Estrogen Promotes Fetal Skeletal Muscle Myofiber Development Important for Insulin Sensitivity in Offspring

Abstract

Using our nonhuman primate baboon model, we showed that offspring born to mothers deprived of estrogen during the second half of gestation exhibited insulin resistance and a deficit in first phase insulin release. Although insulin resistance was not due to an impairment of fetal or offspring growth, nor to an alteration in adipose or hepatic sensitivity to insulin, skeletal muscle microvacularization critical for delivery of nutrients/insulin was significantly reduced in fetuses and offspring deprived of estrogen in utero. Skeletal muscle myofiber maturation occurs in utero and estrogen modulates myofiber growth in adults. Therefore, the current study determined whether fetal skeletal muscle development was altered in baboons in which estradiol levels were suppressed/restored during the second half of gestation by maternal treatment with letrozole ± estradiol benzoate. In estrogen-suppressed animals, fetal skeletal muscle fascicles were structurally less organized, smaller, and comprised of slow type I and fast type II fibers, the size, but not the number of which were smaller than in untreated baboons. Moreover, the proportion of non-muscle fiber tissue was greater and that of muscle fibers lower in estrogen-deprived fetuses. Thus, the maintenance of fetal body weight in estrogen-deprived animals was maintained at the expense of muscle fibers and likely reflected increased deposition of non-muscle proteins. Importantly, fetal skeletal muscle development, including fascicle organization, myofiber size and composition was normal in baboons treated with letrozole and estradiol benzoate. Collectively, these and our previous findings support our proposal that exposure of the fetus to estrogen is important for fetal skeletal muscle development and glucose homeostasis in adulthood.

Introduction

Using our nonhuman primate baboon model, we showed that offspring born to mothers deprived of estrogen during the second half of gestation exhibited insulin resistance and a deficit in first phase insulin release [1, 2], however, fetal and offspring body weights were normal. Thus, insulin resistance was not due to an impairment of growth, nor to any alteration in adipose or hepatic sensitivity to insulin [3]. Based on these findings and the fact that skeletal muscle accounts for more than 70% of total glucose disposal [4, 5], we proposed that estrogen plays an important role in programming mechanisms in fetal skeletal muscle essential for glucose homeostasis in offspring [1].

Recently, we demonstrated that the number of microvessels and the microvessel/skeletal muscle fiber ratio, which is important for delivery of insulin and glucose to myofibers, were 50% lower in near term fetuses deprived of estrogen in utero. Moreover, impaired microvessel/muscle fiber ratio was sustained in offspring that also subsequently developed systemic vascular dysfunction, including reduced endothelial vasodilation and hypertension in addition to insulin resistance [6]. Therefore, we proposed that the elevation in estrogen during the second half of primate pregnancy promotes systemic micro-vascularization essential for insulin sensitivity and vascular homeostasis in adulthood [6].

It is well established [7 for review] that the total number of myofibers that comprise skeletal muscle is established prior to birth in rodents, sheep, other mammals [8-10] and humans [11]. Moreover, myogenesis continues to support myofiber hypertrophy during late gestation and early postnatal life [7, 12-14]. Despite the importance of myofiber development in utero on physiologic and metabolic homeostasis in adulthood, our understanding of the factors regulating fetal skeletal muscle myogenesis and myofiber maturation in late gestation remains incomplete. Studies in humans, transgenic animals, and cell lines support an important role for IGF1 as well as insulin [7, 15]. For example, IGF-1 heterozygous knockouts exhibit reduced muscle mass [16] and homozygous knockouts severe muscle hypoplasia [7, 17, 18]. Moreover, muscle development as well as overall body growth are significantly impaired in animals in which nutrient or oxygen delivery to the fetus is reduced experimentally by maternal nutrient restriction, hypoxia or impairment of placental blood flow [7, 15, 19, 20]. Infants born with low birth weight have lower muscle mass in adulthood [21-23]. Thus, regulation of the expansion of fetal skeletal muscle myofibers appears to be linked in large part to overall growth of the fetus. However, the impairment of fetal skeletal muscle microvascular development and the ontogeny of insulin resistance in offspring in our baboon model of estrogen deprivation in utero occurred in the absence of any change in birth weight of fetuses and their subsequent growth as offspring. Therefore, the current study determined whether the rise in estrogen levels in pregnancy is important for fetal skeletal muscle myofiber development. To examine this possibility, fetal skeletal muscle myofiber size, number and type were determined near term in baboons in which estradiol levels were suppressed or restored during the second half of gestation by maternal administration of letrozole or letrozole and estradiol benzoate, respectively.

Materials and Methods

Animals

The present animal research study was conducted in accordance with the ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines to enable evaluation of the rigor and reproducibility of the methods, statistical analyses and results. Female baboons (Papio anubis), originally obtained from the Southwest National Primate Research Center (San Antonio, TX) were housed individually in large primate cages in air-conditioned rooms and received standard primate chow (Harlan Primate Diet, Madison, WI), fresh fruit twice daily and water ad libitum. Females, 14-18 kg body weight (bw) were paired with male baboons (Papio anubis) for 5 days at mid-cycle as estimated by menstrual cycle history and pregnancy confirmed by ultrasound and day 1 designated as the day preceding perineal deturgesence. Animals were cared for and used strictly in accordance with USDA regulations and the NIH Guide for the Care and Use of Laboratory Animals (8^th edition). The present experimental protocols were approved by the Institutional Animal Care and Use Committees of the University of Maryland School of Medicine and the Eastern Virginia Medical School.

Serum and samples of fetal vastus lateralis skeletal muscle were available from a contemporaneous group of baboons as part of our ongoing studies of the role of estrogen on fetal development [1, 2, 6]. Briefly, pregnant baboons were randomly assigned and either untreated (n = 11; 7 male ♂, 4 female ♀ fetuses) or treated daily beginning on day 100 of gestation (term = 184 days) with the aromatase inhibitor letrozole (4,4’-[1,2,3-triazyol-1-yl-mehylene]-bis-benzonitrite; Novartis Pharma AG, Basel, Switzerland; 115 μg/kg bw/day, maternal sc injection in 1.0 ml sesame oil; n = 10; 5 ♂, 5 ♀ fetuses) or with letrozole (115 μg/kg bw/day) plus estradiol benzoate (25 μg/kg bw/day on day 100 increasing to 115 μg/kg bw/day between days 120 and 165-175; 1.2 ml sesame oil; n = 5; 3 ♂, 2 ♀ fetuses) to replicate the normal increase in maternal estradiol. On days 165-175 of gestation, after an overnight fast, baboons underwent cesarean section and fetuses delivered under isoflurane anesthesia. Blood samples were obtained from the maternal saphenous vein (5 ml) for assay of estradiol and from the umbilical artery (2 ml) for assay of estradiol, insulin and glucose and fetuses euthanized by an iv injection of pentobarbital (100 mg/kg bw). Samples (5-10 mm³) of fetal vastus lateralis skeletal muscle attached to the femur were excised at approximately the midpoint between the hip and patella and biopsies placed in formalin, oriented vertically and paraffin-embedded for histology and quantification of muscle fiber development or snap frozen and stored in liquid nitrogen for assay of glycogen and triglyceride content as described below. Blood samples were also obtained from a maternal saphenous vein (2–3 ml) after brief restraint and sedation with ketamine HCl (10 mg/kg bw, im) at 5-day intervals during the study period for assay of serum estradiol.

Analyses

Estradiol, insulin and glucose levels

Serum levels of estradiol in maternal saphenous vein and serum estradiol, plasma insulin and blood glucose levels in umbilical artery at delivery were determined using an automated chemiluminescent immunoassay system (estradiol, insulin; Siemens Healthcare Diagnostics, Deerfield, IL) and an iStat Portable Clinical Analyzer (glucose; Model #210003, Abbott Labs, East Windsor, NJ) as described previously [1, 24].

Triglyceride and glycogen content

The content of triglyceride in frozen samples of fetal skeletal muscle from baboons untreated (n = 6; 4 ♂, 2 ♀), treated with letrozole (n = 6; 5 ♂, 1 ♀) or treated with letrozole plus estradiol benzoate (n = 2; 1 ♂, 1 ♀) was determined using a Triglyceride colorimetric assay kit (Cayman #10010303; Cayman Chemical, Ann Arbor, MI) and reagents and instructions supplied by the manufacturer. Briefly, fetal skeletal muscle (~100 mg) was homogenized in 0.5 ml of standard diluent and centrifuged at 10,000 x g for 10 min at 4°C. After determination of protein concentrations using the bicinchoninic acid procedure (Millipore Sigma, St. Louis, MO.), aliquots (10 μl in duplicate) of the supernatant or triglyceride standards were mixed with 150 μl of a triglyceride enzyme mixture comprised of lipoprotein lipase, glycerol kinase, glycerol phosphate oxidase and peroxidase and incubated for 15 min at room temperature. The redox-coupled reaction of hydrogen peroxide with 4-aminoantipyrine and N-Ethyl-N-(3-sulfopropyl)-m-anisidine produced a brilliant purple color and absorbance measured at 540 nm. The level of glycogen in these samples was determined using a glycogen assay kit (Cayman # 700480) and reagents/ instruction supplied by the manufacturer (Cayman Chemical). Briefly, aliquots of tissue homogenates (10 μl in duplicate) and glycogen standards were incubated sequentially with amyloglucosidase, glucose oxidase and horseradish peroxidase and levels of hydrogen peroxide reaction product measured fluorometrically as described previously [3]. Triglyceride and glycogen concentrations were expressed per mg tissue wet weight and per mg protein.

Hematoxylin, eosin and collagen staining

Fetal skeletal muscle biopsies from an additional group of baboons untreated (n = 6, 3 ♂, 3 ♀), treated with letrozole (n = 5; 1 ♂, 4 ♀) or letrozole plus estradiol benzoate (n = 3; 2 ♂, 1 ♀) were fixed in 10% formalin overnight, stored in 1X PBS for 24-48 hours and then paraffin-embedded. Tissues sections (5 μm) were deparaffinized in xylene, rehydrated through a series of ethanol washes, stained with hematoxylin and eosin and permanently cover-slipped. For examination of collagen expression, deparaffinized sections (5 μm) were hydrated in distilled water and incubated (10 min) in equal parts of Weigert’s Iron Hematoxylin A and Weigert’s Iron Hematoxylin B (Millipore Sigma). Sections were washed in distilled water and incubated with Van Gieson’s Solution (0.1% acid fuchsin in picric acid; Millipore Sigma) to stain collagen fibers. Sections were dehydrated through graded alcohols, cleared in xylene, permanently mounted, and imaged using an Olympus BX41 fluorescent microscope fitted with a DP70 digital camera and associated software (Olympus America, Inc., Melville, NY).

Immunohistochemistry of slow (Type I) and fast (Type II) muscle fibers

Additional rehydrated sections (5 μm) of fetal skeletal muscle were heated with 1mM EDTA buffer (pH 8.0) for 30 min and then incubated with trypsin for 10 min at 37°C in a humidified chamber for antigen retrieval. The sections were incubated with hydrogen peroxide prior to blocking with 5% normal horse serum (Vector Laboratories, Peterborough, UK) and then incubated overnight at 4°C with human monoclonal anti-mouse antibody to slow myosin (Cat. No. M8421; clone NOQ 7.5.4D; RRID: AB 477248; Millipore Sigma) diluted 1:3000 in 5% normal horse serum, followed by three washes in 1X Tris-buffered saline (1X TBS; pH 7.6; Millipore Sigma.). Sections were then incubated for 1 h with peroxidase-conjugated goat anti-mouse secondary antibody (#12-349; Millipore Sigma) diluted 1:200 in 5% normal horse serum, washed and incubated according to manufacturer’s instructions with a Vector SG peroxidase substrate kit (Vector) to stain the slow (Type I) fibers black. Stained sections were then washed in running tap water, incubated with 1X TBS for 5 min, 5% normal horse serum for 30 min, and with an alkaline phosphatase-conjugated rabbit monoclonal anti-mouse antibody to fast myosin (Cat. No. M4276; clone MY-32; RRID: AB477190, Millipore Sigma) diluted 1:200 in 5% normal horse serum for 1 h. After washing, sections were incubated with Vector red alkaline phosphatase substrate solution (Vector) to stain the fast (type II) fibers red. Finally, sections were washed in running tap water, dehydrated through graded ethanol, cleared in xylene and cover-slipped with Xylene-based Cytoseal XYL (Vector). Controls included absence of slow/fast myosin expression in sections incubated without primary or alkaline phosphatase-conjugated second antibodies. Conventional fluorescence images (original magnification 400x and 1000 x) were obtained using an Olympus BX41 fluorescent microscope fitted with a DP70 digital camera and associated software (Olympus America).

Image analysis

Approximately 25-50 slow fibers and 50-100 fast fibers within each of 5 randomly selected regions/images (original magnification 400 x) of muscle fascicles (myofibrils) or a total of 125-250 slow fibers and 250-500 fast fibers in each fetal baboon were analyzed using Image J software (NIH). The number per fascicle and the size (μm² x 10⁻³) of each individual fast and slow fiber, the total area (μm² x 10⁻³) of the fascicle and the area of the fascicle occupied by slow and fast fibers were determined in each of the 5 regions analyzed and an overall mean calculated for each animal. In addition, the total area and percentage of the fascicle area comprised of fast and slow muscle fibers, i.e. the total number of fibers in the fascicle x area of each fiber divided by the total fascicle area was quantified. The total area and percentage area of the fascicle comprised of non-fiber tissue was determined as the total fascicle area minus the area of the fascicle occupied by fast/slow fibers.

Statistical analysis

Because the apparent distribution and mean values for various parameters were similar in male and female fetuses, values were combined and data expressed as overall means ± SEM. Differences in the number and size of slow and fast muscle fibers and the total fascicle area comprised of slow and fast fibers in untreated, letrozole-treated and letrozole plus estradiol benzoate-treated animals were analyzed by Two-Way Analysis of Variance (ANOVA) and multiple comparison of the means using Tukey-HSD/Kramer post-tests. All other parameters were analyzed by One-Way ANOVA and multiple comparison of the means using Tukey-HSD/Kramer post-tests and GraphPad software (San Diego, CA). The data for umbilical artery estradiol levels were analyzed following log transformation due to marked heterogeneity of variance.

Results

Serum estradiol, glucose and insulin levels and fetal body and organ weights

The pattern of maternal serum estradiol levels in baboons untreated or treated with letrozole ± estradiol benzoate between days 100 and 165-175 of gestation has been well established and previously published [1, 24, 25]. Accordingly, only the levels of estradiol in maternal and fetal (i.e. umbilical artery) samples at the time of delivery are presented (Table 1). Thus, in contrast to the progressive increase in maternal estradiol levels during the second half of gestation in untreated baboons, the administration of letrozole rapidly decreased maternal estradiol to levels which were < 7% of that in untreated animals (P<0.001). The pattern and absolute levels of estradiol were restored to normal by maternal administration of letrozole and estradiol benzoate. Similarly, the level of estradiol in the umbilical artery (0.03 ± 0.01 ng/ml) in letrozole-treated animals was also markedly lower (P<0.001) than that in untreated animals (0.77 ± 0.12 ng/ml). However, although treatment with letrozole plus estradiol benzoate increased (P<0.01) umbilical artery levels of estradiol to a value (0.08 ± 0.01 ng/ml) greater than in animals treated with letrozole alone, levels were significantly lower (P<0.001) than in untreated baboons (Table 1), presumably the result of placental metabolism of maternally administered estradiol.

Table 1:

Serum estradiol (E₂) levels (ng/ml) in maternal saphenous vein and umbilical artery in baboons untreated or treated with letrozole¹.

	Maternal Saphenous	Umbilical Artery
Untreated	3.07 ± 0.13 a	0.77 ± 0.12a
Letrozole	0.29 ± 0.06b	0.03 ± 0.01b
Letrozole + E2	3.47 ± 0.11a	0.08 ± 0.01c

¹Values expressed as Mean ± SEM on the day of delivery (day 165-175) in baboons untreated (n = 11) or treated via maternal sc injection on days 100 to 165-175 of gestation (term = 184 days) with letrozole (115 μg/kg bw/day; n = 10) or letrozole (115 μg/kg bw/day) plus estradiol benzoate (25 μg/kg bw/day on day 100 increasing to 115 μg/kg bw/day between day 120 to 165-175; n = 5).

Values with different letter superscripts differ at P <0.01 - <0.001 (maternal saphenous) and P < 0.05 - <0.01 (umbilical artery). One-Way ANOVA; P < 0.0001 maternal saphenous and P < 0.001 umbilical artery; and Tukey-HSD/Kramer post tests. E₂ values in umbilical artery log transformed.

As seen in Table 2, as previously demonstrated [1, 24] basal (i.e. fasting) umbilical arterial blood glucose and plasma insulin levels as well as placental, fetal body and representative organ weights were similar in baboons untreated or treated with letrozole ± estradiol benzoate.

Table 2:

Blood glucose and plasma insulin levels in umbilical artery and placenta, fetal body and organ weights in late gestation in baboons¹

Treatment	Glucose (mg/dl)	Insulin (IU/ml)		Weight (gm)
			Placenta	Body	Liver	Kidney	Heart
Untreated	74 ± 8	5.0 ± 1.3	201 ± 11	898 ± 25	27 ± 1	5.6 ± 0.3	6.2 ± 0.3
Letrozole	76 ± 3	4.2 ± 1.4	220 ± 14	932 ± 19	28 ± 1	5.5 ± 0.2	6.5 ± 0.4
Letrozole+E2	72 ± 3	7.0 ± 2.2	206 ± 10	914 ± 34	29 ± 2	5.3 ± 0.4	6.9 ± 0.5

¹Values expressed as Mean ± SEM on day of delivery (day 165-175) in baboons untreated (n = 6-11) or treated with letrozole (n = 5-10) or letrozole + estradiol benzoate (E₂; n = 3-5) on days 100 to 165-175 as described in legend to Table 1.

Fetal skeletal muscle structure, fiber subtypes and expression of collagen

Muscle fascicles, a group of muscle fibers (myofibers) surrounded by the perimysium, were highly organized and the size of individual myofibers appeared similar in fetuses of baboons untreated or treated with letrozole plus estradiol benzoate (Fig 1A and C). In contrast, in estrogen-suppressed letrozole-treated animals, fetal skeletal muscle fascicles appeared disorganized, smaller in overall size and comprised of myofibers that also appeared smaller than in estrogen replete animals (Fig 1B). Moreover, the amount of non-muscle fiber tissue between fascicles also appeared to be more extensive in estrogen-deprived fetuses than in animals untreated or treated with letrozole-plus estradiol benzoate (Fig. 1). Thus, as confirmed by image analysis, in untreated animals, 82% ± 2% of the total fascicle area was comprised of muscle fibers (fast plus slow fibers, Fig 2A) and only 18% ± 2% non-muscle fiber tissue (Fig. 2B). In contrast, in fetuses of estrogen-suppressed animals, the total fascicle area comprised of muscle fibers was markedly (P<0.001) reduced (61% ± 3%) and the non-muscle fiber composition (39% ± 3%, P<0.001) increased. Moreover, both parameters were restored to normal (88% ± 4% muscle fibers; 12% ± 3% non-muscle fiber tissue) in baboons treated with letrozole plus estradiol benzoate. In addition, although quantitative analyses were not performed, the expression of collagen, which was detected primarily between muscle fascicles, appeared to be relatively minimal in fetal skeletal muscle of baboons untreated or treated with letrozole-plus estradiol benzoate and more extensive in skeletal muscle of estrogen-suppressed letrozole-treated animals (Fig 3). As seen in Fig 4, fetal vastus lateralis skeletal muscle fascicles are comprised of both slow (Type I, black) and fast (Type II, pink/red) fibers. The relative distribution of slow and fast fibers appeared to be similar in animals untreated and treated with letrozole ± estradiol benzoate, whereas size of the slow and fast fibers appeared to be reduced in estrogen-suppressed baboons.

Fig. 1:

Representative hematoxylin/eosin histology of baboon fetal skeletal muscle on day 165-175 of gestation (term = 184 days) in animals untreated (A) or treated with letrozole (B) or letrozole and estradiol benzoate (C) as described in legend to Table 1. Original magnification 200X (scale bar = 50 μm). Muscle fascicle comprised of individual myofibers highlighted and circled in white.

Fig. 2:

Percent (Mean ± SEM) of fetal skeletal muscle fascicle area comprised of myofibers (A) and non-myofiber tissue (B) in baboons untreated (n = 6) or treated with letrozole (n = 5) or letrozole and estradiol benzoate (n = 3). Total fascicle area (Fig 5) and area comprised of slow and fast myofibers (Fig 6C) determined by quantitative image analysis as outlined in the methods. Values with different letter superscripts differ at P < 0.001 (One-Way ANOVA; P < 0.0001 total fascicle area; P < 0.0001 total fascicle area comprised of slow/fast fibers; and Tukey-HSD/Kramer post-tests).

Fig. 3:

Collagen (pink/red) expression on day 165-175 of gestation in fetal skeletal muscle of baboons untreated (A,D) or treated in utero with letrozole (B, E) or letrozole and estradiol benzoate (C, F). Hematoxylin counterstain = blue. Magnification 100X (A-C; scale bar = 100 μm); 400X (D-F; scale bar = 20 μm).

Fig 4:

Representative immunohistochemistry of fast (Type II, pink/red) and slow (Type I, black) muscle myofibers in fascicles of fetal skeletal muscle on day 165-175 of gestation in baboons untreated or treated with letrozole ± estradiol benzoate (E₂). Absence of staining in sections incubated without primary antibodies (Insert, panel A). Original magnification 400X (A-C; scale bar = 20 μm); 1000X (D-F; scale bar = 10 μm).

Fascicle area, muscle fiber number and size

As seen in Fig 5, the size (μm² x 10⁻³) of individual fetal vastus lateralis fascicles was lower in estrogen-suppressed animals (159 ± 43 μm² x 10⁻³) than in animals untreated (247 ± 15 μm² x 10⁻³; P <0.01) or treated with letrozole plus estradiol benzoate (240 ± 14 μm² x 10⁻³; P < 0.05).

Fig. 5:

Mean ± SEM area (μm² x 10⁻³) of fascicles in fetal skeletal muscle on day 165-175 of gestation in baboons untreated or treated with letrozole ± estradiol benzoate (E₂). Values with different letter superscripts differ at P < 0.05 – 0.01 (One-Way ANOVA; P = 0.0112 and Tukey-HSD/Kramer post-tests).

In fetal vastus lateralis, the number of fast (type II) fibers per fascicle was approximately 2.5 fold greater (P<0.01) than the number of slow (type I) fibers (Fig 6A). Moreover, the respective number of fast and slow fibers were similar in baboons untreated (45 ± 9 fast; 17 ± 3 slow) or deprived of estrogen in utero (35 ± 5 fast; 15 ± 3 slow). However, the size (μm² x 10⁻³) of both the fast and slow fibers in untreated animals (3.9 ± 0.6 μm² x 10⁻³ fast; 3.6 ± 0.5 μm² x 10⁻³ slow) was reduced (P<0.01) in estrogen-deprived fetuses (1.9 ± 0.2 μm² x 10⁻³ fast; 1.9 ± 0.1 μm² x 10⁻³ slow) but not restored by treatment with letrozole plus estradiol benzoate (Fig 6B). However, the size of fast and slow fibers in respective treatment groups were similar (Fig 6B).

Fig. 6:

Mean ± SEM number (panel A) and size (μm² x 10⁻³) of slow and fast muscle fibers (panel B) and the total fascicle area comprised of slow and fast fibers (panel C) in fetal skeletal muscle on day 165-175 of gestation in baboons untreated or treated with letrozole ± estradiol benzoate (E₂). Values with different letter superscripts differ at P < 0.05 - 0.01 (Two Way ANOVA; P < 0.0001 for difference in slow vs fast fiber number; P < 0.01 for effect of in vivo treatment on fiber number; P = 0.602 for difference in slow vs fast fiber size; P < 0.0002 for effect of in vivo treatment on slow and fast fiber size; P < 0.0001 for difference in total fascicle area comprised of slow vs fast fibers; P < 0.0001 for effect of in vivo treatment on total fascicle area comprised of slow and fast fibers and Tukey-HSD/Kramer post-tests).

Because the number of fast fibers always exceeded the number of slow fibers (Fig 6A), the total fascicle area comprised of fast fibers was 2 to 2.5 fold greater (P<0.01) than that of slow fibers (Fig 6C). Most importantly, the area (μm² x 10⁻³) of the fascicles comprised of fast and slow fibers in untreated animals (132 ± 11 μm² x 10⁻³ fast; 56 ± 5 μm² x 10⁻³ slow) were 2-3 fold lower (P<0.001) in estrogen-suppressed animals (67 ± 10 μm² x 10⁻³ fast; 27 ± 3 μm² x 10⁻³ slow) and restored to normal in baboons treated with letrozole plus estradiol benzoate (162 ± 13 μm² x 10⁻³ fast; 45 ± 4 μm² x 10⁻³ slow; Fig 6C).

Triglyceride and glycogen levels in fetal skeletal muscle

As seen in Table 3, the concentrations of triglycerides and glycogen in skeletal muscle of untreated baboons were not significantly altered by treatment with letrozole or letrozole plus estradiol benzoate.

Table 3:

Concentration of triglycerides and glycogen in fetal skeletal muscle in late gestation in baboons¹

Treatment	Triglycerides		Glycogen
	μg/mg protein	μg/mg tissue	μg/mg protein	μg/mg tissue
Untreated	0.73 ± 0.07	0.40 ± 0.03	64 ± 5	3.3 ± 0.4
Letrozole	1.18 ± 0.29	0.57 ± 0.11	46 ± 9	2.2 ± 0.4
Letrozole + E2	0.67 (.74; .60)	0.33 (.34; .33)	72 (71; 74)	3.6 (3.4; 4.0)

¹Values expressed as Mean ± SEM (individual values) on day of delivery (day 165-175) in baboons untreated (n = 6; n = 5 glycogen) or treated with letrozole (n = 6; n = 5 glycogen) or letrozole + estradiol benzoate (E_2; n = 2) on days 100 to 165-175 as described in legend to Table 1.

Discussion

The results of the current study show that fetal skeletal muscle myofiber development was markedly curtailed in baboons in which the levels of estradiol were suppressed during the second half of gestation by maternal administration of the aromatase inhibitor, letrozole. Specifically, in estrogen-suppressed animals, fetal skeletal muscle fascicles were structurally less organized, smaller in overall size and comprised of slow type I fibers and fast type II fibers, the size but not the number of which were also smaller than in untreated baboons. Moreover, the proportion of non-muscle fiber tissue was nearly two fold greater and the amount of slow and fast muscle fibers much lower in estrogen-deprived than in untreated fetuses. However, estrogen deprivation did not impair overall fetal growth and it appears that body weight was maintained at the expense of muscle fibers and likely reflected increased deposition of non-muscle proteins such as collagen. Although the current study did not determine the functional status of fetal skeletal muscle fibers, because fetal skeletal muscle development, including fascicle myofiber organization, size and proportion of muscle comprised of slow and fast myofibers, was normal in baboons treated with letrozole and estradiol benzoate, we suggest that estrogen exposure in utero is required for and thus promotes fetal skeletal muscle myofiber growth and organization.

We recently demonstrated that skeletal muscle microvascularization, which is important for delivery of insulin and glucose, as well as nutrients and growth factors such as IGF-1 to myofibers, was 50% lower in near-term fetuses and offspring deprived of estrogen in utero and restored to normal in animals treated with letrozole and estradiol benzoate (6). Therefore, it is possible but remains to be determined, that the reduced size of muscle fibers in animals of the current study primarily reflects impaired development of the fetal skeletal muscle micro-vasculature. However, there is significant evidence that estrogen plays an important role in controlling skeletal muscle growth in adult female animals and women. For example, estrogen increases muscle mass in adult rodents [26-28], promotes muscle regeneration after injury, maintains muscle mass and strength in aging females [26, 29], alleviates age- and/or disuse-induced muscle atrophy and promotes skeletal muscle regeneration [26, 30]. Several of these actions of estrogen appear to be mediated by activation of estrogen receptor (ER)β, although stimulation of regenerative responses after injury appear to involve ERα-induced myofiber satellite cell activation and proliferation [29, 31, 32]. Moreover, both ER subtypes are expressed in fetal as well as adult skeletal muscle [33-35]. Although the specific site/sites of estrogen action and whether ER expression is altered in letrozole-treated baboons remain to be elucidated, the present study is the first to demonstrate an important role for estrogen in promoting growth of fetal skeletal muscle in the primate.

The placenta is a source of several growth factors, e.g. IGF [36 for review] and thus it is possible that the latter in addition to and/or controlled by estrogen are important to fetal skeletal muscle maturation. Thus, infusion of insulin or IGF1 in fetal sheep and piglets promotes overall fetal as well as muscle-specific protein synthesis [37-39]. Moreover, it is well established that fetal muscle development is significantly impaired in adverse conditions of human pregnancy e.g. placental ischemia, and in experimental animal models in which delivery of nutrients, oxygen or regulatory factors to the fetus are reduced by maternal nutrient restriction, hypoxia or impaired placental blood flow [7]. However, in all of these situations, overall growth of the fetus is significantly reduced [7, 40-42]. In contrast, although we did not ascertain whether fetal IGF 1 levels were altered by treatment with letrozole, fetal body weight and growth were normal in estrogen-suppressed baboons of the current and our previous studies [1-3, 25]. In addition, as shown previously and in the current study placental blood flow [25] and fetal insulin levels [1; Table 2 this study) at term were similar in animals untreated or treated with letrozole ± estradiol benzoate. Thus, while there is certainly a strong relationship between expansion of skeletal muscle myofibers and overall growth of the fetus, the current study indicates a more specific role for estrogen in promoting fetal skeletal muscle myofiber development.

We have shown that offspring delivered to baboon mothers deprived of estrogen during the second half of gestation exhibit a reduction in skeletal muscle micro-vessel expansion and insulin resistance [1, 2, 6] and which was not due to an alteration in sensitivity of adipose tissue to insulin [3]. Interestingly, although we did not perform quantitative analyses, collagen expression appeared to be more prominent and the concentration of triglycerides higher in fetal skeletal muscle of estrogen-deprived baboons of the current study than in muscle of animals untreated or treated with letrozole plus estradiol benzoate. Others have shown that skeletal muscle levels of triglycerides and other lipid molecules are elevated in individuals with insulin resistance [43-46]. Therefore, because of the importance of skeletal muscle to overall utilization of glucose and thus insulin sensitivity, it is likely that the reduced size of skeletal muscle fascicles and the type I and type II fibers in letrozole-treated fetuses is sustained in offspring and as initiated by the impairment of microvascularization accounts for the onset of insulin resistance afterbirth.

In summary, the current study shows that fetal skeletal muscle myofiber development was markedly curtailed in baboons in which the levels of estradiol were suppressed during the second half of gestation and normal after estrogen repletion. However, estrogen deprivation did not impair overall fetal growth, i.e. body weight, and it appears that the latter was maintained at the expense of muscle fiber maturation by increased deposition of non-muscle fiber proteins. Collectively, these results, and those of our previous study showing that skeletal muscle microvascular expansion and insulin sensitivity is impaired in offspring deprived of estrogen in utero, support our proposal that exposure of the fetus to estrogen is important for fetal skeletal muscle development and glucose homeostasis in adulthood.