Glucose regulates protein turnover and growth-related mechanisms in rainbow trout myogenic precursor cells

Abstract

Rainbow trout are considered glucose intolerant because they are poor utilizers of glucose, despite having functional insulin receptors and glucose transporters. Following high carbohydrate meals, rainbow trout are persistently hyperglycemic, which is likely due to low glucose utilization in peripheral tissues including the muscle. Also, rainbow trout myogenic precursor cells (MPCs) treated in vitro with insulin and IGF1 increase glucose uptake and protein synthesis, whereas protein degradation is decreased. Given our understanding of glucose regulation in trout, we sought to understand how glucose concentrations affect protein synthesis, protein degradation; and expression of genes associated with muscle growth and proteolysis in MPCs. We found that following 24 h and 48 h of treatment with low glucose media (5.6 mM), myoblasts had significant decreases in protein synthesis. Also, low glucose treatments affected the expression of both mstn2a and igfbp5. These findings support that glucose is a direct regulator of protein synthesis and growth-related mechanisms in rainbow trout muscle.

1. Introduction

Protein accretion is important for continued muscle growth in salmonids and is dependent on a careful balance between protein synthesis and protein degradation. Protein synthesis is regulated by the mammalian target of rapamycin (mTOR) in humans (Yoshizawa et al., 1998) and ingestion of a meal also activates several of the same kinases in rainbow trout (Oncorhynchus mykiss) (Seiliez et al., 2008). In cultured trout hepatocytes, however, the mTOR pathway is only activated when both amino acids and insulin are added together (Lansard et al., 2010). This differential activation of mTOR can also be achieved in vivo through alteration of the protein to carbohydrate ratio (Seiliez et al., 2011) of the diet with little impact on protein degradation. This differential response of protein synthesis and degradation to macronutrient composition of the diet could also be regulated by glucose availability, as many genes related to glucose metabolism were affected postprandially by a high protein diet (Seiliez et al., 2011).

During the postprandial period, plasma concentrations of insulin, amino acids, and glucose increase, largely due to a nutrient influx and the insulin response to food intake (Seiliez et al., 2008; Huyben et al., 2018; Karlsson et al., 2006; Ok et al., 2001). Insulin is recognized as an endocrine factor with anabolic effects in fish. Salmonid in vitro muscle cultures report that insulin increases protein retention via increasing rates of protein synthesis and decreasing rates of protein degradation (Cleveland and Weber, 2010). Insulin further promotes muscle accretion by stimulating muscle glucose and amino acid uptake (Castillo et al., 2004), the former due in part to stimulation of GLUT4 (Diaz et al., 2009). Amino acids are also recognized to directly promote muscle growth in fish, both by stimulating rates of protein synthesis and reducing rates protein degradation (Brown and Cameron, 1991; Mcmillan and Houlihan, 1988; Seiliez et al., 2008; Cleveland and Radler, 2019). Although glucose can indirectly stimulate growth by promoting hyperinsulinemia, it is unknown whether this energy substrate can directly affect growth-related mechanisms in skeletal muscle. Therefore, one of our objectives was to determine whether glucose can directly regulate protein turnover and expression of growth-related genes in primary rainbow trout myogenic precursor cells (MPCs). These findings are important to understanding the nutrient-gene interactions that regulate muscle growth in fish.

Similar studies to that presented here have been performed in mammalian muscle cultures which supported direct effects of glucose on myogenic mechanisms (Nedachi et al., 2008; Grzelkowska-Kowalczyk et al., 2013). However, findings are somewhat contradictory (see discussion for explanation) and salmonids exhibit persistent hyperglycemia after consumption of high-carbohydrate diets (Wright Jr et al., 1998; Ryman, 1972; Wilson, 1994), a response that would be classified as abnormal in mammals as it reflects glucose intolerance. Therefore, we cannot assume that direct effects of glucose in mammalian muscle also occur in salmonids since the capacity for glucose clearance is very different. While the mechanisms underlying this blunted salmonid response to insulin and glucose have been investigated extensively (del sol Novoa et al., 2004; Panserat et al., 2000; Legate et al., 2001; Skiba-Cassy et al., 2013; Polakof et al., 2012), a consensus has not been reached as the cause is likely multifactorial. In addition to normal fluctuations in glucose levels following meals and even at different times of the day (Holloway et al., 1994), persistent hyperglycemia and transient hyperinsulinemia are the hallmarks of this phenotype in carnivorous fishes.

Interestingly, when culturing fish MPCs, it is routine practice to use high glucose (4500 mg/L) media which is a concentration that falls above the physiological range of plasma glucose found in salmonids. It is not clear what affect lower glucose media concentrations have on protein balance and gene expression in these cell cultures. Therefore, a secondary objective is to examine the impact of culturing rainbow trout MPCs in media containing lower levels of glucose, that are more consistent with natural levels, on mechanisms regulating muscle growth. These findings could contribute to development of novel media formulations specific for promoting differentiation and growth of salmonid MPC cultures.

2. Methods

2.1. Experimental animals

Excess juvenile rainbow trout (5–10 g, all female) produced from the in-house breeding program at the National Center for Cool and Cold Water Aquaculture (NCCCWA) were used for myosatellite cell isolation. Early fish rearing occurred according to standard husbandry protocols in 150 L tanks at biomass density of < 40kg/m³ with flow-through water (4 L per min) at 12.5–13.5 degrees C with 12h:12h light:dark cycle. An automatic feeding system (Arvotec) dispensed a commercially available feed (Finfish G, Zeigler Bros. Inc., Gardners, PA) at or just below satiation. All rearing procedures and protocols were performed according to institutional animal care and use guidelines and received approval from the NCCCWA Institutional Animal Care and Use Committee (IACUC protocol #098).

2.2. Myosatellite cell isolation

Following a protocol developed by Rescan et al. (Rescan et al., 1995) and published by Froehlich et al. (Froehlich et al., 2014) primary MPC’s were isolated from juvenile fish (5–10 g). Fish were anesthetized in tricaine methanesulfonate (> 300 mg/mL; AVMA, Pittsburgh, PA), submerged in 70% ethanol, and de-scaled. Epaxial (white) muscle was dissected (5 g) and placed in chilled (4 °C) suspension media (DMEM, 9 mM NaHCO3, 20 mM HEPES, 100 U/mL penicillin, and 100 μg/mL streptomycin). Muscle tissue was minced using a sterile razor blade and re-suspended in media (5 g/25 mL) before being centrifuged (5 min, 4 °C, 300 g). The supernatant was discarded and a solution of 0.2% collagenase was added (C9891, 1 g tissue/5 mL collagenase solution; Sigma-Aldrich, St. Louis, MO) to the pellet and agitated for 60 min at 20 °C. The suspension was then centrifuged (20 min, 4 °C, 300 g), the supernatant was discarded, and the pellet resuspended in 0.1% trypsin solution (153,571; MP Biomedicals, Irvine, CA). The pellet was dissociated and agitated for 45 min at 20 °C before being diluted 1:5 in suspension media and centrifuged (20 min, 4 °C, 300 g). The supernatant was discarded and the pellet mechanically dissociated before being resuspended and dissociated in suspension media. Cell suspensions were passed sequentially through 100, 70, and 30 μm strainers before being diluted 1:2 in suspension media and centrifuged (10 min, 4 °C, 150 g). The supernatant was discarded and cell pellets were resuspended in complete medium (suspension media, 10% Fetal Bovine Serum (FBS) before being counted, diluted to 1 × 10⁶, and plated onto treated 6 well culture plates. Cells were allowed to adhere for 3 h before being washed with Phosphate Buffered Saline (PBS, pH 8.0) to remove any debris or loosely adherent cells.

2.3. Culture conditions

Six well culture dishes were pretreated with poly-l-lysine (P4832; Sigma-Aldrich) for 3 h at 18 °C before being washed twice with nanopure water and layered with laminin (L2020; Sigma-Aldrich) overnight at 20 °C. Plates were wrapped in parafilm and stored at −80 °C, before plating plates were thawed and washed with PBS (pH 8.0). Cells were cultured at 18 °C under ambient air, and fresh media was supplied every other day.

2.4. Treatment

Cell were cultured from day 0 (day of plating) to day 3 in complete media containing 10% FBS and 4500 mg/L glucose. On day 4 media was replaced with treatment media (DMEM with 2% FBS) containing either routine glucose (RG, 4500 mg/L, 25 mM) or low glucose (LG, 1000 mg/L, 5.6 mM). Protein turnover assays were completed at 24 h and 48 h of treatment while gene expression analyses were done at 48 h and 72 h of treatment. Treatments were applied to triplicate wells and replicated for triplicate isolation events.

2.5. Protein synthesis

Rates of protein synthesis were determined by minor modification of previously published procedures (Cleveland and Weber, 2010). At 24 h and 48 h, treatment media was supplemented with 2.5 μCi [3,5-3H] tyrosine (MP Biomedicals)/mL media for 1 h. Media was removed, and the cells were washed with ice-cold HBSS, layered with 500 μL 10% trichloracetic acid (TCA), and incubated at 4 °C for 1 h. The wells were scraped, and the precipitated protein was pelleted by centrifugation at 9600 g for 5 min. The acid-insoluble protein pellet was resuspended in cell lysis solution (1% Triton-X 100 and 1 N NaOH) and counted for radioactivity using liquid scintillation counting. A subsample of the resuspended protein pellet was retained and analyzed for protein content. Protein synthesis was expressed as cpm per μg of protein before normalizing these values to the control treatment.

2.6. Protein degradation

Rates of protein degradation were determined at 24 h and 48 h by minor modifications to a previously published procedure (Cleveland and Weber, 2010). Two-day old cells were incubated in complete media containing 2.5 μCi [3,5-3H] tyrosine/mL to label intracellular proteins. On day 4 radioactive media was removed, cells were washed two times with HBSS + 2 mM tyrosine, and media was replaced with treatment media. At 24 h or 48 h, media was removed and 500 μL cold 20% TCA was added and incubated at 4 °C for 1 h. The mixture was centrifuged at 9500 g for 5 min, and the radioactivity in the acid-soluble supernatant was quantified using liquid scintillation counting. Immediately after removing 500 μL media, the remaining 1.5 mL media was removed, and the cell layer was washed two times with HBSS. Cells were layered with 500 μL 10% TCA and incubated at 4 °C for 1 h. The wells were scraped, and the precipitated protein was pelleted by centrifugation at 9600 g for 5 min. The acid-insoluble protein pellet was resuspended in cell lysis solution (1% Triton-X 100 and 1 N NaOH) and counted for radioactivity using liquid scintillation counting. Total radioactive protein was calculated as the sum of the radioactivity in the TCA-soluble and -insoluble fractions. Protein degradation was expressed as [3,5-3H] tyrosine released in the media (TCA soluble) as a percentage of total [3,5-3H] tyrosine incorporated in cells (total radioactive protein).

2.7. Multiplex PCR

The GenomeLab GeXP genetic analysis system (Beckman Coulter Inc., Brea, CA) was used to simultaneously analyze eighteen genes involved in three proteolytic pathways (ubiquitin Proteasome, Cathepsin-lysosome, and Calpains) as well as nineteen genes spanning the GH/IGF1 axis, TGFβ superfamily, and muscle specific genes. Methods for multiplex gene expression analysis and primer sequences were previously published (Cleveland and Weber, 2015; Manor et al., 2015); gene symbols, names and accession number for growth-related genes (Table 1) and degradation-related genes (Table 2) are available. Briefly at 48 h and 72 h treatment media was removed from culture dishes and cells were washed twice with PBS (pH 8.0). Trizol (1 mL, Invitrogen, Carlsbad, CA) was applied to cells and total RNA was isolated following the manufacturer’s instructions. Total RNA purity (A260/ A280 > 1.9) was verified by spectrophotometry (NanoDrop, Thermo Fisher). RNA (2000 ng) was DNase treated (Promega, Madison, WI) and reverse transcribed into cDNA using 2 μL 5 × RT buffer, 1 μL gene specific primer mix, 0.5 μL RT, and 1.25 μL kanamycin RNA (internal control) in a 10 μL reaction. An aliquot (4.6 μL) of the RT resultant cDNA was used in the PCR reaction along with 2 μL 25 mM MgCl2, 2 μL 5 × PCR buffer, 1 μL forward primer mix, and 0.35 μL DNA taq polymerase. The PCR was run according to kit instructions and 1.25 μL (optimized) of PCR product was combined with 38.5 μL of sample loading solution. The PCR products were separated by capillary electrophoresis in the GeXP genetic analysis system using a modified Frag-3 protocol with a separation voltage of 6.0 kV for 45 min. Areas for each gene peak within the multiplex were exported for analysis. Data for reference genes were imported into GeNorm software to determine which reference genes were most stable. The normalized expression of each gene transcript is reported as the quantity relative to the geometric mean expression of the selected reference genes.

Table 1.

Muscle growth-related genes targeted in multiplex analysis.

Gene symbol	Gene name	NCBI gene accession no.
Reference genes
rplp2	Acidic ribosomal protein P2	BT074359
actb	Beta actin	NM_001124235
ppia	Peptidyl-prolyl cis–trans isomerase A	CX723368
eef1a	Elongation factor 1 alpha	NM_001124339
GH/IGF axis genes
ghr1	Growth hormone receptor 1	NM_001124535
ghr2	Growth hormone receptor 2	NM_001124731
igfbp4	IGF binding protein 4	DQ146967
igfbp5b	IGF binding protein 5b-1	DQ206713
igfr1b	IGF receptor 1b	AY100460
igf1	Insulin-like growth factor-I	M95183
Muscle specific genes
mstn1b	Myostatin 1b	DQ138300
mstn2a	Myostatin 2a	DQ138301
myod	Myoblast determination protein	NM_001124720
myog	Myogenin	NM_001258337
myf5	Myogenic regulatory factor 5	AY751283
pax7a	Paired box transcription factor 7a	NM_001258337
TGFb superfamily genes
tgfbr1	Transforming growth factor beta receptor 1	FN822751
tgfbr2	Transforming growth factor beta receptor 2	CA382594
acvr1b	Activin receptor type-1B	BX081962
bambi	BMP and activin membrane-bound inhibitor homolog	BX305907
fst	Follistatin	FJ609941
smad2a	Mothers against decapentaplegic homolog 2a	HQ184937
smad2b	Mothers against decapentaplegic homolog 2b	HQ184936
smad7	Mothers against decapentaplegic homolog 7	HQ184941

Table 2.

Protein degradation-related genes targeted in multiplex analysis.

Gene symbol	Gene name	NCBI Gene accession no.
Reference genes
rplp2	Acidic ribosomal protein P2	BT074359
actb	Beta – actin	NM_001124235
gapdh	Glyceraldyhyde phosphate dehydrogenase	NM_001124246
eef1a	Elongation factor 1 – alpha	NM_001124339
Ubiquitin–Proteasome
psma5	Proteasome alpha subunit – type 5	BX078765
psmb3	Proteasome beta subunit – type 3	NM_001124250
psmd6	26S proteasome non-ATPase regulatory subunit 6	NM_001165054
ub	Polyubiquitin	NM_001124306
fbxo32	F-box only protein – 32	HM189693
fbxo25	F-box only protein – 25	NM_001193325
murf1	Muscle RING finger protein – 1	HM357611
Autophagy–Lysosome
ctsd	Cathepsin D	NM_001124711
ctsl	Cathepsin L	NM_001124305
atg4b	Autophagy-related protein 4 homolog B	BX914166
atg12	Autophagy-related protein 12	CB490089
lc3b	Microtubule-associated proteins 1A/1B light chain 3B	CA350545
gabarapl1	Gamma-aminobutyric acid receptor associated protein	CA345480
Calpain system
capn1	Calpain – 1 catalytic subunit (micro)	NM_001124490
capn2	Calpain – 2 catalytic subunit (milli)	NM_001124491
Caspases
casp3	Caspase – 3	CA366923
casp8	Caspase – 8	CU072654
casP9	Caspase – 9	NM_001124647

2.8. Statistical analysis

Fold change values were log2 transformed to calculate final values for statistical analysis with a non-linear mixed effects model with Day and Glucose Status as fixed effects and Repetition as the random effect in R with P < .05 considered significant. A post-hoc Fishers LSD was used to determine significant differences between low and routine glucose means at 48 h and 72 h. Protein synthesis and degradation were analyzed using a PROC GLM in SAS with day, repetition, and glucose status as variables and P < .05 was considered significant.

3. Results

3.1. Protein synthesis/degradation

Protein degradation and synthesis rates were measured after 24 h and 48 h exposure to LG or RG media. Protein degradation was measured by the amount of tritium released into the culture media and was greater at 48 h compared to 24 h however, protein degradation rates were not significantly different between the RG and LG treatments. Protein synthesis was measured by tritium incorporated into protein and was decreased in LG media at by 41.2% at 24 h and 34.8% at 48 h compared to RG media (Fig. 1).

Fig. 1.

Effects of low glucose media on rates of protein synthesis (3H incorporation) and protein degradation (3H release) in 5 and 6 day old cells, 24h and 48h after treatment respectively. Cell were incubated in media containing 2% FBS supplemented with routine glucose (25 mM) or low glucose (5.5 mM). Means ± SE (n = 3 experiments) with an asterisk (*) are defined as significantly different at P < 0.05.

3.2. Gene expression analysis

Out of seven genes examined as part of the ubiquitin-proteasome pathway, only one gene, psma5, (Fig. 2A; Supplementary File 1) was affected by media glucose levels. The expression of psma5 was higher in cells cultured with low glucose, but only after 48 h of exposure. Additionally, the expression of six genes involved in the autophagy lysosome system (ALS) were analyzed and only one gene, gabarapl1, exhibited altered expression levels in response to media glucose levels. The expression of gabarapl1 was lower in myotubes cultured in low media glucose compared to cells cultured in normal glucose after 72 h (Fig. 2B; Supplementary File 1). No effect of media glucose concentration was detected in calpain gene expression, but a main effect of glucose was detected for casp9 expression (Fig. 2C; Supplementary File 1).

Fig. 2.

Effects of low glucose media on genes involved in protein synthesis and degradation after 48h and 72h of treatment. Cells were incubated in media containing 2% FBS supplemented with routine glucose (25 mM) or low glucose (5.5 mM) media for 48h and 72h. Means ± SE (n = 3 experiments) with an asterisk (*) are defined as significantly different between treatments at 48h or 72h with letters denoting a significant difference between 48h or 72h of the same treatment, all at P < 0.05.

Out of 24 growth-related genes analyzed, only the expression of igfbp5, igf-1, and mstn2a were affected by media glucose levels (Supplementary File 2). The expression of Igfbp5 was reduced in low glucose media at 48 h and then increased from 48 h to 72 h (Fig. 2D). Low glucose media reduced mstn2a expression 48 h compared to normal glucose media and levels were decreased from 48 h to 72 h in low glucose media (Fig. 2E). In addition, 48 h of low glucose media exposure resulted in increased igf-1 mRNA levels compared to normal glucose levels, but levels returned to normal level at 72 h (Fig. 2F).

4. Discussion

In fish, protein synthesis and growth-related mechanisms in muscle are directly up-regulated by amino acids and endocrine signals responsive to nutrient intake, such as insulin and IGF1 (Cleveland and Weber, 2010). Findings obtained from this study support that glucose can also be considered as a direct regulator of protein synthesis and gene expression in rainbow trout muscle, although glucose does not appear to directly regulate protein degradation.

Studies performed in fish have demonstrated that protein synthesis rates were influenced by the nutritional and physiological state of the animal (Seiliez et al., 2011; Rolland et al., 2015; Seiliez et al., 2011). Cell culture studies have been valuable for identifying biological factors associated with this response (Cleveland and Weber, 2010; Cleveland and Weber, 2011; Cleveland, 2014; Seiliez et al., 2008; Seiliez et al., 2010; Castillo et al., 2004; Castillo et al., 2002). In vitro and in vivo studies support that insulin increases protein retention in the rainbow trout muscle in part through increasing protein synthesis, decreasing protein degradation, and promoting glucose and amino acid uptake (Castillo et al., 2004; Cleveland and Weber, 2010; Seiliez, Panserat, Lansard, et al., 2011). During the fed state there is an influx of nutrients, including amino acids and glucose, which stimulate insulin secretion (Caruso and Sheridan, 2011; Mommsen and Plisetskaya, 1991) and promote the endocrine mechanisms responsible for increasing muscle growth. In addition to this indirect endocrine effect, in vitro studies indicate that amino acids directly regulate mechanisms in muscle that direct growth, specifically through reducing protein degradation, increasing signaling through TOR, and increasing expression of myogenic genes (Seiliez et al., 2012; Seiliez et al., 2008; Azizi et al., 2016; Velez et al., 2017; Cleveland and Weber, 2010). Adding to this knowledge, our findings provide evidence that glucose was also able to directly regulate growth-related mechanisms in rainbow trout muscle, in part through regulation of protein synthesis.

Direct effects of glucose on myogenic mechanisms have been previously reported in murine C2C12 myoblasts, although the data are inconsistent with respect to myogenesis being enhanced or inhibited by routine glucose media. Glucose treatment of 22.5 mM increased myotube formation in C2C12 cells relative to 5 mM low glucose media (Nedachi et al., 2008), whereas a separate study reported that 15 mM glucose decreased C2C12 myoblast fusion relative to 5 mM glucose treatments (Grzelkowska-Kowalczyk et al., 2013). The Grzelkowska-Kowalczyk et al. (2013) study measured protein synthesis rates and found no differences between C2C12 myoblasts exposed to 15 mM and 5 mM glucose media. However, normal glucose (25 mM) stimulated protein synthesis in murine cardiomyocytes (Yeshao et al., 2005), suggesting that effects of normal glucose on protein synthesis and myogenic mechanisms may depend on glucose concentration and cell type. The concentrations of glucose used in the current study were 5.6 mM (LG) and 25 mM (RG) which are closer to the concentrations that promoted myogenesis in murine cells. In rainbow trout in vivo, average blood plasma glucose concentrations are between 5 and 10 mM (Weber and Shanghavi, 2000; Jentoft et al., 2005; Choi and Weber, 2015), closer to this study’s low glucose treatment. Interestingly, the RG treatment exhibited elevated protein synthesis compared to glucose levels more consistent with natural in vivo concentrations, demonstrating a direct interaction between glucose levels and protein synthesis regulation consistent with previous mammalian in vitro work.

The direct effect of glucose on protein synthesis may reflect a metabolic response to changes in energy substrates that support anabolic processes and maintain homeostasis in muscle. In vitro, muscle cells predominantly rely on glucose and amino acids as energy substrates, both of which can regulate the intracellular signaling pathways involving PI3K and FOXO3 that partially control metabolic and mitogenic activity (Yeshao et al., 2005; Nedachi et al., 2008). In trout, amino acids are not only a potent secretagogue of insulin (Moon, 2001), but are also important regulators of protein synthesis in muscle (Seiliez et al., 2008). In the liver and muscle, amino acids and insulin can activate the mTOR complex and cause an increase in protein synthesis (Lansard et al., 2010). We hypothesize that the low glucose media caused a partial shift in amino acid utilization away from protein synthesis towards oxidative pathways for production of energy substrates, largely to compensate for reductions in glucose availability.

Protein degradation was not differentially regulated by normal and low glucose media treatments in rainbow trout myotube cultures, suggesting that extracellular glucose does not directly regulate this mechanism of protein turnover in muscle. However, the in vitro conditions in the present study do not necessarily reflect the typical postprandial response to nutrient intake in which both circulating glucose and amino acid concentrations increased or decreased during feed deprivation. In vivo rates of protein degradation were initially increased when fasted (Smith, 1981) in both white and red muscle and later reduced following prolonged fasts (Loughna and Goldspink, 1984). In this instance, the cell culture environment cannot fully recapitulate the in vivo systemic response to feed deprivation and involve regulation of endocrine signals like insulin and IGF1. These endocrine signals are critical for indirect effects of nutrients like amino acids and glucose on protein degradation in muscle (Cleveland and Weber, 2010). We specifically targeted direct effects of glucose on protein turnover, which appeared to be limited to an effect on protein synthesis and not protein degradation.

Although no change was seen in protein degradation depending on glucose level, there were a few molecular markers associated with autophagic pathways, such as reduced gabarapl1 expression in response to low glucose. Gabarapl1 is a known autophagy-related gene with undefined functions in autophagosome formation (Lee and Lee, 2016). It has previously been shown that in vitro IGF1 treatments down-regulate gabarapl1 expression whereas two weeks of feed deprivation upregulated expression in vivo (Seiliez et al., 2010). The observed decrease in gabarapl1 could be indirectly due to elevated igf-1 expression observed at 48 h. Protein retention appears to be negatively regulated in MPC’s by extended periods of LG media exposure, although it is unlikely LG media signals an induction of autophagosome formation, as LC3b and several other critical autophagy markers were not differentially regulated by glucose levels (Supplementary File 1). We did not detect changes in any of the genes regulating ubiquitination, suggesting that the capacity for ubiquitination was not affected by glucose availability. However, psma5 expression was dynamically regulated by glucose levels, with a substantially higher expression level at 48 h when exposed to low glucose media compared to normal glucose. These levels were not different at 72 h, suggesting an enhanced compensatory action was triggered when glucose levels were low. The absence of a comprehensive regulation of proteolysis-related genes is in agreement with our protein degradation results, for which there was no response to media glucose levels.

This study also examined the effects of low glucose media on mechanisms regulating growth. Of the 24 growth-related genes analyzed, only three genes, igf-1, mstn2a and igfbp5, were affected by the amount of glucose in the media. Igf-1 expression was elevated by low glucose concentrations at 48 h, but returned to levels observed at normal (routine) glucose concentrations at 72 h. It is likely that this dynamic change in igf-1 expression from 48 to 72 h is a compensatory mechanism to enhance glucose utilization when levels were decreased upon treatment initiation. However, the likelihood of these changes affecting protein synthesis or observed gene expression is low, as high IGF-1 should hypothetically increase protein synthesis levels.

Overall, expression of mstn2a was differentially regulated in low glucose media, with higher expression in low glucose media at 48 h and lower expression at 72 h compared to normal glucose media. Mstn2a is one of three (mstn1a and mstn1b) putatively functional myostatin genes in rainbow trout (Garikipati et al., 2006). Myostatin is a well-known negative regulator of skeletal muscle growth (Seiliez et al., 2012; Gabillard et al., 2013). Previous findings in murine C2C12 myoblasts also report reduced MSTN expression in low glucose media (Grzelkowska-Kowalczyk et al., 2013), suggesting that effects of glucose on MSTN expression may be a conserved mechanism for regulating myogenesis. The down regulation of mstn2a in response to low glucose availability in the media observed from 48 h to 72 h may be a negative feedback for MSTN accumulation prior to 72 h, although there is no direct evidence for this mechanism.

The expression of igfbp5 was reduced in low glucose media at 48 h, although there were no differences observed between normal and low glucose 72 h after exposure. The opposite was observed in murine C2C12 myoblasts; where the abundance of igfbp5 was greater in 5 mM low glucose versus 15 mM normal-glucose media (Grzelkowska-Kowalczyk et al., 2013). Igfbp5 has been reported as the most conserved IGFBP in mammals and fish and is one of the major binding proteins expressed in skeletal muscle (Duan et al., 2010; Macqueen et al., 2013). Nutritional and hormonal stressors have been shown to differentially regulate the expression of igfbp5 in a manner that support a positive role for IGFBP5 in muscle growth. In rainbow trout (Cleveland and Weber, 2014; Gabillard et al., 2006) and Atlantic salmon (Bower et al., 2008) feed deprivation decreased igfbp5 expression and refeeding increased igfbp5 expression. Additionally, igfbp5 was upregulated following amino acid supplementation in cultured salmon muscle cells that had been serum starved (Bower and Johnston, 2010). Therefore, the down-regulation of igfbp5 in low glucose media was likely a response from low glucose availability, rather than an insulin endocrine response. Since this study was performed in vitro, the cells were not interacting with the endocrine system, thus, the upregulation observed at 72 h could be a natural response or return to steady-state levels overtime. However, even at these relatively low levels, expression was sensitive to glucose levels in the media and this study demonstrates the variable effects that nutrition and hormones can have on igfbp5 expression.

5. Conclusion

Muscle growth (protein accretion) in rainbow trout responded in a coordinated way to changes in both nutrient availability and composition, therefore changes in these parameters can affect protein synthesis or protein degradation (Rolland et al., 2015; Seiliez et al., 2011; Seiliez et al., 2011). Protein synthesis is up-regulated in response to endocrine signals like insulin and IGF1 (Cleveland and Weber, 2010) however it is unknown if the same response is present during changes in glucose availability. Here we detail for the first-time how glucose availability can act as a regulator of protein synthesis and gene expression in cultured rainbow trout MPCs. This study highlights the need for further research into glucose related mechanisms regulating protein retention and growth in rainbow trout, particularly as aquafeeds transition towards higher inclusion of plant-based feedstuffs that can have higher carbohydrate contents. Additionally noted is how the field routinely uses culture media glucose levels that are not consistent with natural in vivo serum levels, and this study highlights how these concentrations can directly affect cellular functions and should be considered in greater detail in future studies to improve data interpretation.