Sex Differences in Insulin Regulation of Skeletal Muscle Glycogen Synthase and Changes during Weight Loss and Exercise in Adults

Abstract

Objective:

We sought to understand sex differences in muscle metabolism in 73 older men and women.

Methods:

Body composition, VO₂max, and insulin sensitivity (M) by three-hour-hyperinsulinemic-euglycemic clamp with vastus lateralis muscle biopsies were measured. Results: Women had lower body weight, VO₂max and fat-free mass than men. Men had lower M, lower change (insulin-basal) in muscle GS activity and lower change in AKT protein expression than women. M was associated with the change (insulin-basal) in GS activity and the change in AKT protein expression. Sex differences (n=60) were tested with six-months weight loss (WL) or 3x/week aerobic exercise training (AEX). The postintervention minus preintervention change (insulin-basal) (ΔΔ) in GS activity (fractional, independent, total) was higher in men than women in WL, and the ΔΔ in GS fractional activity was higher in women than men in AEX. In the entire group, the ΔΔ in GS fractional and independent activities were related to ΔΔ in AKT expression and glycogen content.

Conclusions:

Sex differences in insulin sensitivity may be explained at the cellular muscle level, and to improve skeletal muscle insulin action in older adults, it may be necessary to recommend different behavioral strategies depending on the individual’s sex.

2. INTRODUCTION

Many older individuals lead a sedentary lifestyle, with only 24% percent of adults aged 18 and over in the US meeting the minimum physical activity guidelines^[1]. Declines in skeletal muscle mass occur in men after age mid-forties and in women after age mid-fifties whereas increases in body fat occur in both men and women with age^[2], with greater increases in fat mass in men over time. In general, men have a higher muscle mass and less body fat than women. Additionally, aging is associated with a decline in insulin action at the whole body and cellular levels^[3–6] levels with insulin signaling being a negative regulator of longevity in animal models^[4]. Both physical activity and nutrition are modifiable behaviors that can impact insulin sensitivity and muscle metabolism with aging.

The importance of including both sexes in experiments was formally recognized by the National Institutes of Health who in 2016 instituted a policy to encourage the study of sex as a biological variable. It is well-known that body composition differs between men and women throughout the lifespan^[2], yet how age and sex impact adipose tissue and skeletal muscle metabolism is less defined. Previously, we described sex differences in insulin sensitivity, with older men having over 62% lower whole-body glucose utilization and reduced in vivo insulin activation of muscle glycogen synthase activity compared to postmenopausal women^[7]. We have also found that middle-aged and older women have higher adipose tissue expression of adiponectin, an adipokine that regulates lipid and glucose metabolism, and higher adipose and skeletal muscle tissue expression of liver receptor homolog-1 (LRH-1), a nuclear receptor that regulates lipid metabolism, than similar aged men^[8]. Furthermore, we have reported higher skeletal muscle expression of SIRT1 and SIRT3, genes involved in metabolic homeostasis^[9], and lower myostatin expression in women than men^[10]. Taken together, these results suggest that differences between men and women in both adipose tissue and skeletal muscle contribute to the overall observed sex differences regulating metabolism that occur with age.

A holistic approach to human health includes dietary and activity-based strategies. Healthy aging should include the maintenance of a normal body weight through consuming a healthy dietary pattern^[11] and participating in the recommended levels of physical activity^[12]. Knowledge of the various mechanistic pathways in skeletal muscle which lend evidence to improve overall health with aging through non-pharmacologic approaches such as diet and exercise are needed. Sex as a biological variable has been demonstrated to impact drug treatments^[13] and influences equations to estimate body composition and physical activity, thus the consideration of sex when designing interventions should be deliberated. Moreover, whether sex differently alters the effects of behavioral interventions on muscle metabolism is poorly understood. Therefore, the objectives of this study were to first test sex differences in in vivo insulin regulation of skeletal muscle glycogen synthase activity and expression of proteins involved in glucose metabolism and secondly to examine the effects by sex on the response to weight loss and aerobic exercise training on muscle metabolism.

3. EXPERIMENTAL PROCEDURES

3.1. Male and Female Study Participants

Study participants were men and women between 45 and 80 years of age who were generally healthy with a body mass index between 21–47 kg/m². Interested individuals were initially screened by phone. Participants then had a physical exam with comprehensive past medical history and fasting blood profile. Participants were excluded if they showed evidence of heart disease, diabetes, cancer, anemia, dementia, untreated dyslipidemia, and other unstable or chronic liver, lung, or kidney diseases. The women were postmenopausal defined as at least one year past their last reported menstrual cycle. Inclusion criteria included weight stability of less than 2.0 kg change in weight over the prior 12 months and sedentary defined as less than 20 min of structured physical activity such as walking, jogging, or swimming twice per week. Participants underwent a graded exercise treadmill test reviewed by a cardiologist prior to study participation to screen for cardiovascular disease. The study was approved by the University of Maryland Institutional Review Board. Each participant provided their written informed consent to participate after all procedures and risks were explained.

3.2. Dietary and weight stability

Participants obtained instruction by a Registered Dietitian (RD) as to how to implement a weight-stable, Therapeutic Lifestyle Changes (TLC) diet^[14]. This food pattern consisted of guidelines for total fat consumption of less than 30% of total calories as well as saturated fat of 10%. In addition, there were guidelines for cholesterol of 300 mg and sodium of 2400 mg per day. Participants in the study were weight stable (e.g. ±2% body weight) on this diet for two weeks before any initial testing.

3.3. Exercise Testing

Participants underwent a maximal graded treadmill test (Supplement Methods).

3.4. Body Composition

Height (cm), weight (kg), and waist and hip circumferences were determined using standardized protocols and used to calculate body mass index (BMI) and waist-to-hip ratio (WHR). Dual-energy X-ray absorptiometry (iDXA, LUNAR Radiation Corp., Madison, WI) was used to measure percent body fat, fat mass, lean body mass, and fat-free mass for the participants.

3.5. Oral Glucose Tolerance Test (OGTT)

To assess insulin resistance, participants underwent a two-hour OGTT (Supplement Methods).

3.6. Skeletal Muscle Biopsies and Hyperinsulinemic-eugylcemic Clamp

Participants were provided two days of food to consume, which was eucaloric and followed the TLC guidelines. To determine whole body insulin sensitivity, a three-hour hyperinsulinemic-euglycemic clamp^[15] was conducted with an insulin infusion of 80 mU.m⁻².min⁻¹ (Humulin, Eli Lilly Co., Indianapolis, IN, USA), a dorsal heated hand vein^[16] and indirect calorimetry to measure nonoxidative glucose metabolism^[17] (Supplement Methods). Two vastus lateralis muscle biopsies were performed with local anesthesia, one before the start of the glucose clamp and the second at 120-min during the clamp procedure. Wet muscle was immediately snap-frozen and stored. Insulin levels during the clamp were not different before and after the interventions in WL (men 1176±219 vs. 1132±226 pmol/L, women: 1133±161 vs. 1156±224 pmol/L) and AEX (men: 956±489 vs. 1149±237 pmol/L, women: 1136±175 vs. 1184±218 pmol/L). Glucose levels were also not different in men with WL (90.6±5.0 vs. 88.9±5.5, CV: 4.5±0.9 vs. 5.4±1.9), women with WL (89.6±5.2 vs. 88.3±4.1, CV: 6.2±2.6 vs. 5.9±1.5), men with AEX (93.4±4.5 vs. 92.2±3.4, CV:5.0±1.4 vs. 4.3±1.5) and women with AEX (90.8±4.5 vs. 90.4±4.8, CV 5.3±1.8 vs. 6.1±2.5.)

3.7. Skeletal Muscle Assays

Protein Expression

Multi-Plex Protein Assays:

Frozen muscle biopsies were put in a lyophilizer for 48 h to allow the separation of the skeletal muscle fibers from the fat, blood, and connective tissue within the muscle sample under a microscope. Approximately 5 mg of the clean muscle was homogenized, processed, and the muscle lysate was used in s ix multiplex protein assay kits (Supplement Methods).

GLUT4:

Microdissected muscle samples were homogenized (0.67% wt/vol solution) in ice-cold PBS buffer (0.01 mol/l, pH 7.2) The suspension was subjected to three freeze-thaw cycles then centrifuged for 5 min at 5,000 x g. Supernatants were assayed using a GLUT4 ELISA kit (Cloud-Clone, SEC023Hu). GLUT4 expression was corrected by total protein concentration (Pierce, Bradford).

Enzyme Activity

Glycogen Synthase Assay:

Glycogen synthase activity was performed as previously described ^[17].

Protein Phosphatase 1 (PP1) and 2A (PP2A) Assays.

Microdissected muscle samples were homogenized (0.67% wt/vol solution) in ice-cold Tris-HCl buffer (pH 7.5) containing 0.1% 2-mercaptoethanol, 2 mg/ml glycogen, protease inhibitor cocktail, 0.1 mmol/l EDTA, and 0.1 mmol/l EGTA. The homogenate was centrifuged at 2,500 x g for 5 min at 4°C, and the supernatant diluted 5-fold with homogenization buffer before assay with the RediPlate Serine/Threonine Phosphatase Assay Kit (R-33700, Molecular Probes). PP1 activity: Microdissected muscle samples were assayed in the absence and in the presence of 400 nmol/l Protein Phosphatase Inhibitor-2 (LAE Biotech International, #P027) in a reaction buffer containing 2 mmol/l DTT and 200 μmol/l MnCl₂. Inhibitor-2 batches were checked for potency against recombinant PP1 (New England Biolabs, P0754S) before sample assay. Sample PP1 activity was calculated as the activity in the absence of Inhibitor-2 minus the activity in the presence of Inhibitor-2, corrected for background fluorescence (reaction mixture). PP2A activity: Samples were assayed in the presence of 200 nmol/l microcystin-LR (Millipore 475821). Microcystin-LR batches were checked for potency against recombinant PP2A (Cayman, 10011237) before sample assay. Sample PP2A activity was calculated as the activity in the presence of Inhibitor-2 minus the activity in the presence of microcystin-LR. Enzyme activities were corrected by total protein concentration (Pierce, Bradford).

Glycogen content

Lyophilized microdissected skeletal muscle was homogenized in 150 μL of buffer (1:150) containing (in mmol/l) 50 tricine pH 7.4, 100 NaF, 10 EDTA, and protease inhibitors (Roche 11836170001). The homogenate was centrifuged at 13,000 x g for 5 min at 4°C. An aliquot of supernatant was removed for total protein determination and the remaining supernatant was frozen and boiled for 5 minutes, and centrifugation repeated. This last supernatant was used for fluorometric determination of glycogen using the Glycogen Assay Kit (Sigma, MAK016).

3.8. Behavioral Interventions of Weight Loss and Aerobic Exercise Training

To test the effects of sex on GS activity and M with behavioral interventions, participants were randomly assigned to groups of six months of weight loss (WL) or aerobic exercise (AEX) after completion of the above testing procedures in order to facilitate the group nature of the program. Data were combined with two additional studies where participants met criteria for WL or AEX to increase the sample size per sex for the sex by GS interaction analyses^{[7, 17]}. Specifically, 4 men were added (3 to WL, 1 to AEX) ^{[7, 17}] and 23 women (12 to WL, 11 to AEX)^{[7, 17]}. A registered dietician (RD) led weekly weight loss classes for those in the WL group with topics such as Coping with Slips and Binges, Problem Solving, Healthy Habits, Self-Talk, and Stress Management. The RD provided expertise and instruction in “heart healthy” dietary modification guidelines and how to implement a 250–500 kcal/d hypocaloric diet. The goal was an approximate 5–10% weight loss over the six-month intervention. At each weekly class, body weight was determined. In addition, those in the WL class were taught to complete a 7-day food record by providing an estimate of food quantity and recording all food eaten which were analyzed using the American Diabetes Association exchange list system.

For those individuals in the AEX group, they were required to attend exercise sessions 3x/week for six-months. An exercise physiologist monitored supervised each exercise session. Heart rate was monitored during each exercise training session using heart rate monitors (Polar Electro Inc., Lake Success, NY, USA). Participants walked on a motorized treadmill which began at week one at an intensity of ~50–60% of heart rate reserve (HRR) for 30–40 min. At week 2, it progressed to 55–65% HRR for 45–50 min, week 3 at 60–65% HRR for 50 min, and week 4 at 65–75% HRR. By week five to eight, the progression of exercise increased to 50 min at an intensity of ~70–80% of HRR which was continued for the six months. Each exercise session included a 5- to 10-min calisthenics and stretching exercise and 5 min of progressive aerobic activity warmup phase and a 5- to 10-min cooldown phase of slow walking after the training period.

3.9. Statistical Analyses

Baseline comparisons of participant characteristics, protein expression, glycogen content, and GS, PP1 and 2A activity between sexes were performed using unpaired Student’s t-tests. Paired Student’s t-tests were used to compare differences between basal and insulin-stimulated protein expression and GS, PP1 and 2A activity and differences between the change (insulin minus basal) in GS activity at baseline and following intervention.

To determine whether there were differences between the sexes and treatment arms, the effects of treatment arm (WL vs. EX) adjusted for sex on subject characteristics and GS activity were compared using a two-way ANOVA with interactions, followed by post hoc analyses (Scheffe). When the sex x arm treatment was not significant, it was dropped from the model, and the analysis was rerun. If neither term was significant, the overall effect was reported. Pearson correlations were used to assess relationships between key variables, reporting two-tailed probability. All data are presented as mean ± SD, with statistical significance set at p<0.05.

RESULTS

BASELINE SEX DIFFERENCES

4.1. Fitness and Body Composition (Table 1)

Table 1:

Participant Baseline Characteristics

	Males		Females
Characteristic	n	Mean ± SD	n	Mean ± SD
Age (yrs)	31	60.1 ± 8.5	42	60.6 ± 7.0
Weight (kg)	31	101.7 ± 17.1	42	89.9 ± 17.8‡
BMI (kg/m2)	31	32.5 ± 5.5	40	33.4 ± 5.1
Waist circumference (cm)	28	109.7 ± 11.3	36	98.3 ± 9.0‡
Hip circumference (cm)	28	113.5 ± 8.8	36	119.6 ± 12.9*
Waist-to-hip ratio	28	0.97 ± 0.06	36	0.83 ± 0.07 ‡
VO2max (l/min)	31	2.79 ± 0.48	41	1.88 ± 0.43 ‡
VO2max (ml/kg/min)	31	27.7 ± 4.8	41	21.1 ± 4.2‡
Percent body fat (%)	30	35.72±5.32	40	47.05±6.16 ‡
Fat mass (kg)	30	37.2±10.73	40	44.1±13.30*
Lean mass (kg)	30	61.88±7.04	40	44.92±6.58 ‡
Fat-free mass (kg)	30	65.18±7.17	40	47.42±6.80 ‡
Fasting glucose (mg/dl)	31	104.03±11.91	40	95.95±15.42*
Glucose at 120min (mg/dl)	31	156.55±51.65	40	140.5±41.25
Fasting insulin (pmol/l)	30	120.00±58.68	38	111.61±53.34
Insulin at 120min (pmol/l)	30	726.07±473.22	38	691.42±358.66
Nonoxidative glucose metabolism (μmolffmmin)	28	29.88±16.78	34	49.26±22.41 ‡
Insulin Sensitivity (M) (μmolkgffmmin)	30	35.89±16.08	38	51.75±18.28 ‡

Females vs. Males:

^*P<0.05;

^†P<0.01;

^‡P≤0.001

Men and women did not differ by age. Women had lower body weight (p<0.001), lower waist circumference (p<0.005), higher hip circumference (p<0.05) and lower waist-to-hip ratio (p<0.001) than men. Women had lower VO₂max expressed either in l/min or ml/kg/min than men (both p< 0.001). Women had lower lean and fat-free mass and a higher fat mass and percent body fat than the men.

4.2. Glucose Metabolism (Table 1) (Figure 1)

Figure 1:

Insulin Sensitivity (M) (upper panel), GS fractional activity (insulin - basal) (middle panel), and insulin-basal AKT P/T ratio (lower panel) in men and women at baseline. Boxplot legend: height of box = inter-quartile range (IQR); top of box = the 75th percentile (Q3); bottom of box = the 25th percentile (Q1); solid center line = median; dashed red line = mean; vertical line = range of data; open circles = outliers (< Q1-[1.5*IQR] or > Q3+[1.5*IQR]).

Fasting glucose was higher in men than women (p<0.05) whereas the two-hour glucose was not different by sex. Fasting and two-hour insulin levels from the OGTT were not different by sex. There were 21 women with normal glucose tolerance (NGT) and 19 with impaired glucose tolerance (IGT), and two women were missing an OGTT. In the men, 13 had NGT and 18 had IGT. Both nonoxidative glucose metabolism and insulin sensitivity (M) (Figure 1A) were lower in men than women (both p<0.001).

4.3. Basal and Insulin-stimulated Muscle Protein Expression and Enzyme Activity (Table 2) (Figure 1)

Table 2:

Baseline Basal and Insulin-stimulated Muscle Protein Expression and Enzyme Activity

	Men				Women
	n	Basal	Insulin	Change	n	Basal	Insulin	Change
Protein Expression
Sex Differences: Basal
ERK1/2 Thr/Tyr:202/204 P/T	31	21.73±10.04	25.75±14.64	4.02±14.60	41	27.21±12.34 *	39.73±30.55*ᵇ	9.64±22.75 (n=40)
AMPK T172 P/T	22	1.54±0.47	1.48±0.48	−0.07±0.37	18	1.92±0.64*	1.74±0.74	−0.17±0.62
Sex Differences: Insulin
IGF-1R Tyr1316 P/T	29	8.56±3.80	8.7±3.53	0.14±2.28	23	6.96±2.47	6.71±2.00*	−0.25±1.32
Sex Differences: Basal & Insulin
p38 Thr183/Tyr185 P/T	31	3.63±5.54	3.83±5.98	0.2±2.54	41	7.44±6.91†	8.45±10.86*	0.92±7.75 (n=40)
pGSK-3β S9 P/T	31	180.89±89.48	392.28±198.30ᶜ	211.39±157.09	41	261.75±152.98 †	523.09±224.55†ᶜ	266.68±141.93
Sex Differences: Insulin & Change (Insulin-Basal)
pAKT T308/S473 P/T	31	21.13±19.50	72.91±55.08ᶜ	51.79±60.42	41	23.56±8.81	121.13±68.39‡ᶜ	97.56±68.06‡
No Sex Differences
p70S6K T421/S424 P/T	31	3.61±1.48	6.17±3.69ᶜ	2.56±3.63	41	4.19±2.49	7.73±4.01ᶜ	3.28±3.80 (n=40)
IRS Tyr612 P/T	29	31.09±10.11	44.43±23.24ᶜ	13.34±16.38	23	38.11±20.77	53.05±37.47ᵇ	14.94±21.83
IR Tyr1150/1151 P/T	29	0.830.53	2.33±1.37ᶜ	1.5±1.44	23	0.76±0.87	2.34±1.52ᶜ	1.58±1.86
JNK Thr180/Tyr182 P/T	31	29.2±9.50	27.89±9.25	−1.31±9.04	41	31.85±10.67	30.7±8.47	−1.37±9.94 (n=40)
GLUT4	15	8.53±3.86			10	6.64±3.57
Enzyme Activity
GSFV (%)	19	7.99±3.70	16.95±6.13ᶜ	8.96±6.26	18	7.42±2.56	21.38±7.63ᶜ	13.98±7.03*
GSI (nmol/min/mg protein)	19	0.54±0.45	1.04±0.66ᵇ	0.5±0.65	18	0.37±0.29	1.42±0.93ᶜ	1.05±0.88*
GST (nmol/min/mg protein)	19	6.35±3.19	6.18±3.17	−0.17±3.07	18	4.85±2.37	6.42±2.65ᵅ	1.57±2.86
PP1(mU/mg protein	19	38.95±22.31	36.84±15.78	−2.16±15.71	17	26.12±8.18*	30.94±19.27	5.69±15.8
PP2A (mU/mg protein)	19	160±54.05	162.26±56.20	2.32±64.03	17	146.94±48.34	153.41±45.32	6.29±48.43

Values are mean ±SD. Different men and women:

^*P<0.05;

^†P≤0.01;

^‡P≤0.001. Different than Basal:

^ᵅP<0.05;

^ᵇP≤0.01;

^ᶜP≤0.001. P/T = phosphorylated/total. pAKT T308/S473= protein kinase B (Akt), pGSK-3β S9= glycogen synthase kinase 3 beta, p70S6K T421/S424=70 kDa ribosomal protein S6 kinase, IRS= insulin receptor substrate 1, IGF-1R= insulin-like growth factor 1 receptor, IR=insulin receptor, ERK1/2= extracellular regulated kinase ½, JNK=c-jun N-terminal kinases, p38=p38 mitogen-activated protein kinases.−1

Basal phosphorylated to total protein expression of GSK-3β (p<0.001), p38 (p<0.001), ERK1/2 (p<0.05), and AMPK (p<0.05) were lower in men than in women. PP1 activity was higher in men than in women (p<0.05). Insulin-stimulated phosphorylated protein to total protein expression of AKT, GSK-3β, ERK1/2 and p38 were lower in men than in women, and IGF-1R was higher in men than women. Total Glut4 protein expression did not differ between men and women. Neither basal (male, 917 ± 500; female, 796 ± 423 U/mg protein) nor insulin-stimulated (male, 904 ±541; female, 788 ± 316 U/mg protein) glycogen contents were different between men and women.

Note: the change in outcome variables following insulin stimulation (insulin minus basal) is denoted as Δ.

The Δ in GS fractional and independent activities and in AKT P/T protein expression were different (lower) in men than women (Figures 1B and 1C, respectively). Changes (Δ) in GS fractional activity (r=0.44, p=0.0004) and in AKT protein expression (r=0.41, p=0.0005) were both related to insulin sensitivity (M) (Figure S1, Supplementary Materials). There were no sex differences in the Δ in P/T protein levels of AMPK, GSK-3B, p70s6K, IRS, IGF-1R, IR, ERK1/2, JNK, and p38.

The Δ in GS fractional activity was related to the changes in AKT (r=0.65, p<0.0001), GSK-3β (r=0.52, p=0.001), IR (r=0.56, p=0.0004) and IRS (r=0.33, p=0.05) protein expression (Figures S1 and S2, Supplementary Materials). The Δ in GS independent activity was related to the changes in AKT (r=0.56, p=0.0004), IR (r=0.63, p<0.0001) and IRS (r=0.33, p<0.05) protein expression. The Δ in GS total activity was inversely related to the changes in IGF-1R (r=−0.56, p=0.0004) and p38 (r=−0.34, p<0.05) protein expression.

INTERVENTION EFFECTS by Sex

4.4. Metabolic Outcomes (Table 3)

Table 3:

Physical and Metabolic Characteristics of Men and Women by Intervention

	Men WL (n=11)		Men AEX (n=11)		Women WL (n=21)		Women AEX (n=17)		SxI	S	I	O
	Pre	Post	Pre	Post	Pre	Post	Pre	Post
Age (yr)	61 ± 7		63 ± 10		60 ± 8		63 ± 8
Weight (kg)	109.4 ± 18.3	99.1 ± 16.8	96.6 ± 16.2	94.9 ± 17.1	94.9 ± 14.3	87.5 ± 14.2	86.6 ± 10.9	84.0 ± 11.0	†
BMI (kg/m2)	35.1 ± 4.8	31.8 ± 4.6	30.9 ± 5.6	30.4 ± 5.9	35.0 ± 5.0	32.3 ± 5.1	33.0 ± 4.7	32.6 ± 4.8	--	--	¶
Percent body fat	36.9 ± 6.3	32.9 ± 7.1	33.7 ± 4.4	31.4 ± 5.5	48.7 ± 3.8	46.6 ± 8.9	46.2 ± 4.9	44.9 ± 5.3	--	*	*
Fat mass (kg)	41.0 ± 12.3	33.9 ± 11.8	40.9 ± 2.5	34.5 ± 2.2	46.6 ± 8.9	41.1 ± 9.7	40.6 ± 9.3	38.6 ± 9.5	--	--	¶
Fat-free mass (kg)	68.1 ± 7.3	66.8 ± 7.6	64.2 ± 7.8	65.4 ± 7.9	48.7 ± 6.7	47.4 ± 6.7	46.6 ± 3.2	46.3 ± 3.3	*
VO2max (l/min)	2.62 ± 0.46	2.59 ± 0.50	2.57 ± 0.60	2.90 ± 0.66	1.79 ± 0.46	1.72 ± 0.39	1.74 ± 0.42	2.03 ± 0.45	--	--	¶
Fasting plasma glucose (mmol/l)	5.69 ± 0.53	5.22 ± 0.37	5.40 ± 0.65	5.45 ± 0.63	5.48 ± 0.54	5.03 ± 0.37	5.09 ± 0.31	5.13 ± 0.34	--	--	¶
Fasting plasma insulin (pmol/l)	109 ± 48	80 ± 32	93 ± 44	83 ± 39	101 ± 48	78 ± 30	93 ± 33	89 ± 37	--	--	†
Glucose120 (mmol/l)	8.44 ± 1.97	7.83 ± 1.38	7.27 ± 2.21	7.36 ± 2.88	7.31 ± 2.37	6.56 ± 1.92	6.92 ± 1.41	6.73 ± 0.98	--	--	--	†
Insulin120 (pmol/l)	691 ± 466	475 ± 242	562 ± 465	472 ± 237	605 ± 332	462 ± 231	627 ± 498	437 ± 233	--	--	--	‡
Glucose utilization (μmol.kgFFM−1.min−1)	34.5 ± 14.1	47.4 ± 15.3	39.7 ± 18.3	46.0 ± 15.8	58.2 ± 20.5	64.3 ± 24.2	58.2 ± 13.5	65.4 ± 14.3	--	--	--	€

Values are means ± SD.

SxI = sex group x intervention; S = sex group, I=intervention, O=overall.

The effect of the interventions in the two sex groups was compared using the following model: change = sex + intervention + sex*intervention.

Significant effect:

^*p=0.05;

^†p<0.05;

^‡p<0.005;

^€p<0.0005;

^¶p<0.0001.

^--non-significant

There was a significant interaction effect for the change in body weight (p<0.05) with post-hoc tests indicating a difference by sex for WL (p=0.01) and by intervention for both men and women (both p<0.0001). There was no sex difference in body weight change for AEX (men vs. women, p=0.41). There was a sex x treatment arm interaction effect for the change in FFM (p=0.05) with post hoc tests indicating sex differences in the AEX group (p=0.01) but not in the WL group, and intervention differences in the men (p<0.0005) but not in the women. There was a significant intervention effect for changes in BMI, fat mass, and percent body fat with significant changes only after WL.

VO₂max increased in women (p<0.001) and men (p=0.001) after AEX, but not after WL; the changes were significantly different by intervention (p<0.05). Overall, there was a significant effect for reductions in both 120-minute glucose (p<0.05) and insulin levels (p<0.001) and increases in M (p<0.05). M increased (paired Student’s t-test) in the men following WL (p<0.005) and tended to increase in women following AEX (p<0.06).

Insulin Effect on GS Activity (Table 4): In the WL group, the women had higher baseline Δ in GS fractional activity than the men. In the AEX group, the women had greater Δ in GS fractional and independent activities than the men during the intervention.

Table 4:

Insulin Effect (Insulin-Basal) on Glycogen Synthase (GS) Before and After Intervention (mean +/− SD)

	Males (n=11)		Females (n=21)		Males (n=11)		Females (n=17)
Insulin-Basal	Pre WL	Post WL	Pre WL	Post WL	Pre AEX	Post AEX	Pre AEX	Post AEX
GS Fractional Activity (%)	9.4 ± 7.5	16.1 ± 11.8b	16.0 ± 7.3*	15.8 ± 6.9	10.3 ± 5.3	7.2 ± 5.8	12.5 ± 5.8	15.2 ± 6.9 †
GS Independent Activity (nmol/min/mg protein)	0.69 ±1.18	1.53 ± 1.49b	1.52 ± 1.21	1.14 ± 0.85	0.78 ± 0.55	0.66 ± 1.11	0.87 ± 1.14	1.71 ± 1.14 a,*
GS Total Activity (nmol/min/mg protein)	−0.41±2.88	0.99 ± 2.0	1.62 ± 3.18	−0.60 ± 3.66c	0.59 ±3.10	0.24 ± 4.51	1.95 ± 2.9	1.72 ± 3.96

WL= weight loss; AEX= aerobic exercise. Different Males vs. Females within same intervention

^*p<0.05,

^†p<0.005

Different vs. Pre:

^ap<0.06;

^bp<0.05;

^cp<0.01.

The Δ in GS fractional and independent activities were higher during WL than at baseline in the men. The Δ in GS total activity was lower during WL than at baseline in the women. The Δ in GS independent activity was higher during AEX than at baseline in the women.

Intervention (Figure 2):

Figure 2:

The change (insulin-basal) in GS fractional activity (upper panel), independent activity middle panel), and total activity (lower panel) with weight loss (WL) and aerobic exercise training (AEX) in men and women. Boxplot legend: height of box = inter-quartile range (IQR); top of box = the 75th percentile (Q3); bottom of box = the 25th percentile (Q1); solid center line = median; dashed red line = mean; vertical line = range of data; open circles = outliers (< Q1-[1.5*IQR] or > Q3+[1.5*IQR]). ‡, p<0.06; *, p<0.05; **, p<0.01

Note: the postintervention minus preintervention change (insulin minus basal) in outcome variables is denoted as ΔΔ.

There were significant sex x intervention interactions for post-intervention minus pre-intervention changes (ΔΔ) in GS fractional, independent, and total activities in the four groups. The sex x intervention interaction for ΔΔ in GS fractional activity was p<0.007. The ΔΔ tended to be higher (p<0.06) in men than in women in the WL group and was significantly higher in the women than men in the AEX group (p<0.05). The ΔΔ was higher in the men in the WL group than the men in the AEX group (p<0.05).

The sex x intervention interaction for ΔΔ in GS independent activity was p=0.003. The ΔΔ was higher in men than in women in WL (p<0.01). The ΔΔ was higher in the men in the WL group than the men in the AEX group (p<0.05). The ΔΔ was higher in the women in the AEX group than the women in the WL group (p<0.05).

The sex x intervention activation for ΔΔ in GS total activity was p<0.05. The ΔΔ was higher in men than in women in WL (p<0.01).

The basal and insulin-stimulated glycogen contents in the men and women pre- and post-intervention are provided in Table S1 (Supplementary Materials). The sample sizes were too small to test for sex x intervention interactions by ANOVA. However, unpaired Student’s t-test showed the men had higher post-pre change (insulin-basal) in glycogen compared to women in the WL group.

Overall, the post-intervention minus pre-intervention change (insulin-basal) (ΔΔ) in GS fractional activity was related to ΔΔ in glycogen content (r=0.56, p=0.001), P/T AKT (r=0.37, p<0.05), IR (r=0.59, p=0.0006) and IRS (r=0.33, p=0.05) and (inversely) to P/T p70S6K (r=−0.041, p<0.05) protein expression. The ΔΔ in GS independent activity was related to ΔΔ in glycogen content (r=0.62, p=0.0001), P/T AKT (r=0.44, p=0.01), IR (r=0.52, p<0.005) and (inversely) to P/T p70S6K (r=−0.40, p<0.05) and JNK (r=−0.47, p<0.01) protein expression. The ΔΔ in GS total activity was inversely related to the ΔΔ in glycogen content (r=0.62, p=0.0001) and P/T JNK protein expression (r=−0.45, p=0.01). The ΔΔ in glycogen was related to the ΔΔ in AKT protein expression (r=0.37, p<0.05).

DISCUSSION

Based on our results, we extend findings in the literature that there are baseline sex differences in skeletal muscle metabolism ^[7–10], specifically the change in GS fractional and independent activity and AKT protein expression with insulin stimulation. Also, basal AMPK, GSK-3β, ERK1/2 and p38 which are each lower in older men than in older women. Further, the changes in GS fractional activity and in AKT protein expression with insulin are both related to insulin sensitivity by the glucose clamp. When examining sex differences in the changes in muscle metabolism with weight loss or exercise training, we found that the change in GS fractional activity with insulin stimulation was higher in women than men after exercise training and higher in men than women after weight loss which may in part explain sex differences in insulin sensitivity. Interventions designed to address skeletal muscle and whole-body insulin resistance could take into consideration biological sex, among other factors.

Older men exhibited lower non-oxidative glucose metabolism and insulin sensitivity compared to similar aged women. Non-oxidative glucose metabolism is the utilization of glucose that does not result in the production of energy (ATP), which includes glucose uptake, glucose phosphorylation, and glycogenesis. Glucose uptake following insulin stimulation is regulated by the glucose transporter 4 (GLUT4). GLUT 4 was not different between men and women in the current study. Skeletal muscle glucose uptake is dependent on GLUT4 translocation to the cell membrane. Insulin-stimulation of GLUT4 translocation is dependent on the binding and activation of the insulin receptor and the activation of the downstream PKB/AKT pathway. Despite both sexes having similar activation/phosphorylation of the insulin receptor, men had a lower activation/phosphorylation of the downstream pathway AKT in skeletal muscle in response to insulin compared to women. Additionally, our study found that men had reduced insulin sensitivity during the hyperinsulinemic clamp compared to women, suggesting impaired glucose uptake and utilization. Taken together, these finding suggest that diminished insulin-stimulation of the downstream PKB/AKT pathway could contribute to the lower skeletal muscle glucose metabolism observed in men compared to women.

The storage of glucose as glycogen is a major contributor to non-oxidative glucose metabolism and glucose homeostasis. Glycogen synthesis is a major route of glucose disposal in skeletal muscle after insulin stimulation, and we found that the change in glycogen content with insulin stimulation following the interventions was related to the change in glycogen synthase activity. Glycogen synthase is inactivated by phosphorylated glycogen synthase kinase 3. Thus, the enzyme glycogen synthase kinase-3 (GSK-3β) is a negative regulator of glycogen synthase and glycogenesis. GSK-3β inversely correlates with both GS activity and insulin stimulated glucose disposal^[18]. Treatment of skeletal muscle cells with GSK-3 inhibitor results in a dose-dependent activation of glycogen synthase and results in glucose incorporation into glycogen^[19]. Phosphorylation of GSK-3β at serine 9 by the PKB/AKT pathway inactivates GSK-3β allowing glycogen synthase to become activated and promote glycogenesis. Activation of GS is done through phosphorylation by PKB/AKT and the binding of the protein phosphatase 1 (PP1) which results in the dephosphorylation of GS at other inhibitory sites (serine 640). This promotes the formation of glycogen. In the current study, we did not find a significant difference between insulin effect on baseline PP1, PP2A, or glycogen in men and women, even though the effect of insulin to increase GS activity was 2-fold higher in women than men. We do not have an explanation for this other than the sample sizes were relatively low (n=15–19 per group).

Neither basal AKT nor p70 were different in older men and women, and both AKT and p70 phosphorylation increased after a meal similarly in men and women^[20] We too found that both AKT and p70 expression were similar between the sexes under basal conditions and that the phosphorylation of the two proteins increased during the hyperinsulinemic-euglycemic clamp. However, in our study, the change (insulin-basal) in AKT was higher in women than in men. The differences between our findings could be partially due to the smaller sample size in the aforementioned study^[20] and the difference in methodologies (postprandial vs. glucose clamps) between the two studies.

Women demonstrated elevated phosphorylation and activation of p38 mitogen-activated protein kinase during basal and insulin-stimulated conditions compared to men. Studies have shown that p38 can impair insulin signaling and contribute to insulin resistance. A study by Kim et al^[21] showed that p38 activity was elevated in skeletal muscle of obese and diabetic subjects and was associated with a reduction in insulin-stimulated glucose uptake. However, this study was primarily in males with only three female participants. Interestingly, phosphorylated p38 in the skeletal muscle of young healthy women at rest was significantly higher than in age-matched men, suggesting a potential sex difference^[22]. Further research is needed to examine the sex differences regulating the activation of p38 and its potential role in female skeletal muscle.

We believe this is the first paper to examine the interactions of sex (men vs. women) and healthy aging interventions (weight loss vs. aerobic exercise training) on skeletal muscle glycogen synthase (fractional velocity, independent, and total activity) during in vivo insulin stimulation. We found that the change in GS fractional activity with insulin stimulation was higher in women than men after exercise training and higher in men than women after weight loss and further that changes (due to intervention) in the change (due to insulin) in GS fractional activity are higher in women who exercise than men. Moreover, changes in GS fractional activity (insulin-basal) with the interventions were positively associated with changes in insulin-stimulated over basal AKT, IR, and IRS and inversely to p70S6K protein expression.

Despite sex differences in GS fractional, independent, and total activity changes with insulin stimulation by intervention and overall improvements in M with the interventions, we did not find significant sex differences in the changes in whole body insulin sensitivity for the interventions. Future studies with larger sample sizes per intervention group are needed to determine whether M was different by gender and intervention.

Because our study was limited in sample size to detect interactions in the protein expression by sex and intervention, we cannot rule out that there are also sex differences in the P/T ratio of GSK-3β, IGF-1R, ERK1/2, IR, IRS, p70s6K, JNK and AMPK during insulin stimulation after weight loss or exercise training. In one study of acute exercise, basal AMPK and ERK1/2 phosphorylation increased in skeletal muscle at 30 mins after sprint exercise in both men and women, but there was no difference between sexes^[23]. However, strengths of our study include our approach in examining how muscle changes during in vivo insulin delivery with WL and AEX which requires four muscle biopsies and two glucose clamps for each participant and thus, is both invasive and time-consuming, and yet, the gold-standard to answer these questions.

We^{[7, 8]} and others^[24] have reported gender differences in insulin sensitivity. A review of studies using hyperinsulinemic-euglycemic clamps comparing men and women concludes that whole body insulin sensitivity is higher in premenopausal women than in age-matched men.^[24] Our studies in older individuals^{[7, 8]} have shown higher insulin sensitivity in postmenopausal women compared to older men. Underlying biological explanations for reduced insulin sensitivity and activation of AKT and GS fractional activity by insulin in older men could potentially involve differences in sex hormones, muscle perfusion, muscle morphology, and substrate availability between the sexes. Testosterone, estrogen, growth hormone, and insulin-like growth factor are hormones that can influence muscle adaptations.^{[25, 26]} Further, men have significantly larger type II fiber areas in vastus lateralis muscle^[27]; yet, women have higher number of type I muscle fibers in the vastus lateralis than men.^{[28, 29]} It has also been reported that women have a greater capillary density per given muscle area than men.^{[28, 29]} The larger capillary supply and higher percent type I fibers in women may improve nutritive flow and increase oxidative glucose and fatty acid metabolism contributing to gender differences in insulin sensitivity^[24]. Measurement of sex steroids, blood flow, muscle morphology (fiber typing and capillary density) was beyond the scope of the present study, but future investigations could examine these potential mediators. Despite having lower aerobic capacity, lower fat-free mass, and higher body fat, postmenopausal women in our study had greater insulin sensitivity than men. Our study is limited in sample size to investigate the contributions of these phenotypic factors on the changes in insulin sensitivity, insulin activation of GS and other mechanisms of insulin action with exercise and weight loss. In a recent review it was suggested that tailoring treatment approaches based on sex or gender can lead to more effective management of insulin resistance.^[30] Therefore, future studies should control for physiological phenotypes in designing intervention(s) to improve skeletal muscle metabolism in men and women.

Conclusion

We found gender differences in the basal expression of key proteins involved in skeletal muscle glucose and glycogen metabolism, and that insulin activation of GS and AKT in response to a hyperinsulinemic-euglycemic clamp were lower in men than women. Men and women had different insulin activation of GS following weight loss and exercise interventions. Based on these findings, we suggest that sex-specific approaches to examine the effects of healthy lifestyle interventions on skeletal muscle mechanisms in insulin action are warranted.

Supplementary Material

Supinfo3

Figure S2: The change (insulin-basal) in GS fractional activity versus the change in GSK-3β (r=0.52, p=0.001) (upper panel), the change in IR (r=0.56, p=0.0004) (middle panel) and the change in IRS (r=0.33, p=0.05) (lower panel).

Supinfo2

Figure S1: The change (insulin-basal) in GS fractional activity versus insulin sensitivity (M) (r=0.44, p=0.0004) (upper panel), the change (insulin-basal) in AKT versus M (r=0.41, p=0.0005) (middle panel), and the change (insulin-basal) in GS fractional activity versus the change in AKT (r=0.65, p<0.0001) (lower panel).

What is already known about this subject?

• Older men are more insulin resistant (whole-body) than women.

• Insulin sensitivity is related to the change (insulin-basal) in muscle GS activity.

• In vivo insulin activation of skeletal muscle glycogen synthase is lower in older men than in older women.

What are the new findings in the manuscript?

• The change (insulin minus basal) in GS activity was related to the change in AKT expression at baseline.

• The postintervention minus preintervention change (insulin-basal) (ΔΔ) in GS activity was related to the ΔΔ in glycogen content.

• Weight loss was more effective at increasing in vivo insulin action (GS) in men than in women.

• Aerobic exercise training was more effective at increasing in vivo insulin action (GS) in women than in men.

How might our results change the direction of research or the focus of clinical practice?

• Sex differences in insulin sensitivity may be explained at the cellular muscle level; to improve skeletal muscle insulin action of glycogen synthase in older adults, it may be important to study different behavioral strategies depending on the individual’s sex.

• Understanding downstream pathways in skeletal muscle important for glucose metabolism may require a sex specific approach in humans.

Data Availability Statement:

Data will be available upon reasonable request.