Visceral Fat Accumulation During Lipid Overfeeding Is
Related to Subcutaneous Adipose Tissue
Characteristics in Healthy Men
Maud Alligier, Laure Gabert, Emmanuelle Meugnier,
Ste´phanie Lambert-Porcheron, Emilie Chanseaume, Frank Pilleul, Cyrille Debard, Vale´rie Sauvinet, Be´atrice Morio, Antonio Vidal-Puig, Hubert Vidal,*
and Martine Laville*
Unite´ Mixte de Recherche 1060 (M.A., E.M., C.D., H.V., M.L.), Institut National de la Sante´ et de la Recherche Me´dicale Laboratoires CarMeN et CENS, Universite´ Lyon 1, 69921 Oullins, France; Centre de Recherche en Nutrition Humaine Rhoˆne-Alpes (M.A., L.G., S.L.-P., V.S., H.V., M.L.), Centre Hospitalier Lyon-Sud, 69310 Pierre Be´nite, France; Institut National de la Recherche Agronomique Unit 1235 (E.M., H.V., M.L.), 69600 Oullins, France; Institut National de la Recherche Agronomique Unite´ Mixte de Recherche 1019 (E.C., B.M.), Unite´ de Nutrition Humaine and Centre de Recherche en Nutrition Humaine Auvergne, Universite´ d’Auvergne, 63001 Clermont-Ferrand, France; Service de Radiologie (F.P.), Hospices Civils de Lyon, Hoˆpital Edouard Herriot, 69002 Lyon, France; and University of Cambridge, Metabolic Research Laboratories and National Institute for Health Research, Cambridge Biomedical Research Centre (A.V.-P.), Institute of Metabolic Science, Cambridge CB2 1TN, United Kingdom
Context: The hypothesis of a limited expansion of sc adipose tissue during weight gain provides an attractive explanation for the reorientation of excess lipids toward ectopic sites, contributing to visceral adipose depots and metabolic syndrome.
Objective: Our objective was to define whether the characteristics of sc adipose tissue influence the partition of lipids toward abdominal fat depots during weight gain in healthy men.
Research Design and Methods: Forty-one healthy nonobese volunteers performed a 56-day over-feeding protocol (⫹760 kcal/d). Insulin sensitivity was estimated by euglycemic hyperinsulinemic clamp. Changes in abdominal visceral and sc adipose tissue depots were measured by magnetic resonance imaging. The fate of ingested lipids before and after overfeeding was investigated using a [d31]palmitate test meal, and gene expression was measured by real-time PCR in sc fat biopsies. Results: Overfeeding led to a 2.5-kg body weight increase with large interindividual variations in abdominal sc and visceral adipose tissues. There was no relationship between the relative expan-sions of these 2 depots, but the increase in visceral depot was positively associated with the mag-nitude of the postprandial exogenous fatty acid release in the circulation during the test meal. The regulation of lipid storage-related genes (DGAT2, SREBP1c, and CIDEA) was defective in the sc fat of the subjects exhibiting the largest accumulation in visceral depot.
Conclusions: Characteristics of sc adipose tissue appear therefore to contribute to the development of visceral fat depot, supporting the adipose tissue expandability theory and extending it to early stages of weight gain in nonobese subjects. (J Clin Endocrinol Metab 98: 802– 810, 2013)
ISSN Print 0021-972X ISSN Online 1945-7197
Printed in U.S.A.
Copyright © 2013 by The Endocrine Society
doi: 10.1210/jc.2012-3289 Received September 6, 2012. Accepted November 19, 2012. First Published Online January 2, 2013
* H.V. and M.L. contributed equally to this work.
Abbreviations: AUC, Area under the curve; CIDEA, cell death-inducing DNA fragmentation
factor,␣-subunit-like effector A; DGAT2, diacylglycerol O-acyltransferase 2; HOMA,
ho-meostasis model assessment; MRI, magnetic resonance imaging; NEFA, nonesterified fatty acid; SCD1, stearoyl coenzyme A desaturase 1; SREBP1c, sterol regulatory element binding transcription factor 1c.
E n d o c r i n e R e s e a r c h
802 jcem.endojournals.org J Clin Endocrinol Metab, February 2013, 98(2):802– 810
O
besity, a major health problem worldwide, is defined as a fat storage disease. The excessive accumulation of visceral (intraabdominal) adipose tissue correlates with metabolic syndrome features in obese and overweight dividuals (1). According to the overflow hypothesis, in-creased energy intake could primarily promote the filling of the visceral fat compartment that will further contribute to hepatic and peripheral insulin resistance (2, 3). The sc fat appears to play the role of a metabolic sink, buffering dietary fat to limit their deposition in other organs (4 – 6). The adipose tissue expandability hypothesis (6, 7) infers, however, that for some obese individuals, sc adipose tissue may reach its maximal storage capacity, the excess of lip-ids then being reoriented toward other tissues (6). Con-versely, some obese individuals, called metabolically healthy, are characterized by increased lipid deposition in the sc depot and conserved adipose tissue functions (8). Interestingly, it has been reported that sc fat from obese subjects with or without insulin resistance differs in the expression of important proteins related to lipid droplet organization, such as perilipin and cell death-inducing DNA fragmentation factor,␣-subunit-like effector A (CI-DEA) (9). This observation identifies the differences in adipocyte characteristics that may affect sc adipose tissue functions and possibly development of ectopic lipid de-position and insulin resistance (9). Under these conditions, it can be hypothesized that the expansion of intraabdomi-nal adipose depot may be facilitated by the failure of the sc depot.Thus, we aimed in the present study to better under-stand whether and how the characteristics of the sc adi-pose tissue could determine the development of the vis-ceral fat depot during weight gain in humans. Using experimental lipid overfeeding in healthy nonobese men, we show that the magnitude of visceral fat accumulation is correlated to sc adipose tissue dysfunctions, as indicated by impaired regulation of genes involved in lipid storage and increased postprandial fatty acid spillover of ingested lipids. These findings suggest that the metabolic ineffi-ciency of the sc adipose tissue could be critical for the redistribution of the energy surplus, already during the early steps of weight gain in healthy individuals.
Materials and Methods
Subjects and overfeeding protocol
Forty-one healthy male volunteers without a family history of di-abetes and with a stable weight were recruited. They gave written con-sent and the protocol was approved by the local ethics committees (registered as n. NCT00905892 atwww.clinicaltrials.gov).
To induce moderate weight gain, the subjects were asked to add to their usual daily diet 100 g of cheese, 20 g of butter, and
40 g of almonds, representing 760 kcal (3180 kJ) per day during 56 days (10). This supplement represents about 70 g of lipids, mainly composed by saturated (46.3%) and monounsaturated (44.7%) fatty acids. The subjects were also asked to maintain their lifestyle, their regular level of physical activity, and their usual eating behavior. To ensure compliance, they completed 5-day dietary records before the study and twice during the over-feeding period (days 9 –13 and days 51–55). Physical activity was also monitored during these 5-day periods using rT3
accelerom-eters. Explorations were planned at day 0 and day 56. Two days before investigation, standardized food was supplied. After an overnight fast, anthropometric parameters were recorded and hormones and metabolites measured in plasma or serum (10). Body composition was determined by whole-body dual-energy x-ray absorptiometry (Hologic, Inc, Bedford, Massachusetts). Ab-dominal adipose tissue distribution (sc and visceral fat regions) was assessed by magnetic resonance imaging (MRI) (Magnetom Sym-phonie 1.5 Tesla; Siemens AG, Munich, Germany) using single-slice image at the L2-L3 disk level. The anthropometric and metabolic characteristics of the subjects are summarized in Table 1.
Then, because the additional metabolic investigations were rather cumbersome, the studied population was separated into 2 groups so that 29 subjects underwent a hyperinsulinemic eugly-cemic clamp and 12 were submitted to a labeled-test meal. These subgroups did not differ from the whole population regarding body composition and response to overfeeding, and their char-acteristics are detailed in specific tables in the Supplemental Ma-terial, published on The Endocrine Society’s Journals Online web site at http://jcem.endojournals.org.
Hyperinsulinemic euglycemic clamp
Twenty-nine subjects were submitted to a 3-hour hyperinsu-linemic euglycemic clamp at day 0 and day 56 (Supplemental Table 1). The clamp was initiated by the infusion of insulin (Ac-trapid; Novo, Copenhagen, Denmark) at a rate of 40 mU/ m2䡠min, and adapted infusion of 20% glucose solution
(Aguettant, Lyon, France). Glucose infusion rate was used as an estimate of whole body glucose uptake (11).
Test meal with labeled palmitate
To investigate the metabolic fate of exogenous lipids, 12 sub-jects (Supplemental Table 1) performed a breakfast test meal enriched with deuterium-labeled palmitic acid ([d31]palmitate; Eurisotope, Saint-Aubin, France) (12) at day 0 and day 56. Tests started by the consumption of a fixed test meal of 870 kcal (52% carbohydrates, 15% proteins, and 33% lipids) including a bev-erage (Fortimel; Nutricia, Saint Ouen, France) that contained 20 mg/kg body weight of [d31]palmitate. At T300 min, a lunch (72% carbohydrates, 18% proteins, and 10% lipids) without tracer was provided. Blood samples were collected each hour until T480 minutes.
Fatty acid composition and isotopic enrichment of triglycer-ides in chylomicrons and plasma nonesterified fatty acids (NE-FAs) were determined after lipid extraction by the method of Bligh and Dyer. Isotopic enrichment of palmitate in the different fractions was determined by gas chromatography/mass spec-trometry (CPG 6890N; Agilent Technologies, Massy, France) after transmethylation (12). The concentration of the tracer in circulating NEFAs was calculated from the isotopic enrichment and the absolute concentration of palmitate. The areas under the curve (AUCs) were calculated using the trapezoidal method.
Subcutaneous abdominal adipose tissue biopsies and quantification of mRNA by real-time PCR
At day 0 and day 56, sc fat samples were obtained from 20 volunteers (17 performed the hyperinsulinemic clamp and 3 the test meal). They did not differ from the whole population (Sup-plemental Table 1). Biopsies were performed by needle (15 gauge) aspiration under local anesthesia (2% lidocaine) from the periumbilical level as previously described (10). Samples (rang-ing from 100 to 450 mg) were immediately frozen in liquid ni-trogen and stored at⫺70°C. Total RNA was isolated using the RNeasy kit (QIAGEN, Courtaboeuf, France). Real-time PCR assays were performed using a Rotor-GeneTM 6000 (QIA-GEN). Values were normalized using hypoxanthine guanine phosphoribosyl transferase (10). The studied genes and the primer sequences are listed in Supplemental Table 2.
Statistical analysis
All data are presented as mean⫾ SEM. Comparisons between time points (day 0 and day 56) were performed using the paired
t test, except for the data of the labeled test meal study that were
not normally distributed and thus compared using the nonpara-metric Wilcoxon test. Pⱕ .05 was considered significant. Cor-relations were performed using Spearman’s correlation test. Be-cause several correlation tests were performed, adjustment for multiple comparisons was done using Bonferroni’s correction. The adjusted P values are indicated when significant.
Results
Effects of overfeeding on body fat mass and repartition
As previously reported (10), 56 days of overfeeding with a lipid-enriched diet providing an excess of 760
kcal/d in nonobese healthy male volunteers, produced a mild increase in body weight (2.5⫾ 0.2 kg, P ⬍ .001) and minor biochemical alterations, including a slight in-crease in homeostasis model assessment (HOMA), but no change in whole-body insulin sensitivity measured by the hyperinsulinemic clamp method (Table 1). The subjects did not change their level of physical activity, as measured using rT3accelerometers, and based on the dietary records, there was no evidence for a lack of compliance that may have contributed to the individual differences in weight gain (10).
Weight gain was accompanied by significant increases in both fat mass and lean body mass. Estimated by dual-energy x-ray absorptiometry, both upper (trunk fat mass) and lower (leg fat mass) body adipose tissue amounts were higher after overfeeding (Table 1). The changes in abdom-inal adipose tissue distribution were studied using MRI. Both sc and visceral fat tissues increased significantly dur-ing overfeeddur-ing (Table 1). However, one of the unexpected findings was the large interindividual variability in the relative distribution of fat between sc and visceral adipose tissue depots (Figure 1) and the lack of correlation be-tween the expansion of the 2 abdominal depots (r⫽ 0.004,
P⫽ .981, Supplemental Table 3). This lack of association
was also observed after correction of the changes in sc and visceral depots by fat mass (r⫽ 0.140, P ⫽ .381). This indicated that some subjects experienced a marked in-crease of intraabdominal fat deposition, whereas others seemed to be resistant to visceral fat expansion and could
Table 1. Characteristics of the 41 Subjects at Baseline and After 56 Days of Overfeeding
Day 0 Day 56 Change, % P Values
n 41 41
Age, y 33⫾ 1
Body mass index, kg/m2 25.0⫾ 0.5 25.8⫾ 0.5 3.2⫾ 0.3 ⬍.0001
Body weight, kg 78.8⫾ 1.9 81.3⫾ 1.9 3.2⫾ 0.3 ⬍.0001 Waist circumference, cm 88.7⫾ 1.6 91.9⫾ 1.6 4.0⫾ 0.5 ⬍.0001 Lean body mass, kg 57.5⫾ 1.2 58.5⫾ 1.2 1.8⫾ 0.3 .006 Fat mass, kg 15.6⫾ 0.9 16.7⫾ 0.9 8.0⫾ 1.5 ⬍.0001 Fat mass, % body weight 19.4⫾ 0.9 20.1⫾ 0.9 4.6⫾ 1.2 .0012 Trunk fat mass, kg 7.7⫾ 0.6 8.5⫾ 0.6 12.0⫾ 1.8 ⬍.0001 Leg fat mass, kg 5.9⫾ 0.3 6.1⫾ 0.3 4.6⫾ 1.6 .0038 Subcutaneous adipose tissue area, cm2 179⫾ 15 199⫾ 15 13.7⫾ 2.4 ⬍.0001
Visceral adipose tissue area, cm2 183⫾ 22 204⫾ 22 22.2⫾ 5.2 .0016
Fasting glucose, mM 5.09⫾ 0.07 5.21⫾ 0.08 2.4⫾ 1.1 .037 Fasting insulin, pmol/L 60.4⫾ 4.0 62.3⫾ 3.7 5.9⫾ 3.1 ns
HOMA 2.3⫾ 0.2 2.4⫾ 0.2 8.6⫾ 3.5 .072
Leptin, ng/mL 9.4⫾ 0.8 11.8⫾ 1.0 29.0⫾ 5.5 .0003 IL-6 concentration, pg/mL 1.9⫾ 0.1 1.9⫾ 0.1 5.1⫾ 4.6 ns Fasting triacylglycerols,M 1068⫾ 86 1060⫾ 88 2.9⫾ 4.9 ns Fasting free fatty acids,M 411⫾ 23 362⫾ 17 8.0⫾ 4.5 .028 Hyperinsulinemic clamp (n⫽ 29)
Glycemia, mM 4.88⫾ 0.07 4.81⫾ 0.07 ns
Insulinemia, mIU/L 530.4⫾ 21.3 553.7⫾ 21.4 ns Glucose use rate, mg/kg free fat mass per min 9.24⫾ 0.53 9.53⫾ 0.46 ns
Abbreviations: BMI, body mass index; ns, not significant. Data are mean⫾ SEM. P values were determined with the paired Student t test.
even exhibit a reduction in their amount of visceral adi-pose tissue during overfeeding.
To characterize the determinants of fat deposition in the abdominal region, we first searched for correlations with the different anthropometric and metabolic param-eters measured during the protocol (Supplemental Table 3). The increment in abdominal sc adipose tissue was log-ically associated with the increase in body weight (r ⫽ 0.449, P⫽ .003) and trunk fat change (r ⫽ 0.321, P ⫽ .04) but not with leg fat (Supplemental Table 3). Relationships were also observed with initial fat mass (r⫽ 0.322, P ⫽ .04) and initial leptinemia (r ⫽ 0.373, P ⫽ .018). After Bonferroni’s correction, only the correlation between the variations in sc fat and total body weight showed an ad-justed P value near significance (adad-justed P⫽ .097). In contrast to sc, we did not found significant association
between the evolution of the visceral fat and the different parameters measured in the study, including markers of insulin sensitivity (Supplemental Table 3).
Gene expression in sc adipose tissue during overfeeding
We determined by quantitative RT-PCR the mRNA levels of a selection of 15 candidate genes encoding key proteins of components of vascular lipolysis (lipoprotein lipase; glycosylphosphatidylinositol anchored high den-sity lipoprotein binding protein 1; acylation stimulating protein), fatty acid uptake (fatty acid translocase/CD36; fatty acid transporter protein 1; and fatty acid binding protein 4), intracellular fatty acid esterification and stor-age [fatty acid coenzyme A ligase, long chain 1; stearoyl coenzyme A desaturase 1 (SCD1); diacylglycerol O-acyl-6 7 8 ght g) 1 2 3 4 5 Body w e ig gain (k g 0 1 100 150 a 50 0 50 V isceral fat ar ea variation (cm2) -100 -50 150 re a V v 0 50 100 cutaneous fat ar variation (cm2) -100 -50 Subc v Subjects (n = 41)
Figure 1. Individual variations of body weight and visceral and sc fat area during overfeeding. Abdominal adipose tissue depots areas were measure by MRI before and after overfeeding. The same operator performed all the quantifications. The coefficient of variation was 6.6% owing to a strict procedure for data acquisition and analysis of the MRI images. The 41 subjects were ranked according to their gain in body weight (expressed in kilograms) and the individual changes in the visceral and sc adipose tissue depot (in square centimeters) were plotted for each subject.
transferase 2 (DGAT2); and sterol regulatory element binding transcription factor 1c (SREBP1c)] as well as lipid droplet regulation (perilipin 1 and CIDEA) and intracel-lular lipolysis (hormone sensitive lipase; adipose triglyc-eride lipase; and monoglyctriglyc-eride lipase). As shown in Table 2, overfeeding induced significant changes in the expres-sion of genes related to lipid storage and triglyceride syn-thesis in sc fat, including marked inductions of SREBP1c,
DGAT2, and SCD1. In contrast, the mRNA expression of
the lipid droplet associated protein CIDEA was down-regulated during overfeeding. Importantly, as illustrated in Figure 2, we found correlations between the changes in visceral adipose tissue during overfeeding and the varia-tion in the mRNA levels of DGAT2 and CIDEA in sc adipose tissue. There was no relationship with the other studied genes (data not shown), except for a tendency for a negative correlation with SREBP1c (Figure 2). The cor-relations between the changes in the expression levels of these genes in sc fat and the variation in visceral fat gain were similarly observed whether the data of visceral
adi-pose tissue change during overfeeding were corrected (as shown in Figure 3) or not by the fat mass of the subjects. Of importance, the correlation between the variation in visceral fat and DGAT2 mRNA level remained significant after Bonferonni’s correction for multiple testing (P ⫽ .046). Therefore, the subjects with high visceral fat gain appeared to be characterized by a reduced induction of the expression of the rate-limiting enzyme of triglyceride syn-thesis in sc adipose tissue and by a tendency for a dereg-ulation of the master transcription factor SREBP1c and the lipid droplet-associated protein CIDEA. In contrast to the variation in visceral fat, we did not find significant association between the expression of the selected genes and the change in sc adipose tissue during overfeeding (data not shown).
Changes in postprandial lipid metabolism during overfeeding
The gene expression data led us to speculate that sub-jects with the highest gain in visceral fat could exhibit defects in lipid handling during the postprandial phase. Test meal experiments are classically used to investigate exogenous lipid handling (12, 13). Figure 3 shows the appearance of exogenous fatty acids in the pool of circu-lating NEFAs during a test meal labeled with deuterated palmitate. Overfeeding produced a significant increase of labeled fatty acids in circulating NEFAs during the postprandial state, with a significant 30% increase of
AUCT0-T480(Figure 3). When taking into account solely
the postmeal period the AUCT0-T300was about 1.5-fold higher at day 56 when compared with day 0 (416⫾ 38 vs 290⫾ 23, P ⫽ .006). Because [d31]palmitate enrichment in the triglycerides of the chylomicrons and the very low-density lipoprotein fractions was not different before and after overfeeding (Supplemental Figure 1), the observed increase in exogenous-derived fatty acids in the circulation after overfeeding could be due either to a decrease in li-polysis or to a defect in triglyceride-derived fatty acid up-take and storage.
Of importance, the magnitude of this increase of la-beled fatty acid in circulating NEFAs was positively cor-related with the change in visceral adipose tissue volume (r⫽ 0.58, P ⫽ .05). The correlation was even stronger when taking into account solely the postmeal period
(AUCT0-T300: r⫽ 0.66, P ⫽ .02) and was further
strength-ened when the change in visceral fat volume was corrected by the fat mass of the subjects (r⫽ 0.804, P ⫽ .002), as shown in Figure 4. Importantly, this postprandial increase of exogenous labeled fatty acid after overfeeding did not correlate with the change in sc adipose tissue (r⫽ ⫺0.19,
P⫽ .54). Together with the gene expression data, these
results suggest that a defect in the capability of sc adipose
Table 2. Regulation of the Expression of Selected
Genes Related to Lipid Metabolism in Subcutaneous Adipose Tissue During Overfeeding
Day 0 Day 56 P Values
Vascular lipolysis and fatty acid uptake LPL 84⫾ 14 109⫾ 21 .081 GPIHBP1 6.5⫾ 1.0 6.3⫾ 1.1 .705 ASP 19.4⫾ 1.5 21.1⫾ 2.2 .267 FATP1 (SLC27A1) 187⫾ 16 216⫾ 27 .175 FAT/CD36 1434⫾ 157 1546⫾ 206 .278 FABP4 505⫾ 55 591⫾ 87 .191 Lipid storage FACL1 40⫾ 7 52⫾ 10 .078 SCD1 0.14⫾ 0.02 0.34⫾ 0.08 .010 DGAT2 85⫾ 13 142⫾ 23 .006 PLIN 243⫾ 26 235⫾ 29 .590 CIDEA 78⫾ 15 45⫾ 6 .019 SREBP1c 17⫾ 2 27⫾ 5 .006 Lipolysis HSL 41⫾ 4 42⫾ 5 .860 ATGL 57⫾ 7 58⫾ 9 .894 MGL 27⫾ 4 32⫾ 4 .091
Abbreviations: ASP, acylation stimulating protein; ATGL, adipose triglyceride lipase; FABP4, fatty acid binding protein 4; FACL1, fatty acid coenzyme A ligase, long chain 1; FAT, fatty acid translocase; FATP1, fatty acid transporter protein 1; GPIHBP1,
glycosylphosphatidylinositol anchored high density lipoprotein binding protein 1; HPRT, hypoxanthine guanine phosphoribosyl transferase; HSL, hormone sensitive lipase; LPL, lipoprotein lipase; MGL,
monoglyceride lipase; PLIN, perilipin 1. The mRNA levels of the selected genes were measured by quantitative RT-PCR and normalized using HPRT mRNA expression in sc adipose tissue biopsies taken before (day 0) and after (day 56) overfeeding in 20 subjects. Data are mean⫾ SEM. Statistical analyses were performed using a paired t test.
tissue to handle exogenous lipids during weight gain is associated with visceral fat expansion.
Discussion
Because visceral fat accumulation is a major feature of the metabolic syndrome and its related complications, it is of importance to better understand the mechanisms leading to its development in humans. We demonstrate, for the
first time in healthy men, that the metabolic characteristics of the abdominal sc adipose tis-sue are associated with an increased deposition of fat in the visceral depot during lipid overfeeding.
The overfeeding protocol led to a mild in-crease in body weight and minor biochemical alterations. This allowed us to investigate the early events of adipose tissue expansion with-out the confounding effects of marked meta-bolic alterations associated with changes in body weight. Given the healthy state of the vol-unteers, it was not surprising that the moderate increase in body weight did not result in a sig-nificant difference in whole-body insulin sen-sitivity, as assessed by a hyperinsulinemic eu-glycemic clamp. However, the apparent increases in HOMA and fasting glycemia sug-gested that glucose homeostasis, and poten-tially hepatic insulin sensitivity, might be slightly affected after overfeeding. In addition to the large interindividual difference in the changes of the abdominal sc and visceral adi-pose tissue depots during overfeeding, one of the unexpected results was the finding that there was no relationship between their respec-tive evolutions. The increment in sc fat was logically associated with the increase in body weight and fat mass. However, none of the measured parameters, including change in in-sulin sensitivity, was associated with the devel-opment of visceral fat. The variability of vis-ceral fat during overfeeding has already been observed (14, 15) and could not be attributed to a lower compliance to the dietary interven-tion. Indeed, in addition to a careful monitor-ing of compliance usmonitor-ing precise dietary re-cords, the lack of association between the change in visceral fat and the evolution of body weight or total fat mass clearly ruled out this assumption. This important variability of vis-ceral fat during lipid overfeeding suggests therefore that the evolution of this depot is not always related to the energy balance, at least during the dynamic phase of weight gain in healthy men. This could represent an adaptive mechanism to regulate lipid distri-bution and control energy homeostasis in these subjects. An important observation is also the demonstration that the change in visceral fat is associated with specific defects in abdominal sc adipose tissue. Gene expression analysis in sc adipose tissue revealed altered regulation of several genes involved in lipid handling in relation to the 4 5 6 A variation D 56 vs D0) r = - 0.65 , p = 0.002 n = 20 0 1 2 3 DGA T 2 mRN A (fold change D -6 -4 -2 0 2 4 6 8 variation D 56 vs D0) 2 r = 0.45 , p = 0.041 n = 20 CIDEA mRNA fold change D 0 1 3 4 ( v ariation 6 vs D0) 0 -5 0 5 10 r = - 0.41 , p = 0.067 n = 20 1 2 3 E BP1c mRNA v ld change D5 6 0 -5 0 5 10 SR E (fo
Visceral fat variaon during overfeeding (relave units) Figure 2. Associations between the change in the mRNA levels of lipid storage-related genes in abdominal sc fat and the variation in visceral adipose tissue during overfeeding. DGAT2, CIDEA, and SREBP1c expression levels were determined by quantitative RT-PCR in sc adipose tissue taken at day 0 and day 56 in 20 subjects. For this figure, the changes in visceral fat area between day 0 and day 56, measured by MRI as shown in Figure 1, were corrected by the total fat mass of the subjects and thus expressed as relative units. As indicated in Results, these correlations were similarly found when the changes in visceral fat area were not corrected by the fat mass of the subjects (r⫽ 0.64, P ⫽ .003 for CIDEA mRNA changes; r ⫽ 0.46, P ⫽ .041 for DGAT2 mRNA changes and r⫽ 0.38, P ⫽ .065 for SREBP1c mRNA changes). The correlation analysis was performed using Spearman’s test. After Bonferroni’s correction, only the correlation between DGAT2 and visceral fat corrected by the total fat mass remained significant (P⫽ .046).
gain in visceral fat during overfeeding. The subjects with the most important increase in visceral fat were charac-terized by reduced induction of DGAT2, and to a lower extent of SREBP1c. SREBP1c is a key transcription factor regulating the lipogenic pathway (16), and DGAT2 cata-lyzes the last step of triglyceride synthesis (17). Induction of SREBP1c and DGAT2 expression is a classical feature of adipose tissue development during weight gain (10, 18, 19). The reduced induction of these genes strongly sup-ported the hypothesis of a metabolic defect in the storage capability in sc adipose tissue of the high visceral fat gain-ers. This observation is in line with the recent report show-ing that down-regulation of adipose tissue fatty acid
traf-ficking in sc adipose tissue could be a driving force for ectopic fat deposition (20). We also observed an association between change in CIDEA gene expres-sion in the sc depot and the change in visceral adipose tissue. CIDEA encodes a lipid droplet associated protein in-volved in lipolysis (21, 22). Its expres-sion is increased with weight loss (23, 24) and reduced during weight gain (10), suggesting a role in the control of the lipid droplet dynamics in response to variations in energy supply. Interest-ingly, it has been recently speculated that CIDEA contributes to a highly reg-ulated pathway of triglyceride deposi-tion in human adipose tissue and that failure of this pathway could result in ectopic lipid accu-mulation (9).
Due to the limited amount of adipose tissue available, we were unable to verify the change at the protein level and to extend the study to additional genes. The potential im-plication of the key transcription factor SREBP1c sug-gested that other genes and pathways are likely to con-tribute to an altered functioning of sc adipose tissue. Additional investigations, with more subjects and with large-scale analysis of gene expression, are thus required to confirm these results. Although we could also not firmly conclude that the observed deregulation in gene expres-sion in sc adipose tissue was directly causing the difference in visceral fat gain, the data of the labeled test meal supported a physiological relevance of this defect. The increase in [d31]palmitate in the pool of circu-lating NEFAs after the diet reflects a higher spillover of exogenous-de-rived fatty acids (13), suggesting thus that 56 days of lipid overfeeding had modified the capability to use and store exogenous fatty acids. The positive correlation between this re-lease of exogenous-derived fatty ac-ids and the visceral fat gain supports a storage defect in sc adipose tissue of the subjects with the highest pro-pensity to accumulate visceral fat. This finding indicates therefore that alteration in the ability to handle di-etary lipids in sc fat may contribute to accumulation of visceral depot during the early phase of weight 4 0 2.5 3.0 3.5 4.0
*
700 900 concentration ol/L) n = 12 0.5 1.0 1.5 2.0 100 300 500 d31palmitate (µm o 0 60 120 180 240 300 360 420 480 0 AUC T0-T480 D0 D56 Time (min)Figure 3. Kinetics of labeled fatty acid concentration in plasma during test meal. A test meal with [d31]palmitic acid was performed by 12 subjects before (E) and after (F) overfeeding. The appearance of labeled palmitate was measured in plasma showing a higher release of ingested fatty acids after overfeeding. The figure also shows the corresponding AUC. The data are mean⫾ SEM. *, P ⬍ .05 using a nonparametric Wilcoxon test.
400 300 350 a tion change 200 250 1 6] concentr a 100 150 AUC [d31C 1 -50 0 50 -4 -3 -2 -1 0 1 2 3 4 5 (n = 12) -100 -50
Visceral fat area change (relative units) Figure 4. Spearman correlation between the change in visceral fat area and the change in the AUC of exogenous fatty acid released in circulation induced by overfeeding. The appearance of labeled palmitate during a test meal was determined as in Figure 3. The change during overfeeding in the AUC between T0 and T300 minutes after the test meal was calculated and plotted against the variation visceral fat area between day 0 and day 56, measured by MRI and corrected by the total fat mass of the subjects, as for Figure 2. The correlation analysis was performed using Spearman’s test.
gain. Although our observations provide a critical role to the functions of sc fat, thus expanding the adipose tissue expandability theory to the healthy state, they do not con-tradict the overflow hypothesis, which infers a critical and initial role of visceral fat in the complications of obesity (2, 3). We suggest that during the initial phase of weight gain, the partitioning of lipids between abdominal depots and the preferential accumulation in visceral fat could result from defective functions of sc fat to handle appropriately exogenous lipids.
The short duration of the overfeeding regimen due to ethical considerations and thus the low magnitude of the weight gain limit the translatability of our findings to the obesity state. It must be stated also that only men were studied. The well-described gender differences in the de-velopment of abdominal adiposity certainly preclude the direct translation of the present results to women. Another important aspect to be mentioned is the fact that the overfeeding was almost exclusively based on lipid supple-mentation. The respective roles of the high fat intake, the calories, or the weight gain per se in the observed effects remain to be established.
In summary, we provide evidence supporting the con-cept that intraabdominal adipose tissue expansion in healthy men may be, at least in part, associated with a dysfunction of the sc adipose tissue. In this respect our results lend support to the adipose tissue expandability theory. They also provide evidence that the partitioning of fatty acid between depots could occur not only after ex-haustion of the buffering capacity of sc adipocytes as it was initially suggested in the theory, but also can contribute to the regulation of lipid metabolism in healthy individuals during positive energy balance. Further studies are now required to understand the underlying mechanisms of these observations and to validate that controlled meta-bolic challenges such as overfeeding and lipid tolerance tests could provide new insights for the identification of subjects at risk of visceral adipose tissue accumulation and thus of obesity complications.
Acknowledgments
The medical staff and nurses (Rhoˆne-Alpes and Auvergne Hu-man Nutrition Research Centers) are acknowledged for their skillful assistance. We particularly thank Dr Ali Ait Hssain (Au-vergne Centre de Recherche en Nutrition Humaine) for his pre-cious assistance; Emmanuelle Loizon (Institut National de la Sante´ et de la Recherche Me´dicale Unite´ 1060, Lyon) for mRNA quantification by RT-PCR; and Hubert Roth and Julien Dugas (Rhoˆne-Alpes Centre de Recherche en Nutrition Humaine) for helpful advice with statistics. We thank Professors D. Langin (Institut National de la Sante´ et de la Recherche Me´dicale Unite´
Mixte de Recherche 1048, Toulouse, France) and P. Arner (De-partment of Medicine, Huddinge, Sweden) for helpful discus-sions. The clinical study was registered atwww.clinicaltrials.gov
with no. NCT00905892. Contributions by the authors include the following: H.V. and M.L. conceived and designed the study and are the guarantors of this work. M.A., S.L.-P., E.C., and B.M. recruited the patients and performed the overfeeding pro-tocol. M.A., L.G., E.C., and V.S. acquired the clinical and bio-logical data. M.A., E.M., C.D., and H.V. performed the gene expression analysis and interpretations. F.P. realized and inter-preted the magnetic resonance imaging. All authors interinter-preted and discussed the data. M.A., A.V.-P., H.V., and M.L. wrote the manuscript. All authors approved the final version of the manuscript.
Address all correspondence and requests for reprints to: Hu-bert Vidal, Institut National de la Sante´ et de la Recherche Me´di-cale Unite´ 1060, Faculte´ de Me´decine Lyon Sud, BP12, F-69921 Oullins, France. E-mail: hubert.vidal@univ-lyon1.fr; or Martine Laville, Centre de Recherche en Nutrition Humaine Rhone-Alpes, Hoˆpital Lyon-Sud, F-69310 Pierre Be´nite, France. E-mail: martine.laville@chu-lyon.fr.
This work was supported by research grants from the Hos-pices Civils de Lyon (Actions Incitatives), from the Programme Hospitalier de Recherche Clinique Interregional, from the Agence Nationale de la Recherche (Programme de Recherche en Nutrition Humaine and Programme National de Recherche en Alimentation). A.V.-P. is supported by the Medical Research Council Centre for Obesity and Related Metabolic Diseases and the National Institute for Health Research. E.C. was supported by a postdoctoral fellowship from Danone, and M.A. was the recipient of a doctoral fellowship from the Ministe`re de l’Enseignement Supe´rieur et de la Recherche, France.
Disclosure Summary: The authors declare that there is no conflict of interest associated with this work.
References
1. Wajchenberg BL. Subcutaneous and visceral adipose tissue: their relation to the metabolic syndrome. Endocr Rev. 2000;21:697–738. 2. Kabir M, Catalano KJ, Ananthnarayan S, et al. Molecular evidence supporting the portal theory: a causative link between visceral ad-iposity and hepatic insulin resistance. Am J Physiol Endocrinol
Metab. 2005;288:E454 –E461.
3. Bergman RN, Kim SP, Catalano KJ, et al. Why visceral fat is bad: mechanisms of the metabolic syndrome. Obesity (Silver Spring). 2006;14:16S–19S.
4. Porter SA, Massaro JM, Hoffmann U, Vasan RS, O’Donnel CJ, Fox CS. Abdominal subcutaneous adipose tissue: a protective fat depot?
Diabetes Care. 2009;32:1068 –1075.
5. Frayn KN. Adipose tissue as a buffer for daily lipid flux.
Diabeto-logia. 2002;45:1201–1210.
6. Tan CY, Vidal-Puig A. Adipose tissue expandability: the metabolic problems of obesity may arise from the inability to become more obese. Biochem Soc Trans. 2008;36:935–940.
7. Virtue S, Vidal-Puig A. Adipose tissue expandability, lipotoxicity and the metabolic syndrome—an allostatic perspective. Biochim
Biophys Acta. 2010;1801:338 –349.
8. Bluher M. The distinction of metabolically ‘healthy’ from ‘un-healthy’ obese individuals. Curr Opin Lipidol. 2010;21:38 – 43. 9. Puri V, Ranjit S, Konda S, et al. Cidea is associated with lipid droplets
and insulin sensitivity in humans. Proc Natl Acad Sci USA. 2008; 105:7833–7838.
10. Alligier M, Meugnier E, Debard C, et al. Subcutaneous adipose tissue remodeling during the initial phase of weight gain induced by overfeeding in humans. J Clin Endocrinol Metab. 2011;97:E183– E192.
11. Laville M, Auboeuf D, Khalfallah Y, Vega N, Riou JP, Vidal H. Acute regulation by insulin of phosphatidylinositol-3-kinase, Rad, Glut 4, and lipoprotein lipase mRNA levels in human muscle. J Clin Invest. 1996;98:43– 49.
12. Binnert C, Pachiaudi C, Beylot M, et al. Metabolic fate of an oral long-chain triglyceride load in humans. Am J Physiol. 1996;270: E445–E450.
13. Fielding B. Tracing the fate of dietary fatty acids: metabolic studies of postprandial lipaemia in human subjects. Proc Nutr Soc. 2011; 70:342–350.
14. Bouchard C, Tremblay A, Despres JP, et al. The response to long-term overfeeding in identical twins. N Engl J Med. 1990;322:1477– 1482.
15. Votruba SB, Jensen MD. Insulin sensitivity and regional fat gain in response to overfeeding. Obesity (Silver Spring). 2011;19:269 –275. 16. Osborne TF. Sterol regulatory element-binding proteins (SREBPs): key regulators of nutritional homeostasis and insulin action. J Biol Chem. 2000;275:32379 –32382.
17. Takeuchi K, Reue K. Biochemistry, physiology, and genetics of
GPAT, AGPAT, and lipin enzymes in triglyceride synthesis. Am J Physiol Endocrinol Metab. 2009;296:E1195–E1209.
18. Meegalla RL, Billheimer JT, Cheng D. Concerted elevation of acyl-coenzyme A:diacylglycerol acyltransferase (DGAT) activity through independent stimulation of mRNA expression of DGAT1 and DGAT2 by carbohydrate and insulin. Biochem Biophys Res Com-mun. 2002;298:317–323.
19. Gosmain Y, Dif N, Berbe V, et al. Regulation of SREBP-1 expression and transcriptional action on HKII and FAS genes during fasting and refeeding in rat tissues. J Lipid Res. 2005;46:697–705.
20. McQuaid SE, Hodson L, Neville MJ, et al. Downregulation of ad-ipose tissue fatty acid trafficking in obesity: a driver for ectopic fat deposition? Diabetes. 2011;60:47–55.
21. Zhou Z, Yon Toh S, Chen Z, et al. Cidea-deficient mice have lean phenotype and are resistant to obesity. Nat Genet. 2003;35:49 –56. 22. Gong J, Sun Z, Li P CIDE proteins and metabolic disorders. Curr
Opin Lipidol. 2009;20:121–126.
23. Nordstrom EA, Ryden M, Backlund EC, et al. A human-specific role of cell death-inducing DFFA (DNA fragmentation factor-alpha)-like effector A (CIDEA) in adipocyte lipolysis and obesity. Diabetes. 2005;54:1726 –1734.
24. Gummesson A, Jernas M, Svensson PA, et al. Relations of adipose tissue CIDEA gene expression to basal metabolic rate, energy re-striction, and obesity: population-based and dietary intervention studies. J Clin Endocrinol Metab. 2007;92:4759 – 4765.
![Figure 3. Kinetics of labeled fatty acid concentration in plasma during test meal. A test meal with [d31]palmitic acid was performed by 12 subjects before (E) and after (F) overfeeding.](https://thumb-eu.123doks.com/thumbv2/123doknet/14023056.457560/7.877.52.539.104.311/figure-kinetics-labeled-concentration-palmitic-performed-subjects-overfeeding.webp)
