Leucine Supplementation Does Not Restore Diminished Skeletal Muscle Satellite Cell Abundance and Myonuclear Accretion When Protein Intake Is Limiting in Neonatal Pigs

ABSTRACT

Background

Rapid growth of skeletal muscle in the neonate requires the coordination of protein deposition and myonuclear accretion. During this developmental stage, muscle protein synthesis is highly sensitive to amino acid supply, especially Leu, but we do not know if this is true for satellite cells, the source of muscle fiber myonuclei.

Objective

We examined whether dietary protein restriction reduces myonuclear accretion in the neonatal pig, and if any reduction in myonuclear accretion is mitigated by restoring Leu intake.

Methods

Neonatal pigs (1.53 ± 0.2 kg) were fitted with jugular vein and gastric catheters and fed 1 of 3 isoenergetic milk replacers every 4 h for 21 d: high protein [HP; 22.5 g protein/(kg/d); n= 8]; restricted protein [RP; 11.2 g protein/(kg/d); n= 10]; or restricted protein with Leu [RPL; 12.0 g protein/(kg/d); n= 10]. Pigs were administered 5-bromo-2’-deoxyuridine (BrdU; 15 mg/kg) intravenously every 12 h from days 6 to 8. Blood was sampled on days 6 and 21 to measure plasma Leu concentrations. On day 21, pigs were killed and the longissimus dorsi (LD) muscle was collected to measure cell morphometry, satellite cell abundance, myonuclear accretion, and insulin-like growth factor (IGF) system expression.

Results

Compared with HP pigs, postprandial plasma Leu concentration in RP pigs was 37% and 47% lower on days 6 and 21, respectively (P < 0.05); Leu supplementation in RPL pigs restored postprandial Leu to HP concentrations. Dietary protein restriction reduced LD myofiber cross-sectional area by 21%, satellite cell abundance by 35%, and BrdU+ myonuclear abundance by 25% (P < 0.05); Leu did not reverse these outcomes. Dietary protein restriction reduced LD muscle IGF2 expression by 60%, but not IGF1 or IGF1R expression (P < 0.05); Leu did not rescue IGF2 expression.

Conclusions

Satellite cell abundance and myonuclear accretion in neonatal pigs are compromised when dietary protein intake is restricted and are not restored with Leu supplementation.

Introduction

During the neonatal period, skeletal muscle relative growth velocity is greater than at any other stage of postnatal life and, at this stage, it is the most rapidly growing protein pool in the body (1, 2). The dietary energy and protein necessary to sustain muscle growth represent a significant component of infant nutrient requirements. In turn, muscle growth is particularly vulnerable if its nutrient needs are not met. Postnatal skeletal muscle growth reflects primarily the hypertrophy of existing muscle fibers and requires the coordinated activation of protein synthesis for the accretion of muscle proteins and the proliferation and differentiation of satellite cells, the resident muscle stem cell, for the addition of myonuclei. Muscle protein synthesis is highest in the postprandial state and is driven by feeding-induced changes in insulin and amino acid concentrations (3, 4). The indispensable amino acid Leu plays a key role in the postprandial response by potentiating mechanistic target of rapamycin complex 1 (mTORC1) signaling and translation initiation (5). The accretion of new myonuclei during postnatal muscle fiber hypertrophy is essential to support the increasing metabolic volume and activity of the growing muscle fibers (6–8). Because muscle nuclei are postmitotic and unable to divide, new nuclei are derived from satellite cell proliferation and the differentiation and fusion of the resulting myogenic precursor cells into fibers (9). The immature muscle of the neonate is highly enriched in activated satellite cells (10), but their proliferation and relative abundance diminish with age (6, 11, 12).

Satellite cell activity is subject to extensive regulation by local and systemic growth factors (13, 14). However, studies investigating the impact of specific nutritional factors on satellite cell function, particularly in the rapidly growing muscle of the neonate when it is exceptionally sensitive to feeding, are limited and results are inconsistent. When newborn calves were fed both inadequate energy and protein from birth to 8 wk of age, there was no difference in the number of satellite cells per fiber compared with calves with high energy and protein intakes (15), although in vitro satellite cells from calves on the low plane of nutrition did not exhibit the expected developmental decline in proliferation rate. Early posthatch starvation of chicks and turkey poults diminished satellite cell proliferation and muscle growth, but starvation imposed at later periods had progressively fewer detrimental effects on satellite cell activity (16, 17). Postnatal undernutrition can compromise myonuclear abundance and muscle growth in mice (18, 19), and muscle mass does not fully recover with the provision of adequate nutrition after weaning. In pregnant sheep, limited nutrient and oxygen supply to the fetus that results in intrauterine growth restriction reduces the numbers of proliferating Pax7+ satellite cells and MyoD+ and myogenin+ myoblasts (20). Collectively, these studies indicate that satellite cell fate during muscle growth in the neonate is affected by nutritional status, but the exact response may depend on stage of development, species, and the nature and severity of the nutritional insult. It is critical to identify whether the response is the consequence of a deficiency of specific nutrients or the result of general metabolic perturbations that inevitably occur when nutrient intake is chronically limited (21).

In addition to promoting translation initiation, Leu can increase the proliferation and differentiation of satellite cells in an mTORC1-dependent manner in vitro (22–25). In contrast, Leu starvation inhibits differentiation of primary murine satellite cells in vitro through changes in myogenic regulatory factor expression independently of mTORC1 activity (26). It is uncertain if this extends to the satellite cell in vivo. In adult rats, long-term Leu supplementation combined with exercise leads to a marginal increase in myonuclear abundance compared with either Leu or exercise alone (27). Previously, we demonstrated that when neonatal pigs were fed a low-protein but energy-sufficient diet, short-term Leu supplementation promoted mTORC1 activation, translation initiation, and protein synthesis in skeletal muscle (28). However, in the long term the Leu supplementation was not sufficient to sustain muscle protein deposition compared with pigs fed adequate protein (29). In these pigs, feeding a protein-restricted diet reduced longissimus dorsi (LD) muscle protein synthesis by 10%, body weight (BW) by 9%, and lean mass gain by 12%. Leu supplementation of the low-protein diet had minimal effects on BW and lean mass gain and had no effect on adiposity despite increasing postprandial plasma Leu concentration, mTORC1 signaling, and protein synthesis in the LD muscle to the same extent as the pigs fed a high-protein diet. Whether satellite cell proliferation and differentiation to support myonuclear accretion is affected by variations in Leu availability should be established. In this study, we tested the hypothesis that dietary protein restriction reduces myonuclear accretion and that Leu supplementation would reverse any reduction in myonuclear accretion in neonatal pigs.

Methods

Animals and housing

The experimental protocol was approved by the Institutional Animal Care and Use Committee of Baylor College of Medicine and was conducted in accordance with the National Research Council's Guide for the Care and Use of Laboratory Animals as described in Columbus et al. (29). Neonatal pigs (Yorkshire × Landrace × Duroc × Hampshire) were obtained from a commercial farm (Burton, TX) at 3 d of age and an average initial BW of 1.53 ± 0.2 kg. Piglets were housed individually in stainless steel cages in an environmentally controlled room at 29°C at the Children's Nutrition Research Center (Houston, TX). Piglets were fed a milk replacement formula (Soweena Litter Life; Merrick's) ad libitum for 3 d. At 4 d of age, a gastric catheter and a jugular vein catheter were surgically placed. Surgical procedures were performed using sterile technique under general anesthesia as described previously (30).

Diets and feeding

At 7 d of age, piglets were randomly assigned to 1 of 3 milk replacement diets: high protein [HP; 22.5 g protein/(kg/d); n = 8, 6 males and 2 females]; restricted protein [RP; 11.2 g protein/(kg/d); n = 10, 9 males and 1 female]; and restricted protein with Leu [RPL; 12.0 g protein/(kg/ d); n = 10, 8 males and 2 females]; diets are described in Supplemental Table 1. The HP diet was formulated to meet or exceed the estimated nutrient requirements of neonatal pigs (31) and supplied protein at a concentration 30% above their estimated requirement [∼16.5 g/(kg/d)]. The RP diet was formulated to contain half the amount of protein present in the HP diet, and supplied protein at a concentration 30% less than their estimated requirement. The RPL diet was formulated to contain the same Leu content as the HP diet. Dietary fat content was adjusted to keep the diets isoenergetic. Dietary lactose content also was adjusted to keep daily carbohydrate intake constant and avoid a confounding effect of carbohydrate on postprandial plasma insulin concentrations. All piglets were fed their respective diet for 21 d through the gastric catheter at 240 mL/(kg/d) in 6 equal meals every 4 h, with feed delivered over 20 min. Piglets were weighed every 3 d and the amount of diet provided daily was adjusted to account for the change in BW.

Plasma hormone and substrate assays

On days 6 and 21, blood samples were collected into heparinized tubes immediately before and 60 min after feeding to determine postabsorptive and postprandial insulin, glucose, and branched-chain amino acid (BCAA) concentrations. Whole-blood glucose (Infinity Glucose Oxidase TR-15521; Thermo Scientific) and plasma insulin (Porcine Insulin Radioimmunoassay No. PI-12K; Millipore Sigma) were determined with commercially available kits. Plasma BCAA concentrations were determined by HPLC (PICO-TAG reverse-phase column; Waters) after deproteinization and derivatization with phenylisothiocyanate, as described previously (3).

5-Bromo-2’-deoxyuridine labelling and tissue collection

To measure myonuclear accretion, 5-bromo-2’-deoxyuridine (BrdU; 15 mg/kg; Calbiochem), a synthetic thymidine analog that is incorporated into newly synthesized DNA of proliferating cells, was injected via the jugular catheter every 12 h from days 6 to 8 of the study, inclusive, for a total of 6 injections. On day 21, samples of the LD muscle were taken between the third and fifth ribs, mounted in gum tragacanth, flash-frozen in liquid nitrogen–cooled isopentane, and stored at −80°C until further processing. The contralateral LD muscle was quantitatively dissected and weighed.

Muscle histology and morphometry

Samples were cryosectioned (10 μm) at −20°C perpendicular to the muscle fibers. For each pig, serial sections were immunostained for BrdU, dystrophin, and total nuclei as described previously (32), or for Pax7, laminin, and total nuclei as described by Gokulakrishnan et al. (33). Briefly, after fixation in 3% paraformaldehyde, sections to be stained for BrdU underwent antigen retrieval followed by incubations with primary antibodies to dystrophin (H300; 4 μg/mL; Santa Cruz Biotechnology) and BrdU (clone G3G4; 2.5 μg/mL; Developmental Studies Hybridoma Bank, University of Iowa). Subsequently, sections were incubated with Alexa Fluor 488–conjugated secondary antibody (10 μg/mL; ThermoFisher Scientific) and biotinylated goat anti-mouse antibody (6 μg/mL; Vector Laboratories), followed by streptavidin-conjugated Alexa Fluor 647 (10 μg/mL; ThermoFisher Scientific). For satellite cell visualization, after antigen retrieval, sections were incubated with primary antibodies to Pax7 (8 μg/mL; Developmental Studies Hybridoma Bank, University of Iowa) and laminin (1 μg/mL; L9393, Millipore Sigma). Subsequently, sections were incubated with Alexa Fluor 488–conjugated goat anti-rabbit IgG (1 μg/mL; ThermoFisher Scientific) and biotin-conjugated anti-mouse (Fcγ subclass 1) IgG (1.2 μg/mL; Jackson ImmunoResearch), followed by streptavidin-conjugated Alexa Fluor 647 (2 μg/mL; ThermoFisher Scientific). Nuclei were visualized by staining with Sytox Orange (0.25 μM; ThermoFisher Scientific). Section images were captured by confocal microscopy (FluoView 300 Laser Scanning Confocal Microscope; Olympus). For each pig, 20–25 nonoverlapping images were taken of each of the BrdU- and Pax7-stained sections, so that 900–1400 fibers were analyzed per pig. Myonuclei were defined as nuclei within the muscle fibers in which ≥50% of the circumference was not in contact with the dystrophin-positive sarcolemma; satellite cells were defined as Pax7+ sublaminal nuclei (Supplemental Figure 1). Because myonuclei are derived from the proliferation of satellite cells, the number of subsarcolemmal BrdU+ myonuclei provides a measure of satellite cell proliferative activity. Fiber cross-sectional area (CSA) and minimum fiber feret diameter were estimated using Image-Pro software (Image-Pro 6.2; MediaCybernetics). Myonuclear domain size was calculated as the mean fiber CSA divided by the number of myonuclei per fiber. When we refer to either myonuclei or satellite cells “per fiber” or their abundance throughout the article, it is understood that this means per cross-sectional profile of a fiber and not the whole fiber or per fiber unit area.

Muscle insulin-like growth factor gene expression

Flash-frozen LD muscle was powdered in liquid nitrogen and homogenized (T25 Ultra-Turrax; IKA) in Trizol reagent (ThermoFisher Scientific); RNA was isolated according to the manufacturer's instructions. Primers for amplification of porcine insulin-like growth factor 1 (IGF1), IGF2, IGF1R, and GAPDH were designed using the Primer-Blast tool from NCBI (Supplemental Table 2). Relative gene expression was quantified on a CFX96 Real-Time PCR detection system (Bio-Rad). Optimal annealing temperature and primer efficiency were assessed, and the absence of primer dimers was determined by melting curve analysis. Data for IGF1, IGF2, and IGF1R are expressed as 2^−ΔΔCt with GAPDH as the reference gene (Δ1) relative to HP values (Δ2). The mean Ct values for GAPDH in LD muscle were not different between the HP, RP, and RPL dietary groups.

Statistical analysis

The study was powered based on anticipated differences between treatments in skeletal muscle protein synthesis (29); secondary analysis of muscle morphometry and histology measurements is presented in the current report. Data are presented as least-square means ± SEs. Data were analyzed using the mixed model, generalized linear model, and correlation procedures of the SAS statistical program (version 9.4; SAS Institute). Myofiber CSA, total myonuclear abundance, myonuclear domain size, Pax7+ satellite cell abundance, BrdU+ myonuclear abundance, and IGF system expression were analyzed by 1-factor ANOVA with diet as the main effect using a mixed model. Insulin, glucose, Leu, Ile, and Val concentrations were analyzed by 3-factor ANOVA with diet, time, and age as the main effects using a mixed model; time was considered as a repeated measure. The difference in minimum muscle fiber feret diameters was analyzed by 1-factor ANOVA with diet as the main effect using a generalized linear model; this analysis accounts for the distribution of fiber diameters within each pig. When interactions were significant, differences among main effects were determined post hoc using the Tukey–Kramer test and were considered significantly different at P ≤ 0.05 and trends at P ≤ 0.10. Pearson correlation coefficients were used to assess correlations between mean fiber CSA and LD weight, mean fiber CSA and myonuclei number, myonuclei and satellite cell numbers, and BrdU+ myonuclei and satellite cell numbers.

Results

The following results are derived from a subset of pigs from a larger study published previously by Columbus et al. (29). Only data from pigs that were administered BrdU for the determination of satellite cell proliferation are included. This subset of pigs (61% of the total) did not differ from the complete set with regards to all common variables that were analyzed.

Plasma insulin, whole-blood glucose, and plasma BCAAs

Plasma insulin, whole-blood glucose, and plasma BCAA concentrations were determined on days 6 and 21 in the postabsorptive state (before feeding, time = 0 min) and in the postprandial state when plasma concentrations peak 60 min after feeding (Table 1). Insulin did not change with age (Age, P = 0.14), but increased with feeding at both ages (Time, P < 0.001) with no effect of diet on the response to feeding (Diet × Time, P = 0.52). Glucose increased with age (Age, P < 0.001), with a feeding effect only at day 6 (Time × Age, P = 0.073); there was no diet effect at either age (Diet, P = 0.12). Leu increased with age only in the postabsorptive state (Time × Age, P = 0.002); nonetheless, there was an increase in Leu with feeding at both ages (Time, P < 0.001), and Leu concentrations were greater for the HP and RPL groups than for the RP group in the postprandial state (Diet × Time, P < 0.001). Ile, like Leu, increased with age (Age, P < 0.001), but largely in the postabsorptive state (Time × Age, P = 0.006); there was an increase in Ile with feeding for all diets (Time, P < 0.001), with the HP group attaining higher concentrations than the RP and RPL groups in the postprandial state (Diet × Time, P = 0.052). Valine also increased with age (Age, P < 0.001) and with feeding (Time, P < 0.001); Val was higher in the postprandial state for the HP than for the RP and RPL groups at both ages (Diet × Time, P = 0.029).

TABLE 1.

Postabsorptive (before feeding) and postprandial (60 min after feeding) plasma insulin, whole-blood glucose, and plasma branched-chain amino acid concentrations in neonatal pigs fed HP, RP, and RPL diets before BrdU administration (study day 6) and at the end of the study (day 21)¹

	Treatment
Diet	HP	RP	RPL	HP	RP	RPL
Time after feeding, min	0	0	0	60	60	60
Day 6
Insulin, μU/mL	1.2 ± 5.9	0.3 ± 5.1	0.7 ± 5.5	29.4 ± 5.1	28.5 ± 4.6	31.9 ± 4.6
Glucose, mg/dL	47.3 ± 9.8	63.9 ± 8.8	64.0 ± 8.8	77.1 ± 9.8	87.0 ± 8.8	85.2 ± 8.8
Leu, μmol/L	237 ± 74c,d	172 ± 66d	335 ± 66c,d	862 ± 74a,b	545 ± 66b,c	1178 ± 66a
Ile, μmol/L	188 ± 35b,c	130 ± 31c	106 ± 31c	540 ± 35a	321 ± 31b	289 ± 31b
Val, μmol/L	445 ± 79b,c	239 ± 70b,c	200 ± 70c	1034 ± 79a	569 ± 70b	454 ± 70b,c
Day 21
Insulin, μU/mL	8.0 ± 5.1	5.1 ± 4.6	6.9 ± 5.1	35.9 ± 5.1	23.4 ± 4.6	36.7 ± 5.1
Glucose, mg/dL	110 ± 10	108 ± 9	121 ± 10	112 ± 10	107 ± 9	136 ± 10
Leu, μmol/L	543 ± 74b	266 ± 66b	387 ± 74b	852 ± 74a	451 ± 66b	1058 ± 74a
Ile, μmol/L	387 ± 35a,b	222 ± 31c	139 ± 34c	519 ± 35a	315 ± 31b,c	358 ± 34a,b
Val, μmol/L	799 ± 89a	435 ± 70b,c	230 ± 78c	1152 ± 79a	555 ± 70b,c	610 ± 78b

¹Values are least square means ± SE calculated from a 3-factor ANOVA; n = 8–10 per dietary group. Relevant P values are reported in the text; all P values are reported in Supplemental Table 3. Labeled means in a row without a common superscript letter differ, P ≤ 0.05. HP, high protein; RP, restricted protein; RPL, restricted protein with Leu.

LD muscle morphometry

An estimate of muscle fiber size was obtained from both the measurement of the minimum feret diameter and mean CSA. The minimum feret diameter for the HP group was 11% higher than for the RPL group (P < 0.05) and 9% higher than for the RP group (P = 0.10; mean diameters of the HP, RP, and RPL groups were 16.1, 14.8, and 14.5 ± 0.54 μm, respectively; Figure 1A). There was a left shift in the distribution of fiber diameters across the whole range of fiber sizes for the RP and RPL groups, indicating that all fibers were affected proportionally by protein restriction. There was no difference among groups in the mean variance, range, kurtosis, and skewness of the distribution; mean kurtosis was −0.45 ± 0.05 (P < 0.001 compared with 0), whereas skewness was not different from 0. Feeding the RP diet reduced mean CSA by ∼21% compared with the HP group (P < 0.05) and this was not reversed when the RP diet was supplemented with Leu (Figure 1B). Among pigs, the variability in mean fiber CSA was highly correlated with the corresponding LD weight (r = 0.73, P < 0.001; Figure 1C), with no separate effect of dietary group on this relation.

FIGURE 1.

Longissimus dorsi muscle morphometry of neonatal pigs fed either an HP, an RP, or an RPL diet for 21 d. (A) Distribution of fiber minimum feret diameters for each dietary group; each data point represents the percentage of fibers within a 2.5-μm bin. (B) Mean fiber CSA. Means without a common letter differ (P ≤ 0.05). Values are least square means ± SE; n = 8–10 per dietary group. (C) Correlation between LD weight of individual pigs and the mean CSA of the fibers analyzed from the corresponding muscle. CSA, cross-sectional area; HP, high protein; LD, longissimus dorsi; n.s., not significant; RP, restricted protein; RPL, restricted protein with Leu.

The number of myonuclei per fiber was ∼10% lower for the RP and RPL muscles than for the HP muscle (P < 0.05; Figure 2A). The number of myonuclei per fiber was highly correlated with mean fiber CSA (r = 0.76, P < 0.001; Figure 2B). Myonuclear domain size was not different among groups (Figure 2C).

FIGURE 2.

Myonuclear (subsarcolemmal nuclei) abundance in the longissimus dorsi muscle of neonatal pigs fed either HP, RP, or RPL diets for 21 d. (A) Mean number of myonuclei per dietary group (averaged for 1000 fibers). Means without a common letter differ (P ≤ 0.05). (B) Correlation between myonuclei per fiber and fiber CSA, showing values for individual animals. (C) Mean myonuclear domain size per dietary group. Values are least square means ± SE; n = 8–10 pigs/group. CSA, cross-sectional area; HP, high protein; n.s., not significant; RP, restricted protein; RPL, restricted protein with Leu.

Feeding the RP diet decreased the abundance of Pax7+ satellite cells per fiber by ∼35% compared with the HP diet (P < 0.001; Figure 3A), and this deficit was not rescued with Leu supplementation (P < 0.001). The reduction in satellite cell number in the RP and RPL compared with the HP muscles was proportionally greater than the reduction in CSA (199 ± 13, 192 ± 13, and 245 ± 14 satellite cells/mm² fiber area, respectively; P < 0.05). Satellite cells were also highly correlated with the abundance of myonuclei per fiber (r = 0.72, P < 0.001; Figure 3B), with no separate effect of dietary group on this relation.

FIGURE 3.

Satellite cell (sublaminal Pax7+ nuclei) abundance in the longissimus dorsi muscle of neonatal pigs fed either HP, RP, or RPL diets for 21 d. (A) Mean number of satellite cells by dietary group (averaged for 1000 fibers). Values are least square means ± SE; n = 8–10 pigs per dietary group; means without a common letter differ (P ≤ 0.001). (B) Correlation between myonuclei and satellite cells, showing values for individual animals. HP, high protein; n.s., not significant; RP, restricted protein; RPL, restricted protein with Leu.

To assess the proliferative activity of satellite cells in vivo, the abundance of BrdU+ myonuclei was determined. When expressed per fiber, LD muscles from RP and RPL pigs were not different but contained 25% fewer BrdU+ myonuclei than HP muscles (P < 0.01; Figure 4A). The number of BrdU+ myonuclei was moderately correlated with the number of satellite cells (r = 0.60, P < 0.001; Figure 4B) and the correlation was similar for all dietary groups.

FIGURE 4.

BrdU+ myonuclei (subsarcolemmal nuclei) per fiber in the longissimus dorsi muscle of neonatal pigs fed either HP, RP, or RPL diets for 21 d. (A) Mean number of BrdU+ myonuclei by dietary group (averaged for 1000 fibers). Values are least square means ± SE; n = 8–10 pigs per dietary group; means without a common letter differ (P ≤ 0.001). (B) Correlation between BrdU+ myonuclei and satellite cells, showing values for individual animals. BrdU, 5-bromo-2’-deoxyuridine; HP, high protein; n.s., not significant; RP, restricted protein; RPL, restricted protein with Leu.

There was no difference in muscle IGF1 and IGF1R mRNA expression among treatment groups, whereas IGF2 mRNA expression for the RP group was decreased by 61% compared with the HP group (P < 0.05; Figure 5). Supplementing the RP diet with Leu did not restore IGF2 expression, which remained similar to the RP group.

FIGURE 5.

Relative expression of IGF1, IGF2, and IGF1R mRNA in the longissimus dorsi muscle of neonatal pigs fed either HP, RP, or RPL diets for 21 d. Values are least square means ± SE; n = 8–10 pigs per dietary group; means without a common letter differ (P ≤ 0.01). HP, high protein; IGF, insulin-like growth factor; RP, restricted protein; RPL, restricted protein with Leu.

Discussion

Skeletal muscle undergoes rapid growth in early postnatal life and reflects the hypertrophy of existing muscle fibers. Undernutrition during this critical stage of muscle development can result in a permanently impaired anabolic response and, therefore, reduced growth even after nutritional rehabilitation (18). Reduced muscle mass has long-term repercussions for muscle function and overall health (34). In this study, we investigated the contribution of myonuclear accretion to the reduced muscle growth in RP pigs and if long-term supplementation of the RP diet with Leu could rescue any impairment to myonuclear accretion incurred by inadequate protein intake.

Skeletal muscle hypertrophy is the product of an increase in both protein deposition and myonuclear accretion. Skeletal muscle protein synthesis in the neonate is exceptionally sensitive to nutritional signals, but little is known about the response of satellite cells to nutrient supply. At weaning, total DNA (a proxy for myonuclear number) was reduced in muscles of rats when their total nutrient intake, including energy, was restricted from birth (35). However, the extent to which the response can be attributed to the resulting metabolic and growth factor adaptations rather than a deficiency in specific nutrients could not be determined. More recently, we demonstrated that by restricting only the protein intake of neonatal pigs, myonuclear accretion was decreased to ∼60% of control levels within 24 h of the diet change (32). It was unclear if this reduction in satellite cell activity might be mitigated by Leu supplementation, an indispensable amino acid and potent activator of the mTORC1 signaling pathway that coordinates anabolic processes (36) and cell cycle progression (37). We anticipated that Leu supplementation of a protein-restricted diet would promote greater muscle fiber hypertrophy and increase satellite cells and myonuclei per fiber.

The blunted growth of the LD muscle in response to protein restriction was reflected in the treatment differences in fiber dimensions. Although the LD muscle is largely comprised of glycolytic type IIB fibers, which have greater CSA and feret diameter than other muscle fiber types in the pig (38), the reduction in size was proportionally similar across the range of fiber areas, suggesting that there was no preferential effect on larger compared with smaller fibers. Supplementation with Leu was not beneficial for muscle growth in the context of its overall mass or fiber CSA. The abundance of myonuclei was highly correlated with mean LD fiber CSA. The absence of any difference between dietary groups in the ratio of CSA to myonuclear abundance, a measure of myonuclear domain size, provides strong evidence that the regulation of myonuclear and muscle protein accretion in the neonatal period is highly interdependent. This contrasts with the response of mature skeletal muscle where the myonuclear domain size appears to be more flexible and possibly fiber type–dependent (39). Leu supplementation during protein restriction did not restore myonuclear abundance.

The lower abundance of myonuclei per fiber in the protein-restricted groups was correlated with the reduced frequency of satellite cells per fiber, and there was also no gain with Leu supplementation. A lower abundance of BrdU+ myonuclei in the RP and RPL muscle fibers indicates that satellite cell proliferative activity was blunted when protein intake was limiting. Although accelerated myonuclear production in the post-BrdU administration period would also result in a reduction in the proportion of myonuclei that are BrdU+ by the end of the study in RP and RPL muscles, this would be associated with a relative increase in total myonuclei per fiber, which was not the case. The moderate positive correlation between Pax7+ nuclei and BrdU+ myonuclei suggests that the differentiation of the satellite cells into myonuclei is likely not altered. If differentiation were impaired, we would anticipate a greater deficit in total myonuclei relative to satellite cells in RP and RPL pigs, which also was not the case. However, we cannot exclude the possibility that satellite cell proliferation was reduced while differentiation of myogenic precursor cells into myonuclei was greater in pigs with restricted protein intake. This would accelerate the normal developmental decline in satellite cell numbers. Indeed, the greater deficit in Pax7+ satellite cells relative to total myonuclei or to mean fiber CSA in RP and RPL pigs is consistent with this possibility. A relative increase in satellite cell differentiation together with a reduction in satellite cell abundance may have implications for the establishment of the quiescent satellite cell pool and, thus, affect the capacity for long-term muscle growth, muscle repair, and the remodeling of its associated extracellular matrix (40).

The reduction in satellite cell activity during dietary protein restriction indicates that protein intake is a key determinant of myonuclear accretion and muscle growth in the early postnatal period. Unlike much previous data derived from rodent models of postnatal malnutrition, the neonatal pig model enabled us to maintain adequate intakes of all other nutrients, including energy, and thus to exclude reduced energy intake as a possible contributing factor. Accordingly, to assess if the predominant hormonal and nutritional signals were affected by dietary treatments, we measured blood glucose and plasma insulin and BCAA concentrations in the postabsorptive and postprandial states before the period of BrdU administration on day 6 and at the end of the study on day 21. There was no difference among groups at either age in postabsorptive or postprandial insulin or glucose concentrations and, therefore, they are unlikely to have contributed to the observed responses. The effect of protein intake, however, was manifested by the BCAA concentrations primarily in the postprandial state at both ages when values for the RP group were 50–60% of those for the HP group. Compared with the RP group, the RPL group differed only in Leu concentration, reflecting its supplementation in the diet. These differences were less pronounced in the postabsorptive state, especially on day 6. As observed previously (41), BCAA concentrations in the postabsorptive state increase with age for all groups and may reflect a diminished rate of substrate utilization in older pigs consistent with the developmental decline in protein turnover. However, the overall pattern of Leu substrate availability among groups, and to which the satellite cells were exposed, remained relatively unchanged over the duration of the study. In a similar study, albeit one in which energy intake was restricted in addition to protein with and without Leu supplementation, the differences between groups after 21 d of treatment were similar to those present at 9 d (42). In the present study, there also was no difference in postabsorptive and postprandial plasma Ile and Val concentrations between the RP and RPL groups on days 6 and 21. This excludes the possibility that a Leu-induced increase in BCAA catabolism effectively lowered the Ile and Val available to the LD muscle. Nonetheless, it is unclear whether satellite cells in vivo respond to Leu in the same manner as muscle fibers, where a distinct rise and fall in Leu is needed to maximally stimulate translation initiation and protein synthesis through mTORC1 activation (43). It is possible that sustained elevation in plasma Leu concentrations, as would occur when myoblasts are studied in vitro, may be required to show a clear role of Leu on satellite cells when protein is limiting.

The IGF system, including IGF-I, IGF-II, and their associated binding proteins and receptors, plays a critical role in the regulation of myogenesis and postnatal muscle growth through the autocrine/paracrine actions of locally produced peptides (44–46). IGF-I enhances the proliferation of satellite cells, but also promotes differentiation in a time- and concentration-dependent manner (47). IGF-II has been considered a key regulator of fetal rather than postnatal growth and promotes differentiation but not proliferation of satellite cells in vitro (48). A genetic variant in pigs has also been identified in which IGF-II expression is sustained postnatally and increases muscle mass, possibly by promoting the formation of muscle fibers in the fetus (49#x2013;51).

At the whole-body level, the IGF system is sensitive to energy, protein, and amino acid status (52–54), but the responsiveness of locally produced peptides in muscle to nutrients is poorly defined. In the present study, dietary protein restriction decreased the expression of IGF2 but not IGF1 or IGF1R transcripts in the LD muscle. Leu supplementation of the RP diet did not restore IGF2 expression. Some studies have reported no effect of chronic protein restriction on muscle IGF1 expression (52, 55, 56), whereas we and others have observed a reduction in muscles of younger animals (18, 57). In the latter studies, however, the muscles were sampled at a younger developmental age after a shorter duration of nutrient restriction. Thus, it is possible that earlier in the current study, during the period of BrdU administration, IGF1 expression could have been lower and contributed to lower satellite cell proliferation. Conversely, IGF2 expression was markedly affected by protein restriction. In globally undernourished suckling mice, we found no difference in hind limb muscle IGF2 expression, suggesting that it may be less sensitive than the IGF1 response to short-term global nutrient restriction (18, 58). Amino acids have been reported to upregulate IGF2 expression via an mTORC1 kinase-independent mechanism and to promote myoblast differentiation in vitro (59), but whether this also occurs in vivo has not been established. Our findings, which are inconsistent with the response of C2C12 myoblasts, could reflect differences in species and/or the regulation of IGF2 expression by differentiated muscle in vivo compared with that of muscle cell lines in vitro. However, of pertinence to our observation is a recent report that demonstrates that IGF-II plays a vital role in the function of intestinal and neuronal adult stem cell niches (60). These observations are consistent with the parallel reductions we observed in IGF2 expression, satellite cell numbers, and myonuclear accretion.

These findings support the conclusion that the Leu-induced activation of protein synthesis through mTORC1 is not sufficient to sustain muscle growth in meal-fed neonatal pigs for a prolonged period when protein intake is suboptimal (29). Our results also support an essential role for satellite cells in postnatal muscle growth. We have identified that satellite cells require an optimal amount of dietary protein to sustain the level of proliferative activity needed for the accretion of myonuclei during postnatal muscle fiber hypertrophy. The decrease in myonuclear accretion in pigs fed a protein-restricted diet provides a mechanism that, in parallel to reduced protein synthesis rates, explains why many low-birth-weight infants who fail to receive adequate nutrition in early life have less lean mass in adulthood (61, 62). It remains to be determined whether the skeletal muscle of such infants is capable of responding appropriately to other anabolic stimuli to promote the rapid growth rate typical of the neonatal period.

Notes

Supported by NIH grants AR-044474, HD-072891, and HD-085573 (to TAD), National Institute of Food and Agriculture grant 2013-67015-20438 (to TAD), and Agricultural Research Service grants 3092-51000-060 (to TAD) and 3092-51000-056 (to MLF).

Author disclosures: MR, DAC, JS-W, AS, HVN, RF, TAD, and MLF, no conflicts of interest.

Supplemental Tables 1–3 and Supplemental Figure 1 are available from the “Supplementary data” link in the online posting of the article and from the same link in the online table of contents at https://academic.oup.com/jn.

Abbreviations used: BCAA, branched-chain amino acid; BrdU, 5-bromo-2’-deoxyuridine; BW, body weight; CSA, cross-sectional area; HP, high protein; IGF, insulin-like growth factor; LD, longissimus dorsi; mTORC1, mechanistic target of rapamycin complex 1; RP, restricted protein; RPL, restricted protein with Leu.