Twenty-Four Hour Energy Expenditure and Skeletal Muscle Gene Expression Changes After Bariatric Surgery

Abstract

Context:

Obesity is characterized by decreased insulin-stimulated glucose uptake in muscle and shift from glucose to lipid oxidation, the so-called metabolic inflexibility. Biliopancreatic diversion (BPD), a mainly malabsorptive bariatric operation, determines a prompt improvement of insulin resistance, but the mechanisms are still unclear.

Objective:

We aimed to estimate the response of glucose transporter 4 (GLUT4) and hexokinase-II (HKII) gene expression to hyperinsulinemia before and after surgical treatment with a BPD or dietary-induced weight loss. The association with 24-hour energy expenditure and its different components—diet-induced thermogenesis (DIT), resting energy expenditure, physical activity (PA) of daily living, and physical exercise—was also determined.

Design, Setting, and Main Outcome Measures:

Case-control study: 20 subjects, BPD vs diet-induced weight loss. Four subjects withdrew in the surgical arm and 1 subject withdrew in the dietary arm. Insulin sensitivity was measured by the euglycemic-hyperinsulinemic clamp. Energy expenditure was assessed by indirect calorimetry over 24 hours. Muscle biopsies were obtained during the clamp to measure gene expression: GLUT4 and HKII.

Results:

Insulin sensitivity increased significantly (P < .01) only after BPD (0.101 ± 0.012 to 0.204 ± 0.033 μmol/kg/min/pM). Enhanced GLUT4 and HKII mRNA levels were observed after surgery (P < .0001 and P = .021, respectively), whereas they were not affected by diet-induced weight loss. Carbohydrate oxidation (P = .0027), DIT (P = .033), PA (P = .036), and energy expenditure during the exercise (P = .017) increased only in the BPD group.

Conclusions:

BPD improved impaired glucose metabolism and insulin resistance through increased glucose uptake, glycogen synthesis, and glucose oxidation. Furthermore, the concomitant increase in DIT, PA, and exercise in BPD patients may partly explain their ability to sustained long-term weight loss and may contribute to the improved insulin sensitivity.

In the skeletal muscle of lean healthy individuals, insulin increases the efficiency of glucose metabolism by recruiting the glucose transporter, GLUT4, from an intracellular vesicle pool to the plasma membrane, and by enhancing glucose phosphorylation to glucose-6-phosphate, catalyzed by hexokinase-II (HKII) (1–3). An impairment in glucose transport associated with a reduced efficiency of glucose phosphorylation plays a critical role in the pathogenesis of insulin resistance of skeletal muscle in obesity and type 2 diabetes.

Kelley et al demonstrated with dynamic positron emission tomography imaging a defect in glucose transport and phosphorylation in diabetes, but only a defect in glucose transport in obesity (3), whereas Williams et al showed the defect in glucose phosphorylation was present in both obesity and type 2 diabetes (4). A defective response of GLUT4 gene expression to insulin was also demonstrated during a euglycemic hyperinsulinemic clamp in insulin resistant, obese subjects (5).

Insulin increased HKII transcription in L6 cells (6) and enhanced its muscle gene expression during the euglycemic-hyperinsulinemic clamp in rats (7). Insulin infusions increased HKII mRNA and protein expression (other than its enzymatic activity) in percutaneous muscle biopsies of healthy control subjects (8). Hyperinsulinemia during a euglycemic clamp failed to increase HKII mRNA in the skeletal muscle of obese individuals (9). Therefore, an imbalance between glycolytic and oxidative enzyme capacity in skeletal muscle of obese subjects has been suggested as a feature of insulin resistance (10). Increased energy expenditure (EE) has been associated with improvements in insulin resistance (11).

In humans, substrate oxidation is usually measured by indirect calorimetry. Twenty-four-hour indirect calorimetry is optimal for measuring substrate oxidation because the subject can freely move, consume meals, and perform physical exercise.

We hypothesize that lipid malabsorption, as it occurs after biliopancreatic diversion (BPD), can determine a shift from lipid to glucose oxidation through an increased glucose uptake and phosphorylation. To this extent, 24-hour substrate oxidation was measured in an indirect calorimetry chamber and insulin sensitivity was assessed by the euglycemic hyperinsulinemic clamp, whereas the gene expression of GLUT4 and HKII was measured in skeletal muscle biopsies.

Materials and Methods

Subjects

Twenty morbidly obese (body mass index >40 kg/m²) patients (6 men and 14 women) were included in the study (Table 1). None of the patients had diabetes mellitus or any other endocrine or nonendocrine disease. After an initial assessment, 10 patients chose to undergo BPD (consisting of a partial gastrectomy with a distal Roux-en-Y reconstruction resulting in an alimentary limb of 250 cm and a common channel of 50 cm) (12), while 10 patients opted for a balanced, hypocaloric dietary regimen (1200 kcal/d, 30% fat, 15% protein, and 55% carbohydrate). All 20 patients were studied before and 15 patients (9 in the diet group and 6 in the BPD group) were reevaluated 6 months after the start of the interventions.

The study protocol was approved by the Institutional Review Board and conducted according to the principles of the Helsinki declaration with all subjects providing written consent.

Euglycemic-hyperinsulinemic clamp

Peripheral insulin sensitivity was evaluated by a 3-hour euglycemic-hyperinsulinemic clamp procedure. After inserting a cannula in a dorsal hand vein for sampling arterialized venous blood, and another cannula in the antecubital fossa of the contralateral arm for infusions, the subjects rested in the supine position for at least 1 hour. They were placed with 1 hand warmed in a heated air box set at 60°C to obtain arterialized blood samples. Whole-body glucose uptake (M value; in μmol · kg_FFM⁻¹ [fat-free mass] · min⁻¹) was determined during a primed constant infusion of insulin (at the rate of 6 pmol kg⁻¹ min⁻¹) (13). The fasting plasma glucose concentration was maintained throughout the insulin infusion by means of a variable glucose infusion and blood glucose determinations every 5 minutes. Whole-body peripheral glucose use was calculated during the last 40-minute period of the steady-state insulin infusion.

Muscle biopsy

A muscle needle biopsy was taken from the midvastus lateralis muscle under local anesthesia with 1% lidocaine. Muscle samples were snap-frozen in liquid nitrogen and stored at −80°C for subsequent analysis.

RNA extraction and real-time PCR analysis

RNA from skeletal muscle (approximately 50 mg of muscle) was extracted using the RNeasy kit (Qiagen GmbH, Hilden, Germany) and 40 ng RNA was reverse transcribed to cDNA using Sensiscript enzyme (Invitrogen, Carlsbad, California). Real-time PCR was used to measure specific mRNAs (ABI-PRISM 7700 Sequence Detector; Perkin-Elmer Applied Biosystems, Foster City, California). All reactions were performed in 384-well MicroAmp Optical plates. Amplification mixes (10 μL) contained the diluted cDNA sample, 2× TaqMan Universal PCR Mastermix (Qiagen) or SYBR Green PCR Mastermix (Invitrogen), forward and reverse primers, and probe for the specific mRNAs, as well as cyclophilin A mix. Thermal cycling conditions included 10 minutes at 95°C before the onset of the PCR cycles, which consisted of 40 cycles at 95°C for 15 seconds and 65°C for 1 minute. As endogenous control to correct for potential variation in RNA loading and quantification, human PPIA (peptidylprolyl isomerase A; cyclophilin A) was used. The oligonucleotide sequences for the primer pairs used were asa follows: for SLC2A4 (GLUT4) (5′-GCTACCTCTACATCATCCAG-3′ and 5′-TGTCTCGAAGATGCTGGTC-3′) and HKII (5-ACAGGTGCTCTCAAGCCCTAAG-3 and 5-FCGAGGCCGCATCTCAGAGCGG-3).

These mRNA levels were obtained by measuring fluorescence from the progressive binding of SYBR green I dye to double-stranded DNA. mRNA expression was calculated using the ΔCT (threshold value) method (14). Briefly, the ΔCT was calculated by subtracting the CT for cyclophilin A, from the CT for the gene of interest. The relative expression of the gene of interest was calculated using the expression 2-ΔCT and reported as arbitrary units. All PCR runs were performed in duplicate.

Body composition

The day preceding the start of the intervention, body weight was measured to the nearest 0.1 kg by a beam scale, and height was measured to the nearest 0.5 cm using a stadiometer (Holatin; Crosswell, Wales, United Kingdom).

Total body water measurement

Total body water (TBW) was determined using 0.19 Bq tritiated water in 5 mL saline solution administered as an iv bolus injection (15). Blood samples were drawn before and 3 hours after the injection. Radioactivity was determined in duplicate on 0.5 mL plasma using a β-scintillation counter (Model 1600TR; Canberra-Packard, Meriden, Connecticut). A 5% correction was made for nonaqueous hydrogen exchange (16); water density at body temperature was assumed to be 0.99371 kg/L. TBW (kg) was computed as ³H₂O dilution space (L) × 0.95 × 0.99371. The within-subject coefficient of variation for this method is 1.5% (17).

Indirect calorimetry chamber

At the time of the preintervention study, all subjects were on an ad libitum diet, which consisted of 60% carbohydrate, 30% fat, 10% protein (at least 1 g per kg of body weight). This dietary regimen was maintained for 1 week prior to the study and was similar between the groups.

Subjects spent 1 day (starting at 08:00) in the indirect calorimetry chamber. Twenty-four-hour EE and resting EE (REE) were measured as previously described (18, 19). Briefly, the carbon dioxide (CO₂) concentration was measured by a 2% full-scale (0%–2%) infrared absorption analyzer (URAS 3G; Hartmann & Braun, Frankfurt, Germany), while the oxygen (O₂) concentration was assessed by a 2% full-scale (19%–21%) paramagnetic analyzer (Magnos 4G; Hartmann & Braun). Both gas analyzers operated with a precision of 0.02 vol%. The zero values of both analyzers were calibrated by allowing fresh air to flow through the sample and the reference lines simultaneously, whereas the span values were calibrated using commercially available gas mixtures (Rivoira, Torino, Italy). The composition of the gas mixture used to calibrate the O₂ analyzer was 19.48% O₂ in N₂. The composition of the gas mixture used to calibrate the CO₂ analyzer was 1.5% CO₂ in N₂. Every day at the beginning of each experimental session, the chamber was calibrated. The algorithm used for computing O₂ consumption (VO₂ mL/min) and CO₂ production (VCO₂ mL/min) approximated the gas consumption/production during a time interval by adding 2 independent quantities: dynamic (open) and static (closed) gas production (20), consumption being negative production. Values were corrected for temperature, barometric pressure, and humidity. Energy expenditure was calculated according to Ferrannini (21). Protein oxidation was determined from 24-hour urinary nitrogen excretion (on a BUN Analyzer; Beckman Instruments, Fullerton, California); carbohydrate and lipid oxidation rates were determined from the nonprotein respiratory quotient (RQ). Calibration procedures, precision, and variability of the respiratory chamber have been reported elsewhere (20). Physical activity (PA) was monitored by means of 2 orthogonal ultrasound sensors: the sensitivity of the receivers was set so that respiratory movements were not detected.

Exercise testing

On the day before the experimental session, the patients entered the Energy Metabolism Research Unit (Department of Internal Medicine, Catholic University, Rome, Italy) at 08:00 after an overnight fast. They were allowed to become familiar with the testing equipment and walk on the motorized treadmill. During the day spent in the calorimetric chamber, the patients performed physical exercise at 16:00 by walking for 30 minutes up a 10% gradient at a constant speed of 3 km/h.

Diet in the indirect calorimetry chamber

While in the indirect calorimetry chamber, all subjects were given a diet of 30 kcal/kg of fat-free mass (kg_FFM) consisting of 55% carbohydrate, 30% fat, and 15% protein. This amount was divided as follows: 20% at breakfast, 40% at lunch, 10% as an afternoon snack, and 30% at dinner. The 4 meals served in the chamber were the same for all patients and were prepared by a dietitian using common foods such as meat, fish, vegetables, bread, fruit, and so on. The food given and returned was weighed to the nearest gram on precision scales (KS-01; Rowenta, Berlin, Germany). The nutrient content of all food items was calculated by using computerized tables (Food Processor II; Hesha Research, Salem, Oregon, modified according to the food tables of the Istituto Nazionale di Nutrizione, Rome, Italy). The energy content of food was calculated as follows: 4.3 kcal/g for protein, 4.2 g for starch (or starch equivalent), and 9.3 kcal/g for fat (22). The net caloric intake was computed by subtracting the energy fecal loss to the gross caloric intake.

24-hour nitrogen and lipid output

Twenty-four-hour urine and stool were collected while in the indirect calorimetry chamber. Total urinary nitrogen was analyzed by the BUN Analyzer. Stool aliquots were homogenized and analyzed in triplicate for nitrogen, carbohydrate, and lipid content using a Fenyr analyzer (PerCon Prüfgeräte, Hamburg, Germany). The interassay variation of total fecal fat content was found to be <0.5 g.

Analytical methods

Blood samples were drawn into EDTA-evacuated tubes. The plasma was immediately separated by centrifugation at 4°C and stored at −80°C until assay. Plasma glucose was measured by the glucose-oxidase method (Beckman Instruments). Plasma insulin was assayed by microparticle-enzyme immunoassay (Abbott, Pasadena, California) with a sensitivity of 1 μU/mL and an intra-assay coefficient of variation of 6.6%.

Statistical analysis

All of the data are expressed as means ± SEM unless otherwise specified. The Wilcoxon paired-sample test and ANOVA for repeated measurements, followed by the Tukey test, were used for intragroup and intergroup comparisons, respectively. Correlation analyses were performed with the Spearman correlation coefficient. Regression analysis was performed to identify weight loss determinants.

The details of this model are found in the Supplemental Appendix published on The Endocrine Society's Journals Online web site at http://jcem.endojournals.org.

Two-sided P < .05 was considered significant. The correlation coefficients and the relative significance between variables were evaluated by SPSS version 13 (IBM, Armonk, New York).

Results

Four of the 10 subjects in the surgical arm dropped out of the study. The reasons for withdrawal were pulmonary thromboembolism (n = 1), incisional hernia (n = 1), and refusal to undergo muscle biopsy (n = 2). One of the controls also refused to have the second biopsy. The data reported below and in the tables only pertain to those subjects who completed the study.

The anthropometric characteristics of the study subjects are reported in Table 1. After 6 months, the subjects in the BPD arm lost 29.8 ± 11.6 kg and those that had dietary restriction lost 14.7 ± 22.2 kg (P < .01). As shown in Table 1, the loss of fat mass accounted for most of weight lost, especially after BPD. However, FFM also was significantly (P < .05) reduced.

Table 1 shows the fasting plasma glucose, insulin and alanine aminotransferase concentrations, and profiles with significant decreases in plasma cholesterol and triglycerides after BPD.

Insulin sensitivity (Table 2), in absolute terms or normalized by the steady-state plasma insulin concentration, improved significantly only in the BPD group.

The gene expression of GLUT4 and HKII increased significantly after BPD (P < .001 and P = .025), respectively, but not after diet (Figure 1). Changes in GLUT4 mRNA and HKII mRNA correlated with the changes in M/I (M normalized by the plasma insulin concentration averaged during the last 40 minutes of the clamp (4 time point mean) (β = .971, P < .0001 and β = .885, P = .021, respectively).

Individual fits of 24-hour EE time course were generally good (R² ranging from 0.50 to 0.89). An example of the experimental EE time course and its fitting function is shown in Supplemental Figure 1 of the Supplemental Appendix. Preoperative REE and weight loss were associated in BPD patients (R² = 0.88, P = .029). In the subjects undergoing dietary restriction, weight loss was correlated to both the initial REE (R² = 0.89, P = .0018) and the initial weight (R² = 0.86, P = .0039). However, no relationship between change in weight loss and change in REE was found.

Energy expenditure measurements and model-derived parameter estimates for diet-induced thermogenesis (DIT) are reported in Table 2. DIT increased in the BPD group (P = .033), but did not change in the dieting group. Furthermore, both groups increased carbohydrate oxidation after weight loss; however, this increase was only significant in the BPD group (P = .0027) with nonprotein RQ increasing by 0.14 ± 0.05, which indicates a shift from lipid to carbohydrate oxidation.

The kilocalories spent during the physical exercise increased significantly (P < .05) only after BPD. PA increased in the BPD group from 3.35 ± 0.33 to 4.65 ± 0.66 a.u. (P = .036) and in the dieting group from 3.65 ± 0.36 to 4.39 ± 0.34 a.m.u. (P = not significant). Weight loss after BPD was also associated with increased PA, as suggested by the regression analysis (R² = 0.88, P = .015). REE did not change in either group.

Discussion

Obesity results in resistance to insulin-promoted uptake of glucose into muscle. Although insulin should stimulate skeletal muscle to take up glucose from the circulation as well as actively synthesizes glycogen (24–26), in the insulin resistance state defective glucose transport and glycogen synthesis are observed (27, 28). We showed that 3 hours of hyperinsulinemia significantly increased both GLUT4 and HKII mRNA levels by 1.5 and 0.6, respectively, after BPD, but not after diet-induced weight loss. The BPD also increased glucose oxidation, DIT, 24-hour PA, and exercise-induced EE. Therefore, overall, our data suggest that at least BPD reduces the deposition of lipids into the skeletal muscle fibers (29) and improves defective glucose metabolism and insulin resistance through an increased EE, glucose uptake, glycogen synthesis, and glucose oxidation. Other metabolic operations, such as gastric bypass, seem to act mainly through the regulation of food consumption in the frame of the gut–brain axis in both humans (30) and animals (31). Furthermore, gastric bypass, sleeve gastrectomy, and BPD have been proven to be effective in the treatment of type 2 diabetes (32–34).

The DIT increase after BPD might be explained by the BPD-associated intestinal hypertrophy and intestinal mass (35), which requires more energy for postprandial processes and/or by the lower metabolizable lipid intake. The fecal lipid loss after BPD can be as high as 80% of the amount consumed. Westerterp et al (36) estimated that the DIT after an isocaloric combined high-protein, high-carbohydrate, and low-fat diet is 90 kcal higher over 24 hours than a high-fat, low-protein, and low-carbohydrate diet regimen that matches energy needs of the individual, increasing carbohydrate oxidation and decreasing lipid oxidation. Another possibility is increases in gut energy requirements due to intestinal hypertrophy after BPD, which has been described in humans (37) and in rats (38). Either an increase in intestinal wall thickness, a lengthening of the gut, or an increased mass may be involved. In a different experimental model (ie, a very long limb gastric bypass on Wistar rats [35]), DIT was greater in gastric bypass rats than either sham-operated ad libitum fed animals or sham-operated weight-matched controls. In contrast with our findings, Bueter et al (35) showed in gastric bypass rats a lower RQ than sham-operated rats, which may be explained by the gastric bypass rat model not having increased fecal calorie content or fecal lipid loss.

The increased EE and PA after BPD may partly explain the long-term weight loss maintenance of this procedure, although our study was not powered to show changes in 24-hour EE. Two-thirds of the compensatory increase in EE associated with overfeeding can be attributable to increased PA (ie, fidgeting, maintenance of posture, and other PAs of daily living, which have been termed “nonexercise activity thermogenesis [NEAT]” [39]). Differences in NEAT accounted for much of the variability in fat gained during overfeeding (39). We were not able to measure NEAT accurately, but our data suggest that the BPD group had an increase in PA due to increased movements.

Differences in the energy needed to move a larger or smaller body mass seem to account for some of the differences in 24-hour EE, as shown by replacing the lost weight with backpack loads (40). One possibility is that the efficiency with which skeletal muscle performs mechanical work differs depending on body weight. Changes in skeletal muscle composition after weight loss may alter EE and could explain why exercise is helpful in maintaining weight loss. Wade et al (41) demonstrated an inverse relationship between the percentage of type I (slow twitch) fibers and percentage of body fat, suggesting that muscle fiber type may contribute to obesity.

In conclusion, the present study shows that biliopancreatic diversion improves insulin sensitivity through the restoration of the insulin stimulatory effect on GLUT4 and HKII transcription and increased glucose oxidation. Furthermore, BPD increases DIT, exercise-induced EE, and PA, features that may explain the ability of this bariatric operation to sustained long-term weight loss and which may in its own right contribute to improved insulin sensitivity.

Acknowledgments

The authors thank Anna Caprodossi for her valuable technical support and declare no conflict of interest in relation to the present study.

Disclosure Summary: The authors have nothing to disclose.

Abbreviations

 
• BPD

  Biliopancreatic diversion

 
• ΔCT

  threshold cycle

 
• DIT

  diet-induced thermogenesis

 
• EE

  energy expenditure

 
• FFM

  fat-free mass

 
• GLUT4

  glucose transporter 4

 
• HKII

  hexokinase-II

 
• NEAT

  nonexercise activity thermogenesis

 
• PA

  physical activity

 
• REE

  resting energy expenditure

 
• RQ

  respiratory quotient

 
• TBW

  Total body water.

References