Branched-chain amino acids (BCAA) belong among nutrients strongly linked with insulin sensitivity (IS) measures. Their exact effect on IS appears to be dependent on various conditions that remain to be better defined . The aim of the current study was to investigate effects of chronic increase of BCAA intake on IS in two groups of healthy subjects differing in their basal consumption of BCAA, i.e. vegans and omnivores. Interventional trial was designed where vegans and omnivores were separately interveined with 15 (women) or 20 (men) grams of BCAA daily. Examinations of participants are intended to be done at baseline, after the intervention and wash-out period. Examinations (anthropometry, hyperinsulinaemic-euglycaemic clamp, arginine test) and blood analyses will be performed at baseline, after the intervention and after a 6 month wash-out period. Samples of subcutaneous abdominal adipose tissue (AT) and skeletal muscle (SM, vastus lateralis) obtained before and after the intervention will be used for various analyses (mRNA levels of selected metabolic markers, fatty acid composition, mitochondrial activity).

Subjects and methods Listed elsewhere Intervention Listed elsewhere

Clinical examination Each subject underwent a basic medical check-up with anthropometric examination (height (m), weight (kg), BMI (kg/m2), waist circumference (cm), waist-hip ratio). Body composition was measured using bioimpedance analysis (BIA, Nutriguard-M, Data Input GmbH, Germany). Each of above mentioned measurements was performed three times and mean was recorded.

Dietary assessment Each participant filled in a prospective questionnaire, where dietary data from 3 days were collected (2 working days, 1 weekend day). For dietary intake calculations Nutridan program was used. As nutritional data of some special vegan products are not available in the database, vegans were asked to collect packages from these products with producer declared nutritional content, and these data were used for calculations. Saccharides, lipid and protein intake were calculated separately. BCAA content of dietary proteins was estimated using nutritional database.

Physical activity assessment Physical activity was assessed in the whole sample using Baecke questionnaire for habitual physical activity [1], that correlates well with maximum oxygen consumption (VO2max)[2]. Maximal exercise test was performed in each subject on an electromagnetically braked bicycle ergometer (Ergoline 800, Bitz, Germany) to determine peak oxygen uptake (Vo2peak). An initial workload of 50W was increased by 25 W every minute continuously until fatigue despite the verbal encouragement. Oxygen uptake was measured using Vmax, Sensor Medics (Yorba Linda, CA). Heart rate was monitored continuously.

Laboratory analysis Peripheral venous blood was drawn from each subject after 12 hours of fasting. Parameters of glucose homeostasis were assessed in certified University hospital laboratory: plasma glucose using hexokinase reaction (kit KONELAB, Germany), HbA1c using high-pressure liquid boronate afinit chromatography (Primus corporation), insulin using solid phase competitive chemiluminescent enzyme immunoassay (Immulite 2000). Lipid profile: total cholesterol and triglycerides were measured using enzymatic method (kit KONELAB, Germany), HDL-cholesterol using PEG modified enzymatic measurement (kit ROCHE, Switzerland). Plasma levels of free fatty acids (FFA) were measured using method already described [12]. Briefly, FFA were extracted together with neutral lipids using isooctan and cleaned by reverse extraction. FA obtained this way were derivatized to methylesters and subsequently analyzed using gas chromatography (GC). Serum AA levels were determined using capillary electrophoresis (CE) with contactless conductivity detection, which has been already described in details [3, 4]. CE measurements were carried out using HP3DCE system (Agilent Technologies, Waldbronn, Germany) equipped with a built-in contactless conductivity detector.

Insulin sensitivity and secretion IS was assessed using 2-hours hyperinsulinaemic euglycaemic clamp, method described elsewhere[5]. Clamp was performed after 12-hours fasting in a standard insulin dose 1mIU/kg/min and an infusion of 15 % glucose solution was used. Mean infusion rate in steady-state of the clamp (6 consecutive measurements) was used for measurements. Glucose disposal was expressed as metabolic clearance rate (MCR, ml.kg-1.min-1) after correction for changes in glucose pool in extracellular fluid (space correction) and insulin sensitivity index (MCR divided by stady state insulinaemia, MCR/I, ml.kg-1.min-1/mU.l-1). Insulin secretion was assessed by IV arginine test as already described [6] performed on a different day (at about 7 days in between) from glucose clamp. Acute insulin response was calculated as the incremental trapezoidal area for insulin during the 30 min of the test.

Muscle biopsy, respiratory chain, and citrate synthase enzymes activities Skeletal muscle sample from vastus lateralis muscle was performed in every participant using a standard Bergström technique[7, 8]. Biopsy was performed at fasting conditions. About 200 mg of wet weight was obtained. The sample was immediately microdissected, weighed, divided for respective analyses and snap-frozen in liquid nitrogen and stored in -80oC until analyses.

Mitochondrial respiratory chain (RC) enzymatic activity and citrate synthase (CS) activity were measured spectrophotometrically in muscle homogenates. Muscle homogenates were prepared as described [9]. Briefly, about 50 mg of SM was cut into small fragments and homogenized (glass-glass grinder) in 20 volumes of ice-cold homogenization buffer (250 mM sucrose, 20 mM Tris, 40 mM KCl, 2 mM EGTA, protease inhibitor cocktail, pH 7.4). The homogenates were then centrifuged for 1 min at 600g at 4°C and the supernatants were immediately used for analysis. Protein concentration was measured in aliquots using the bicinchoninic acid (BCA) assay (Sigma). Activity of complex I-IV and CS was determined accordingly to previously published protocols [10-11], that were modified for measurement in a microplate reader (Infinite M200 PRO, Tecan). All measurements were made in tetraplicates. Activities of RC and CS were expressed as nmol/min/mg of total proteins, except for complex IV, which was expressed as Δ log (A550)/min/mg of total proteins. The enzymatic activities of the RC complexes were also normalized to the activity of CS, which is used as a marker of the abundance of mitochondria within a tissue.

Adipose tissue biopsy Subcutaneous abdominal adipose tissue (SAAT) biopsy was performed in every participant. Bergström needle was used to obtain samples of subcutaneous fat from paraumbilical area as already described [12]. Biopsy was performed at fasting conditions. About 300 mg of wet weight was obtained. The sample was immediately microdissected, weighed, divided for respective analyses and snap-frozen in liquid nitrogen and stored in -80oC until analyses.

Quantitative real time PCR (RT-qPCR) Total RNA from tissues was isolated using Lipid Tissue and Fibrous Tissue RNeasy Mini Kits (Qiagen). RNA concentration was measured by Nanodrop1000 (Thermo Fisher Scientific, Wilmington, USA). RNA was treated with DNAse I (Invitrogen, Carlsbad CA, USA) to remove any contaminating genomic DNA. cDNA was prepared from 200- 600 ng of RNA using High Capacity cDNA archive kit (Applied Biosystems, Carlsbad CA, USA). Equivalent of 5 ng of RNA was used for Real Time PCR reactions using Fast Advanced master mix and Gene expression assay (IRS1, GLUT4, BCKDHA, BCKDHB, ACOX, CPT1b, PLIN1, PLIN2, PLIN5, FASN, SCD1, DGAT2, PPAR-γ; Applied Biosystems). All samples were run in duplicates. Gene expression of target genes was normalized to expression of RPS13 (glucuronidase, beta) and fold change of expression was calculated using delta delta Ct method.

Fatty acid spectrum in AT Fatty acid spectrum in AT was assessed using gas chromatography, a method already described by Lepage and Roy [13] with modifications by Rodriguez-Palmero et al. [14] In brief, the method involves a chloroform/methanol extraction of freeze-dried AT to isolate lipids and subsequent transesterification or esterification of FA bound to lipids to form methyl esters, which were then analyzed using gas chromatography.

Statistical analysis Data are presented in text, tables, and figures as means ± SD with 95 % CI and p-values <0.05 were considered statistically significant. Student t-test was used in observational samples for normally distributed data. Wilcoxon matched pairs test was used when data were not normally distributed. Intervention samples were compared using a general linear model to test the statistical significance of differences between groups. A mixed model ANOVA was used to assess group × time interaction. Within each group time effects of the intervention were assessed using repeated measures ANOVA with Bonferroni's multiple comparison. Where there were only two data sets within each group (i.e. only baseline and intervention data as in data obtained from tissue biopsies) paired t-tests were used for normally distributed data and Wilcoxon test when data were not normally distributed. Pearson's correlation coefficient was calculated to express relationship between changes from baseline to intervention for normally distributed data and Spearman's correlation coefficient was calculated when data were not normally distributed. Statistica 9.0, StatSoft, Inc. USA was used to perform all statistical procedures.

Inclusion Criteria:

• healthy volunteers

Exclusion Criteria:

• age under 18 years, any chronic disease related to energy metabolism (particularly diabetes, thyreopathy, hypertension, dyslipidemia, atherosclerosis etc.), family history of type 2 diabetes (first degree relatives), any chronic medication hormonal contraception included, smoking (even in past medical history), regular alcohol consumption and BCAA supplementation