The natural life cycle of many fish species often includes long periods of low winter temperatures and restricted feeding opportunities or prey availability that lead to a depletion of energy reserves and a reduction in growth rate( Reference Jobling 1 ). The restoration of adequate nutrition or favourable environmental conditions results in rapid weight gain and compensatory growth relative to continuously fed controls. Compensatory growth occurs at various stages in salmonid fish and is an important adaptation that allows fish to remain on target in a fluctuating and unpredictable environment( Reference Nicieza and Metcalfe 2 – Reference Ali, Nicieza and Wootton 5 ). From a practical point of view, the compensatory strategy is of great interest to the aquaculture industry because feeding programmes can be designed to improve food conversion and growth rates, thereby minimising production costs( Reference Jobling 1 ).
Periods of nutrient restriction are associated with changes in metabolism to provide cellular energy via catabolic processes( Reference Peragón, Barroso and García-Salguero 3 ). In carnivorous fish, nutrient restriction enhances the release of amino acids from muscle fibres which are used by hepatocytes as the main gluconeogenic precursors( Reference Ballantyne 6 ). During refeeding and compensatory growth an accelerated turnover takes place resulting in an increased protein synthesis:degradation ratio( Reference Peragón, Barroso and García-Salguero 3 , Reference Carter, Houlihan and Kiessling 7 ). In salmon, the main mechanism underlying compensatory growth after a nutrient-restriction period is an increase of feed intake rates( Reference Nicieza and Metcalfe 2 ). Food contains ligands for retinoic acid receptors, PPAR, vitamin D receptors and other nuclear transcription factor receptors and can directly affect signal transduction pathways( Reference Kaput, Klein and Reyes 8 , Reference Kaput and Rodriguez 9 ). Branched-chain amino acids, particularly leucine, have a major role in stimulating protein synthesis( Reference Kimball and Jefferson 10 ). These phenomena seem to be the result of endocrine alterations, although in most species it is difficult to detect differences in the endocrine status between animals undergoing compensatory growth and control animals( Reference Hornick, Van Eenaeme and Gérard 11 ). In fish, recent molecular tools enable us to gain deeper insight into how growth responses are regulated by dietary factors( Reference Bower and Johnston 12 , Reference Panserat and Kaushik 13 ). In Atlantic salmon( Reference Bower, Taylor and Johnston 14 ) and rainbow trout( Reference Rescan, Montfort and Ralliere 15 ) a genomic approach was used to identify nutritionally regulated genes involved in muscle growth and revealed a complex response. The principal groups of up-regulated transcripts post-refeeding were genes involved in transcription, ribosomal biogenesis, translation, chaperone activity, ATP production, cell division and muscle remodelling. These genes possibly play a role in the stimulation of myogenesis during the transition from a catabolic to anabolic state in skeletal muscle.
Several in vivo and in vitro studies evaluated the impact of feeding protocols on muscle growth. Using a candidate gene approach, it was observed that switching onto fast growth induced by a fasting–refeeding schedule involves the local up-regulation of several components of the insulin-like growth factor (IGF) system, a major hormone axis regulating the cellular dynamics of muscle growth( Reference Duan 16 – Reference Bower, Li and Taylor 19 ). The role of plasma IGF-I during compensatory growth is not clear and must be explained in connection with changes of its binding proteins, which act on the phosphoinositide-3-kinase/Akt/mammalian-target-of-rapamycin (PI3K/Akt/mTOR) pathway( Reference Montserrat, Sánchez-Gurmaches and García de la Serrana 20 , Reference Seiliez, Gabillard and Skiba-Cassy 21 ). The autocrine IGF-II transcription required for skeletal myocyte differentiation is regulated by mTOR and the availability of amino acids( Reference Erbay, Park and Nuzzi 22 , Reference Codina, García de la Serrana and Sánchez-Gurmaches 23 ). Thus the mTOR–IGF axis provides a molecular link between nutritional levels and protein synthesis leading to muscle fibre hypertrophy. In salmonids, fasting decreased the expression and plasma levels of IGF, and up-regulated IGF-I binding, whereas the plasma level of growth hormone was shown to increase( Reference Gabillard, Kamangar and Montserrat 24 , Reference Montserrat, Gabillard and Capilla 25 ). Switching to fast growth in Atlantic salmon muscle involved the up-regulation of IGF-I, IGF binding protein (IGFBP)-5·2 and IGFBP-4( Reference Bower, Li and Taylor 19 ), whereas 1 d of refeeding completely restored plasma growth hormone levels in rainbow trout( Reference Gabillard, Kamangar and Montserrat 24 ). IGF-I induced the activation of the PI3K/Akt pathway, which causes an increase in protein translation via activation of p70S6K and inhibition of 4E-BP (also known as PHAS-1)( Reference Gabillard, Kamangar and Montserrat 24 ). In cell lines, it was recently shown that in addition to its hypertrophic effects, Akt can dominantly inhibit induction of the atrophy genes the muscle-specific E3 ubiquitin ligases, MuRF1 (muscle RING finger protein 1) and MAFbx/atrogin-1 (muscle atrophy F box), by phosphorylating and thereby inhibiting the function of the forkhead box O (FOXO) family of transcription factors( Reference Glass 26 , Reference Cleveland and Weber 27 ).
The genetic networks mobilised in muscle following recovery from fasting are likely to be dependent on nutrient availability. The experimental protocols employed to investigate compensatory growth often involve prolonged fasting followed by continuous refeeding with transcript abundance monitored over several days or weeks( Reference Bower, Taylor and Johnston 14 , Reference Chauvigné, Gabillard and Weil 17 , Reference Terova, Rimoldi and Chini 28 , Reference Hagen, Fernandes and Solberg 29 ). In contrast, a single satiating meal design allows the evolution and decay of transcriptional responses to nutrient input to be studied with relatively high temporal resolution. Using this approach the aim of the present study on Atlantic salmon was to investigate the transcript abundance of muscle growth-related genes to changes in nutrient supply. The expression of genes involved in myogenesis (muscle regulatory factors (MRF) and paired box protein 7 (Pax7)), growth signalling (the IGF system), myofibrillar protein degradation and synthesis pathways (the PI3K/AKT/mTOR pathway and the muscle-specific E3 ubiquitin ligases) and metabolic genes (CrebA) shown to be critical modulators of fish myotomal muscle growth were analysed using quantitative real-time PCR (qPCR).
Materials and methods
Experimental conditions and sampling
All experiments were approved by the Animal Welfare Committee of the University of St Andrews and fish were humanely killed following Schedule 1 of the Animals (Scientific Procedures) Act 1986 (Home Office Code of Practice, H. M. Stationery Office: London, January 1997).
Two homogeneous groups of Atlantic salmon (Salmo salar L.) juveniles (average body weight 70 g) were reared in duplicate tanks (500 litres; thirty fish per tank). Each tank was supplied with fresh water with an average water temperature of 12°C and dissolved oxygen >80 %. Fish were exposed to an artificial photoperiod of 12 h light–12 h dark (08.00 to 20.00 hours) and provided a commercial salmon feed (EWOS Innovation) during 3 weeks (1·5 % body weight). Fish were then fasted for 1 week, and fed a single meal distributed to all fish to visual satiation.
Sampling of fish occurred at − 12 and 0 h (before the meal), and at 1, 3, 6, 12, 24, 48 and 96 h (following the single meal), with seven fish sampled at each time point and individual mass and fork lengths measured. The intestine and stomach content was determined for each fish and photographs were taken. Fast muscle was then dissected from the dorsal myotome and snap-frozen in liquid N2. Samples were kept at − 80°C until analysed.
RNA extraction and cDNA synthesis
Total RNA was extracted by the addition of 100 mg of salmon muscle to Lysing matrix D (Qbiogene) with 1 ml TRI reagent (Sigma) and homogenised for 40 s using a Fast Prep instrument (Qbiogene). Total RNA was quantified based on absorbance at 260 nm using a NanoDrop spectrophotometer (ThermoFisher Scientific) and its integrity was confirmed by agarose gel electrophoresis on a 1·2 % gel (w/v) after 3–5 min denaturation at 65°C. Genomic DNA contamination was removed and first-strand cDNA was synthesised from 1 μg total RNA with the QuantiTect reverse transcription kit (Qiagen) following the manufacturer's instructions.
Quantitative real-time PCR
The following procedures were performed in order to comply with the Minimum Information for Publication of Quantitative Real-Time PCR experiments MIQE guidelines( Reference Bustin, Benes and Garson 30 ).
Several muscle growth-related genes (Table 1) (IGF-I, IGF-II, IGF-IRa, IGF-IRb, IFG-IIR, IGFBP-2·1, IGFBP-4, IGFBP-5·1, IGFBP-5·2, IGFBP-6, IGFBP-rP1, Myogenin, MyoD1a, MyoD1b, MyoD1c, myf5 (myogenic factor 5), MRF4 (also known as myogenic factor 6; myf6), the E3 ubiquitin ligases MuRF1 and MAFbx/atrogin-1, fibroblast growth factor 2 (FGF2), CrebA, MEF2A, Pax7 and five reference genes, HPRT1 (hypoxanthine phosphoribosyltransferase 1), elongation factor 1-α (EF1-α), 60S ribosomal protein L13 (rpl13), 40S ribosomal protein S29 (rps29), beta-actin and RNA polymerase II) were selected, and primers were designed to have a melting temperature (Tm) of 60°C as previously reported by Bower et al. ( Reference Bower, Taylor and Johnston 14 , Reference Bower, Li and Taylor 19 ). Quantification of gene expression was performed by qPCR using a Stratagene MX3005P QPCR system (Stratagene) with SYBR Green chemistry (Brilliant II SYBR green, Stratagene). The cDNA were diluted 80 × before using them as templates for the qPCR reactions. Each qPCR reaction mixture contained 7·5 μl 2 × Brilliant II SYBR green master mix, 6 μl cDNA (80-fold dilution), 500 nm each primer and RNAse-free water to a final volume of 15 μl. Amplification was performed in duplicate in ninety-six-well plates with the following thermal cycling conditions: initial activation 95°C for 10 min, followed by forty cycles of 15 s at 95°C, 30 s at 60°C and 30 s at 72°C. Dissociation analysis of the PCR products was performed by running a gradient from 60 to 95°C to confirm the presence of a single PCR product. The PCR amplification efficiency of each primer pair was calculated using LinregPCR 2009 (http://LinRegPCR.HFRC.nl)( Reference Ruijter, Ramakers and Hoogaars 31 ).
Table 1 Summary of parameters for quantitative real-time PCR (qPCR)
E, PCR efficiency; f, forward; r, reverse; IGF-I, insulin-like growth factor I; IGF-II, insulin-like growth factor II; IGF-IRa, insulin-like growth factor I receptor a; IGF-IRb, insulin-like growth factor I receptor b; IGF-IIR, insulin-like growth factor II receptor; IGFBP-2·1, insulin-like growth factor binding protein 2 paralogue 1; IGFBP-4, insulin-like growth factor binding protein 4; IGFBP-5·1, insulin-like growth factor binding protein 5 paralogue 1; IGFBP-5·2, insulin-like growth factor binding protein 2 paralogue 2; IGFBP-6, insulin-like growth factor binding protein 6; IGFBP-rP1, insulin-like growth factor binding protein-related protein 1; CrebA, cyclic AMP response element binding protein; FGF2, fibroblast growth factor 2; MAFbx, muscle atrophy F box; MEF2A, myocyte enhancer factor 2A; MuRF1, muscle RING finger protein 1; myf5, myogenic factor 5; MyoD1a, myoblast determination factor 1a; MyoD1b, myoblast determination factor 1b; MyoD1c, myoblast determination factor 1c; Pax7, paired box protein 7; MRF4, myogenic factor 6; EF1-α, elongation factor 1-α; Rpl13, 60S ribosomal protein L13; Rps29, 40S ribosomal protein S29; RNA pol II, RNA polymerase II; HPRT1, hypoxanthine phosphoribosyltransferase 1.
Data analysis
Evaluation of expression stability of several potential housekeeping genes including elongation factor 1-α (EF1-α), 60S ribosomal protein L13, rpl13, 40S ribosomal protein S13 and S29 (rps13 and rps29), beta-actin and RNA polymerase II was done using the statistical application called geNorm (http://medgen.ugent.be/~jvdesomp/genorm/)( Reference Vandesompele, De Preter and Pattyn 32 ). Analysis revealed both rpl13 and RNA polymerase II as the most stable genes in this experiment. Hence, the reference gene rpl13 was used for normalisation of qPCR data and QGene was used for normalisation and calculation of relative expression data( Reference Muller, Janovjak and Miserez 33 ).
All data were tested for normality and homogeneity of variances by Kolmogorov–Smirnov and Bartlett tests, and then submitted to a one-way ANOVA using STATISTICA software (version 10; StatSoft, Inc.). When the data did not meet the normality and/or equal variance requirements, a Kruskal–Wallis one-way ANOVA on ranks was performed instead. When these tests showed significance, individual means were compared using Tukey's honestly significant difference test or Dunn's test. Correlation of gene expression was analysed by the Spearman rank order correlation test. Hierarchical clustering was performed using Cluster3 software( Reference de Hoon, Imoto and Nolan 34 ).
Results
Fish mass and length ranged between 90–117 g and 20–23 cm, respectively, and were similar among sampling points (P < 0·05). The gut content of the salmon increased significantly 1 h after distributing the meal and remained high for 6 h (Fig. 1). These results confirm that all sampled fish had ingested food, and the intestine was fully evacuated between 48 and 96 h after the meal.
Fig. 1 (A) Gut content (g) of fish sampled before (T–24 and T0) and several hours (T1–T96) after feeding a single meal to satiation. Values are means (seven fish per sample), with their standard errors represented by vertical bars. a,b,cMean values with unlike letters were significantly different (P < 0·05). (B) Photographs of gut and intestine, which were dissected to ensure the presence of food in the stomach and to follow their fullness during the course of the single meal experiment. Scale bar = 2 cm.
An up-regulation of several components of the IGF system was observed immediately after the meal. IGF-I transcripts increased significantly 1 h after the single meal (P < 0·05), decreasing thereafter (Fig. 2). Expression of IGF-II was significantly reduced in response to feeding (P < 0·05) with significant down-regulation observed at 3 and 6 h after the single meal, returning to initial values after 12 h.
Fig. 2 Normalised expression of the insulin-like growth factor (IGF) system genes (IGF-I (A), IGF-II (B), IGF-IRa (C) and IGF-Irb (D)) in the fast myotomal muscle at different times postprandially. Values are means (seven fish per sample), with their standard errors represented by vertical bars. a,b,c Mean values with unlike letters were significantly different (P < 0·05).
Expression levels of IGF-IRb (Fig. 2) and IGF-IIR did not vary significantly in response to feeding (P>0·05), but IGF-IRa was minimal 12 h after the single meal. Several IGFBP were detected in fast muscle of Atlantic salmon (Fig. 3). IGFBP-related protein 1 (IGFBP-rP1), IFGBP-2·1 and IGFBP-5·2 expression did not vary significantly with the single meal. IGFBP-4 and IGFBP-5·1 had increased expression 1 h after the meal administration, but then decreased. IGFBP-6 expression decreased 12 h after the meal, reaching maximal expression levels 48 h postprandially. IGFBP-6 was significantly correlated with IGF-II (P < 0·05). A positive correlation was found between IGF-I, IGFBP-rP1 and IGFBP-4 expression.
Fig. 3 Expression profiles of insulin-like growth factor binding protein (IGFBP) transcripts (IGFBP-rP1 (A), IGFBP-2·1 (B), IGFBP-4 (C) and IGFBP-6 (D)) in the fast myotomal muscle at different times postprandially. Values are means (seven fish per sample), with their standard errors represented by vertical bars. a,b Mean values with unlike letters were significantly different (P < 0·05).
Expression of MRF revealed that MRF4 and myf5 expression were significantly correlated with the gut content of the fish (P < 0·05). These two genes were up-regulated until 3 h after the single meal, returning to initial values 6 h after feeding (Fig. 4), and their expression clustered together when compared with other genes (Fig. 5). The three paralogues of MyoD1 responded differently to the meal distribution. MyoD1a showed no variation following the meal, whereas MyoD1b and MyoD1c peaked 1 h after feeding a single meal, although without statistical significance in the case of MyoD1b. MyoD1c mRNA expression correlated with gut content (P < 0·05) and clustered together with IGF-I (Fig. 5). Myogenin expression was not significantly affected by feeding, although a slight decrease was observed after feeding.
Fig. 4 Normalised expression of muscle-regulatory factors (MRF) (MyoD1c (A), Myogenin (B), myogenic factor 5 (myf5) (C) and MRF4 (D)) and the atrophy genes (MAFbx/atrogin-1 (E) and muscle RING finger protein 1 (MuRF1) (F)) at different times postprandially. Values are means (seven fish per sample), with their standard errors represented by vertical bars. a,b,c Mean values with unlike letters were significantly different (P < 0·05).
Fig. 5 Heat map summary and unsupervised hierarchical clustering analysis of the genes in muscle according to the similarity in their expression across different postprandial times ( − 24 to 96 h). Yellow and blue shading represents the maximal (Max) and lowest (Min) transcript abundance, respectively. IGFBP-2·1, insulin-like growth factor binding protein 2 paralogue 1; IGFBP-5·1, insulin-like growth factor binding protein 5 paralogue 1; IGF-II, insulin-like growth factor II; MEF2A, myocyte enhancer factor 2A; FGF2, fibroblast growth factor 2; CrebA, cyclic AMP response element binding protein; IGFBP-6, insulin-like growth factor binding protein 6; IGF-IIR, insulin-like growth factor II receptor; IGF-IRa, insulin-like growth factor I receptor a; MuRF1, muscle RING finger protein 1; IGF-IRb, insulin-like growth factor I receptor b; IGFBP-rP1, insulin-like growth factor binding protein-related protein 1; IGFBP-4, insulin-like growth factor binding protein 4; Pax7, paired box protein 7; MyoD1b, myoblast determination factor 1b; IGF-I, insulin-like growth factor I; MyoD1c, myoblast determination factor 1c; MRF4, myogenic factor 6; myf5, myogenic factor 5; myog, myogenin; MAFbx, muscle atrophy F box; IGFBP-5·2, insulin-like growth factor binding protein 2 paralogue 2; MyoD1a, myoblast determination factor 1a.
The ubiquitin ligase MAFbx/atrogin-1 mRNA levels were significantly down-regulated until 12 h after feeding, returning to levels observed in fasted fish by 24 h (Fig. 4). MuRF1 expression was initially up-regulated at 1 h, before being down-regulated at 3 and 12 h postprandially. MAFbx/atrogin-1 expression was positively correlated (R 0·66) with IGF-II. FGF2, CrebA and Pax7 expression was not significantly affected by the single meal distribution.
Discussion
Nutrient availability is amongst the most important environmental variable altering muscle growth. The genetic network that is mobilised in the stimulation of myogenesis during the transition from a catabolic to anabolic state in skeletal muscle has not been exhaustively described, but seems to be a nutrient-sensing pathway( Reference Bower and Johnston 12 , Reference Panserat and Kaushik 13 , Reference Rescan, Montfort and Ralliere 15 , Reference Rehfeldt, Te Pas and Wimmers 35 ). The present study was designed to explore the postprandial regulation of growth-related genes shortly after feeding a single meal.
The effects of fasting and subsequent continuous refeeding protocols following transcript abundance over time have been studied in several fish species such as Atlantic salmon( Reference Bower, Taylor and Johnston 14 , Reference Duan and Plisetskaya 36 ), rainbow trout( Reference Chauvigné, Gabillard and Weil 17 , Reference Montserrat, Gabillard and Capilla 25 ), Atlantic halibut( Reference Hagen, Fernandes and Solberg 29 ), sea bass( Reference Terova, Rimoldi and Chini 28 ) and seabream( Reference Montserrat, Gomez-Requeni and Bellini 37 ). However, the only study describing early transcriptional changes during the postprandial period was recently performed in zebrafish( Reference Amaral and Johnston 38 ). The present results indicated that a single meal affects the expression of several growth-related genes in Atlantic salmon juveniles shortly after ingestion, confirming data in zebrafish( Reference Amaral and Johnston 38 ). A 3·5-fold increase of IGF-I mRNA expression was observed 1 h after refeeding, indicating a fast response to nutrient availability. An increased expression of IGF-I in fast skeletal muscle was registered 3–4 d after refeeding in Atlantic salmon( Reference Bower, Li and Taylor 19 ) and rainbow trout( Reference Chauvigné, Gabillard and Weil 17 ), and after 1 week in sea bass( Reference Terova, Rimoldi and Chini 28 ), although no earlier point was analysed. Duan, Plisetskaya( Reference Duan and Plisetskaya 36 ) also reported a significant increase in hepatic IGF-I mRNA levels in salmon after refeeding, suggesting an endocrine/autocrine/paracrine growth stimulation of myotomal muscle induced by food intake. In myogenic cell culture, Atlantic salmon IGF-I mRNA levels increase in response to IGF and amino acid stimulation. Seiliez et al. ( Reference Seiliez, Gabillard and Skiba-Cassy 21 ) showed that in rainbow trout insulin levels peak 0·5 h after feeding, whereas increased amino acid levels were observed after 2·5 h. Based on this, the increased IGF-I mRNA levels we observed 1 h postprandially are likely to be in response to hormonal stimulation. IGF-I regulates many anabolic pathways in skeletal muscle, stimulating cell proliferation and differentiation( Reference Castillo, Codina and Martinez 39 ) and myocyte hypertrophy( Reference Florini, Ewton and Coolican 40 ) through the subsequent activation of the PI3K/AKT/mTOR pathway and prevention of atrophy mediators( Reference Stitt, Drujan and Clarke 41 ).
IGF-I exerts its effects on cells through binding to IGF-IR. The expression of IGF-IRa was minimal 12 h after refeeding, whereas IGF-IRb was not significantly affected by feeding, which is consistent with previous results in Atlantic salmon( Reference Bower, Li and Taylor 19 ). In trout, IGF-IRa was shown to be maximal in fasted fish and declined after refeeding, but no changes were reported in IGF-IRb ( Reference Chauvigné, Gabillard and Weil 17 ). Montserrat et al. ( Reference Montserrat, Gabillard and Capilla 25 ) pointed towards a different regulation of these two genes by nutritional status, with isoform a responding to refeeding and isoform b responding to fasting. During periods of nutrient restriction, sensitivity to IGF-I seems to be increased in muscle by increasing the abundance of IGF-IRa.
The expression of several IGFBP in response to a single meal followed distinct patterns. IGFBP-rP1, IGFBP-2·1 and IGFBP-5·2 expression did not vary significantly up to 96 h after feeding the single meal. Although IGFBP-rP1 did not seem to be modulated by feeding, its expression was positively correlated with that of IGF-I, confirming previous results in Atlantic salmon( Reference Bower, Li and Taylor 19 ) and trout( Reference Gabillard, Kamangar and Montserrat 24 ). In a previous study with salmon starved for 22 d and refed to satiation thereafter, IGFBP-2·1 was significantly down-regulated from 14 d onwards, which was attributed to an increased availability of IGF-I to the IGF-I receptor( Reference Bower, Li and Taylor 19 ). IGFBP-4 showed maximal expression 1 h after feeding and like IGFBP-rP1, was significantly correlated with IGF-I expression, suggesting a coordinated regulation of these genes towards resumption of myogenesis soon after refeeding. IGFBP-6 was significantly correlated with IGF-II, but not with IGF-I. As mammalian IGFBP-6 has a 10- to 100-fold higher affinity for IGF-II than IGF-I( Reference Bach 42 ), the present results suggest a role for Atlantic salmon IGFBP-6 in IGF-II regulation. Amaral & Johnston( Reference Amaral and Johnston 38 ) pointed to lineage-specific differences in IFGBP function and regulation among teleosts, suggested by the apparent lack of IGFBP-4 in zebrafish, so caution is needed when comparing results between different fish species.
The role of IGF-II in fish metabolism is unclear, but it seems to be implicated in the autocrine/paracrine regulation of growth( Reference Codina, García de la Serrana and Sánchez-Gurmaches 23 , Reference Montserrat, Gabillard and Capilla 25 ). In the present study, IGF-II mRNA expression showed a dramatic decrease in muscle of refed salmon until 6 h after feeding, but levels were restored 12 h postprandially. Likewise, Bower et al. ( Reference Bower, Li and Taylor 19 ) reported a significant decrease of IGF-II expression in a time-dependent fashion after at least 7 d continuous feeding. Hevrøy et al. ( Reference Hevrøy, El-Mowafi and Taylor 43 ) reported an up-regulation of IGF-II in the muscle of fish fed high lysine levels, suggesting a role as an anabolic stimulatory agent. In juvenile rainbow trout, IGF-II mRNA levels in myotomal muscle tissue increased 34 d after refeeding( Reference Chauvigné, Gabillard and Weil 17 ), but in a different study Montserrat et al. ( Reference Montserrat, Gabillard and Capilla 25 ) could not observe any effect of fasting or refeeding on IGF-II mRNA expression. These distinct responses could be due to distinct developmental stages and/or nutritional status of the fish.
The expression of several MRF has been reported to be modulated by the nutritional status of the fish. The three paralogues of MyoD1 responded differently to the meal distribution. MyoD1a showed no variation following the meal, whereas MyoD1b and MyoD1c peaked 1 h after feeding a single meal. The up-regulation of MyoD1c following the single meal was positively correlated with IGF-I. No clear change in either MyoD isoform could be observed during compensatory growth in trout( Reference Montserrat, Gabillard and Capilla 25 ), but amino acid withdrawal led to a down-regulation of both MyoD1b and MyoD1c in salmon myogenic cells culture, and increased levels of Pax7 mRNA, suggesting that serum and amino acid withdrawal leads to cell cycle arrest and the production of quiescent cells( Reference Bower and Johnston 44 ).
MRF4 and myf5 showed a dramatic and simultaneous up-regulation immediately after feeding a single meal. MRF4 and myf5 are closely linked genes that clustered together (Fig. 5). Myf5 is the first MRF to be expressed during embryonic development and is considered a specification factor that determines the muscular lineage, whereas MRF4 functions later and can be considered as both a specification and a differentiation factor( Reference Chen and Tsai 45 ). An early peak in myf5 expression was correlated with increased MyoD1b transcript abundance during the maturation of an Atlantic salmon primary myogenic cell culture( Reference Bower and Johnston 44 ), and so it is interesting that we see myf5 and MyoD1b clustering together (Fig. 5). To our knowledge there are few studies reporting the nutritional modulation of myf5 and MRF4 in fish, but the present results point towards a possible nutritional regulation of muscle fibre number. Lower growth due to high dietary lipid levels in Senegalese sole was associated with reduced expression of muscle MRF4, but not myf5 ( Reference Campos, Valente and Borges 46 ).
In the present study no modulation of Pax7 or FGF2 was observed in refeeding, although Chauvigné et al. ( Reference Chauvigné, Gabillard and Weil 17 ) indicated FGF2 was a critical modulator of trout myotomal muscle growth 4 d after refeeding. Expression of the muscle-specific gene Myogenin decreased soon after a single meal but without statistical significance (P>0·05). Similarly, Bower et al. ( Reference Bower, Li and Taylor 19 ) described a Myogenin reduction in response to feeding. In rainbow trout, Myogenin mRNA was unchanged 4 d after refeeding, but increased significantly after 12 d( Reference Chauvigné, Gabillard and Weil 17 ). However, increased Myogenin expression was reported in trout during feed restriction, suggesting a role in muscle maintenance( Reference Montserrat, Gabillard and Capilla 25 ).
The genes regulated by atrophy include the E3 ubiquitin ligase MAFbx/atrogin-1 and MuRF1 that are up-regulated during catabolism and atrophy and down-regulated during fibre hypertrophy. Fasted individuals showed increased expression of MAFbx/atrogin-1 and MuRF1 and both genes were strongly down-regulated after feeding in Atlantic salmon( Reference Bower, Taylor and Johnston 14 , Reference Bower, de la serrana and Johnston 47 ) and zebrafish( Reference Amaral and Johnston 38 ). The present results showed that a single meal was capable of promoting MAFbx/atrogin-1 depression, though its effect on MuRF1 was less clear. The increased expression observed 1 h postprandially for MuRF1 at first glance is puzzling. However, it is noteworthy that during the fasting period, fish were inactive, but became active once feed was distributed in the tank. This sudden increase in activity could lead to depletions in muscle glucose reserves, leading to metabolic stress in the muscle, and MuRF1 is known to regulate responses to metabolic stress in muscle of mice( Reference Hirner, Krohne and Schuster 48 ). MAFbx/atrogin-1 is regulated by both IGF signalling and amino acid availability( Reference Bower, de la serrana and Johnston 47 ), and the expression patterns we observed following a single meal is consistent with this. In trout, circulating amino acids remained high from 2·5 to 12 h following feeding and returned to those of a fasted state by 24 h( Reference Seiliez, Gabillard and Skiba-Cassy 21 ) and the expression profile we observed for MAFbx/atrogin-1 is inversely proportional to this. The decrease in MAFbx/atrogin-1 expression (within 1 h) suggests that hormonal stimulation of the AKT/mTOR pathway via endocrine signalling or through local production of IGF-I is responsible for this rapid transcriptional response that could result in increased protein synthesis. Recent findings demonstrated that MAFbx/atrogin-1 contributed to muscle wasting by down-regulating protein synthesis whereas MuRF1 is mostly involved in the breakdown of myofibrillar proteins( Reference Attaix and Baracos 49 ).
In conclusion, the present results show that the transcription of several growth-related genes in the fast skeletal muscle of Atlantic salmon responds quickly to a single meal. In muscle, our observations indicate that refeeding induced a coordinated regulation of several genes involved in a strong resumption of myogenesis with feeding. IGF-I, MyoD1c, MRF4 and myf5 transcripts in muscle were sharply up-regulated in response to refeeding, being promising candidate genes involved in a cellular-level signalling system that regulates fish myotomal muscle growth. It is also suggested that local production of IGF-I within the muscle might suppress catabolic pathways depressing MAFbx/atrogin-1.
Acknowledgements
We thank Dr Vera Vieira-Johnston, Attia Anwar and Cristina Salmeron for their assistance in the sampling and laboratorial analysis. L. M. P. V. carried out the main experimental work and wrote the draft under the direction of the project designer and leader I. A. J.; N. I. B. assisted with the experimental design, the laboratory work and draft writing. L. M. P. V. was supported by a Foundation for Science and Technology (FCT) grant during the sabbatical licence in St Andrews. The present study was supported by Biotechnology and Biological Research Council grant no. BB/D015391/1. There are no conflicts of interest.
