ARTICLE IN PRESS

Clinical Nutrition (2005) 24, 1019–1028

KEYWORDGuarana;Lipid metaExercise trMuscle;Carnitine

0261-5614/$ - sdoi:10.1016/j.c

�Correspondifax: +5511 3091

E-mail addr

http://intl.elsevierhealth.com/journals/clnu

ORIGINAL ARTICLE

Lipid metabolism in trained rats: Effect of guarana(Paullinia cupana Mart.) supplementation

Waldecir P. Limaa, Luiz C. Carnevali Jra, Robson Edera, Luis FernandoB.P. Costa Rosaa, Elfriede M. Bacchib, Marılia C.L. Seelaendera,�

aDepartment of Cell and Developmental Biology, Institute of Biomedical Sciences,Av. Prof. Lineu Prestes 1524, Sao Paulo, SP, CEP 05508-900, BrazilbDepartment of Pharmaceutical Sciences, University of Sao Paulo, Sao Paulo, Brazil

Received 14 April 2005; accepted 5 August 2005

S

bolism;aining;

ee front matter & 2005lnu.2005.08.004

ng author. Tel.: +5511 37402.ess: seelaend@icb.usp.

SummaryBackground and aims: Guarana is widely consumed by athletes, either insupplements or in soft drinks, under the belief that it presents ergogenic and ‘‘fatburning’’ effects. We examined the effect of guarana supplementation (14 days)upon aspects of lipid metabolism in sedentary (C) and trained rats (T).Methods: To isolate the effect of caffeine from that of other components ofguarana, we adopted two different doses of whole extract (G1—0.130 g/kg;G2—0.325 g/kg) or decaffeinated extract (DG1, DG2). Body weight, food and waterintake; muscle fat content, oleate incorporation, glycogen content, and carnitinepalmitoyltransferase I (CPT I) activity and mRNA expression; along with plasmalactate concentration, were assessed.Results: Muscle oleate incorporation was decreased in rats receiving decaffeinatedguarana in relation to G1 and G2; as was CPT I mRNA expression in thegastrocnemius. Whole extract supplementation, but not DG induced reduced plasmalactate concentration in trained rats. G1 showed higher muscle glycogen contentcompared with all other groups. The results show an effect of guarana on aspects oflipid metabolism, which is abolished by decaffeination.Conclusion: The changes in lipid metabolism of supplemented rats herein reportedare associated with the methylxanthine content of guarana.& 2005 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. Allrights reserved.

Elsevier Ltd and European So

0918039;

br (M.C.L. Seelaender).

Introduction

Guarana (Paullinia cupana Mart. var. sorbilis), asprawling shrub or woody vine, found in northernBrazil,1 is widely used as a flavouring for soft

ciety for Clinical Nutrition and Metabolism. All rights reserved.

ARTICLE IN PRESS

W.P. Lima et al.1020

drinks, and in the composition of a variety ofdietary supplements throughout the world. It isclaimed to have stimulant and ergogenic propertiesand to be of therapeutic value for a variety ofconditions, showing alleged antidiarrheic, diuretic,and antineuralgenic properties.2,3 Guarana has beenshown to have an antioxidant effect,3 to inhibitplatelet aggregation,4,5 and to have a gastroprotec-tive action.6 In Brazil, the use of guarana byathletes is widespread, with claims of performanceimprovement, and it has also been adopted as asupplement in the diet of racing horses.7 Guaranaand Ma Huang mixtures are commercialized asweight reducing supplements, and seem to effec-tively promote fat loss in overweight subjects,8

despite considerable adverse effects.9

The role of changes in lipid metabolism in theanedoctal weight loss induced by guarana con-sumption has never, to our knowledge, however,been previously investigated. The existing studiesassume that the effects of guarana equal those ofcaffeine, what is not necessarily true.

Guarana seeds contain caffeine (2.5–5%), theo-bromine and theophylline (small amounts), andtannins (up to 16%).10 Although the methylxantinecontent of guarana may explain many of the effectsattributed to the plant, some studies demonstratedthat treatment of rats with caffeine in similar dosesto those found in guarana fails to induce many of theresponses observed after guarana supplementation.2

It was thus suggested2,3 that tannins could play a partin these responses to guarana supplementation.

The aim of this study was to examine the effectsof guarana supplementation upon tissue lipidmetabolism in rats receiving different doses ofaqueous extract of guarana (GE), and compare theresults with those presented by animals whose dietwas supplemented with decaffeinated guaranaextract (DG), hence isolating the influence of themethylxantine content from that of other compo-nents present in the extract. Trained rats wereincluded in the study as to allow the comparison ofthe effects of exercise and supplementation uponlipid metabolism. An intermittent exercise trainingprotocol was chosen, since anecdotal evidenceshows frequent consumption of guarana is common(67%) among male teenager (14–18 years-old)volleyball and other court sports players in Brazil.

Materials and methods

Materials

Solvents, buffer reagents, and Tween 20 werepurchased from LABSYNTH (Brazil); palmitoyl CoA,

carnitine, albumin, lipid standards, and solvents forHPLC and mass spectrometry from the Sigma Che-mical Co., USA. 14C-oleate and 3H-carnitine werepurchased from Amersham, UK.

Guarana powder (batch GUAR04/01, P. cupanaHBK, Sapindacea) was the kind gift from SantosFlora Ervas Medicinais Ltd. (Brazil).

Obtainment and analysis of guarana extracts

One litre of an ethanol:water solution (6.6:3.4 v/v)was added to each 2 kg of guarana crude powderand then, percolated with 6 l of the same solution,as in Prista et al.11 After 7 days the percolate wasevaporated and the resulting extract lyophilized,yielding 327 g of dry extract. The extract wassubmitted to HPLC analysis (Shimadzu), followingthe method of Salvadori et al.7 that determined acaffeine content of 0.153 g/g of extract.

The decaffeinated extract was obtained afterchloroform extraction,12 which eliminated all themethylxantines present in the extract. Tannins andcatechins were not removed with this process beingtherefore present in the DG extract. Each 5 g of dryGE was mixed with 40.0ml of chloroform and 5ml10% ammonium hydroxide. After vigorous agitation,followed by a decanting period of 15min, themixture was filtered with cotton pads. The finalfiltrate was mixed with 15ml distilled water and0.5ml of a 10% aqueous solution of sulphuric acid.The mixture was heated for 2min, filtered andbrought to alkaline pH with 10% ammoniumsulphate. Caffeine was extracted three times withchloroform. The obtained decaffeinated extractwas submitted to HPLC analysis to ensure theeffectiveness of the process.

Animals and supplementation

Male adult Wistar rats (160–250 g), obtained fromthe Institute of Biomedical Sciences, University ofSao Paulo, were maintained under a 12 h light/12 hdark cycle (lights on at 7:00 a.m.), and controlledtemperature conditions (2371 1C), receiving waterand food (commercial chow, Nuvilabs, Nuvital,Brazil) ad libitum. Weight gain and food intakewere assessed daily by measuring the amount ofchow and the volume of water consumed by eachanimal, kept in a metabolic cage. The animals werekilled by decapitation, between the interval of 8:00and 11:00 a.m. The Biomedical Sciences Institute/USP Ethical Committee for Animal Research appro-ved all the adopted procedures, which were carriedout in accordance with the ethical principles statedby the Brazilian College of Animal Experimentation.

ARTICLE IN PRESS

Guarana and lipid metabolism 1021

The doses of supplemented guarana were calcu-lated based on the caffeine content of the extract.Guarana supplemmentation was carried out for 14days. The following scheme was adopted (Table 1).

A lower dose was also adopted (0.065 g/kg ofguarana, corresponding to 0.010 g/kg of caffeine),but found in preliminary studies (data not shown)to be insufficient to promote any significantchanges in the studied parameters. The trainedrats were supplemented with 0.130 g/kg bodyweight, after the results with sedentary rats wereobtained, showing that this dose would be enoughto promote changes in the investigated para-meters, without presenting deleterious side effects(diarrhoea).

Training protocol

All non-sedentary animals were submitted to anadaptation period of 6 days prior to the beginning ofthe training protocol. Training consisted of swim-ming in individual PVC tanks (100� 60� 60 cm)filled with circulating water at 31.071.0 1C.13

The training program was the following (Table 2).

Table 1 Experimental design.

Group GE (g/kg bodyweight)

SedentaryC 0G1 0.130G2 0.325DG1 0DG2 0

TrainedT 0TG1 0.130TDG1 0

Table 2 Training program.

Period Number ofbouts

Activityduration (m

1st day (adaptation) 5 12nd day (adaptation) 5 13rd day (adaptation) 5 14th day (adaptation) 8 15th day (adaptation) 8 16th day (adaptation) 8 12nd–5th week (training) 10 1

During the training period exercise intensity wasequivalent to 100% VO2 max, as the lactatethreshold is attained with an extra load of 5–6%of body weight.14 Body weight was assessed daily asto ensure proper calculation of the extra load.During the resting times the rats were kept outsidethe water.

Measurement of fat content and 14C-oleateincorporation into the skeletal muscle

Animals received 0.5ml of [14C]-triolein (approxi-mately 2.5 mCi) intragastrically, as in Oller doNascimento and Williamson15 After 5 h the soleus,the gastrocnemius and the digestive tract wereremoved and submitted to the method described inStansbie et al.16: duplicate samples were digestedwith 30% KOH (w/v) for 15min, and then, afterabsolute ethanol was added, incubated for 2 h at70 1C. The free fatty acids from the saponifiablelipid fraction were extracted three times withpetroleum ether and, after evaporation; the massof the lipid present in the sample was assessed.Scintillation fluid was added to samples, whoseradioactivity was determined in a scintillation

Caffeine content(g/kg body weight)

DG (g/kg bodyweight)

0 00.020 00.050 00 0.1300 0.325

0 00.020 00 0.130

in)Rest duration(min)

Load (% bodyweight)

Total time(min)

1 3 101 5 101 7 101 7 161 10 161 10 161 10 20

ARTICLE IN PRESS

W.P. Lima et al.1022

counter (Packard, TRICARB 2100). To assess thetotal radioactivity in the tissue, 0.5ml of NaOH(1N) was added to 300mg of either liver ordigestive tract, and the samples incubated for30min at 70 1C. An aliquot of 100 ml was transferredto a vial containing scintillation fluid, 100 ml of HCland some drops of hydrogen peroxide (130 V). Thetotal radioactivity present in the tissue wasmeasured to provide a control value for comparisonwith the amount of radiolabelled oleate incorpo-rated in the form of lipid.

Measurements of carnitinepalmitoyltransferases I and II activities

To isolate the mitochondria17 the muscles wereminced with scissors and homogenized manually inisolation buffer (mannitol 220mM, sucrose 70mM,Hepes 2mM, EDTA 0.1mM, pH 7.4). The homo-genate was filtered and centrifuged twice at1000 rpm (12min). The supernatant was thencentrifuged twice at 8500 rpm for 15min. Theisolated mitochondria were suspended in a bufferconsisting of 0.15mM KCl and 5mM Tris–HCl (pH7.5), centrifuged (10,000g, 15min), resuspended in10mM phosphate buffer (pH 7.5), frozen in liquidnitrogen and thawed. Samples were then ultracen-trifuged (100,000g, 1 h—Hitachi). The resultingpellet was suspended in phosphate buffer to whichTween 20 (1% w/v) had been added, and stirred onice for 30min, in order to separate CPT I(membrane bound) from CPT II.18 Another ultra-centrifugation followed, after which the fractionscontaining CPT I (pellet) and CPT II (supernatant)were obtained.

CPT activity was measured with the methodof Bremer,19 which was modified by Seelaenderet al.20 Assay medium consisted of 60mM KCl,40mM mannitol, 20mM Hepes, 0.15mM EGTA,1.5mM KCN, defatted bovine serum albumin(0.5%), 42 mM palmitoyl CoA, 0.35mM carnitine(0.6 Ci 3H-methylcarnitine) and approximately0.03mg of the isolated enzyme fraction or distilledwater (blanks). The final volume of the assaymixture was 0.5ml, and the pH, 7.3. The assaywas stopped by addition of 1.5ml of 7% perchloricacid, and the acylcarnitine formed was extractedwith n-butanol, as described previously.21 CPTactivity is expressed as nmol/min per mg of proteinin the isolated enzyme fraction.

Anaysis of gene expression

Total RNA was obtained from aliquots of 100mg ofthe muscle of the animals after TRIZOLs reagent

extraction, accordingly to Chomzynski and Sac-chi.22 RNA concentration was determined spectro-photometrically (Beckman DU 640, Fullerton, CA,USA).

A semi-quantitative reverse transcriptase–poly-merase chain reaction method was used for theestimation of the concentration of CPT I and CPT IImRNA. A 33 ml assay mix containing 3 mg RNA, 10units of placental RNAse inhibitor, 2 ml oligo(dt),2 ml dNTP (10 nmol), 2 ml DTT, 10 units of Moloney-murine leukaemia virus reverse transcriptase (In-vitrogen, USA), and 4 ml 10� reaction buffer(100mM Tris–HCl, pH 8.3, 500mM KCl, 150mMMgCl2 in nuclease free water) was used to producecDNA. The RT-mixture was incubated at 80 1C for3min, followed by 21 1C for 10min, 42 1C for 30minand then 99 1C for 10min 2 ml of the productobtained were fractionated in 1% agarose andethidium bromide gel to assess the quality of thereaction. The primers were designed after thepublished Genebank sequences: CPT I [sense:CAAGGCCCTGGCTGATGATGTG; antisense: AGTCTCTGTCCGCCCCTCTCG]23; CPT II [sense: GATAAGCA-GAATAAGCACACC; antisense: GGAGGAACAAAGCGAATGAGT].24

Amplification of cDNA was carried out in 30 cyclesof 35 s at 94 1C, 60 s at 70 1C, and 60 s at 72 1C. TheRPL19 gene was used as the internal control. Each3 ml PCR mixture contained 40, 8 or 4 ng (threedifferent cDNA dilutions were used) of cDNA, 0.5units AmpliTaq Gold Polymerase (Perkin Elmer,Foster City, CA, USA), 2.5 nmol each dNTP, and1.0 mM of the primers in reaction buffer (10mMTris–HCl, pH 9.0, 50mM KCl, 0.1% Triton X-100,1.5mM MgCl2). Five microlitres of the PCR mixturewere fractionated in polyacrilamide gel (agarose1.2% and ethidium bromide. Semi-quantitativeanalysis was performed after image obtainmentwith the Typhoons (Molecular Dynamics). Imageanalysis was carried out with the program ImageQuaNT TM.25,26

Sample protein content was assessed with themethod of Lowry et al.27

Lactate concentration was measured (lactateanalyzer 1500 YSI Sport) in the plasma obtainedon the last day of training (at rest, and after the 5thand the last bouts of exercise).14

Glycogen content of the gastrocnemius wasevaluated by a histological method: After samplefixation (4% paraphormaldehyde) for 18 h, thespecimens were kept (24 h) in 70% ethanol, andthen immersed into 96% and 100% ethanol solu-tions, xylol and Paraplasts. The samples were keptunder 60 1C for 5 h. After rehydration they wereincubated with a solution containing perchloricacid, Schiff reagent, and sodium sulphide. Sections

ARTICLE IN PRESS

Guarana and lipid metabolism 1023

were examined with a Nikon Eclipse E600 micro-scope. Images were captured using a digital camera(cool Snap—PROcf color), and image Pro Plussoftware (Media Cybernetics, Silver Spring, MD)was applied for the morphometric analysis.

Results are expressed as mean+standard error ofmean. Statistical analysis was performed with one-way ANOVA, followed by Tukey0s post-test. Signifi-cance level was set at, at least, Po0:05.

Results

Guarana supplementation with the higher dosestudied caused, after 14 days, the decrease of thetotal intake of food (Table 3), which was not foundwhen the decaffeinated extract was supplemented.Although the total weight gain in the same periodwas not significantly changed by treatment withwhole GE, both doses of DG induced lower weightgain in the period as compared with controls,despite the lack of difference in food consumption.Training reduced the weight gain rate of all groups.Water consumption was unaffected by the supple-mentation and training protocols.

The relative weight (percentage of total bodyweight represented by the tissue) and the neutrallipid content of both soleus and gastrocnemiusmuscles was not different among the studiedgroups (Table 4). Animals receiving the higher doseof guarana showed, nevertheless, decreased car-cass fat content (9.1470.40% of total tissueweight, n ¼ 7), as compared with controls (10.77+0.2%, n ¼ 7, Po0:05) and with G1 (12.5+0.36%,n ¼ 6, Po0:05). Exogenous oleate incorporation

Table 3 Weight, total food and water variations after t

Group Weight variation after14 days (g)

Food c(g/14d

C 42.0873.47 360.00G1 35.0073.52 381.00G2 37.0872.98 329.50DG1 29.1073.29 361.66DG2 27.2273.61** 357.00T 20.1371.66*,a,d 391.66TG1 17.1371.71*,b,e 418.00TDG1 11.5071.55*,c,f,g,h 376.50

C, control sedentary, G1, sedentary rats supplemented with 0.130guarana extract; G2, sedentary rats supplemented with 0.325 g/guarana extract; T, trained control rats; TG1, trained rats supple(TDG1) decaffeinated guarana extract.Results are mean7s.e.m. of 12 (sedentary) or 8 (trained) animalaPo0:05 in relation to G1; bPo0:01 in relation to G1; cPo0:001 into G2; fPo0:001 in relation to G2; gPo0:01 in relation to DG1; hP

was decreased in the soleus and gastrocnemiusmuscles of DG1 and DG2 in relation to G1 and G2.This parameter was not affected by the trainingprotocol adopted.

Maximal long-chain fatty acid mitochondrialtransport capacity, measured as the maximalactivity of the CPT I system was not changed dueto supplementation or training with the intermit-tent exercise protocol (Table 5), nor was the mRNAexpression for CPT I in the muscles of rats receivingthe different guarana treatments (Table 6). How-ever, chronic consumption of DG led to decreasedCPTI mRNA expression in the gastrocnemius oftrained rats (Table 7).

When plasma lactate was measured (Table 8) atrest, after 5 or after 10 bouts, a significantly lowerconcentration was found for rats receiving guaranasupplementation as compared with controls, afterthe 5th bout. This difference did not persist untilthe 10th bout, as the result for TG1 was notdifferent from T. The decaffeinated extract did notinduce a different response compared with con-trols. Figure 1A–C show the Periodic Acid of Schiff(PAS) reaction in the gastrocnemius of C, G1 andDG1. A more intense reaction was obtained for G1,suggesting higher glycogen content in this group.

Discussion

Guarana (Paullinia cupana) consumption is increas-ing in the world, as it takes part in the compositionof many commercial dietary supplements,28 as aweight loss-promoting adjuvant.8 In Brazil, GE hasbeen also used as a stimulant and in the therapeu-tics of depression, fatigue, and migraine.29

he 14 days of supplementation.

onsumptionays)

Water consumption(ml/14days)

74.39 583.00731.37713.29 476.00722.67713.30** 473.00724.32715.91 505.00719.09712.72 497.00726.87718.42 542.00732.60720.69 e 528.00738.96714.75 507.00732.54

g/kg body weight of guarana extract; or (DG1) decaffeinatedkg body weight of guarana extract; or (DG2) decaffeinatedmented with 0.130 g/kg body weight of guarana extract; or

s. *Po0:001 in relation to C; **Po0:05 in comparison with C;relation to G1; dPo0:05 in relation to G2; ePo0:01 in relationo0:01 in relation to DG1.

ARTICLE IN PRESS

Table 5 Maximal activity of carnitine palmitoyltransferase I (CPT I) and II (CPT II).

CPT I CPT II

GAS SOL GAS SOL

C 3.0670.29 3.3970.30 2.1470.20 1.3370.12G1 3.3470.35 2.9170.36 2.0170.39 1.4870.20G2 3.1070.23 4.4170.51a 2.4370.30 1.4670.18DG1 3.0370.29 3.4570.38 2.1770.45 1.1970.17DG2 2.6870.33 2.6370.15* 1.8970.10 0.9370.05T 3.2170.21 3.2470.23 1.9770.32 1.0970.14TG1 3.3170.44 2.9570.45 2.4870.37 0.9870.09TDG1 2.3970.41 3.0370.39 2.0370.10 1.2770.19

nmol/min/mg of protein in the isolated enzyme fraction.C, control sedentary, G1, sedentary rats supplemented with 0.130 g/kg body weight of guarana extract; or (DG1) decaffeinatedguarana extract; G2, sedentary rats supplemented with 0.325 g/kg body weight of guarana extract; or (DG2) decaffeinatedguarana extract; T, trained control rats; TG1, trained rats supplemented with 0.130 g/kg body weight of guarana extract; or(TDG1) decaffeinated guarana extract.Results are mean7s.e.m. of 12 (sedentary) or 8 (trained) animals. aPo0:05 in relation to G1; *Po0:01 in comparison with G2.

Table 4 Relative weight, fat content and oleate incorporation of gastreocnemius (GAS) and soleus (SOL)muscles.

Weight (% body weight) Fat content (mg fat/g tissue) Incorporation (%)

GAS SOL GAS SOL GAS SOL

C 0.58570.072 0.04070.004 16.1370.53 28.1272.41 0.23270.032 0.18170.020G1 0.56470.042 0.03970.004 15.1970.71 27.1272.98 0.39170.042 0.37370.070G2 0.60970.068 0.04170.005 15.6971.06 29.1271.75 0.29670.030 0.28970.041DG1 0.64070.077 0.04170.005 17.2270.75 30.1273.48 0.20570.020* 0.17670.044*DG2 0.59370.082 0.03970.006 18.5971.22 29.1271.75 0.15370.029** 0.14270.029**T 0.71370.098 0.04470.006 15.1570.80 30.1272.93 0.29570.046 0.27870.064TG1 0.69270.083 0.04270.004 17.1070.48 31.4573.02 0.34570.067 0.29370.037TDG1 0.66370.079 0.04170.005 16.5871.08 30.2374.13 0.30270.062 0.26970.058

C, control sedentary, G1, sedentary rats supplemented with 0.130 g/kg body weight of guarana extract; or (DG1) decaffeinatedguarana extract; G2, sedentary rats supplemented with 0.325 g/kg body weight of guarana extract; or (DG2) decaffeinatedguarana extract; T, trained control rats; TG1, trained rats supplemented with 0.130 g/kg body weight of guarana extract; or(TDG1) decaffeinated guarana extract.Results are mean7s.e.m. of 12 (body weight), 15 (fat content) sedentary animals; or 8 (body weight and fat content) for trainedrats. For the oleate incorporation experiments 6 animals from each group were used. *Po0:05 in relation to G1; **Po0:01 inrelation to G2.

Table 6 mRNA expression for CPT I in the gastrocnemius (gastro) and soleus muscles of sedentary groups.

C G1 G2 DG1 DG2

Gastro 0.90570.116 0.92970.245 1.19270.276 0.97070.116 1.09270.381Soleus 1.25270.136 1.33870.300 1.01970.103 1.50570.389 1.47370.156

Results are mean7s.e.m. of the ratio CPT I/RPL 19 of expression for 7 (gastrocnemius) or 3 (soleus) animals.C, control sedentary, G1, sedentary rats supplemented with 0.130 g/kg body weight of guarana extract; or (DG1) decaffeinatedguarana extract; G2, sedentary rats supplemented with 0.325 g/kg body weight of guarana extract; or (DG2) decaffeinatedguarana extract.

W.P. Lima et al.1024

Although caffeine is considered to be the activecomponent of guarana, there is evidence that itsother components may also be involved in the

response to supplementation.2,3 Although there isextensive knowledge on the effects of caffeineupon lipid metabolism, to our knowledge no study

ARTICLE IN PRESS

Table 7 mRNA expression for CPT I in the gas-trocnemius and soleus muscles of trained groups.

TG1 TDG1

Gastrocnemius 0.76770.076 0.59070.041*Soleus 1.12070.093 1.06870.235

Results are mean7s.e.m. of the ratio CPT I/RPL 19 ofexpression for 7 (gastrocnemius) or 4 (soleus) animals.TG1, trained rats supplemented with 0.130 g/kg bodyweight of guarana; TDG1, trained rats supplemented with0.130 g/kg body weight of decaffeinated guarana.*Po0:005 for comparison with TG1.

Table 8 Plasma lactate of trained rats (mmol/l),supplemented with 0.130 g/kg body weight ofguarana (TG1) or decaffeinated guarana (TDG1),and control (T).

Rest 5th bout 10th bout

T 1.8470.26 6.1470.48* 9.8470.68*,**TG1 1.7970.23 3.6870.59a 9.0270.89*,**TDG1 1.5270.24 7.8470.54*,b 11.8770.60*,**

Results are mean7s.e.m. of the plasma samples obtainedfrom 9 (T) or 8 (TG1) and 6 (TDG1) rats.*Po0:001 in relation to rest; **Po0:001 in relation to the5th bout; aPo0:01 in relation to T; bPo0:001 in relationto TG1.

Figure 1 PAS reaction in the gastrocnemius of animaltreated with guarana extract (G1—A), decaffeinatedguarana extract (DG1—B) and controls (C—C).

Guarana and lipid metabolism 1025

has addressed so far the effect of guaranasupplementation, except in relation to lipid perox-idation.3 We have thus examined aspects of lipidmetabolism in sedentary and trained rats, sub-mitted to guarana or DG supplementation, in orderto isolate the effects of caffeine from those of theother components.

Food intake was found to be decreased in thegroup receiving the higher dose of guarana/caffeine (0.325/0.05 g/kg, respectively). Racottaet al.30 also reported diminished food consumptionafter caffeine injection in rats, although othersfound no differences in this parameter31 betweencaffeine treated animals and controls, or still, anincrease.32 The discrepancies among studies maybe related to the different doses of caffeine andmodels adopted. Andersen and Fogh33 showed thatin overweight human subjects gastric emptying wasdelayed by the consumption of a herbal preparationconsisting of a mixture of guarana, yerbe mate anddamiana, consequently modifying food energyintake and inducing weight loss. Guarana, whencombined with Ma Huang, has been shown to induceweight loss in overweight men, women and

adolescents, as reviewed by Carlini.10 In the samereview the author stresses the anorectic propertiesof guarana, and emphasizes the role of constituentsof the extract, other than caffeine, upon many ofthe reported physiological alterations induced bysupplementation. Indeed, the final weight gain ofthe animals consuming 0.130 g/kg DG was presently

ARTICLE IN PRESS

W.P. Lima et al.1026

found to be decreased, as compared with controlsand animals submitted to whole guarana supple-mentation. Theophylline, known to be present inGE, a potent adenosine receptor antagonist, hasbeen linked to increased lipolysis.34,35 In thepresent study carcass fat content was also reducedas a response to guarana supplementation, aneffect which may be attributed to both caffeineand theophylline. However, the fact that the finalweight gain of animals consuming the decaffei-nated extract was lower than that of the groupreceiving whole guarana suggests that tannins andcatechins may also contribute to the result,especially after the removal of methylxanthines.The tannin content of GE may reach 8.5%,33 andthe consumption of tannins has been associated36

with decreased body weight in various animalspecies.

As the use of mixtures containing guarana isincreasing among athletes it was also our aim toexamine the effect of the supplementation uponmuscle lipid metabolism in sedentary and trainedrats. Training caused the weight of the animals tobe reduced in all groups compared with thesedentary counterparts, but guarana supplementa-tion (TG1 and TDG1) did not allow further significantdecreases of this parameter. Interestingly, never-theless, a trend towards reduced weight gain inTDG1 was detected, similarly to what was observedfor sedentary animals. Although the fat content ofthe soleus and gastrocnemius muscles was un-changed by either guarana consumption or training,oleate incorporation was decreased in both musclesfrom the animals receiving the decaffeinatedextract. This method measures the amount ofexogenous oleate that is found in tissues 5 h afterthe administration of an intragastric bolus of long-chain fatty acid-rich triacylglycerol.15 Therefore,the results with the sedentary, DG consuminggroups reflect either a decreased capacity of fattyacid uptake by the muscle or still, enhancedoxidation of this substrate associated with reducedreesterification, in comparison with G1 and G2,which showed a trend towards increased oleateincorporation. Mittal et al.37 reported decreasedtissue fat level in mice fed proanthocyanidins,attributing this effect to possible changes inlipolysis or lipogenesis. Our results suggest that adiminished capacity for fatty acid uptake from theplasma may be involved in one such effect. The factthat trained rats did not show the same decrease inoleate incorporation is very possibly related to theincrease in intrafibrilar triacylglycerol contentpromoted by training reported by our group in aprevious study,38 when we found an importantcontribution of very low density lipoproteins

(VLDL)-derived triacylglycerol to the replenishmentof these stores.

The major regulatory step in the oxidation oflong-chain fatty acids is the entry of this substrateinto mitochondria, whose rate is a reflex of theactivity of CPT I39 This enzyme catalyzes theformation of long acyl–carnitine complexes, which,according to the most recent models40 enter theintermembrane space through a porine, being thentransported by carnitine–acylcarnitine translocase(CACT) to the matrix compartment in which theyare exposed to the catalytic action of carnitinepalmitoyltransferase II (CPT II), regenerating carni-tine and acyl CoA. Whole GE supplementation (G2)induced an increase of the activity of CPT I in thesoleus, but not in DG2, suggesting caffeine takespart in the observed response. The literatureprovides no evidence of a direct effect of caffeineon the activity of this enzyme. However, the effectsof caffeine upon lipolysis, and consequently, onsubstrate availability, could account, at leastpartially, for the increased activity observed, as itis also associated with an increase in resting oxygenconsumption in humans.9 The other methylxanthinesin the extract could also interfere with CPT activity,as Alhomida41 demonstrated an increase in ratheart carnitine palmitoyltransferase activity aftertheophylline administration; while Greer et al.42

reported improved performance in the cycle ergo-meter after theophylline supplementation.

Measuring CPT activity in an isolated systempresents the advantage of abolishing the influenceof alterations in malonyl CoA concentration and inthe sensitivity to malonyl CoA-induced inhibition. Itis thus possible to discard the influence of theseaspects in the present results of activity, whichregard the catalytic capacity of CPT I, but not itsconcentration. On the other hand, CPT I mRNAexpression was not modified by supplementation insedentary rats, a parameter that, if differentamong the studied groups, could point out toeventual differences in the concentration of theenzyme (although a semi-quantitative analysis wascarried out). These results suggest that theincrease in the catalytic capacity herein reportedcould be linked to post-translational modificationof CPT I.

Although CPT I activity was not altered byintermittent exercise training, nor by the combina-tion of training and the supplementation protocols,CPT I mRNA expression was reduced in the gastro-cnemius of trained rats receiving DG in comparisonwith TG1. These results suggest that, while in thesoleus caffeine is able to induce changes in thecatalytic capacity of CPT I, in the gastrocnemiusthe regulation would rest upon the control of

ARTICLE IN PRESS

Guarana and lipid metabolism 1027

enzyme expression. This latter effect would be onlyapparent in trained animals. A possibility whichcannot be discarded, nevertheless, is that the othercomponents of guarana play a part in the discre-pancy of CPT I mRNA expression found betweenTDG1 and TG1. We have not been able to findstudies in the literature concerning the effect oftannins upon CPT activity.

In order to investigate the eventual occurrenceof a glycogen-sparing effect in the gastrocnemiusinduced by the treatment with guarana, weexamined histological sections after PAS reaction.The stronger positivity of the reaction obtained forG1 suggests an influence of the caffeine content ofguarana upon this parameter. This hypothesis iscorroborated by the fact that plasma lactate wasreduced in TG1 as compared with C and TDG1, afterthe 5th bout of exercise. Greer et al.42 found noeffect of either caffeine of theophylline ingestionin exercising subjects regarding muscle glycogencontent and plasma lactate concentration. Thepresent muscle glycogen results were obtained 24 hafter the last exercise bout, while Greer et al.42

collected the biopsies immediately after theexercise session.

Taken together the results show that guaranaconsumption is able to induce changes in lipidmetabolism, but the predominant element inducingthe alterations reported seems to be the methyl-xanthine content of the extract.

References

1. Henman AR. Guarana (Paullinia cupana): Ecological andsocial perspectives on an economic plant of the CentralAmazon basin. J Ethnopharmacol 1982;6:311–38.

2. Espınola EB, Dias RF, Mattei R, Carlini EA, Barros SBM.Pharmacological activity of Guarana (Paullinia cupanaMart.) in laboratory animals. J Ethnopharmacol 1997;55:223–9.

3. Mattei R, Dias RF, Espınola EB, Carlini EA, Barros SBM.Guarana (Paullinia cupana): toxic behavioral effects inlaboratory animals and antioxidant activity in vitro.J Ethnopharmacol 1998;60:111–6.

4. Bydlowski SP, Yunker RL, Subbiah MTR. A novel property ofan aqueous guarana extract of (Paullinia cupana): inhibitionof platelet aggregation in vitro and in vivo. Braz J Med BiolRes 1988;21:535–8.

5. Bydlowski SP, D’Amico EA, Chamone DA. An aqueous extractof guarana (Paullinia cupana) decreases platelet thrombox-ane synthesis. Braz J Med Biol Res 1991;24:421–4.

6. Campos AR, Barros AIS, Santos FA, Rao VSN. Guarana(Paullinia cupana Mart.) offers protection against gastriclesions induced by ethanol and indomethacin in rats.Phytother Res 2003;17:1199–202.

7. Salvadori MC, Rieser EM, Ribeiro Neto LM, Nascimento E.Determination of xanthines by high-performance liquidchromatography in horse urine after ingestion of guaranapowder. Analyst 1994;119:2071–3.

8. Boozer CN, Nasser JA, Heymsfield SB, Wang V, Chen G,Solomon JL. An herbal supplement containing Ma Huang-Guarana for weight loss: a randomized double-blind trial. IntJ Obes 2001;25:316–24.

9. Greenway FL. The safety and efficacy of pharmaceutical andherbal caffeine and ephedrine use as a weight loss agent.Obes Rev 2001;2:199–211.

10. Carlini EA. Plants and the central nervous system. Pharma-col Biochem Behav 2003;75:501–12.

11. Prista LN, Alves AC, Morgado RMR. Tecnica Farmaceutica eFarmacia Galenica, 3th ed. Lisboa: Fundac- ao ColousteGulbenkian; 1981.

12. Evans WC. Trease and Evan0s pharmacognosy, 14th ed.London: Saunders; 1996. p. 343–4.

13. Harri M, Kuusela P. Is swimming exercise or cold exposure forrats? Acta Physiol Scand 1986;126:189–97.

14. Gobatto CA, de Mello MA, Sibuya CY, de Azevedo JR, dosSantos LA, Kokubun E. Maximal lactate steady state in ratssubmitted to swimming exercise. Comp Biochem Physiol AMol Integr Physiol 2001;130:21–7.

15. Oller do Nascimento CM, Williamson DH. Evidence forconservation of dietary lipid in the rat during lactation andthe immediate period after removal of the litter. Biochem J1986;239:233–6.

16. Stansbie D, Brownsey RW, Cretazz M, Denton RM. Acuteeffects in vivo of anti-insulin serum on rates of fatty acidsynthesis and activities of acetyl-coenzime A carboxylaseand pyruvate dehydrogenase in liver and epididymal adiposetissue of fed rats. Biochem J 1976;160:413–6.

17. Curi R, Newsholme P, Newsholme EA. Metabolism ofpyruvate isolated rat mesenteric lymphocytes, lymphocytemitochondria and isolated mouse macrophage. Biochem J1988;250:383–93.

18. Woeltje KF, Kuwagima M, Foster DW, McGarry JD. Char-acterization of the mitochondrial carnitine palmitoyltrans-ferase enzyme system. Ll. Use of detergents and antibodies.J Biol Chem 1987;262:9822–7.

19. Bremer J. The effect of fasting on the activity of liverpalmitoyltransferase and its inhibition by malonyl-CoA.Biochem Biophys Acta 1981;665:628–31.

20. Seelaender MCL, Curi R, Colquhoun A, Williams JF, ZammittVA. Carnitine palmitoyltransferase II activity is decreased inliver mitochondria of cachectic rats bearing the walker 256carcinosarcoma: effect of indomethacin treatment. Bio-chem Mol Biol Int 1998;44:185–93.

21. Bremer J, Norum KR. Palmityl-CoA: carnitine O-palmityl-transferase in the mitochondrial oxidation of palmityl-CoA.Eur J Biochem 1967;1:427–33.

22. Chomczynski P, Sacchi N. Single-step method of RNAisolation by acid guanidinium thiocyanate–phenol–chloro-form extraction. Anal Biochem 1987;162:156–9.

23. Hildebrandt AL, Darrell P. Exercise attenuates the fasting-induced transcriptional activation of metabolic genes in skeletalmuscle. Am J Physiol Endocrin Metab 2000;278:E1078–86.

24. Woeltje KF, Esser V, Weis BC, et al. Inter-tissue and inter-species characteristics of the mitochondrial carnitine palmi-toyltransferase enzime system. J Biol Chem 1990;265:10714–9.

25. Davies B, Fried M. The L19 ribosomal protein gene (RPL19):gene organization, chromosomal mapping, and novel pro-moter region. Genomics 1995;25:372–80.

26. Altschul SF, Madden TL, Schaffer AA, et al. Gapped BLASTand PSI-BLAST: a new generation of protein database searchprograms. Nucl Acids Res 1997;25:3389–402.

27. Lowry OH, Rosenborough NJ, Farr AL, Randall RJ. Proteinmeasurements with the folin phenol reagent. J Biol Chem1951;193:265–9.

ARTICLE IN PRESS

W.P. Lima et al.1028

28. Haller CA, Duan M, Benowitz NL, Jacob P. Concentrations ofephedra alkaloids and caffeine in commercial dietarysupplements. J Anal Toxicol 2004;28:145–51.

29. Caldas ED, Machado LL. Cadmium, mercury and lead inmedicinal herbs in Brazil. Food Chem Toxicol 2004;42:599–603.

30. Racotta IS, Leblanc J, Richard D. The effect of caffeine onfood intake in rats: involvement of corticotropin-releasingfactor and the sympatho-adrenal system. Pharmacol Bio-chem Behav 1994;48(4):887–92.

31. Cox RH, Maickel RP. Interactions of caffeine with variousamphetamines on rat food consumption and avoidanceresponding. Neuropharmacology 1976;15:767–71.

32. Casas M, Prat G, Robledo P, Barbanoj M, Kulisevsky J, Jane F.Methylxanthines reverse the adipsic and aphagic syndromeinduced by bilateral 6-hydroxydopamine lesion of thenigrostriatal pathway in rats. Pharmacol Biochem Behav2000;66:257–63.

33. Andersen T, Fogh J. Weight loss and delayed gastricemptying following a South American herbal preparation inoverweight patients. J Hum Nutr Diet 2001;14:243–50.

34. Peters EJ, Klien S, Wolfe RR. Effect of short-term fasting onthe lipolytic response to theophylline. Am J Physiol1991;261:E500–4.

35. Raguso CA, Coggan AR, Sidossis LS, Gastaldelli A, Wolfe RR.Effect of theophylline on substrate metabolism duringexercise. Metaboilism 1996;45:1153–60.

36. Awika JM, Rooney LW. Sorghum phytochemicals and theirpotential impact on human health. Phytochemistry 2004;65:1199–221.

37. Mittal A, Elmets CA, Katiyar SK. Dietary feeding ofproanthocyanidins from grape seeds prevents photocarcino-genesis in SKH-1 hairless mice: relationship to decreased fatand lipid peroxidation. Carcinogenesis 2003;24(8):1379–88.

38. Belmonte MA, Aoki MS, Tavares FL, Seelaender MCL. Ratmyocellular and perimysial intramuscular triacylglycerol:a histological approach. Med Sci Sports Exerc 2004;36:60–7.

39. Ramsay RR, Gandour RD, Van Der Leij FR. Molecularenzymology of carnitine transfer and transport. BiochimBiophys Acta 2001;1546:21–43.

40. Kerner J, Hoppel C. Fatty acid import into mitochondria.Biochim Biophys Acta 2000;1486:1–17.

41. Alhomida AS. Theophylline-induced changes in the activityof carnitine palmitoyltransferase in rat cardiac tissues.Toxicology 2000;145:185–93.

42. Greer F, Friars D, Graham TE. Comparison of caffeine andtheophylline ingestion: exercise metabolism and endurance.J Appl Physiol 2000;89:1837–44.