Mitochondrial Function in Skeletal Muscle Is Normal and Unrelated to ...

ORIGINAL

ARTICLE

E n d o c r i n e

C a r e

Mitochondrial Function in Skeletal Muscle Is Normal and Unrelated to Insulin Action in Young Men Born with Low Birth Weight Charlotte Brøns, Christine B. Jensen, Heidi Storgaard, Amra Alibegovic, Stine Jacobsen, Emma Nilsson, Arne Astrup, Bjørn Quistorff, and Allan Vaag Steno Diabetes Center (C.B., C.B.J., H.S., A.Al., S.J., E.N., A.V.), 2820 Gentofte, Denmark; Department of Human Nutrition (C.B., S.J., A.As.), Faculty of Life Sciences, University of Copenhagen, 1870 Frederiksberg C, Denmark; and Department of Medical Biochemistry and Genetics (B.Q.), Panum Institute, University of Copenhagen, DK-2200 Copenhagen, Denmark

Objective: Low birth weight (LBW) is an independent risk factor of insulin resistance and type 2 diabetes. Recent studies suggest that mitochondrial dysfunction and impaired expression of genes involved in oxidative phosphorylation (OXPHOS) may play a key role in the pathogenesis of insulin resistance in aging and type 2 diabetes. The aim of this study was to determine whether LBW in humans is associated with mitochondrial dysfunction in skeletal muscle. Methods: Mitochondrial capacity for ATP synthesis was assessed by 31phosphorus magnetic resonance spectroscopy in forearm and leg muscles in 20 young, lean men with LBW and 26 matched controls. On a separate day, a hyperinsulinemic euglycemic clamp with excision of muscle biopsies and dual-energy x-ray absorptiometry scanning was performed. Muscle gene expression of selected OXPHOS genes was determined by quantitative real-time PCR. Results: The LBW subjects displayed a variety of metabolic and prediabetic abnormalities, including elevated fasting blood glucose and plasma insulin levels, reduced insulin-stimulated glycolytic flux, and hepatic insulin resistance. Nevertheless, in vivo mitochondrial function was normal in LBW subjects, as was the expression of OXPHOS genes. Conclusions: These data support and expand previous findings of abnormal glucose metabolism in young men with LBW. In addition, we found that the young, healthy men with LBW exhibited hepatic insulin resistance. However, the study does not support the hypothesis that muscle mitochondrial dysfunction per se is the underlying key metabolic defect that explains or precedes whole body insulin resistance in LBW subjects at risk for developing type 2 diabetes. (J Clin Endocrinol Metab 93: 3885–3892, 2008)

ndividuals born with low birth weight (LBW), a surrogate marker of impaired intrauterine growth, are at increased risk for developing insulin resistance and type 2 diabetes later in life (1– 4). In previous studies we have shown that LBW is associated with a redistribution of body fat to the abdominal region (5), impaired insulin-stimulated glucose uptake by the forearm tissue (6), and decreased whole body insulin-stimulated glycolytic flux (GF), even before the development of whole body peripheral insulin resistance (7) in 19-yr-old healthy men. Moreover, we

I

recently showed that LBW is associated with altered muscle fiber distribution (8) and reduced basal expression in muscle and adipose tissue of the glucose transporter 4 (GLUT-4), as well as several key insulin-signaling proteins (9, 10). These findings suggest the presence of multiple abnormalities in insulin-sensitive tissues before the onset of whole body insulin resistance and type 2 diabetes, which could represent key metabolic defects linking LBW with the later development of insulin resistance and the metabolic syndrome, including type 2 diabetes.

0021-972X/08/$15.00/0

Abbreviations: BW, Birth weight; FFM, fat-free mass; FM, fat mass; GF, glycolytic flux; HGP, hepatic glucose production; IVGTT, iv glucose tolerance test; LBW, low birth weight; Mvalue, insulin stimulated glucose uptake; NBW, normal birth weight; OXPHOS, oxidative phosphorylation; PCr, phosphocreatine; PGC, prostaglandin; Pi, inorganic phosphor; 31PMRS, 31phosphorus magnetic resonance spectroscopy; Ra, rate of unlabeled glucose appearance; rt-PCR, real-time PCR; Vmax, maximum velocity; VO2 max, maximal rate of oxygen consumption.

Printed in U.S.A. Copyright © 2008 by The Endocrine Society doi: 10.1210/jc.2008-0630 Received March 18, 2008. Accepted July 7, 2008. First Published Online July 15, 2008

J Clin Endocrinol Metab, October 2008, 93(10):3885–3892

jcem.endojournals.org

3885

3886

Brøns et al.

Mitochondrial Function in LBW Men

Evidence has emerged over recent years that impaired mitochondrial function may represent a primary pathogenic defect involved in the development of insulin resistance in type 2 diabetes (11–13) as well as in aging (14 –16). Impaired ATP synthesis due to reduced oxidative phosphorylation (OXPHOS) capacity, leading to accumulation of intramyocellular fat in skeletal muscle and subsequently muscle insulin resistance, appears to be particularly important in this respect. Obese individuals and subjects with type 2 diabetes display alterations in mitochondrial morphology (17), and patients with type 2 diabetes and their prediabetic offspring exhibit decreased expression of OXPHOS genes in skeletal muscle (18). In addition, mitochondrial dysfunction in liver and pancreatic ␤-cells has been linked to impaired insulin secretion in growth-retarded rats (19, 20). Although challenged by other recent studies (21–23), the hypothesis that mitochondrial dysfunction is a general mechanism contributing to a variety of known key metabolic defects in type 2 diabetes remains attractive. Following this line of thinking, we hypothesized that programming in utero might be a potentially important etiological factor determining mitochondrial function. To test our hypothesis, that LBW in humans is associated with impaired muscle mitochondrial function, we assessed in vivo mitochondrial function by 31phosphorus magnetic resonance spectroscopy (31P-MRS) after energy depleting exercise in young, prediabetic men born with LBW, and in matched controls. To clarify further the role of specific key OXPHOS genes involved in the electron transport chain and previously associated with insulin resistance and type 2 diabetes, quantitative real-time PCR (rt-PCR) was performed on muscle biopsies for selected genes.


family history of diabetes in two generations were excluded in this study to exclude major genetic confounding. Subjects with a body mass index greater than 30 kg/m2 and a high physical activity level were also excluded from participation.

Experimental protocol Study activities were performed over 3 d. Subjects were requested to refrain from strenuous physical activity and consuming alcohol 3 d before examination. To ensure standardized conditions, all meals the day before and throughout the study period were provided to the subjects.

Dual-energy x-ray absorptiometry scanning On d1, fat-free mass (FFM) and fat mass (FM) were assessed by dual-energy x-ray absorptiometry (Lunar Radiation, Madison, WI). 31

P-MRS

An Otsuka Electronics VivoSpec spectrometer was used, interfaced with a 2.9-tesla magnet (Magnex Scientific, Oxford, UK) with a 26-cm bore. 31P-MRS was performed on the 2nd day on two different muscle groups in separate experiments, and data were acquired as previously described (24). Initially, maximal voluntary contraction was determined as the best of three 1-sec maximal contractions. Thereafter, 31P-MRS recording at a time resolution of 10 sec was performed for 3 min rest, 3 min exercise, and 6 min recovery. The protocol involved 18 successive, intermittent isometric contractions at 50% maximal voluntary contraction, each lasting 7 sec, interspersed by 3 sec rest. The 31P-MRS spectra during rest and the changes in response to work are depicted in Fig. 1. The contraction force was monitored throughout the experiment. The protocol was selected to obtain steady-state aerobic exercise, in which pH changes are minimal, and all measurements were performed on the right arm and leg.

Hyperinsulinemic euglycemic clamp

On d 3 the clamp procedure was initiated at 0700 h after an overnight fast. A polyethylene catheter was placed in the antecubital vein for blood sampling. The hand was placed in a heated Plexiglas Subjects and Methods box to ensure arterialization of the venous blood. A second catheter was placed in the antecubital vein of the contralateral arm for test Ethical approval infusions. A primed-continuous infusion of [3-3H] tritiated glucose (bolus 10.9 ␮Ci, 0.109 ␮Ci/min) was initiated at 0 h and continued The protocol conformed to the Declaration of Helsinki and was apthroughout the examination. proved by the ethics committee for Copenhagen County. All subjects After a 120-min basal period, a 30-min iv glucose tolerance test (IVGTT) signed an informed consent before participation. was initiated to determine ␤-cell function. A glucose bolus of 0.3 g/kg body weight was infused over 1 min. Blood samples for glucose, insulin, and Subjects C-peptide were collected at 0, 2, 4, 6, 8, 10, 15, 20, and 30 min. A total of 46 healthy males was recruited from the Danish National Birth After the IVGTT a primed-continuous insulin infusion was initiated Registry according to birth weight (BW). There were 20 men who had LBW and fixed at 80 mU/m2䡠min throughout the 180-min clamp. Steady-state (BW ⱕ10th percentile), and 26 were age-matched controls with normal BW was defined as the last 30 min of the basal and insulin clamp period, when (NBW) (50th ⱕBWⱕ90th percentile). Of the 46 subjects, 14 had particitracer equilibrium was anticipated. Variable infusion of “cold” glucose pated in a previous study by our group (7). All men were singletons, born in (180 g/liter) enriched with tritiated glucose (110 ␮Ci/500 ml) was used 1979 –1980 at term in Copenhagen. Because a family history of type 2 to maintain euglycemia during insulin infusion. Blood glucose concendiabetes is a risk factor for developing the disease in itself, subjects with a tration was monitored every 5 min during steadystate using a blood glucose meter (OneTouch; LifeScan Inc., Milpitas, CA). The target blood glucose concentration was 5 mmol/liter, and the infusion rate was adjusted if necessary immediately after each blood glucose assessment. During the clamp period, blood samples for measuring tritiated glucose and water were drawn every 10 –15 min and determined as previously described (25). Samples for measuring insulin and C-peptide were drawn every 30 min. At the end of each steady-state period, a vastus lateralis muscle biopsy was taken using a Bergstro¨m needle under local anesthesia. The tissue was immediately froFIG. 1. 31P-MRS spectra of human skeletal muscle at rest and during exercise. Only part of the spectra, zen in liquid nitrogen and stored at ⫺80 C. including from the left Pi, phosphodiester (PDE), PCr, and the three phosphate groups (␣, ␤, and ␥) of ATP, is included. ppm, Parts per million.


jcem.endojournals.org

Quantitative rt-PCR Extraction of total RNA from the muscle biopsies was performed with TRI reagent (Sigma-Aldrich, St. Louis, MO). cDNA was synthesized using the QuantiTect Reverse Transcription kit (QIAGEN, Inc., Valencia, CA). rt-PCR was performed using the ABI 7900 Sequence Detection System (Applied Biosystems, Foster City, CA). Assays from Applied Biosystems were used: NDUFB6 (Hs00159583_m1), UQCRB (Hs00559884_m1), COX7A1 (Hs00156989_m1), ATP5O (Hs00426889_m1), and peroxisome proliferator-activated receptor gamma, coactivator 1 alpha (PGC-1␣) (Hs00173304_m1). All samples were run in duplicate; values were calculated using the standard curve method and normalized to the mRNA level of Cyclophilin A (4326316E; Applied Biosystems).

Calculations 31

3887

pearance, and hepatic glucose production (HGP) were calculated using Steele’s nonsteady-state equation (33). During the steady-state period, the rate of unlabeled glucose disappearance, and HGP were calculated at 10min intervals, and the distribution volume of glucose was set at 200 ml/kg body weight and the pool fraction as 0.65 (34). The GF was calculated from the appearance rate of tritiated water, and the total plasma water was assumed to be 93% of the total plasma volume (35, 36). HGP during the insulin-stimulated steady-state period was calculated as the difference between Ra and the glucose infusion rate. In cases in which Ra was lower than the exogenous glucose infusion (HGP negative), the M-value calculated from the glucose infusion rate was used in the calculations. The area under the curve was calculated using a trapezoidal method for glucose and insulin during the first-phase insulin response (0 –10 min) of the IVGTT. The hepatic insulin-resistance index was calculated as the product of mean fasting plasma insulin concentration and basal HGP (37).

P-MRS

The spectra were subjected to a 5-Hz line broadening and a baseline correction, involving a semi-manual, cubic spline procedure. Subsequently, the phosphocreatine (PCr), inorganic phosphor (Pi), and the three ATP peaks were fitted by a least-squares routine, assuming a Lorentzian or Gaussian curve (26). The peak areas were corrected for partial saturation measured at rest, i.e. 1.22, 1.14, and 1.11 for PCr, Pi, and ATP, respectively. Metabolite concentrations were calculated, assuming a resting ATP concentration of 5.8 mM (27). Intracellular pH was calculated from the difference in the chemical shift between Pi and PCr (28). The concentration of free ADP was estimated from the creatine kinase reaction, assuming equilibrium (29, 30). The aerobic capacity for ATP production was estimated from the PCr recovery kinetics, assuming a monoexponential model, and the maximum velocity (Vmax) was calculated from these values as described by Ratkevicius and Quistorff (31), assuming Michaelis-Menten kinetics and a Michaelis-Menten constant for ADP for OXPHOS of 30 ␮M (32). Similarly, the recovery kinetics of Pi after exercise was modeled by monoexponential kinetics and the recovery half-times presented.

Hyperinsulinemic euglycemic clamp and IVGTT Glucose turnover rates During the predefined insulin stimulated steady-state period (150 –180 min), rates of unlabeled glucose appearance (Ra), unlabeled glucose disap-

Statistics Statistical analysis was performed using SAS software (version 9.1; SAS Institute Inc., Cary, NC). An unpaired Student’s t test was used to identify statistically significant differences between NBW and LBW subjects. Correlations were calculated using Spearman’s or Pearson’s (normally distributed data) correlation coefficient. A P value of less than 0.05 was considered significant. Data are presented as mean values ⫾ SD, except gene expression data, which are presented as means ⫾ SEM. In a post hoc power calculation with recovery rates of PCr and Pi as endpoints, we had an 80% chance of detecting differences between NBW and LBW subjects of 23–35% in arm and leg muscles.

Results Clinical characteristics (Table 1) At birth, the LBW subjects were approximately 1200 g lighter than the control group (3893 ⫾ 207 vs. 2688 ⫾ 269 g; P ⬍ 0.0001), current height was less (1.83 ⫾ 0.07 vs. 1.77 ⫾ 0.05 m; P ⫽ 0.004), as were the plasma high-density lipoprotein levels (1.38 ⫾ 0.22 vs. 1.17 ⫾ 0.23 mmol/liter; P ⫽ 0.003). LBW subjects had a higher trunk FM (g) to total FM (g) ratio (0.50 ⫾

TABLE 1. Clinical characteristics of the study participants

BW (g) Age (yr) Weight (kg) Height (m) BMI (kg/m2) W/H ratio HDL (mmol/liter) LDL (mmol/liter) Triglyceride (mmol/liter) Cholesterol (mmol/liter) Systolic BP (mm Hg) Diastolic BP (mm Hg) VO2 max (l/min)a Trunk FM/total FM (g) Leg FM/total FM (g) Trunk FM/leg FM (%)

NBW subjects (n ⴝ 26) Mean ⴞ SD

LBW subjects (n ⴝ 20) Mean ⴞ SD

P value

3893 ⫾ 207 24.6 ⫾ 1.1 78.3 ⫾ 9.1 1.83 ⫾ 0.07 23.4 ⫾ 2.4 0.88 ⫾ 0.05 1.38 ⫾ 0.22 2.42 ⫾ 0.16 0.92 ⫾ 0.34 4.36 ⫾ 0.83 137 ⫾ 13 72 ⫾ 7 3.7 ⫾ 0.5 0.50 ⫾ 0.05 0.37 ⫾ 0.04 1.10 ⫾ 0.19

2688 ⫾ 269 24.2 ⫾ 0.5 77.7 ⫾ 10.9 1.77 ⫾ 0.05 24.8 ⫾ 3.7 0.89 ⫾ 0.05 1.17 ⫾ 0.23 2.65 ⫾ 0.73 1.10 ⫾ 0.37 4.31 ⫾ 0.76 135 ⫾ 12 73 ⫾ 7 3.5 ⫾ 0.7 0.53 ⫾ 0.04 0.34 ⫾ 0.03 1.24 ⫾ 0.16