Maternal Undernutrition Alters The Skeletal Muscle Development and Methylation of Myogenic Factors in Progeny: A Randomized Controlled Trial

Maternal malnutrition alters protein synthesis and muscle development in offspring. Epigenetic modication is one of the mechanisms involved in this developmental programming. However, the role of the methylation of myogenic factors remains unclear. Twenty-four pregnant goats were assigned to a control (100% of the nutrients requirement, n = 12) or a restricted group (60% of the nutrients requirement, n = 12) between 0.3 and 0.7 gestation (G). Descendants were harvested at 0.7 G and at 91 days after birth to analyze the ber area and ber type of femoris muscle and to examine the effects on the mTOR signal pathway and methylation of the promoter regions of myogenic factors. Maternal undernutrition increased the ber area of the vastus muscle in the fetuses (P < 0.05). Undernutrition enhanced (P < 0.01) the proportions of MyHCI and MyHCIIA bers in the vastus muscle of fetuses and semitendinosus (ST) muscle of kids, while the proportion of MyHCIIX bers were decreased (P < 0.01) in the restricted fetuses and kids. The mRNA expression levels of EYA1 and MYOZ2 in the restricted fetuses were downregulated (P < 0.05), but no difference was observed in the restricted kids (P > 0.05). The percentage of DNA methylation at the + 530 CpG site of the MYF5 gene in the restricted fetuses was increased (P < 0.05), but the methylation percentages of the MYF5 gene at the +274,280 CpG site and of the MYOD gene at the +252 CpG site in the restricted kids were reduced (P < 0.05). The mTOR protein signals were downregulated (P < 0.05) in the restricted fetuses and kids. These results indicated that maternal undernutrition altered the ber type of femoris muscle in offspring, but its relationship with the methylation in the promoter regions of myogenic genes needs to be further studied.


Abstract
Maternal malnutrition alters protein synthesis and muscle development in offspring. Epigenetic modi cation is one of the mechanisms involved in this developmental programming. However, the role of the methylation of myogenic factors remains unclear. Twenty-four pregnant goats were assigned to a control (100% of the nutrients requirement, n = 12) or a restricted group (60% of the nutrients requirement, n = 12) between 0.3 and 0.7 gestation (G). Descendants were harvested at 0.7 G and at 91 days after birth to analyze the ber area and ber type of femoris muscle and to examine the effects on the mTOR signal pathway and methylation of the promoter regions of myogenic factors. Maternal undernutrition increased the ber area of the vastus muscle in the fetuses (P < 0.05). Undernutrition enhanced (P < 0.01) the proportions of MyHCI and MyHCIIA bers in the vastus muscle of fetuses and semitendinosus (ST) muscle of kids, while the proportion of MyHCIIX bers were decreased (P < 0.01) in the restricted fetuses and kids. The mRNA expression levels of EYA1 and MYOZ2 in the restricted fetuses were downregulated (P < 0.05), but no difference was observed in the restricted kids (P > 0.05). The percentage of DNA methylation at the + 530 CpG site of the MYF5 gene in the restricted fetuses was increased (P < 0.05), but the methylation percentages of the MYF5 gene at the +274,280 CpG site and of the MYOD gene at the +252 CpG site in the restricted kids were reduced (P < 0.05). The mTOR protein signals were downregulated (P < 0.05) in the restricted fetuses and kids. These results indicated that maternal undernutrition altered the ber type of femoris muscle in offspring, but its relationship with the methylation in the promoter regions of myogenic genes needs to be further studied.

Background
During myogenesis, the numbers of myoblast, primary muscle ber and secondary muscle ber are determined before birth in humans [1] and domestic animals [2,3]. The skeletal muscle is highly susceptible to nutritional de ciencies due to the low priority of nutrient acquisition in mammals [4,5]. Nutritional de ciency in early gestation lessens the number of primary muscle bers [6], but the effects are generally minor due to low nutrients requirement by early fetuses. Secondary muscle ber proliferates from around 0.3 G (30% of gestation stage), and this process ceases at about 0.7 G [1][2][3], which constitutes nearly 95% of the nal ber population [2]. Insu cient nutrition during midgestation (0.3 -0.7 G) affects the number, size, and type of secondary bers before [2,7] and after birth [8][9][10]. It is widely accepted that skeletal muscle development is nutritionally programmed in utero [4,5], and understanding the underlying mechanisms of this programming helps to prevent nutritional sarcopenia in humans and improve meat production in livestock.
Alteration of muscle phenotype can partly be interpreted by the changes in hormone response [11], mTOR signaling pathway [7], myogenic regulatory factors expression [12,13], and mitochondrial function [14] in model animals. However, an unresolved question is how this phenotype is programmed during ontogenetic muscle development. Epigenetic modi cation is one of the key mechanisms responsible for the persistent programming [15]. Myogenic factors including MYOD, myogenin, MYF5, and MYF6 (or MRF4) govern skeletal muscle determination and differentiation [16], among which MYF5 and MYOD are the master regulators. Epigenetic modi cation, including DNA methylation of enhancer chromatin [17,18], histone acetylation, and long noncoding RNA in mice [19,20] of the MYF5 and MYOD genes, is associated with the expression of muscle-speci c genes during myogenesis and differentiation. Methylation of the promoter region is a regulation target in intrauterine skeletal programming, for example, the GLUT4 promoter in pigs [21]. However, the relationship between methylation of the promoter regions of the Myf5 and MYOD genes and the muscle developmental programming under nutrient stress remains uncharacterized. DNA methylation generally reduces the expression of muscle-related genes [22]. We thus hypothesized that maternal undernutrition during midgestation would decrease the expression of MYF5 and MYOD genes via increasing the methylation of their promoter regions.
Small ruminants are optimum animal models for intrauterine programming [23], and 40% energy restriction in pregnant goats leads to hepatic metabolic programming in offspring in our previous study [24]. In the present study, we investigated the effects of 40% maternal undernutrition during mid-gestation on the muscle ber histology, myosin heavy chain expression, methylation modi cation of myogenic factors, and the mTOR signal pathway in the skeletal muscles of prenatal and postnatal goats.

Ethical approval
All of the protocols for this study were approved by the Animal Care Committee according to the Animal Care and the Use Guidelines of the Institute of Subtropical Agriculture, the Chinese Academy of Sciences, Changsha, China (No. KYNEAAM-2015-0009).

Experimental design and animal management
Twenty-four goats (45 ± 3 d of gestation, Liuyang black goats, a local breed) were selected and tested by portable ultrasonography (Aloka SSD-500 with a 5-MHz linear probe, Aloka, Shanghai, China). Dams were randomly assigned into the control [100% of the nutrients requirements suggested in the Chinese Meat Goat Requirements (2004), n = 12] or restricted group (60% of the nutrients requirements, n = 12) according to body weight (BW) and litter size. All dams were placed within individual pens and fed a diet twice (0800 h and 1600 h) per day with a 50:50 ratio of concentrate to roughage, with free access to drinking water. The ingredients and composition of the experimental diet on a dry matter basis are shown in additional Table S1. The feeding of the restricted group was conducted by providing 60% of the feed allowance of the control group during days 45 -100 of gestation, and the actual restriction level (1.04 kg/d for the control group vs 0.62 kg/d for the restricted group) was 60.2% after measurement of the daily feed allowance and refusal. At day 100 of gestation, six pregnant dams per group were selected for harvesting, and ten fetuses (three singlets, two sets of twins, and one set of triplets) in each group were obtained. The ratio of females to males was 7:3 in the control group and 6:4 in the restricted group.
After day 100 of gestation, feed restriction was removed, and the remaining dams were fed to supply 100% of the nutrients requirements and were managed as before during the following experimental period. After parturition, neonatal kids were nursed by their dams until preweaning between days 50 and 60. Preweaning was conducted by separating offspring from their dams in the daytime from 0800 to 1600 h, and a mixed diet of starter and fresh Miscanthus spp. was provided at a ratio of 20:80 during this period. Complete weaning was performed on day 60. Then, all kids were given ad libitum access to the above diet and had free access to drinking water. The ingredients and composition of the kid diet are shown in a Table S1. At 90 days of age, eight kids in each group were obtained for harvesting. The ratio of females to males was 3:5 in the control and 4:4 in the restricted group.
Body weight measurement and muscle tissue sampling At day 100 of gestation and at day 90 after birth, the feeding of all selected animals was withdrawn for 24 h, while clear water was offered freely. The empty BW of fetuses and kids were reported previously [24]. Following electric shock and exsanguination, samples of semitendinosus (ST) and vastus lateralis (VL) muscle were collected from the kids, while the gross vastus muscle in the fetuses was sampled because it was di cult to dissect into individual muscles. The sliced tissue samples were immersed in 10% formalin xing solution for histology measurement, while another aliquot of samples was quick-frozen in liquid nitrogen and then stored at -80°C for further analysis.

Histology
After xation, the muscle samples were embedded in para n, sliced at a thickness of 8 μm using a rotary microtome (RM2016, Leica Microsystems Inc., Wetzlar, Germany), and stained with hematoxylin (AS1055A, Aspen Biological, Inc., Wuhan, China) and eosin (AS1094, Aspen Biological, Inc.) according to Zou et al. [12]. Ten different microscopic elds of each section and ve sections per sample were randomly chosen to determine the muscle ber area using the Image Pro Plus 6.0 software (Media Cybernetics, Bethesda, USA).

Quantitative real-time PCR
The extraction and quali cation of total RNA and the analyses of messenger RNA (mRNA) expression levels were performed using the SYBR green-based method with gene-speci c primers (Table 1) according to Zhou et al. [24]. A melt curve analysis was conducted to con rm speci c product ampli cation. ACTG1 was used as the reference gene, and the real-time PCR data were calculated using the 2 −ΔΔCt method [25]. The expression of mRNA was presented as the fold change relative to the reference gene.

Immunoblotting analysis
The extraction and quali cation of total protein and the immunoblotting analysis were carried out according to the method described [26]. ) and covered 55 CpG sites, among which 42 CpGs could be effectively quanti ed. The spectral data were pre-processed and analyzed according to the method of Suchiman et al. [27]. The methylation level of the restricted group in each CpG site was expressed as the value relative to the control group.

Statistical analysis
Data were analyzed by a mixed model with treatment, gender and litter size as xed factors, and initial BW of dams as the covariate. Statistical signi cance was considered at P < 0.05, and Sidak method was applied to compare means. All results are presented as the means and standard errors (SE).

Body weight (BW) and area of muscle bers
The effect of maternal undernutrition on the muscle ber area of fetuses and kids is presented in Fig. 1. The BW of fetuses were unaffected (P > 0.05) by undernutrition, but the BW of restricted kids was decreased relative to controls (P = 0.027). The muscle ber area of the vastus muscle of restricted fetuses was increased (P = 0.032), while it was unaffected (P > 0.05) in the ST and VL muscles of restricted kids. Gender and litter size did not affect the area of muscle bers in the fetuses or kids (P > 0.05).

Myosin heavy chain expression
The effect of maternal undernutrition on the expression of myosin heavy chain in the muscle tissues of fetuses and kids is presented in Fig. 2. The expression of MyHCIIB mRNA in the fetuses and kids was not detectable, so the expression of MyHCIIB protein in these samples was not analyzed further. Compared to the control group, maternal undernutrition did not change (P > 0.05) the expression of MyHCI, MyHCIIA, or MyHCIIX mRNA in the vastus muscles of fetuses ( Fig. 2A), or in the ST (Fig. 2B) and VL (Fig. 2C) muscles of kids. However, the protein expression levels of MyHCI and MyHCIIA were increased (P < 0.05) both in the vastus muscles of fetuses (Fig.  2D) and in the ST muscles of kids (Fig. 2E), while the MyHCIIX protein expression in the fetuses and in the ST muscles of kids was decreased (P < 0.01). Protein expression levels of MyHCI, MyHCIIA, and MyHCIIX in the VL muscles of kids (Fig. 2F) were not affected (P > 0.05). Gender and litter size did not affect (P > 0.05) the mRNA or protein expression of MyHCI, MyHCIIA, and MyHCIIX genes in the fetuses or kids.

mRNA expression and DNA methylation of myogenic factors
Since no signi cant differences in myosin heavy chain expression of muscle resulting from maternal undernutrition were observed, we further determined the mRNA expression and DNA methylation of myogenic factors in the muscles of fetuses and kids (Fig. 3). Maternal undernutrition did not in uence (P > 0.05) the expression of MYF5, MYF6, MYOD, or MYOG mRNA in the vastus muscles of fetuses (Fig. 3A) or in the ST muscles of kids (Fig. 3B). The expression levels of EYA1 (P = 0.044) and MYOZ2 (P = 0.038) mRNA in the vastus muscles of restricted fetuses were downregulated, while no difference was observed in the ST muscles of kids (P > 0.05). Maternal undernutrition increased (P = 0.040) the methylation level of +530 CpG site of MYOD gene in the fetuses, while the methylation levels of the +274, 280 CpG site of the MYF5 gene and the +252 CpG site of the MYOD gene in the restricted kids were decreased (P < 0.05). The methylation levels of the remaining CpG sites of MYOD and MYF5 genes were not affected (P > 0.05) in the fetuses or kids. Gender and litter size did not affect (P > 0.05) the mRNA expression of these genes or the CpG methylation of MYF5 and MYOD promoters in the fetuses. Gender affected (P < 0.05) the MYF5 and SIX1 mRNA expression in kids, which were greater in the females than that in the males. And the methylation of the MYF5 +32 CpG site in the female kids was higher (P = 0.041) than that of the males.
mTOR Signaling Pathway Maternal undernutrition reduced (P < 0.01) the protein expression of mTOR, p-mTOR, and the ratio of p-mTOR/mTOR in fetal vastus muscle. Maternal undernutrition also reduced (P = 0.037) the mTOR protein expression in kids and tended to reduce (P = 0.088) the PI3K phosphorylation level, while the effects on the PI3K, AKT, and p-AKT proteins and the ratio of p-AKT/AKT in kids were not signi cant (P > 0.05). Gender and litter size did not affect the above protein expression in the fetuses or kids (P > 0.05).

Discussion
During the past two decades, there has been increased research interest in the effect of maternal undernutrition on muscle development of offspring. In this study, maternal undernutrition altered the muscle bers and composition of the vastus muscle in fetuses and kids, and downregulated the mRNA expression of myogenic factors in the fetuses. DNA methylation levels of several sites in the MYF5 and MYOD promoter regions in the fetuses and kids were also affected by maternal undernutrition. Many studies have con rmed that maternal undernutrition in the second trimester is closely associated with the thrifty phenotype of offspring in human epidemiological surveys [28,29] and in animal model trials [8,30,31]. In the present study, no changes in the BW of fetuses and the diminished BW in the restricted kids were consistent with a compensation effect by the mother in utero and a thrifty phenotype after birth. However, the increment of muscle ber area in restricted fetuses in this study was not consistent with the ndings (no signi cant change) in the fetuses of cattle [13] and pigs [12] during midgestation or in sheep (the decreased size) [7]. According to Wigmore and Stickland [2], the primary bers begin to lose the central myo bril-free region at around 0.65 G, and then the diameter of the muscle bers decreases rapidly. The nutrient-restricted fetuses can prolong the maturation of the primary muscle bers and delay the time of disappearance of the central myo bril-free cytoplasm. In this study, the observation period of fetal muscle bers coincided with this stage (0.67G), and this may be the reason for the increase in the area of muscle bers of the restricted fetuses.
Skeletal muscle is generally divided into four types according to myosin heavy chains (MyHC): slow-twitched type I (MyHCI), fast-twitched type IIA (MyHCIIA), IIB (MyHCIIB), and IIX (MyHCIIX). Previously, Zhu et al. [10] reported that 50% maternal undernutrition increased the MyHCIIB ratio and decreased the percentage of MyHCIIA of the longissimus dorsi (LD) muscle in the restricted lambs at eight months of age, while the type composition of LD and ST muscles in the 50% restricted lambs at 150 days of age was unaffected [8]. In contrast, Fahey et al. [32] reported fewer fast bers and more slow bers in the LD and VL muscles of 50% maternally restricted 14-day-old lambs. In this study, the types of MyHCI and MyHCIIA were increased, but MyHCIIX was decreased both in the vastus muscle of 40% restricted fetuses and in the ST muscles of restricted kids, while the VL muscles in offspring were not affected. These results suggest that the effects of similar maternal undernutrition during midgestation on the muscle ber development vary, and the cause needs further research. Moreover, comparing LD and ST bers from the same muscle position, a similar degree of feeding restriction (such as 50%) exerted different effects on muscle ber development [8,10,32]. It has been speculated that this discrepancy is related not only to the muscle ber composition in different muscle positions but also to the timing of phenotypic plasticity induced by different types of nutrient restriction [33,34].
In this study, we noted an inconsistency in the expression between mRNA and protein levels of MyHC isoforms in the fetal vastus and kid ST muscles. Similar differences in mRNA and protein abundances of MyHC genes have been reported in other species [35]. Generally, the protein accretion rate of hindlimb muscle is approximately 0.17% per day in ovine fetuses [36] and 0.8% in young lambs [37]. However, transcriptional and posttranscriptional regulation of MyHC isoforms mRNA in muscle tissue is active and sensitive to environmental cues (e.g., hormones or metabolites) [38,39]. MyHC expression at the protein level represents the existing ber type composition within the muscle. Therefore, the effect of protein expression was highlighted.
Consistent changes in the MyHC types in the vastus muscle and the ST muscle of the goats hinted that metabolic programming could have taken place in the present study. DNA methylation is one of the universal mechanisms regulating this phenotypic change [40]. MYF5 and MYOD gene methylation is important in regulating muscle differentiation. Increased methylation of three important CpG sites in the MYOD1 gene and decreased MYOD1 mRNA expression have been observed during the differentiation of human primary muscle bers from myoblasts to myotubes, while the methylation and mRNA levels of the MYF5 gene were decreased [41]. However, there are few reports on the methylation modi cation of these two MRF factors under intrauterine nutrition restriction. Our results revealed that maternal nutritional restriction during midgestation affected the methylation of individual sites in the MYOD and MYF5 promoter regions of the fetuses and kids. These affected sites may not be closely correlated with the regulation of MYOD and MYF5 mRNA expression, because the mRNA expression of MYOD and MYF5 was unaffected. Previous studies have shown that maternal low-amino acid diets in the middle and late stages of pregnancy reduced the methylation level of the CpG island in IGF2R gene intron 2 in the longissimus muscles of sheep fetuses at 130 days [42]. A similar study found that the average CpG island methylation in the PGC1A promoter sequence in the skeletal muscle of introuterine growth retardation rats with protein restriction during pregnancy increased, while the GLUT4 and PGC1A mRNA expression and glucose tolerance decreased [43]. In our study, we have not yet identi ed the key methylation modi cation sites that can explain the phenotypic changes in offspring, thus further research is needed.
The mRNA expression levels of myogenic factors MYOG and MRF4 in fetuses and kids were unaffected, but the levels of EYA1 and MYOZ1 mRNA were decreased in the restricted fetuses. SIX and EYA1 genes are speci cally involved in the differentiation of limb buds [44]. The SIX1 and EYA1 proteins accumulate preferentially in the nuclei of fast-twitch muscles and induce a ber-type transition with the replacement of myosin heavy chain I and IIA isoforms by the faster IIB and/or IIX isoforms [45]. Furthermore, MYOZ2 is expressed in slow-twitch skeletal muscle, and the downregulation of MYOZ2 is associated with increased slow-twitch muscle bers [46]. The downregulation of EYA1 and MYOZ2 mRNA in the restricted fetuses is consistent with the increases in MyHCI and MyHCIIA types and the decrease in MyHCIIx in the present study.
The development of skeletal muscle is a highly complicated process involving nutritional factors and numerous signaling pathways to regulate muscle-speci c transcription factors, and mTOR signaling is a key regulator of skeletal muscle development at distinct stages of myogenesis [47]. In this study, mTOR protein was downregulated in the femoris muscles of fetuses and kids. We considered that compared to the normal energy requirement, a 40% maternal undernutrition aggravated the overall lack of proteins and amino acids, resulting in the downregulation of the mTOR signals in the offspring, thereby affecting the protein synthesis of muscle tissues. Furthermore, the mTOR protein controls the muscle-speci c miR-1 transcription in mice that regulates the stability of MYOD and MYF5 expression [48]. In addition, other miRNAs such as miR-133, miR-206, and miR125b are also linked to the mTOR pathway to regulate muscle ber differentiation [48,49], and the musclespeci c expression of miRNAs can be regulated by nutrients through epigenetic mechanisms [47]. This provides new ideas for studying the mechanism of muscle ber development programming under maternal nutrition restriction.

Conclusion
The maternal undernutrition during midgestation increased the expression of MyHCI and MyHCIIA proteins in the vastus muscles of fetuses and the ST muscles of kids and reduced MyHCIIX protein expression, hinting at the developmental programming of skeletal muscle bers. Transcripts of EYA1 and MYOZ2 genes in the restricted fetuses were downregulated, and the mTOR protein signals declined in the restricted fetuses and kids. Maternal feeding restriction may prompt the switch from MyHCIIX to MyHCI and IIA in the femoris muscles of fetuses and kids associated with repression of the mTOR pathway. This programming was associated with the reduction of EYA1 and MYOZ2 transcripts, but the association between the mRNA expression and the DNA methylation in the promoter region of the myogenic factors MYF5 and MYOD needs to be further investigated.

Declarations
Ethics approval and consent to participate