Pathways regulating equine skeletal muscle protein synthesis respond in a dose-dependent manner to graded levels of protein intake

Abstract

Activation of the mechanistic target of rapamycin (mTOR)-controlled anabolic signaling pathways in skeletal muscle of rodents and humans is responsive to the level of dietary protein supply, with maximal activation and rates of protein synthesis achieved with 0.2 to 0.4 g protein/kg body weight (BW). In horses, few data are available on the required level of dietary protein to maximize protein synthesis for maintenance and growth of skeletal muscle. To evaluate the effect of dietary protein level on muscle mTOR pathway activation, five mares received different amounts of a protein supplement that provided 0, 0.06, 0.125, 0.25, or 0.5 g of crude protein (CP)/kg BW per meal in a 5 × 5 Latin square design. On each sample day, horses were fasted overnight and were fed only their protein meal the following morning. A preprandial (0 min) and postprandial (90 min) blood sample was collected and a gluteus medius muscle sample was obtained 90 min after feeding the protein meal. Blood samples were analyzed for glucose, insulin, and amino acid concentrations. Activation of mTOR pathway components (mTOR and ribosomal protein S6 [rpS6]) in the muscle samples was measured by Western immunoblot analysis. Postprandial plasma glucose (P = 0.007) and insulin (P = 0.09) showed a quadratic increase, while total essential amino acid (P < 0.0001) concentrations increased linearly with the graded intake of the protein supplement. Activation of mTOR (P = 0.02) and its downstream target, rpS6 (P = 0.0008), increased quadratically and linearly in relation to the level of protein intake, respectively. Comparisons of individual doses showed no differences (P > 0.05) between the 0.25 and 0.5 g of protein intake for either mTOR or rpS6 activation, indicating that protein synthesis may have reached near maximal capacity around 0.25 g CP/kg BW. This is the first study to show that the activation of muscle protein synthetic pathways in horses is dose-dependent on the level of protein intake. Consumption of a moderate dose of high-quality protein resulted in near maximal muscle mTOR pathway activation in mature, sedentary horses.

Introduction

Adequate dietary protein intake is essential for maintaining muscle mass and function in horses. While general crude protein (CP) requirements for horses have been defined to avoid severe deficiencies and muscle loss (National Research Council, 2007), there is still a gap in our knowledge regarding the quantity of dietary protein necessary to stimulate maximal muscle protein synthesis (MPS).

In humans, dietary protein levels have been well studied in relation to both athletic performance and their impact on sarcopenia in elderly subjects. Several studies have illustrated that MPS increases in a dose-dependent manner with the level of protein intake. Maximal muscle fractional synthesis rates (FSR) in human adults are reportedly achieved with the consumption of 0.2 to 0.4 g protein/kg body weight (BW)/meal, depending on the quality of the protein source and the age of the individual (Cuthbertson et al., 2005; Moore et al., 2009). Protein intake greater than these levels does not seem to result in any further increase in MPS and, therefore, has a little additional benefit in either young or elderly humans (Symons et al., 2009). In fact, an excessive influx of amino acids is known to trigger negative feedback mechanisms, resulting in cellular insulin resistance that decreases muscle protein synthetic capacity and can lead to health problems (Bohé et al., 2001; Krebs et al., 2002; Newgard, 2012). Recent work showed similar results in horses, where a large protein meal resulted in hyperinsulinemia in horses with equine metabolic syndrome, indicating that elevated dietary protein intake could exacerbate insulin dysregulation (Loos et al., 2019). Considering protein is commonly overfed in the equine diet (Bott et al., 2016), it is warranted to further investigate the mechanisms underlying MPS in horses and how these respond to the levels of dietary protein.

MPS is activated and regulated by the mechanistic target of rapamycin (mTOR) pathway, which is highly sensitive to anabolic growth and nutritional factors, including insulin and amino acids (Yoon, 2017). Leucine, in particular, is known to strongly stimulate MPS by modulating mTOR pathway components (Crozier et al., 2005). It is well known that the mTOR pathway is upregulated after consumption of a protein-rich meal and that an increase in signal transduction (i.e., phosphorylation of key components) coincides with a rise in muscle FSR in humans (Atherton et al., 2010). Studies in other species also report that mTOR pathway activation is sensitive to the level of dietary protein intake, showing a dose-dependent increase in phosphorylation of its anabolic signaling components (Norton et al., 2009; D’Souza et al., 2014). Similar to MPS rates, mTOR signal transduction is saturable and reports in rodents suggest that phosphorylation of regulatory components in the muscle reaches a plateau at 50% of the daily recommended levels of dietary leucine (Yoshizawa et al., 2013).

In horses, several studies have provided clear evidence that muscle mTOR pathways are activated in response to feeding; however, these data are based on supraphysiological doses of protein (Urschel et al., 2011; Wagner and Urschel, 2012). It is, therefore, still unclear as to whether the activation of mTOR pathway components could be obtained at lower levels of protein consumption. Moreover, there is a paucity of data regarding protein levels needed to stimulate maximal mTOR activation and thus the potential for optimal muscle accretion in horses. While data from human and rodent models provide useful information, the distinct differences in digestive anatomy compared with the horse warrant further species-specific research.

The present study was designed to evaluate dose–response effects in skeletal muscle mTOR pathway activation following the graded levels of protein intake in mature, sedentary horses. It was hypothesized that mTOR pathway activation would exhibit a Michaelis–Menten-like response to the level of dietary protein.

Materials and Methods

The University of Kentucky Institutional Animal Care and Use Committee approved all procedures (approval no. 2018-3110).

Animals and housing

Five mature, Thoroughbred mares from the University of Kentucky Maine Chance farm herd were selected for the study (19 ± 1.6 yr old; 556 ± 54 kg BW; body condition score 5.2 ± 0.8; mean ± SD; Henneke et al., 1983). Horses were weighed the day prior to each sample collection day. All horses were housed individually in 3 × 15 m, partially covered, dry-lot pens for the duration of the study with ad libitum access to water and a salt block. Horses were fitted with grazing muzzles and turned out into outdoor paddocks for approximately 6 h each day. They remained in their dry-lot pens overnight and on sampling days to allow for individual feeding and handling during experimental procedures. Horses were fed an alfalfa-timothy mixed hay (2.12 Mcal/kg digestible energy [DE]; 19.1% CP; 3.3 % ethanol-soluble carbohydrates (ESC); 0.4% starch; on dry matter [DM] basis; 91.4% DM) at 1.5% of BW per day (as-fed) and received two concentrate meals (at 0800 and 1600 hours) consisting of 0.750 kg (as-fed) of whole oats (3.51 Mcal/kg calculated DE; 12.7% CP; 2.2% ECS; 50.1% starch; on DM basis; 90.4% DM) and 50 g of a custom-made vitamin/mineral premix pellet (Cavalor Feeds and Supplements, Deinze, Belgium). This feeding regimen intended to meet all nutrient requirements for mature, idle horses (National Research Council, 2007). Hay was provided in hay nets and secured above a large tub to catch any spillage. Leftover hay was weighed and recorded every morning after which fresh hay was provided. Prior to the study, horses were removed from pasture and adapted to all housing, management, and feeding regimes for 5 d prior to the first sample collection day.

Experimental procedures

In a 5 × 5 Latin square design, horses randomly received each of the five doses of dietary protein on five different sample collection days. Each sample day was separated by two recovery days with the exception of the last sample day, which was separated by 5 d from the previous collection day due to weather conditions. Approximately 14 h prior to the start of sample procedures, all feed was removed to ensure horses were in a postabsorptive state. Horses had continuous access to water at all times. On the morning of sample days, horses were aseptically fitted with an intravenous jugular catheter (14 G, 2 inches, Nipro, Bridgewater, NJ) and allowed to recover from the procedure for approximately 1 h. Next, two baseline blood samples (10 mL; −15 and 0 min pre-feeding) were collected after which each horse received its protein meal. Variable amounts of a commercially available high protein, vitamin, mineral supplement (Cavalor VitAmino; Cavalor Feeds and Supplements, Deinze, Belgium; Table 1) were fed to provide 0, 0.06, 0.125, 0.25, or 0.5 g of CP/kg BW. This equated to a supplement intake of 0, 0.16, 0.33, 0.66, and 1.32 g DM/kg BW. The protein levels were chosen based on dose–response studies evaluating MPS in humans (Moore et al., 2009; Witard et al., 2014). A second blood sample (10 mL) was collected 90 min after the protein meal was administered. All blood samples were collected into heparinized vacutainers, which were immediately centrifuged (1,500 × g; 10 min) and plasma harvested and frozen (−20 °C) until further analysis. After collection of the 90-min blood sample, horses were lightly sedated with approximately 3 mL of i.v. xylazine hydrochloride (AnaSed, Lloyd Inc., Shenandoah, IA) and brought into holding stocks. Biopsy sites were aseptically scrubbed and anesthetized subcutaneously with 3 mL of 2% lidocaine (Henry Schein, Melville, NY). A percutaneous sample of the gluteus medius muscle was collected through a single incision, using the Bergstrom needle technique as previously described (Urschel et al., 2011). After the biopsy was completed, horses were allowed to recover in their pens after which catheters were removed, and horses received their daily allotment of hay and evening ration of concentrate. Muscle samples were immediately processed, as described below, and stored at −80 °C until further analysis. All sample procedures were repeated for each of the treatments with blood and muscle samples collected from alternating sides of the horse, to allow healing from the previous collection period.

Table 1.

Nutrient composition of the protein supplement on a DM basis

Nutrient	% DM
DE, Mcal/kg1	3.2
CP	38
EAA
Lysine	2.3
Leucine	3.0
Isoleucine	1.8
Valine	2.0
Threonine	1.6
Methionine	0.6
Histidine	0.9
Phenylalanine	1.9
Tryptophan	0.5
Arginine	2.3
Acid detergent fiber	13.9
Neutral detergent fiber	21.3
Water-soluble carbohydrates	9.5
Ether-soluble carbohydrates	8.4
Starch	7.6
Non-fiber Carbohydrates	31

¹DE calculated value (Pagan, 1998); Pellet DM: 87.8%. Additional nutritional elements: calcium: 1.2%; phosphorus 0.9%; magnesium: 0.81%; potassium: 2.03%; sodium: 0.14%; Iron: 569 ppm; zinc: 654.5 ppm; copper 154 ppm; manganese: 487 ppm; molybdenum 2 ppm; Main ingredients: Soybean meal, alfalfa, potato protein, and wheat bran.

Mild inflammatory processes were expected during the natural healing of the biopsy sites and, therefore, each muscle sample was taken at least 2 inches from the previous site. Although mTOR signaling can be influenced by inflammation, a previous study showed that such effects are mitigated when horses received phenylbutazone, a nonsteroidal anti-inflammatory drug (NSAID) (Wagner, 2011). In an attempt to minimize any influence of inflammation on the mTOR response to feeding, all horses received 2 g of a phenylbutazone paste (Butler Animal Health Supply, Dublin, OH) daily for the entire study period, starting 2 d prior to the first collection period. Horses were carefully monitored throughout the study and no negative effects from NSAID administration or biopsies were observed at any point. Oral NSAID administration did not affect the mTOR response to feeding in the aforementioned study (Wagner, 2011) and was, therefore, not expected to interfere with the results of the current study. All horses returned to the University of Kentucky, Department of Animal and Food Sciences, herd after the completion of the study.

Plasma glucose and insulin analyses

Plasma glucose concentrations were determined using an enzymatic assay (Konelab 20XTi, Thermo Electron Corp., Waltham, MA). Plasma insulin levels were assayed using a commercially available radioimmunoassay kit (PI-12K, Millipore Sigma, Burlington, MA) as validated for horses and previously described (Loos et al., 2019). All plasma samples were run in duplicate within a single assay. The average intra-assay variation was 1.2% and 4.8% for glucose and insulin, respectively.

Plasma amino acid analysis

Plasma free amino acid concentrations were determined on individual samples by reverse-phase high-performance liquid chromatography (HPLC) of phenylisothiocyanate derivatives, as previously described (Urschel et al., 2011). Feed amino acid concentrations were determined by cation-exchange chromatography (cIEC-HPLC) coupled with post-column ninhydrin derivatization and quantitation by a commercial laboratory (ESCL, University of Missouri, MO).

Western immunoblot analysis

To isolate intracellular proteins for Western immunoblot analyses, freshly harvested muscle tissue (~100 mg) was immediately rinsed with ice-cold phosphate-buffered saline solution and placed in cold buffer solution (20 mM hydroxyethyl piperazineethanesulfonic acid, 2 mM ethylene glycol tetraacetic acid, 50 mM NaF, 100 mM KCl, 0.2 mM ethylenediaminetetraacetic acid, and 50 mM β-glycerophosphate) that contained a protease/phosphatase inhibitor cocktail (1:100 ratio with buffer; Cell Signaling Technology, Danvers, MA). Samples were homogenized over ice using a handheld homogenizer for 15 s at a time after which they were centrifuged at 10,000 × g for 10 min. The supernatant was then removed and stored at −80 °C until further analysis. A Bradford assay (Cat. No. 97065-020, VWR International, Indianapolis, IN) was performed to determine the total protein content of the homogenate samples. Based on these results, a new aliquot of each sample was then diluted with 50 µL of a 3X Laemmli buffer (Bio-Rad, Hercules, CA), followed by variable amounts of a 1X Laemmli buffer to obtain a final protein concentration of 2 µg/µL. Laemmli-diluted samples were boiled for 5 min at 95 °C and loaded (20 µg of protein) on an 8% or 12% sodium dodecyl sulfate-polyacrylamide gels for the detection of mTOR and rpS6 (ribosomal protein S6) proteins, respectively. Proteins were separated by use of sodium dodecyl sulfate-polyacrylamide gel electrophoresis (Laemmli, 1970) and then transferred to a 0.45-µm polyvinylidene fluoride membrane (MilliporeSigma, Burlington, MA) for 1 h at 300 mA. Membranes were stained with Fast Green FCF (1% fast green, 50% methanol, and 1% glacial acetic acid), and a digital image was obtained for the quantification of the total amount of protein in each lane (Luo et al., 2006). After removal of the stain, membranes were blocked with 5% (wt:vol) nonfat milk in tris-buffered saline with 0.1 % Tween (vol:vol) (TBST) and incubated with appropriate primary antibodies (Cell Signaling Technology, Danvers, MA) in a 5% milk (for rpS6) or bovine serum albumin (for mTOR) TBST solution overnight at 4 °C. Rabbit polyclonal antibodies were used against total (1:1,000) and Ser²⁴⁴⁸-phosphorylated (1:1,000) mTOR and Ser^235/236- and Ser^240/244-phosphorylated (1:2,000) rpS6, while a monoclonal antibody was used against total rpS6 (1:10,000). The following day, membranes were washed in TBST and incubated with a goat anti-rabbit IgG antibody conjugated to horseradish peroxidase (1:2,000; Cell Signaling Technology, Danvers, MA) for 1 h, washed in TBST, and proteins visualized using a chemiluminescent detection kit (Amersham enhanced chemiluminescence Plus Western blotting detection reagent; GE Healthcare Bio-Sciences, Pittsburgh, PA). Images were captured using a digital imager (Azure c600, Azure biosystems, Dublin, CA). The membranes were then stripped with a buffer solution (Tris 62.5 mM, 2% w/v sodium dodecyl sulfate (SDS), and 0.1 M β-mercaptoethanol), reprobed, and imaged for the appropriate total form of each protein using the same procedures.

The rpS6 antibody has previously been validated in our lab for use with equine muscle tissue (Urschel et al., 2011). For validation of the mTOR antibody, the chemiluminescent signal of positive control extracts (Cell Signaling, Danvers, MA), known to react with mTOR antibody, was compared with the signal obtained from the equine muscle samples using a molecular weight ladder. Subsequently, antibody specificity against mTOR protein in the equine muscle samples was confirmed using blocking peptides (Cell Signaling, Danvers, MA). The primary mTOR antibody was mixed with a blocking peptide in a 1:1 ratio, diluted in a 5% bovine serum albumin solution, and incubated for 30 min at room temperature. Next, duplicate membranes were incubated either with the mTOR antibody or with the antibody/blocking peptide mixture at 4 °C overnight and imaged the following day. The primary antibody specificity for the equine form of mTOR was confirmed as the chemiluminescent signal disappeared when blocking peptide was added to the primary antibody mixture.

All gels were run in duplicate and results are expressed as an average of two gels. Band densities of all blots and Fast Green-stained membranes were quantified using a densitometry software (AzureSpot, Azure Biosystems, Dublin, CA). As a loading control, each sample band was normalized to the value of total protein in each lane, as determined by the Fast Green stain. To account for inter-gel variation, each sample was also normalized to a positive control (four randomly pooled samples), which was loaded onto each gel.

Statistical analysis

Data analysis was performed using mixed procedures of SAS 9.3 statistical software (SAS Institute, Cary, NC). All data were analyzed using a repeated measures two-way ANOVA, with baseline values (0 min) used as a covariate in the model for all blood metabolites. Based on appropriate fit statistics, a “variance components” covariance structure was chosen for the repeated measures analysis. Dose and day were considered fixed effects with day as the repeated variable and horse as the random subject. There was no effect (P ≥ 0.2) of day on any of the variables. For each variable, the overall dose–response curve to the level of protein intake was evaluated using orthogonal polynomial contrasts. Coefficients were determined using the interactive matrix language (IML) procedure for unequally spaced dose treatments. Additionally, individual differences between doses for each variable were detected by separating the means using protected (P < 0.05) Fisher’s least significant differences. The relationship between amino acids and muscle proteins were evaluated using Pearson’s correlation coefficient. Due to a slight nonnormal distribution as evident from a Shapiro–Wilk’s test, insulin values were log (ln)-transformed for all analyses. Studentized residuals for all response variables were analyzed graphically (using Q/Q plots) and with the SAS univariate procedure to ensure adherence to normality assumptions. All data are presented as least squares means and standard error of the mean (SEM).

Results

Animals

There were no refusals for any of the daily concentrate meals and hay intake (as-fed) was similar across sampling periods (P = 0.7) with horses consuming approximately 1.2 ± 0.1% of BW daily. BW was maintained (P = 0.6) for all horses throughout the study, and all protein meals were consumed in a timely manner (3.6 ± 2.6 min) on each sample day.

Blood metabolites

Postprandial plasma glucose (Figure 1C; P = 0.007) and insulin (Figure 1 D; P = 0.09) concentrations showed a quadratic increase in relation to the level of protein intake. Comparison of individual doses showed that there was no difference in either glucose (P = 0.3) or insulin (P = 0.2) concentrations between the 0.25 and 0.5 g/kg BW protein dose. Despite this, there was a positive correlation observed for glucose and insulin concentrations and abundance of phosphorylated mTOR (glucose, P < 0.0001; R = 0.8; insulin, P = 0.005, R = 0.5) and rpS6 (glucose, P < 0.0001; R = 0.7; insulin P < 0.0001, R = 0.7). Concentrations of essential plasma amino acids (sum of lysine, leucine, methionine, threonine, valine, isoleucine, histidine, phenylalanine, and tryptophan; Figure 1 E) showed a linear increase (P < 0.0001) relative to protein dose with a tendency for a quadratic relationship (P = 0.1). With exception of leucine and isoleucine, the postprandial response of the individual plasma essential amino acids (EAA) to protein intake (data not shown) had both linear (P < 0.0001) and quadratic terms (P ≤ 0.07). Postprandial plasma isoleucine (data not shown) and leucine (Figure 1 F) concentrations increased linearly (P < 0.0001) with the level of protein intake. Plasma leucine concentrations were positively correlated to mTOR (P = 0.002; R = 0.6) and rpS6 (P = 0.005; R = 0.5) activation. There was a positive correlation between concentrations of plasma EAA and abundance of phosphorylated mTOR (P = 0.0007; R = 0.6) and rpS6 (P = 0.0001; R = 0.6). Plasma EAA (P = 0.07) and leucine (P = 0.1) concentrations tended to differ between the 0.25 and 0.5 g/kg BW protein doses, based on individual dose comparisons.

Figure 1.

The abundance of phosphorylated skeletal muscle signaling proteins and plasma glucose, insulin, EAA, and leucine concentrations 90 min post consumption of a commercial protein supplement providing different levels of protein. Five mature mares were fed different levels of the protein supplement to provide 0, 0.06, 0.125, 0.25, and 0.5 g CP/kg BW. (A) Phosphorylation of mTOR at Ser²⁴⁴⁸ 90 min post feeding each protein dose (arbitrary units); (B) phosphorylation rpS6 at Ser^235/236 and Ser^240/244 90 min post feeding each protein dose (arbitrary units); (C) plasma glucose concentrations at 90 min post feeding each protein dose (mg/dL); (D) plasma log (ln) insulin concentrations at 90 min post feeding each protein dose (µIU/mL); (E) plasma EAA (sum of lysine, leucine, methionine, threonine, valine, isoleucine, histidine, phenylalanine, and tryptophan) concentrations at 90 min post feeding each protein dose (µmol/L); and (F) plasma leucine concentrations at 90 min post feeding each protein dose (µmol/L). Phosphorylated mTOR and phosphorylated rpS6 showed a quadratic (P = 0.02) and linear (P = 0.0008) increase relative to the level of protein intake, respectively, with a tendency for rpS6 to plateau at the 0.25 g/kg BW dose (P = 0.16). Glucose (P = 0.007) and insulin (P = 0.086) increased quadratically with the level of protein intake. EAA and leucine increased linearly (P < 0.0001) with increased protein intake with a strong tendency for EAA to show a quadratic relationship (P = 0.1). ^a–dDoses with no letters in common were statistically different according to individual comparisons (P < 0.05). Data presented as least squares means ± SEM; n = 5.

Muscle mTOR signaling pathway activation

The degree of muscle mTOR protein phosphorylation increased quadratically (P = 0.02) in relation to the level of protein intake (Figure 1 A). Individual dose comparisons showed that phosphorylation of mTOR was not different (P = 0.99) between the 0.25 and 0.5 g/kg BW doses, indicating a plateau was reached. Muscle rpS6 protein phosphorylation increased linearly (P = 0.0008) relative to protein dose (Figure 1 B) with a tendency to reach a plateau in a quadratic manner (P = 0.16). Similar to mTOR, there was no difference (P = 0.6) between the 0.25 and the 0.5 g/kg BW doses for rpS6 phosphorylation. The total abundance of mTOR and rpS6 proteins was not affected (P ≥ 0.3; data not shown) by protein feeding.

Discussion

This study provides novel data showing that the activation of the signaling pathway regulating MPS is dose–responsive to the level of protein intake in horses. We show that the intake of graded levels of a high-protein pellet results in a stepwise increase in phosphorylation of mTOR signaling proteins and that this regulatory pathway is saturable, as evidenced by a clear Michaelis–Menten-like response (Johnson and Goody, 2011).

The dose–responsiveness of MPS is made possible by highly sensitive underlying regulatory mechanisms that respond to subtle changes in cellular nutrient and energy status. The major metabolic sensor, mTOR, modulates diverse cellular growth-promoting processes. Transduction through the mTOR pathway is widely accepted to be central in the activation of protein translational machinery leading to MPS. Different anabolic stimuli such as amino acids, insulin and exercise will trigger an intracellular signaling cascade resulting in the phosphorylation of mTOR and several of it’s downstream effectors, including S6 kinase 1 (S6K1). This kinase will subsequently phosphorylate its target, the rpS6, ultimately leading to the assembly of the ribosomal units necessary to initiate mRNA translation (Goodman, 2013). The stimulation of MPS following a meal or exercise has been shown to be accompanied by the activation of mTOR signaling components (Norton et al., 2009). Moreover, in the presence of rapamycin, a potent inhibitor of mTOR, the postprandial increase in MPS is completely blocked, further confirming that mTOR is a key regulator and that changes in the phosphorylation status of its signaling protein are indirect markers for MPS (Dickinson et al., 2011).

We have previously shown that feeding a high-protein meal results in the activation of mTOR pathway signaling components in young and adult horses (Urschel et al., 2011; Wagner and Urschel, 2012). We have now extended these findings to show that mTOR activation in horses is dose–responsive to the intake of graded levels of a protein supplement. There was a clear quadratic relationship between mTOR phosphorylation and the level of protein intake. Doses of 0.06, 0.125, 0.25, and 0.5 g CP/kg BW resulted in 1.6-, 2-, 3.7-, and 3.7-fold increase in mTOR phosphorylation, respectively, compared with fasted values (0 g protein). Although mTOR phosphorylation was increased at lower doses of protein intake, 0.25 g CP/kg BW was necessary to induce significant activation over fasted levels. Interestingly, doubling that dose to 0.5 g CP/kg BW did not result in a further increase in mTOR phosphorylation, indication a plateau was reached at levels around 0.25 g CP/kg BW. Signal transmission from mTOR to its downstream components was effective as rpS6 activation showed a very similar dosing response. RpS6 activation increased linearly with the levels of protein intake with a trend for a quadratic relationship. Visual assessment of the data clearly shows that this relationship resulted in two slopes, with a much smaller increment in phosphorylation between the 0.25 and 0.5 g protein doses compared with lower doses. Phosphorylation of rpS6 increased 2.2-, 4.4-, 8-, and 9-fold over fasted values following the consumption of 0.06, 0.125, 0.25, and 0.5 g CP/kg BW, respectively. This illustrates that while there was a 84% and 97% increase in rpS6 phosphorylation when the intake increased from 0.06 to 0.125 g CP/ kg BW and from 0.125 to 0.25 g CP/kg BW, respectively, there was only a 14% increase when the latter dose was doubled. Additionally, individual dose comparisons showed that there was no significant difference between the two highest doses, indicating that rpS6 activation was near maximal at the levels of 0.25 g CP/kg BW. A larger sample size would likely have resulted in a statistically stronger quadratic relationship. Similar to mTOR, a protein dose of 0.25 g CP/kg BW was necessary to result in a substantial increase in rpS6 activation over fasted levels.

Similar dose–responses in MPS to the level of protein intake have been well-described in other species. Several studies illustrate an increase in rates of MPS in adult men after consumption of 5 or 10 g (~0.06 to 0.125g protein/kg BW) of protein, with maximal levels reached with the intake of 20 g (0.25 g protein/kg BW; Moore et al., 2009; Witard et al., 2014). This was also confirmed at the cellular level, where the consumption of 20 g of whey protein after exercise resulted in a greater degree of phosphorylation of mTOR and its upstream (Akt) and downstream (S6K1) components compared with 10 g of protein in young healthy men (Kakigi et al., 2014). Similar to our findings, phosphorylation of S6K1, that is, kinase responsible for rps6 activation, shows a linear increase in relation to the level of whey protein intake in healthy older men with a significant elevation over baseline levels with consumption of 20 g of protein with no further increase in S6K1 phosphorylation with the intake of 30 or 40 g of protein (D’Souza et al., 2014). Likewise, the intake of three levels of protein in a mixed meal (10%, 20%, and 30 % energy from protein) in rats resulted in a stepwise increase of S6K1 phosphorylation but no difference was seen between the two highest levels of protein intake (Norton et al., 2009). Although it is difficult to compare protein sources between studies, or between purified and mixed protein feedstuffs, the effect of level of protein intake on the mTOR pathway observed in the current study was very similar to those reported in other species. Under the conditions of this experiment, maximal mTOR stimulation seems to be achieved with the intake of approximately 0.25 g of protein/kg BW.

As previously described, the activation of the mTOR signaling pathway occurs in the presence of various anabolic stimuli, with amino acids being the predominant dietary-related factor. The protein supplement used in the current study mainly consisted of ingredients commonly used as protein sources in the equine feed industry, including soybean and alfalfa meal. Additionally, it also contained isolated potato protein. Soy and potato protein are considered high-quality plant-based protein sources based on their amino acid composition and digestibility (Willis, 2003; Alting et al., 2011). Moreover, potato protein has higher scores for various nutritional parameters compared with other vegetable protein sources, and its EAA content is, in fact, greater than that of certain animal proteins such as casein and egg (Alting et al., 2011; Gorissen et al., 2018). Therefore, the pre-cecal protein digestibility and amino acid availability of the treatment pellet were assumed to be high, which was supported by the significant postprandial rise in plasma EAA concentrations within 90 min. Human studies report that approximately 10 g of EAA, of which 3 g of leucine, is necessary to maximally stimulate mTOR pathway activation and subsequent MPS in young adult men (Cuthbertson et al., 2005; Witard et al., 2014). Horses consumed approximately 13, 27, 55, and 109 g of EAA, including 2, 5, 10, and 19 g of leucine, for the 0.06, 0.125, 0.25, and 0.5 g CP/kg BW dose, respectively. Regardless of potential differences in digestibility between the mixed protein pellet used in this study and sources used in human work, the level of EAA and leucine needed to maximally stimulate the mTOR pathway was similar between horses (0.48 g EAA/BW^0.75 and 0.09 g leucine/BW^0.75) and humans (0.37 g EAA/BW^0.75 and 0.11 g leucine/BW^0.75) when expressed on a metabolic BW. The dose–response for plasma EAA was in concordance with the measured activation response for the mTOR signaling components. As expected, plasma EAA concentrations increased with the level of protein intake in a linear manner, albeit with a tendency to reach a plateau at 0.25 g CP/kg BW. A similar dose–response was observed for plasma leucine. As mentioned previously, of all dietary EAA, leucine is known to be the most potent stimulator of MPS (Anthony et al., 2000) and enhances phosphorylation of mTOR signaling components in a dose-dependent manner in other species (Crozier et al., 2005). In fact, leucine is known to directly allow mTOR protein kinase phosphorylation in skeletal muscle by binding its negative regulators (Wolfson et al., 2016). In an equine in vitro model, the administration of leucine to equine satellite cell myotube cultures increased the protein synthesis in a dose-dependent manner and enhanced phosphorylation of downstream mTOR signaling components (DeBoer et al., 2018). Plasma leucine in the current study showed a strong correlation with the activation of mTOR and rpS6, demonstrating its significance in the stimulation of the equine muscle protein synthetic pathways. Taken together, these data provide evidence that the activation of protein translation initiation in horses is responsive to the influx of graded amounts of dietary amino acids.

Although treatments were based on dietary protein intake, the nature of the treatment pellet makes it difficult to distinguish between the effects of the protein per se and the nonprotein matrix in which it was ingested. This explains the moderate elevation in blood glucose concentrations, which showed a quadratic response to the level of protein intake. Similarly, plasma insulin levels showed a quadratic response with increasing protein intake. However, glucose intake was relatively low compared with protein intake. Additionally, it is well known that certain amino acids have strong insulinotropic effects (Floyd et al., 1966), making it difficult to distinguish what caused the rise in insulin. Although glucose, amino acids, and insulin exhibited a positive correlation with mTOR pathway activation to some degree, it has been well established that EAA are the main stimulatory factor for MPS, even in a mixed meal. For example, despite increasing the circulating insulin concentrations, a protein-deficient meal does not result in the activation of muscle protein translation initiation compared with an isocaloric meal containing 20% CP in rodent models (Yoshizawa et al., 1998). Similar studies in humans show that while carbohydrate co-ingestion with protein increased blood insulin level to a greater extent, it did not further stimulate MPS compared with protein alone (Gorissen et al., 2014). Furthermore, when clamping plasma insulin at basal levels, ingestion of amino acids is able to increase MPS and mTOR pathway activation in a dose-dependent manner in young and old men, illustrating that the anabolic effect of amino acids is independent to that of insulin (Cuthbertson et al., 2005). Similarly, a different study using stable isotope infusions elegantly demonstrated that a graded increase of insulin concentration above basal values under the conditions of hyperaminoacidemia did not further stimulate MPS. Conversely, in the same study, increasing the amino acid availability while keeping insulin at fasted levels resulted in a significantly higher rate of MPS (Greenhaff et al., 2008). Based on these and other studies, it is suggested that the primary function of insulin in muscle protein turnover is inhibiting muscle protein breakdown and its role in MPS is likely more permissive rather than stimulatory (Phillips, 2008). Thus, although different dietary components work together to stimulate a general postprandial anabolic response, the presence of dietary protein is key to protein translation initiation. Consequently, while glucose and amino acids likely had a synergistic effect on the insulin response in the current study, we believe that the observed dosing response in the activation of muscle protein synthetic pathways was mainly due to the increased availability of amino acids.

This preliminary study intended to investigate the potential for MPS to be dose–responsive in horses. The next step would be to use purified protein sources to define the required level of dietary protein to maximize mTOR activation and compare the effect of different types of protein sources on the activation of protein synthetic pathways in horses. The protein supplement used in this study provided a mixture of protein sources widely used in horse feeds today (i.e., soybean and alfalfa meal) and would translate to applicable information for modern equine feeding practices. Protein-rich, vitamin–mineral supplements are becoming more prevalent in the equine industry, as feeding practices are shifting away from providing large quantities of cereal-based concentrates to smaller amounts of nutrient-dense feeds. Additionally, these types of feeds are frequently used as muscle-building supplements or complementing forage-based diets for horses with lower energy needs or those sensitive to high levels of dietary non-structural carbohydrates (NSC). Knowledge on the required level of protein intake to maximize MPS in horses would improve the proper utilization of such protein supplements.

In conclusion, this study is the first to provide evidence that the mTOR pathway responds in a dose-dependent manner to graded levels of protein intake in skeletal muscle of mature, nonexercised horses. The consumption of 0.25 g CP/kg BW of a high-protein supplement resulted in near maximal activation of mTOR signaling components with no further increase with the intake of 0.5 g CP/kg BW. These data indicate that the regulatory mechanisms underlying MPS in horses are saturable and that specific amounts of protein needed to maximize synthetic capacity under different nutritional and/or physiological conditions can be determined to further optimize protein feeding in horses.

Glossary

Abbreviations

BW

body weight

CP

crude protein

DE

digestible energy

DM

dry matter

EAA

essential amino acids

ESC

ethanol soluble carbohydrates

FSR

fractional synthesis rates

HPLC

high-performance liquid chromatography

MPS

muscle protein synthesis

mTOR

mechanistic target of rapamycin

NSAID

non-steroidal anti-inflammatory drug

rpS6

ribosomal protein S6

TBST

tris-buffered saline with 0.1% tween

Conflict of interest statement

The coauthor D.A.D. was hired as a consultant by the funder of this project (Versele-Laga, Deinze, Belgium). He also receives a portion of the profits from Cavalor VitAmino. The funding company did not participate in analysis or the decision to publish. All authors declare that they had full autonomy and independency in research and publishing.