT) by N-acetylation by serotonin N -acetyltransferase (arylalkylamine N -acetyltransferase, AANAT) and O-methylation by hydroxyindole-Omethyltransferase. One mechanism by which stress may
E.T. Larson et al. / General and Comparative Endocrinology 136 (2004) 322–327
325
Fig. 2. The relationship between plasma cortisol and melatonin concentrations control, dominant, and subordinate rainbow trout sampled during the day (A) and night (B). Pearson correlation r2 and p values are displayed.
modulate melatonin synthesis is by affecting availability of the serotonin precursor, the essential amino acid L -tryptophan (TRP). It is quite likely that the rate of serotonin synthesis in the pineal organ is affected by TRP availability, like it is in 5-HT producing neurons in the brain (Aldegunde et al., 2000; Lepage et al., 2002; Winberg et al., 2001), and acute social stress has been shown to increase brain TRP levels and brain serotonergic activity in another salmonid species, the Arctic char (Salvelinus alpinus) (Winberg and Nilsson, 1993b). In rats, the effect of TRP on melatonin synthesis appears to depend on circadian rhythms. Oral administration of TRP during the day caused a fourfold increase in serum melatonin compared to that of saline treated rats (Yaga et al., 1993), while nighttime tryptophan loading led to a substantial reduction in AANAT activity and melatonin levels (Reiter et al., 1990). Time- and context-dependent effects of TRP manipulation on pineal function was also
observed by Guchhait and Haldar (2001). Investigating the effects of long-term exogenous TRP treatment in a nocturnal bird, the Indian spotted owlet (Athene brama), these authors observed increased pineal gland weights and elevated plasma melatonin concentrations in birds receiving TRP during the breeding phase. The opposite effect was seen during the reproductively quiescent phase, at which time similar amounts of TRP significantly decreased the plasma melatonin levels in the same species (Guchhait and Haldar, 2001). Another possible mechanism mediating effects of stress on melatonin production is an influence of circulating glucocorticoid hormones on pineal function. Recently, Benyassi et al. (2001) showed that glucocorticoid receptors are expressed in the trout pineal organ, and the synthetic glucocorticoid dexamethazone inhibited AANAT activity after 6 h of incubation in darkness, an effect which was reversed by the glucocorticoid
326
E.T. Larson et al. / General and Comparative Endocrinology 136 (2004) 322–327
antagonist RU 486. These findings, along with observations in other species (Barriga et al., 2002) demonstrate that glucocorticoids can modulate melatonin production. The results of the current study suggest that correlations between circulating glucocorticoids and melatonin depend on circadian rhythms as well as social context (cf. Fig. 2). The nocturnal increase in melatonin production depends on protein synthesis in trout as well as other teleosts (Falc on et al., 1998), and a delayed effect of the sustained daytime increase in cortisol on nighttime melatonin synthesis can not be excluded. In the experiment by Heinzeller et al. (1988), animals subjected to repeated social subordination demonstrated a temporal delay in the rise of both AANAT activity and melatonin in comparison with untreated controls, and sampling at a single time point does not give information about the true shape of the plasma melatonin profile during the dark period. Thus, depending on the time the measurement is taken, results may indicate an increase, decrease, or no effect of stress or hormonal treatment on melatonin levels. Further research is needed to determine the causal link between social stress and melatonin production, and the possibility that melatonin affects glucocorticoid production should also be considered (Yamada, 1990). In agreement with a limited number of reports in mammals (Fuchs and Schumacher, 1990; Heinzeller et al., 1988), the present results suggest that altered melatonin production may be added to the list of physiological reactions to social stress. Altered melatonin production may be part of the mechanism behind previously observed changes in the diurnal activity rhythm of subordinate fish (Alan€ ar€ a et al., 2001; Kadri et al., 1997), which may serve to limit competition with dominant fish. Social status and other determinants of the stress level of experimental animals should therefore be taken into consideration as potential factors influencing the results from in vivo research on this hormone.
Acknowledgments We thank Ken Renner, Wayne Korzan, and Jamie Scholl for valuable help with the writing process. ETL would like to dedicate this paper to the late Milt Stetson. This study was supported financially by the Swedish Agricultural Research Council (SJFR), the FACIAS foundation, and an NIH COBRE grant to the University of South Dakota (NIH P 20 RR15567).
References Alan€ ar€ a, A., Burns, M.D., Metcalfe, N.B., 2001. Intraspecific resource partitioning in brown trout: the temporal distribution of foraging is determined by social rank. J. Anim. Ecol. 70, 980–986.
Albonetti, M.E., Farabollini, F., 1994. Social stress by repeated defeat: effects on social behaviour and emotionality. Behav. Brain Res. 62, 187–193. Aldegunde, M., Soengas, J.L., Rozas, G., 2000. Acute effects of L tryptophan on tryptophan hydroxylation rate in brain regions (hypothalamus and medulla) of rainbow trout (Oncorhynchus mykiss). J. Exp. Zool. 286, 131–135. Barriga, C., Marchena, J.M., Lea, R.W., Harvey, S., Rodrıguez, A.B., 2002. Effect of stress and dexamethasone treatment on circadian rhythms of melatonin and corticosterone in ring dove (Streptopelia risoria). Mol. Cell. Biochem. 232, 27–31. Benyassi, A., Schwartz, C., Ducouret, B., Falc on, J., 2001. Glucocorticoid receptors and serotonin N -acetyltransferase activity in the fish pineal organ. Neuroreport 12, 889–892. Borjigin, J., Li, X., Snyder, S.H., 1999. The pineal gland and melatonin: molecular and pharmacologic regulation. Annu. Rev. Pharmacol. Toxicol. 39, 53–65. Cassone, V.M., 1998. MelatoninÕs role in vertebrate circadian rhythms. Chronobiol. Int. 15, 457–473. Chen, W.M., Naruse, M., Tabata, M., 2002. The effect of social interactions on circadian self-feeding rhythms in rainbow trout Oncorhynchus mykiss Walbaum. Physiol. Behav. 76, 281–287. Ekstr€ om, P., Meissl, H., 1997. The pineal organ of teleost fishes. Rev. Fish Biol. Fish. 7, 199–284. Falc on, J., 1999. Cellular circadian clocks in the pineal. Prog. Neurobiol. 58, 121–162. Falc on, J., Barraud, S., Thibault, C., Begay, V., 1998. Inhibitors of messenger RNA and protein synthesis affect differently serotonin arylalkylamine N -acetyltransferase activity in clock-controlled and non clock-controlled fish pineal. Brain Res. 797, 109–117. Fraser, S., Cowen, P., Franklin, M., Franey, C., Arendt, C., 1983. Direct radioimmunoassay for melatonin in plasma. J. Clin. Chem. 29, 396–397. Fuchs, E., Schumacher, M., 1990. Psychosocial stress affects pineal function in the tree shrew (Tupaia belangeri). Physiol. Behav. 47, 713–717. Fuchs, E., Fl€ ugge, G., 2002. Social stress in tree shrews: effects on physiology, brain function, and behavior of subordinate individuals. Pharmacol. Biochem. Behav. 73, 247–258. Guchhait, P., Haldar, C., 2001. A reproductive phase-dependent effect of dietary L -tryptophan on pineal gland and gonad of a nocturnal bird, Indian spotted owlet Athene brama. Acta Biol. Hung. 52, 1–7. Heinzeller, T., Joshi, B.N., N€ urnberger, F., Reiter, R.J., 1988. Effects of aggressive encounters on pineal melatonin formation in male gerbils (Meriones unguiculatus, Cricetidae). J. Comp. Physiol. 164, 91–94. H€ oglund, E., Kolm, N., Winberg, S., 2001. Stress-induced changes in brain serotonergic activity, plasma cortisol and aggressive behavior in Arctic charr (Salvelinus alpinus) is counteracted by L -DOPA. Physiol. Behav. 74, 381–389. Jenkins, T.M., 1969. Social structure, position choice and microdistribution of two trout species (Salmo trutta and Salmo gairdneri) resident in mountain streams. Anim. Behav. Monogr. 2, 57–123. Kadri, S., Metcalfe, N.B., Huntingford, F.A., Thorpe, J.E., 1997. Daily feeding rhythms in Atlantic salmon II: size-related variation in feeding patterns of post-smolts under constant environmental conditions. J. Fish Biol. 50, 273–279. Klein, D.C., Moore, R.Y., 1979. Pineal N -acetyltransferase and hydroxyindole-O-methyltransferase: control by the retinohypothalamic tract and the suprachiasmatic nucleus. Brain Res. 174, 245– 262. Larson, E.T., Summers, C.H., 2001. Serotonin reverses dominant social status. Behav. Brain Res. 121, 95–102. Lepage, O., Tottmar, O., Winberg, S., 2002. Elevated dietary intake of L -tryptophan counteracts the stress-induced elevation of plasma cortisol in rainbow trout (Oncorhynchus mykiss). J. Exp. Biol. 205, 3679–3687.
E.T. Larson et al. / General and Comparative Endocrinology 136 (2004) 322–327 LeSauter, J., Silver, R., 1998. Output signals of the SCN. Chronobiol. Int. 15, 535–550. Mayer, I., 2000. Effect of long-term pinealectomy on growth and precocious maturation in Atlantic salmon, Salmo salar parr. Aquat. Living Res. 13, 139–144. Meerlo, P., Overkamp, G.J., Koolhaas, J.M., 1997. Behavioural and physiological consequences of a single social defeat in Roman highand low-avoidance rats. Psychoneuroendocrinology 22, 155–168. ~ ez, J., Korf, H.W., 1996. Regulation of Meissl, H., Kroeber, S., Yan melatonin production and intracellular calcium concentrations in the trout pineal organ. Cell Tissue Res. 286, 315–323. Olsen, Y.A., Falk, K., Reite, O.B., 1992. Cortisol and lactate levels of Atlantic salmon Salmo salar developing infectious anaemia (ISA). Dis. Aquat. Organ. 14, 99–104. Okimoto, D.K., Stetson, M.H., 1999. Presence of an intrapineal circadian oscillator in the teleostean family poeciliidae. Gen. Comp. Endocrinol. 114, 304–312. Øverli, Ø., Winberg, S., Damsg ard, B., Jobling, M., 1998. Food intake and spontaneous swimming activity in Arctic charr (Salvelinus alpinus): role of brain serotonergic activity and social interactions. Can. J. Zool. 76, 1366–1370. Øverli, Ø., Harris, C.A., Winberg, S., 1999. Short-term effects of fights for social dominance and the establishment of dominant–subordinate relationship on brain monoamines and cortisol in rainbow trout. Brain Behav. Evol. 54, 263–275. Reiter, R.J., King, T.S., Steinlechner, S., Steger, R.W., Richardson, B.A., 1990. Tryptophan administration inhibits nocturnal N acetyltransferase activity and melatonin content in the rat pineal
327
gland. Evidence that serotonin modulates melatonin production via a receptor-mediated mechanism. Neuroendocrinology 52, 291– 296. Sloman, K.A., Metcalfe, N.B., Taylor, A.C., Gilmour, K.M., 2001. Plasma cortisol concentrations before and after social stress in rainbow trout and brown trout. Physiol. Biochem. Zool. 74, 383– 389. Summers, C.H., Greenberg, N., 1994. Somatic correlates of adrenergic activity during aggression in the lizard, Anolis carolinensis. Horm. Behav. 28, 29–40. Yaga, K., Reiter, R.J., Richardson, B.A., 1993. Tryptophan loading increases daytime serum melatonin levels in intact and pinealectomized rats. Life Sci. 52, 1231–1238. Yamada, K., 1990. Effects of melatonin on adrenal function in male rats. Res. Commun. Chem. Pathol. Pharmacol. 69, 241–244. Winberg, S., Nilsson, G.E., 1993a. Roles of brain monoamine neurotransmitters in agonistic behaviour and stress reactions, with particular reference to fish. Comp. Biochem. Physiol. 106C, 597– 614. Winberg, S., Nilsson, G.E., 1993b. Time course of changes in brain serotonergic activity and brain tryptophan levels in dominant and subordinate juvenile Arctic charr. J. Exp. Biol. 179, 181–195. Winberg, S., Lepage, O., 1998. Elevation of brain 5-HT activity, POMC expression, and plasma cortisol in socially subordinate rainbow trout. Am. J. Physiol. R 43, 645–654. Winberg, S., Øverli, Ø., Lepage, O., 2001. Suppression of aggression in rainbow trout (Oncorhynchus mykiss) by dietary L -tryptophan. J. Exp. Biol. 204, 3867–3876.
ENDOCRINOLOGY General and Comparative Endocrinology 136 (2004) 322–327 www.elsevier.com/locate/ygcen
Social stress affects circulating melatonin levels in rainbow trout Earl T. Larson,a,b Svante Winberg,c Ian Mayer,d,e Olivier Lepage,c Cliff H. Summers,f and Øyvind Øverlif,g,* a Department of Neuroscience, Uppsala University, Box 593, SE-75124 Uppsala, Sweden Departments of Biology and Psychology, Northeastern University, Boston, MA 02115, USA Department of Comparative Physiology, Evolutionary Biology Centre, Norbyv€agen 18A, Uppsala University, SE-75236 Uppsala, Sweden d Department of Zoology, Stockholm University, SE-10691 Stockholm, Sweden e Department of Fisheries and Marine Biology, University of Bergen, Box 7800, N-5020 Bergen, Norway f Department of Biology and Neuroscience Group, University of South Dakota, 414 East Clark Street, Vermillion SD 57069, USA g Division of General Physiology, Department of Biology, University of Oslo, P.O. Box 1051, N-0316 Oslo, Norway b
c
Received 9 September 2003; revised 25 November 2003; accepted 13 January 2004
Abstract In salmonid fishes there are indications that socially subordinate individuals avoid competition with larger, dominant fish by adjusting daily feeding and activity cycles. As in other vertebrates, the pineal organ and its hormone melatonin act as synchronizers of daily rhythms to the external light/dark cycle in salmonids. Social defeat may act as a potent stressor; inducing elevated glucocorticoid secretion and a general behavioral inhibition. Here, we show that social stress also affects circulating melatonin levels in rainbow trout, a species known to display strong dominance hierarchies both in the wild and under captive rearing. Subordinate individuals had significantly higher nighttime melatonin levels than dominant fish or controls. There was no effect of social rank on the much lower melatonin levels observed in animals sampled during the day. Correlations between circulating glucocorticoids and melatonin depended on circadian cycles as well as social context. This study suggests that altered melatonin production contributes to the physiological and behavioral profile of subordinate animals. Social status, and other determinants of the stress level of experimental animals, therefore should be taken into consideration as potential factors influencing the results from in vivo research on this hormone. Ó 2004 Elsevier Inc. All rights reserved. Keywords: Aggression; Behavior; Hierarchy; Melatonin; Pineal function; Stress response
1. Introduction In a range of vertebrate species socially subordinate individuals often exhibit behavioral inhibition, such as suppressed aggressive behavior, reduced feeding, and decreased locomotor activity and exploration (Albonetti and Farabollini, 1994; Fuchs and Fl€ ugge, 2002; H€ oglund et al., 2001; Larson and Summers, 2001; Meerlo et al., 1997; Øverli et al., 1998; Summers and Greenberg, 1994; Winberg and Nilsson, 1993a). Salmonid fishes are widely used species in research and aquaculture, and in groups of these fish there are indications that unsuccessful com* Corresponding author. Fax: 47-22854664. E-mail address: [email protected] (Ø. Øverli).
0016-6480/$ - see front matter Ó 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.ygcen.2004.01.005
petitors might limit interactions with dominant individuals by adjusting their diurnal activity pattern, rather than by a general behavioral inhibition (Alan€ ar€ a et al., 2001; Chen et al., 2002; Kadri et al., 1997). The pineal organ and its hormone melatonin are regarded as synchronizers of daily rhythms to the external light/dark cycle in most vertebrates (Cassone, 1998; Falc on, 1999), including teleost fishes (Ekstr€ om and Meissl, 1997; Okimoto and Stetson, 1999). A limited number of reports indicate that social stress leads to altered melatonin release in mammals (Fuchs and Schumacher, 1990; Heinzeller et al., 1988). The current study was designed to investigate whether the formation of dominance–subordination relationships also affects circulating melatonin levels in rainbow trout, a species
E.T. Larson et al. / General and Comparative Endocrinology 136 (2004) 322–327
known to display strong dominance hierarchies both in the wild (Jenkins, 1969) and when reared in the laboratory (Øverli et al., 1999; Sloman et al., 2001; Winberg and Lepage, 1998).
2. Materials and methods Experimental fish were juvenile (1.5 year old) rainbow trout, Oncorhynchus mykiss, weighing between 125 and 160 g. Prior to the experiment fish were maintained inside in a 1 m3 holding tank at a density of approximately 0.03 kg L1 for more than 1 month. The holding tank was continuously supplied with aerated Uppsala tap water at 8–12 °C. The light/dark regime was adjusted to conditions at 51° N latitude. In the holding tanks and throughout the experiments, the fish were fed with commercial trout pellet food (EWOS ST40) at 1% body weight. All procedures were approved by the Uppsala University Animal Care Committee. The experiment was conducted in glass aquaria (100 50 50 cm) continuously supplied with aerated tap water (0.9 L min1 , 8–12 °C). Experiments were performed in April and May, and light/dark cycles were set at 12:12 with the light period beginning at 06:00 and ending at 18:00 h. Light was provided by 2 30 W Lumilux daylight fluorescent tubes placed 10 cm above the water surface. Each aquarium was divided into four compartments by removable PVC walls. Rainbow trout were transferred from the holding tank, weighed, and marked for individual recognition of pair members by small cuts (<3 mm) in the tail fin. The fish were then held in isolation in observation aquaria for 1 week. During this acclimation period, fish were fed pellet food once a day between 13:00 and 14:00 h. After 1 week of acclimation, paired interactions were started by removing walls separating isolated fish. As in previous experiments (Øverli et al., 1999), distinct dominance–subordination relationships were formed rapidly after pairing. After initial fights lasting between 1 and 45 min, aggressive behavior was unidirectional in all pairs with only one fish (the dominant individual) performing any aggressive acts. Fish were then allowed to interact in pairs for 30 h prior to sampling. This time span slightly exceeds the 24 h shown to be sufficient for a complete reversal of the stress response induced by fighting in dominant fish, while subordinate pair members were chronically stressed after 24 h of paired interactions in a previous experiment (Øverli et al., 1999). At either 12:00 or 24:00, dominant and subordinate fish were netted simultaneously, anesthetized in 0.5 g mL1 tricaine methanesulfonate (MS 222), and 1 mL blood was drawn from the caudal sinus within 3 min after netting. Capture of fish at night was conducted by briefly using a darkroom safe red light in order to minimize disturbance. Plasma was separated by centrifugation and kept at
323
)80 °C until analysis. In all, seven pairs of fish were sampled at night, and eight pairs were sampled during the day. In addition, nine isolated and undisturbed fish were sampled as controls at each time point. A small number of samples (no more than one in any given experimental group) were lost during analysis. Plasma levels of melatonin were measured by a direct radioimmunoassay (RIA) method as previously described by Mayer (2000), based on that of Fraser et al. (1983). Melatonin was measured in duplicate 250 lL plasma samples. The lower limit of detection was 8 pg melatonin mL1 plasma. An inhibition curve obtained from a serial dilution (1:2) of salmon plasma collected during darkness showed good parallelism with the standard curve. The intra- and interassay coefficient of variance of a pooled plasma sample containing ca. 200 pg mL1 was 6.9 and 10.4%, respectively. Cortisol analysis was performed directly on rainbow trout plasma (duplicate 20 lL plasma samples) without extraction, using a validated RIA modified from Olsen et al. (1992) as described by Winberg and Lepage (1998). The lower detection limit was approximately 0.5 ng mL1 and intra- and interassay coefficient of variance of a pooled plasma sample was 2.1 and 7.1%, respectively. Group differences in hormone concentrations were analyzed statistically by ANOVA followed by least significant difference (LSD) post hoc test. Day- and nighttime data were analyzed separately, for two reasons: 1. Circadian cycles in melatonin and cortisol levels have been published numerous times in the past. 2. The differences in hormone levels were so large that normality and homogeneity of variance criteria of parametric statistics were not fulfilled even after extensive transformations and removal of outliers in the complete data set. Data on cortisol concentrations were logtransformed prior to analysis, while melatonin concentrations fulfilled normality and homogeneity of variance criteria without transformation when day- and nighttime data were analyzed separately. Dallal and Wilkinson approximation to LillieforsÕs method were used as normality test and homogeneity of variance was checked by LeveneÕs test. Spearman correlation was used to test relationships between plasma cortisol and melatonin within experimental groups. Statistical software was GraphPad Instat (GraphPad Software, San Diego, CA) and Statistica for Windows (StatSoft, Tulsa, OK).
3. Results and discussion In agreement with previous data (Øverli et al., 1999), social interaction led to significantly increased plasma cortisol concentrations in subordinate pair members sampled during the day (ANOVA, Fð2;21Þ ¼ 13:33, p < 0:001), as compared to dominant fish as well as isolated controls (Fig. 1A). There was no significant effect of
324
E.T. Larson et al. / General and Comparative Endocrinology 136 (2004) 322–327
Fig. 1. Plasma cortisol and melatonin in control, dominant, and subordinate rainbow trout sampled during the day (A) and night (B) (means + SEM, group n between 7 and 9). Experimental groups assigned different letters displayed statistically significant differences (p < 0:05, ANOVA followed by least significant difference post hoc test).
social rank on plasma cortisol in fish sampled at night (ANOVA, Fð2;18Þ ¼ 1:43, p ¼ 0:24) (Fig. 1B). Most likely, fish did not behave aggressively in darkness, since neither dominant nor subordinate fish were observed to move or even try to evade dip-netting during sampling at night. In fish sampled during the day, plasma melatonin levels were similarly low in all experimental groups (ANOVA, Fð2;21Þ ¼ 0:17, p ¼ 0:85) (Fig. 1A). Subordinate individuals, however, showed significantly higher nighttime melatonin levels than dominant fish and controls (ANOVA, Fð2;19Þ ¼ 3:96, p ¼ 0:03) (Fig. 1B), while there was no difference in circulating melatonin levels between dominant individuals and controls at either of the two time points. Significant positive correlations between plasma cortisol and melatonin concentrations were seen in dominant and subordinate individuals sampled during the day (Fig. 2A), but not in fish sampled during the night (Fig. 2B). No relationship between melatonin and cortisol was seen in unstressed controls at any time point. These data demonstrate that social subordination is associated with increased nighttime levels of circulating melatonin in rainbow trout. The mechanism by which stressful experiences affect changes in circulating melatonin levels, however, remains unclear. Over a 10-day experimental period, Fuchs and Schumacher (1990) observed a sustained increase in levels of the melatonin metabolite 6-sulfatoxymelatonin (aMT6s) in morning urine of subordinate male tree shrews (Tupaia belangeri). Thus, it seems unlikely that the effect of social stress on
plasma melatonin levels is due to decreased melatonin metabolism. In an experiment with male gerbils (Meriones unguiculatus), Heinzeller et al. (1988) found that a single short-term (3 min) subjugation in a resident–intruder test was sufficient to induce an immediate doubling of pineal melatonin content, further supporting the suggestion that social interactions affect melatonin synthesis in the pineal gland. The rapidity of this response, and the fact that it was dependent on intact sympathetic innervation (Heinzeller et al., 1988), suggests that catecholamine modulation may have mediated this effect of social stress in gerbils. In mammals, pineal synthesis of melatonin is indirectly controlled by photoperiod via the hypothalamic circadian oscillator, the supraschiasmatic nucleus (Borjigin et al., 1999; Klein and Moore, 1979; LeSauter and Silver, 1998). In salmonid fish, on the other hand, melatonin synthesis in the pineal gland is directly controlled by ambient illumination (reviewed by Ekstr€ om and Meissl, 1997). Norepinephrine, the b-adrenergic receptor agonist isoproterenol, and dopamine were all found to be without effect on melatonin synthesis in the rainbow trout pineal (Meissl et al., 1996), indicating that catecholaminergic mechanisms are not of great importance for signal transduction in the pineal organ of salmonid fishes. Melatonin is synthesized from serotonin (5-hydroxytryptamine, 5-HT) by N-acetylation by serotonin N -acetyltransferase (arylalkylamine N -acetyltransferase, AANAT) and O-methylation by hydroxyindole-Omethyltransferase. One mechanism by which stress may
E.T. Larson et al. / General and Comparative Endocrinology 136 (2004) 322–327
325
Fig. 2. The relationship between plasma cortisol and melatonin concentrations control, dominant, and subordinate rainbow trout sampled during the day (A) and night (B). Pearson correlation r2 and p values are displayed.
modulate melatonin synthesis is by affecting availability of the serotonin precursor, the essential amino acid L -tryptophan (TRP). It is quite likely that the rate of serotonin synthesis in the pineal organ is affected by TRP availability, like it is in 5-HT producing neurons in the brain (Aldegunde et al., 2000; Lepage et al., 2002; Winberg et al., 2001), and acute social stress has been shown to increase brain TRP levels and brain serotonergic activity in another salmonid species, the Arctic char (Salvelinus alpinus) (Winberg and Nilsson, 1993b). In rats, the effect of TRP on melatonin synthesis appears to depend on circadian rhythms. Oral administration of TRP during the day caused a fourfold increase in serum melatonin compared to that of saline treated rats (Yaga et al., 1993), while nighttime tryptophan loading led to a substantial reduction in AANAT activity and melatonin levels (Reiter et al., 1990). Time- and context-dependent effects of TRP manipulation on pineal function was also
observed by Guchhait and Haldar (2001). Investigating the effects of long-term exogenous TRP treatment in a nocturnal bird, the Indian spotted owlet (Athene brama), these authors observed increased pineal gland weights and elevated plasma melatonin concentrations in birds receiving TRP during the breeding phase. The opposite effect was seen during the reproductively quiescent phase, at which time similar amounts of TRP significantly decreased the plasma melatonin levels in the same species (Guchhait and Haldar, 2001). Another possible mechanism mediating effects of stress on melatonin production is an influence of circulating glucocorticoid hormones on pineal function. Recently, Benyassi et al. (2001) showed that glucocorticoid receptors are expressed in the trout pineal organ, and the synthetic glucocorticoid dexamethazone inhibited AANAT activity after 6 h of incubation in darkness, an effect which was reversed by the glucocorticoid
326
E.T. Larson et al. / General and Comparative Endocrinology 136 (2004) 322–327
antagonist RU 486. These findings, along with observations in other species (Barriga et al., 2002) demonstrate that glucocorticoids can modulate melatonin production. The results of the current study suggest that correlations between circulating glucocorticoids and melatonin depend on circadian rhythms as well as social context (cf. Fig. 2). The nocturnal increase in melatonin production depends on protein synthesis in trout as well as other teleosts (Falc on et al., 1998), and a delayed effect of the sustained daytime increase in cortisol on nighttime melatonin synthesis can not be excluded. In the experiment by Heinzeller et al. (1988), animals subjected to repeated social subordination demonstrated a temporal delay in the rise of both AANAT activity and melatonin in comparison with untreated controls, and sampling at a single time point does not give information about the true shape of the plasma melatonin profile during the dark period. Thus, depending on the time the measurement is taken, results may indicate an increase, decrease, or no effect of stress or hormonal treatment on melatonin levels. Further research is needed to determine the causal link between social stress and melatonin production, and the possibility that melatonin affects glucocorticoid production should also be considered (Yamada, 1990). In agreement with a limited number of reports in mammals (Fuchs and Schumacher, 1990; Heinzeller et al., 1988), the present results suggest that altered melatonin production may be added to the list of physiological reactions to social stress. Altered melatonin production may be part of the mechanism behind previously observed changes in the diurnal activity rhythm of subordinate fish (Alan€ ar€ a et al., 2001; Kadri et al., 1997), which may serve to limit competition with dominant fish. Social status and other determinants of the stress level of experimental animals should therefore be taken into consideration as potential factors influencing the results from in vivo research on this hormone.
Acknowledgments We thank Ken Renner, Wayne Korzan, and Jamie Scholl for valuable help with the writing process. ETL would like to dedicate this paper to the late Milt Stetson. This study was supported financially by the Swedish Agricultural Research Council (SJFR), the FACIAS foundation, and an NIH COBRE grant to the University of South Dakota (NIH P 20 RR15567).
References Alan€ ar€ a, A., Burns, M.D., Metcalfe, N.B., 2001. Intraspecific resource partitioning in brown trout: the temporal distribution of foraging is determined by social rank. J. Anim. Ecol. 70, 980–986.
Albonetti, M.E., Farabollini, F., 1994. Social stress by repeated defeat: effects on social behaviour and emotionality. Behav. Brain Res. 62, 187–193. Aldegunde, M., Soengas, J.L., Rozas, G., 2000. Acute effects of L tryptophan on tryptophan hydroxylation rate in brain regions (hypothalamus and medulla) of rainbow trout (Oncorhynchus mykiss). J. Exp. Zool. 286, 131–135. Barriga, C., Marchena, J.M., Lea, R.W., Harvey, S., Rodrıguez, A.B., 2002. Effect of stress and dexamethasone treatment on circadian rhythms of melatonin and corticosterone in ring dove (Streptopelia risoria). Mol. Cell. Biochem. 232, 27–31. Benyassi, A., Schwartz, C., Ducouret, B., Falc on, J., 2001. Glucocorticoid receptors and serotonin N -acetyltransferase activity in the fish pineal organ. Neuroreport 12, 889–892. Borjigin, J., Li, X., Snyder, S.H., 1999. The pineal gland and melatonin: molecular and pharmacologic regulation. Annu. Rev. Pharmacol. Toxicol. 39, 53–65. Cassone, V.M., 1998. MelatoninÕs role in vertebrate circadian rhythms. Chronobiol. Int. 15, 457–473. Chen, W.M., Naruse, M., Tabata, M., 2002. The effect of social interactions on circadian self-feeding rhythms in rainbow trout Oncorhynchus mykiss Walbaum. Physiol. Behav. 76, 281–287. Ekstr€ om, P., Meissl, H., 1997. The pineal organ of teleost fishes. Rev. Fish Biol. Fish. 7, 199–284. Falc on, J., 1999. Cellular circadian clocks in the pineal. Prog. Neurobiol. 58, 121–162. Falc on, J., Barraud, S., Thibault, C., Begay, V., 1998. Inhibitors of messenger RNA and protein synthesis affect differently serotonin arylalkylamine N -acetyltransferase activity in clock-controlled and non clock-controlled fish pineal. Brain Res. 797, 109–117. Fraser, S., Cowen, P., Franklin, M., Franey, C., Arendt, C., 1983. Direct radioimmunoassay for melatonin in plasma. J. Clin. Chem. 29, 396–397. Fuchs, E., Schumacher, M., 1990. Psychosocial stress affects pineal function in the tree shrew (Tupaia belangeri). Physiol. Behav. 47, 713–717. Fuchs, E., Fl€ ugge, G., 2002. Social stress in tree shrews: effects on physiology, brain function, and behavior of subordinate individuals. Pharmacol. Biochem. Behav. 73, 247–258. Guchhait, P., Haldar, C., 2001. A reproductive phase-dependent effect of dietary L -tryptophan on pineal gland and gonad of a nocturnal bird, Indian spotted owlet Athene brama. Acta Biol. Hung. 52, 1–7. Heinzeller, T., Joshi, B.N., N€ urnberger, F., Reiter, R.J., 1988. Effects of aggressive encounters on pineal melatonin formation in male gerbils (Meriones unguiculatus, Cricetidae). J. Comp. Physiol. 164, 91–94. H€ oglund, E., Kolm, N., Winberg, S., 2001. Stress-induced changes in brain serotonergic activity, plasma cortisol and aggressive behavior in Arctic charr (Salvelinus alpinus) is counteracted by L -DOPA. Physiol. Behav. 74, 381–389. Jenkins, T.M., 1969. Social structure, position choice and microdistribution of two trout species (Salmo trutta and Salmo gairdneri) resident in mountain streams. Anim. Behav. Monogr. 2, 57–123. Kadri, S., Metcalfe, N.B., Huntingford, F.A., Thorpe, J.E., 1997. Daily feeding rhythms in Atlantic salmon II: size-related variation in feeding patterns of post-smolts under constant environmental conditions. J. Fish Biol. 50, 273–279. Klein, D.C., Moore, R.Y., 1979. Pineal N -acetyltransferase and hydroxyindole-O-methyltransferase: control by the retinohypothalamic tract and the suprachiasmatic nucleus. Brain Res. 174, 245– 262. Larson, E.T., Summers, C.H., 2001. Serotonin reverses dominant social status. Behav. Brain Res. 121, 95–102. Lepage, O., Tottmar, O., Winberg, S., 2002. Elevated dietary intake of L -tryptophan counteracts the stress-induced elevation of plasma cortisol in rainbow trout (Oncorhynchus mykiss). J. Exp. Biol. 205, 3679–3687.
E.T. Larson et al. / General and Comparative Endocrinology 136 (2004) 322–327 LeSauter, J., Silver, R., 1998. Output signals of the SCN. Chronobiol. Int. 15, 535–550. Mayer, I., 2000. Effect of long-term pinealectomy on growth and precocious maturation in Atlantic salmon, Salmo salar parr. Aquat. Living Res. 13, 139–144. Meerlo, P., Overkamp, G.J., Koolhaas, J.M., 1997. Behavioural and physiological consequences of a single social defeat in Roman highand low-avoidance rats. Psychoneuroendocrinology 22, 155–168. ~ ez, J., Korf, H.W., 1996. Regulation of Meissl, H., Kroeber, S., Yan melatonin production and intracellular calcium concentrations in the trout pineal organ. Cell Tissue Res. 286, 315–323. Olsen, Y.A., Falk, K., Reite, O.B., 1992. Cortisol and lactate levels of Atlantic salmon Salmo salar developing infectious anaemia (ISA). Dis. Aquat. Organ. 14, 99–104. Okimoto, D.K., Stetson, M.H., 1999. Presence of an intrapineal circadian oscillator in the teleostean family poeciliidae. Gen. Comp. Endocrinol. 114, 304–312. Øverli, Ø., Winberg, S., Damsg ard, B., Jobling, M., 1998. Food intake and spontaneous swimming activity in Arctic charr (Salvelinus alpinus): role of brain serotonergic activity and social interactions. Can. J. Zool. 76, 1366–1370. Øverli, Ø., Harris, C.A., Winberg, S., 1999. Short-term effects of fights for social dominance and the establishment of dominant–subordinate relationship on brain monoamines and cortisol in rainbow trout. Brain Behav. Evol. 54, 263–275. Reiter, R.J., King, T.S., Steinlechner, S., Steger, R.W., Richardson, B.A., 1990. Tryptophan administration inhibits nocturnal N acetyltransferase activity and melatonin content in the rat pineal
327
gland. Evidence that serotonin modulates melatonin production via a receptor-mediated mechanism. Neuroendocrinology 52, 291– 296. Sloman, K.A., Metcalfe, N.B., Taylor, A.C., Gilmour, K.M., 2001. Plasma cortisol concentrations before and after social stress in rainbow trout and brown trout. Physiol. Biochem. Zool. 74, 383– 389. Summers, C.H., Greenberg, N., 1994. Somatic correlates of adrenergic activity during aggression in the lizard, Anolis carolinensis. Horm. Behav. 28, 29–40. Yaga, K., Reiter, R.J., Richardson, B.A., 1993. Tryptophan loading increases daytime serum melatonin levels in intact and pinealectomized rats. Life Sci. 52, 1231–1238. Yamada, K., 1990. Effects of melatonin on adrenal function in male rats. Res. Commun. Chem. Pathol. Pharmacol. 69, 241–244. Winberg, S., Nilsson, G.E., 1993a. Roles of brain monoamine neurotransmitters in agonistic behaviour and stress reactions, with particular reference to fish. Comp. Biochem. Physiol. 106C, 597– 614. Winberg, S., Nilsson, G.E., 1993b. Time course of changes in brain serotonergic activity and brain tryptophan levels in dominant and subordinate juvenile Arctic charr. J. Exp. Biol. 179, 181–195. Winberg, S., Lepage, O., 1998. Elevation of brain 5-HT activity, POMC expression, and plasma cortisol in socially subordinate rainbow trout. Am. J. Physiol. R 43, 645–654. Winberg, S., Øverli, Ø., Lepage, O., 2001. Suppression of aggression in rainbow trout (Oncorhynchus mykiss) by dietary L -tryptophan. J. Exp. Biol. 204, 3867–3876.