carbohydrate-responsive gene expression in the adipose

Endocrinology 2010 151:153-164 originally published online Oct 30, 2009; , doi: 10.1210/en.2009-0840

Kartik Shankar, Amanda Harrell, Ping Kang, Rohit Singhal, Martin J. J. Ronis and Thomas M. Badger

Carbohydrate-Responsive Gene Expression in the Adipose Tissue of Rats

Society please go to: http://endo.endojournals.org//subscriptions/ or any of the other journals published by The EndocrineEndocrinologyTo subscribe to

Copyright © The Endocrine Society. All rights reserved. Print ISSN: 0021-972X. Online

Carbohydrate-Responsive Gene Expression in theAdipose Tissue of Rats

Kartik Shankar, Amanda Harrell, Ping Kang, Rohit Singhal, Martin J. J. Ronis,and Thomas M. Badger

U.S. Department of Agriculture-Arkansas Children’s Nutrition Center (K.S., A.H., P.K., R.S., M.J.J.R.,T.M.B.), Little Rock, Arkansas 72202; and the Departments of Pediatrics (K.S., M.J.J.R., T.M.B.),Pharmacology and Toxicology (R.S., M.J.J.R.), and Physiology and Biophysics (T.M.B.), University ofArkansas for Medical Sciences, Little Rock, Arkansas 72205

Although obesity is often associated with high-fat diets, it can develop from a variety of mealpatterns. Excessive intake of simple carbohydrates is one consistent eating behavior leading toobesity. However, the impact of overconsumption of diets with high carbohydrate to fat ratios (C/F)on body composition and global adipose tissue gene expression remains unclear. We used totalenteral nutrition to evaluate the effects of caloric intake and C/F on body weight gain and devel-opment of obesity. Female Sprague Dawley rats were fed diets with either low C/F or high C/F (HC)(reflecting a 19.5-fold increase in C/F) at two levels of caloric intake: 187 or 220 kcal/kg3/4 � d (15%excess) for 4 wk. At the end of the study period, rats fed HC diets had about 20% higher body weightat either caloric intake compared with rats fed low C/F diets (P � 0.05). Body composition (assessedby nuclear magnetic resonance, computerized tomography, and adipose tissue weights) revealedhigher percent fat mass (P � 0.05) in HC rats. Obesity was associated with increased serum resistin,leptin, fasting hyperinsulinemia, and insulin resistance after an oral glucose challenge (P � 0.05).Microarray analyses of adipose tissues revealed HC diets led to changes in 270 and 464 transcriptsat 187 and 220 kcal/kg3/4 � d intakes. Genes regulating glucose transport, glycolysis, fatty acid andtriglyceride biosynthesis, desaturation and elongation, adipogenesis, and adipokines were af-fected by HC diets. These results suggest that C/F and interactions with excessive caloric intake perse may regulate body composition and play important roles in the development of obesity andmetabolic syndrome. (Endocrinology 151: 153–164, 2010)

Dietary carbohydrates exert profound influence on sev-eral aspects of body weight accretion, endocrinology,

and appetite (1, 2). Mechanisms to use carbohydrates di-rectly for energy via oxidation (glycolysis) or energy stor-age via triglyceride synthesis (lipogenesis) are highly con-served. In mammals, both the liver and adipose tissues areequipped to carry out these processes (3, 4). After a high-carbohydrate meal, plasma glucose rises, triggering arapid release of insulin from pancreatic �-cells. Insulin,mainly via the insulin receptor/phosphatidylinositol 3-ki-nase pathway, acts on its target tissues and orchestrates a

multitude of anabolic effects: inhibiting gluconeogenesisin the liver, enhancing glucose uptake and oxidation inskeletal muscle, and suppressing lipolysis and increasinglipogenesis in adipose tissues (5). The lipogenic actions ofinsulin are in large part mediated via transcriptional ac-tivation of target genes controlled by sterol regulatory el-ement binging protein (SREBP)-1c. In recent years, it hasbecome evident that glucose itself is an important regula-tor of mRNA transcription of an increasing number oftargets genes via the carbohydrate-response element bind-ing protein (ChREBP) (6–11).

ISSN Print 0013-7227 ISSN Online 1945-7170Printed in U.S.A.Copyright © 2010 by The Endocrine Societydoi: 10.1210/en.2009-0840 Received July 17, 2009. Accepted October 2, 2009.First Published Online October 30, 2009

Abbreviations: ACC, Acetyl CoA carboxylase; ChREBP, carbohydrate response elementbinding protein; C/F, carbohydrate to fat ratio; CT, computerized tomography; GO, geneontology; HC, high carbohydrate to fat ratio; H&E, hematoxylin and eosin; IPA, ingenuitypathway analysis; LC, low carbohydrate to fat ratio; NEFA, nonesterified fatty acid; NMR,nuclear magnetic resonance; OGTT, oral glucose tolerance test; PPAR, peroxisome prolif-erator-activated receptor; SREBP, sterol regulatory element binding protein; TEN, totalenteral nutrition.

E N E R G Y B A L A N C E - O B E S I T Y

Endocrinology, January 2010, 151(1):153–164 endo.endojournals.org 153

The majority of studies examining carbohydrate-in-duced transcriptional regulation have focused on the liver(7, 8, 11). Relatively little is known about the effect of dietswith high carbohydrate to fat ratios (C/F) on global geneexpression profiles in the adipose tissue (12). Further-more, few studies have examined the effect of high C/F(HC) on development of obesity, insulin resistance, andother metabolic/endocrine parameters (13). Self-limitingconsumption of diets due to satiety has been the primarylimitation in examining the effects of overfeeding HC dietson development of obesity in animal models. In the presentstudy, we used controlled enteral feeding of liquid diets viatotal enteral nutrition (TEN) as a mechanistic tool to over-come this limitation.

The present studies had two objectives. First, we ex-amined the interaction between caloric intake and carbo-hydrate content on body weight and adiposity. Specifi-cally, we examined how increasing dietary C/F, either inthe context of normal or overfed states, impacts bodyweight and composition. Second, we elucidate global tran-scriptomic changes in the adipose tissues and the biochem-ical and endocrine alterations that occur in response tohigh dietary carbohydrates. In addition, we examine therelationship between changes in body composition causedby HC diets and development of whole-body insulin re-sistance. Our data strongly suggest that HC diets tran-scriptionally regulate a diverse suite of genes that leads tocoordinated regulation of glucose transport, glycolysis,and lipid biosynthesis in the adipose tissue leading togreater obesity and insulin resistance.

Materials and Methods

Animals and chemicalsFemale Sprague Dawley rats (150–175 g) were purchased

from Charles River Laboratories (Wilmington, MA). Animalswere housed in an Association Assessment and Accreditation ofLaboratory Animal Care-approved animal facility. Animalmaintenance and experimental treatments were conducted inaccordance with the ethical guidelines for animal research es-tablished and approved by the Institutional Animal Care and UseCommittee at the University of Arkansas for Medical Sciences.Unless otherwise specified, all chemicals were purchased fromSigma-Aldrich Chemical Co. (St. Louis, MO).

Experimental protocolRats were intragastrically cannulated and allowed to recover

for 10 d as previously described (14–17). Rats (n � 7–9/group)were fed liquid diets via TEN. Diets were either low C/F [LC;35% carbohydrate and 45% fat (corn oil), as percent of totalcalories] or HC (75% carbohydrate and 5% fat, as percent oftotal calories), reflecting a 19.5-fold increase in C/F in the HCgroup. Both diets were isocaloric, met National Research Coun-cil nutrient recommendations, including essential fatty acids,

and provided 20% of total calories from protein (casein). Ratswere fed via TEN for 23 h/d at either 187 kcal/kg3/4 � d (ResearchCouncil recommendations) or 220 kcal/kg3/4 � d (15% overfed).Compositions of diets have been described previously (14, 15,17). Animals had unlimited access to drinking water throughoutthe experiment. TEN feeding continued for 4 wk throughoutwhich body weights were monitored twice weekly. At the end of4 wk of diet infusion, body composition was noninvasively es-timated using nuclear magnetic resonance (NMR; Echo MedicalSystems, Houston, TX) and x-ray computerized tomography(CT; LaTheta LCT-100; Echo Medical Systems) as detailed be-low. At the end of the study, rats were euthanized under anes-thesia and blood, livers, kidneys, and adipose tissues (retroper-itoneal, gonadal, and perirenal depots) were weighed andcollected. Samples were fixed in alcoholic-formalin for histolog-ical analyses, and remaining tissues were frozen in liquid nitro-gen and stored at �70 C for RNA and protein analyses. Serumwas obtained by centrifugation of blood samples and stored at�20 C for endocrine and metabolic assessments. In a separate setof animals (n � 5/group), which included an ad libitum chow-fedcontrol group, insulin sensitivity was examined after an oralglucose tolerance test (OGTT).

Body composition analysesBody composition was assessed via three independent meth-

ods, namely whole animal body composition by NMR (EchoMedical Systems), x-ray CT (LaTheta LCT-100; Echo MedicalSystems) and postmortem dissected weights of retroperitoneal,perirenal, and gonadal fat pads from rats (17). NMR was per-formed in conscious rats and did not require the animals to re-main still. Each NMR measurement took about 1 min/rat and allmeasurements were performed in duplicate (17). Indices of per-cent fat and lean mass were derived using this technique. For CTanalyses, approximately 90 sections, 1 mm apart, were acquiredencompassing the entire visceral region of the animal underisoflurane anesthesia. Densitometric calculations of fat and mus-cle were performed using Aloka CT software (Tokyo, Japan)using attenuation number thresholds of �120 to �500 for fatand �120 to �350 for muscle. Indices of percent fat mass andpercent lean mass were calculated. Subcutaneous and visceral fattissues were distinguished via manual tracing of the abdominalwall in each of the sections (17).

OGTT and other endocrine parametersTo assess development of insulin resistance in relationship

with adiposity changes, rats were challenged with an OGTT. Allrats were fasted for 6 h (from 0900 to 1500 h) before receivinga 3.5 g/kg challenge of glucose (0.5 g/ml solution) (17). Blood(�50 �l) from the tail vein was collected in capillary tubes at thebeginning of the fast and at 0, 15, 30, 60, 90, and 150 min afterthe glucose challenge. Serum glucose was measured using glucoseoxidase method (Synermed, Westfield, IN). Serum insulin con-centrations were assayed using ELISA for rat insulin (Millipore,Billerica, MA). To determine status of adipose-related biochem-ical and endocrinological parameters, we assessed concentra-tions of serum leptin, adiponectin, resistin, nonesterified fattyacids (NEFAs), and triglycerides. Hormones were assayed usingELISA (leptin; Linco Research, St. Charles, MO; adiponectin andresistin; B-Bridge International, Sunnyvale, CA). Serum triglyc-erides were estimated using colorimetric assay (Synermed).NEFA levels were measured using NEFA C kit (Wako Chemi-cals, Richmond, VA).

154 Shankar et al. Carbohydrate-Responsive Gene Expression Endocrinology, January 2010, 151(1):153–164

Adipose gene expression analyses

RNA isolation and microarray analysesTotal RNA was isolated from retroperitoneal adipose tissues

of rats (n � 7–9/group) fed LC or HC diets at 187 and 220kcal/kg3/4 � d using TRI reagent (Molecular Research Center,Cincinnati, OH) and cleaned using RNeasy kit (QIAGEN, Va-lencia, CA). Three microarrays (GeneChip Rat 230 2.0) wereused for each group. Pools of equal amounts of RNA from twoto three rats per microarray were used for analyses. Thus, sevento nine rats per group were represented over the three microar-rays. Briefly, 5 �g of purified RNA were used to synthesizecDNA. Labeled cRNA was synthesized from double-strandedcDNA using the GeneChip IVT labeling kit (Affymetrix, SantaClara, CA) according to the manufacturer’s instructions. Theprobe array was scanned after the wash and staining protocolswith GeneChip Scanner 3000 (Affymetrix). CEL files containingthe 31,099 transcripts on the GeneChip Rat genome 230 2.0array were generated using GCOS (Affymetrix).

Microarray data normalization and analysisMicroarray data analyses were carried out using GeneSpring

version 7.3X software (Agilent Technologies, Santa Clara, CA)(18–20). The CEL files containing the probe level intensitieswere processed using the robust multiarray analysis algorithm,for background adjustment, normalization, and log2-transfor-mation of perfect match values (21). Subsequently the data weresubjected to normalization by setting measurements less than0.01 to 0.01 and by per-chip and per-gene normalization usingGeneSpring normalization algorithms (Agilent Technologies).The normalized data were subjected to pairwise comparisons asfollows: 187-HC vs. 187-LC and 220-HC vs. 220-LC. In eachcomparison, genes were filtered based on minimum �1.8-foldchange (HC vs. LC) and P � 0.05 using Student’s t test. Correc-tions for multiple testing were performed using the False Dis-covery Rate (FDR) method (22). Volcano plots and a list of alltranscripts that were differentially expressed as a function of HCwas generated and correlation-based hierarchical clustering be-tween treatment groups was performed. Known biological func-tions of genes were queried using Affymetrix NetAffx and geneontology (GO) analyses performed using GeneSpring (AgilentTechnologies) (18–20).Abbreviations forgenesymbolscanbeque-riedfromtheNetAffxDataAnalysesCenter(http://www.affymetrix.com/analysis/index.affx). Furthermore, the list of genes affected byHC was analyzed using ingenuity pathway analysis (IPA).

Real-time RT-PCROne microgram of total RNA was reverse transcribed (n �

7–9/group) using IScript cDNA synthesis kit, and subsequentreal-time PCR analysis was performed using an ABI Prism 7000sequence detection system (Applied Biosystems, Foster City,CA). Gene-specific primers were designed using Primer ExpressSoftware (Applied Biosystems) (supplemental Table S1, pub-lished as supplemental data on The Endocrine Society’s JournalsOnline web site at http://endo.endojournals.org). The relativeamounts of mRNA were quantitated using a standard curve andnormalized to the expression of cyclophilin A mRNA (18).

Histology and histomorphometrySamples of liver tissues were fixed in either buffered formalin

or optimum cutting temperature compound and processed using

routine histological techniques (17, 20). For histomorphometricanalyses of adipose tissue, 3- to 4-mm pieces from the retroper-itoneal fat depots were fixed in buffered alcoholic formalin for4 d and embedded in paraffin. Six-micrometer-thick sectionswere stained with hematoxylin and eosin (H&E). Diameters ofadipocytes were measured using an Axiovert microscope (CarlZiess Inc., Thornwood, NY) with ZiessVision software (CarlZiess). A minimum of 300 cells at random were measured foreach slide (n � 7–9/group) and percentage of cells in each sizerange was computed using MS Excel (Microsoft, Redmond, WA)(17, 20).

Cell lysate and immunoblottingTotal lysates from liver and retroperitoneal adipose tissues

were prepared in radioimmunoprecipitation assay buffer con-taining 1 mM phenylmethylsulfonyl fluoride and protease inhib-itor cocktail. Nuclear protein from adipose tissues was preparedusing NE-PER reagents (Thermo Fisher, Rockford, IL). Proteinswere resolved by SDS-PAGE and immunoblotting was carriedout using standard procedures (17, 18, 23). Membranes wereincubated with rabbit anti-FAS (Abcam, Cambridge, MA), rab-bit antiacetyl CoA carboxylase (ACC; Cell Signaling, Beverly,MA), rabbit anti-ChREBP (Cayman Chemicals, Ann Arbor,MI), antimouse �-actin, or lamin A (Sigma) antibodies for 16 hat 4 C. Proteins were visualized using ECL-Plus (GE HealthcareBio-sciences, Piscataway, NJ) and detected by autoradiographyfollowed by densitometric scanning.

Data and statistical analysisData are expressed as means � SEM. Associations between the

variables, serum leptin, insulin at 15 min after glucose challenge,and percent fat mass, respectively, were examined by linear re-gression. Similarly, linear regression was carried out betweenpercent fat mass and normalized mRNA expression in adiposetissues for selected genes quantitated via real-time RT-PCR. Thedata were tested for equality of variance. A two-way ANOVAfollowed by all-pair wise comparison by the Student-Neuman-Keuls method was used to compare the effects of carbohydratediets and caloric intake (overfeeding). P � 0.05 was consideredstatistically significant. Statistical analyses were performed usingSigmaStat 3.3 software (Systat Software Inc., San Jose, CA).Graphical representation performed using SigmaPlot version10.0 for Windows (Systat Software).

Results

HC diets lead to changes in body weight, fataccretion, and endocrine profiles

Rats fed 220-kcal diets for 4 wk were heavier and ex-hibited greater adiposity compared with 187 kcal-fedcounterparts (Fig. 1, A and B, P � 0.05). However, bothat 187 and 220 kcal intakes, HC diets led to greater weightgain (125 and 116%) compared with rats fed LC diets (Fig.1A, P � 0.05). Percent fat mass in HC-fed animals was 153and 161% greater when compared with the LC diet-fedrats (Fig. 1B). Quantitation of percent fat mass and leanmass in the truncal region using CT revealed increasedadiposity (P � 0.05) in both visceral and sc compartments


(Fig. 1, C and D). Finally, postmortem quantitation ofdissected fat depots showed a robust effect (P � 0.001) ofHC diets and overfeeding on total fat weights (retroperi-toneal, gonadal, and perirenal depots) (Table 1). HC dietsalso significantly increased relative liver weights (P �0.001) (Table 1). In separate preliminary studies, we de-termined that body weight gain and body composition ofrats fed 187-LC diets via TEN were similar to that of adlibitum chow-fed rats (body weight at 4 wk, 267 � 16 inchow fed vs. 271 � 4 in 187-LC, respectively; percent leanmass at 4 wk, 63 � 0.7 in chow fed vs. 62.7 � 1.1 in187-LC, respectively; percent fat mass at 4 wk, 14 � 1 inchow fed vs. 17 � 1.1 in 187-LC, respectively). Both fedand fasting concentrations of insulin, leptin, and glucosewere also not significantly different between the twogroups (data not shown). Hence, the 187-LC fed rats rea-sonably serve as a lean control group to examine effects ofHC diets.

The impact of overfeeding and HC diets on endocrineand biochemical parameters was investigated (Table 1).Serum glucose was only modestly increased by HC dietsbut robustly increased by overfeeding per se (P � 0.005).Serum insulin levels increased 300 and 130% at 187 and220 kcal, respectively, in rats after HC diets, indicatingdevelopment of insulin resistance. As anticipated, serumleptin levels positively correlated with degree of adiposity(Table 1, r2 � 0.84, P � 0.0001). Moreover, despite nor-malization of serum leptin levels to fat mass, higher serumleptin concentrations were observed in the rats fed HCdiets at 187 kcal, indicative of leptin resistance (0.45 �0.03 vs. 0.74 � 0.08 in LC and HC-187 groups, respec-tively, P � 0.01). Serum adiponectin concentrations weredecreased by about 50% after overfeeding of LC diets.However, feeding of HC diets led to increased serum adi-ponectin levels at both caloric intakes (P � 0.01), suggest-ing that carbohydrate-driven adiposity in this model was

FIG. 1. A, Body weights of female rats fed diets with LC or HC via TEN at 187 kcal/kg3/4 � d (n � 7 and 9 in LC and HC, respectively) or 220 kcal/kg3/4 � d (n � 9 and 9 in LC and HC, respectively) for 4 wk. Infusion of diets was carried out for 23 h/d via computer-controlled syringe pumps. B,Body composition analyses (n � 7–9/group) of rats at end of diet infusion (4 wk). Fat mass and lean mass expressed as percent of body weight asestimated noninvasively by NMR analyses (echo magnetic resonance imaging). C, X-ray CT-based analyses of body composition was carried outusing LaTheta LCT-100 scanner. Scout view of the entire rat (top) and transverse slice view (bottom) of representative rats are depicted. Vis,Visceral adipose tissue; Sc, subcutaneous adipose tissue. Approximately 90 transverse slices, 1 mm apart, were acquired encompassing the entirevisceral region of the animal under isoflurane anesthesia (n � 5/group). Densitometric calculations of fat and muscle were done using Aloka CTsoftware using attenuation number thresholds of �120 to �500 for fat and �120 to �350 for muscle. Subcutaneous and visceral fat tissues weredistinguished via manual tracing of the abdominal wall in each of the sections. D, Percent lean mass, total, visceral, and sc fat mass (as a percentof body weight) in rats fed LC or HC diets as computed using Aloka CT software Data are expressed as means � SEM. Statistical differences weredetermined using a two-way ANOVA for the effects of high carbohydrate and overfeeding, followed by Student-Neuman-Keuls post hoc analyses.Differing superscripts indicate significant differences (P � 0.05).


not associated with hypoadiponectinemia. Serum resistinconcentrations increased with HC diets at both caloricintakes (P � 0.05). Serum triglyceride levels were in-creased by HC (P � 0.05), and NEFA concentrations werespecifically increased by overfeeding (P � 0.05) (Table 1).

Histology and adipose histomorphometryHistological examination of H&E- and Oil Red O-stained

liver sections revealed significant hepatic lipid accumulationin HC-fed rats (Fig. 2A). Overfeeding of LC diets causeda modest increase in hepatic lipid levels. Feeding of HCdiets increased the mean diameter of retroperitoneal adi-pocytes, indicating hypertrophy (P � 0.05, Fig. 2, A andB). At 187 kcal, the percentage of adipocytes in the 75- to100-�m size range increased from 12% in LC rats to 27%in the HC-fed rats (P � 0.05). Similarly, feeding the HCdiet at 220 kcal led to dramatic hypertrophy of adipocytes.At 220 kcal, approximately 33% cells were in the greaterthan 100 �m size range (P � 0.05) compared with 8% cellsin the LC rats (Fig. 2B).

HC diets lead to systemic insulin resistanceTo assess whether increased obesity and adipose hy-

pertrophy in HC-fed rats were associated with insulin re-sistance, we performed OGTT. Rats fed the HC diets at187 and 220 kcal showed fasting hyperinsulinemia (P �0.05; Fig. 2C). Marked changes in serum insulin responsewere observed after glucose challenge along with the gainin body weight and development of adiposity. Overfedrats (both LC and HC) were more insulin resistant com-pared with controls. Whereas insulin responses of 187-LCrats were similar to chow-fed controls (supplemental Fig.

S1), serum insulin levels were greater in HC-187 rats com-pared with LC controls (Fig. 2, D and E). Hyperglycemiawas observed in the overfed HC group, suggesting wors-ening of the glucose intolerance phenotype (Fig. 2F). Cor-relation analyses of the peak insulin response (15 min) andpercent fat mass (from NMR) revealed a strong positivecorrelation (P � 0.0001, r2 � 0.87, Fig. 2G), indicatingthat development of obesity was associated with progres-sive insulin resistance.

We further examined the status of two critical lipogenicenzymes in the liver and adipose tissues. Immunoblot anal-yses of total lysates revealed significant induction of bothhepatic and adipose tissue, fatty acid synthase, and ACC-1in HC-fed groups (Fig. 2, H and I). Furthermore, levels ofChREBP were significantly increased in adipose tissue nu-clear extracts from HC-fed rats (Fig. 2J). These data areconsistent with increased hepatic steatosis and adiposehypertrophy in the HC-fed rats.

HC diets lead to coordinated transcriptomicchanges in the adipose tissue

Microarray data were analyzed after robust multiarrayanalysis normalization using GeneSpring Gx version 7.3(Agilent Technologies). Unsupervised global conditionclustering of microarrays revealed significant associationof gene expression profiles between samples from the samegroup, indicating significant treatment effects before anygroup comparisons (supplemental Fig. S2A). Pairwisecomparisons of 187-HC vs. 187-LC and 220-HC vs.220-LC showed that 270 and 464 transcripts were altered,respectively (�1.8-fold, P � 0.05, Fig. 3A). The resultingunion of these comparisons containing 581 transcripts re-

TABLE 1. Characteristics of rats fed LC or HC diets

Parameter

187 kcal/kg3/4 � d 220 kcal/kg3/4 � d

P values

Effectof HC

Effect ofoverfeedingLC HC LC HC

Body weight at 4 wk 226 � 6.1 266 � 6.8 249 � 5.2 302 � 4.9 �0.001 �0.001Percent liver weight 3.2 � 0.06 3.5 � 0.08 3.0 � 0.08 3.5 � 0.10 �0.001 0.08Percent total fat 1.5 � 0.25 3.8 � 0.15 2.6 � 0.17 5.1 � 0.28 �0.001 �0.001Percent RP fat 0.58 � 0.10 1.4 � 0.08 0.9 � 0.08 1.9 � 0.18 �0.001 �0.001Percent gonadal fat 0.64 � 0.13 1.7 � 0.10 1.2 � 0.14 2.3 � 0.19 �0.001 �0.001Percent perirenal fat 0.32 � 0.06 0.68 � 0.09 0.54 � 0.09 0.54 � 0.09 0.002 0.049Percent kidney weight 0.72 � 0.02 0.70 � 0.02 0.74 � 0.02 0.70 � 0.02 0.16 0.76Glucose (mg/dl) 110 � 4.9 123 � 8.4 133 � 10.3 152 � 7.3 0.06 0.005Insulin (ng/ml) 1.7 � 0.15 5.1 � 1.0 8.6 � 1.4 11.2 � 1.0 �0.001 �0.001Leptin (ng/ml) 8.0 � 0.6 20.6 � 2.6 31.4 � 2.4 36.4 � 5.1 0.01 �0.001Adiponectin (�g/ml) 18.1 � 5.5 21.6 � 8.0 8.6 � 2.4 38.1 � 6.2 0.01 0.57Resistin (ng/ml) 39.9 � 5.9 73.1 � 18.4 38.2 � 6.9 61.3 � 10.6 0.02 0.57Triglycerides (mg/dl) 63 � 15 151 � 13 144 � 28 144 � 25 0.04 0.08NEFA (�M) 156 � 12 233 � 21 319 � 61 277 � 46 0.66 0.01

Data were obtained at 4 wk after consumption of LC or HC diets at 187 or 220 kcal/kg3/4 � d (n � 7, 9, 9, and 9 in LC and HC groups at 187 and220 kcal, respectively). Rats were fed diets via TEN to as described in Materials and Methods. Data are expressed as means � SEM. Statisticaldifferences were determined using a two-way ANOVA to examine the effects of HC and overfeeding, followed by Student-Neuman-Keuls post hocanalyses. RP, Retroperitoneal adipose tissue.


flects the effect of HC diets at either caloric intake (sup-plemental Table S2). Two-way correlation-based hierar-chical clustering of the carbohydrate-responsive genes isdepicted in supplemental Fig. S2B. These transcripts wereused for GO analyses based on molecular function, bio-logical function-based hierarchical clustering, and path-way analyses. Most prominently altered genes were clas-sified as having binding or catalytic functions representing39 and 27% of the 581 transcripts altered by HC (Fig. 3B).Approximately 8% of genes possessed had transporter orsignal transduction functions and about 2.5% genes hadbona fide roles in regulating transcription (transcriptionfactors, coactivator/repressors) (Fig. 3B). Furthermore,we interrogated known biological functions of genes al-tered by HC. Of the 581 transcripts altered, 100 tran-scripts were expressed sequence tag sequences with poorly

defined annotations. Approximately 55 transcripts wereidentified via the Rat Genome Database based on se-quence similarity. Using GeneSpring Gx version 7.3 (Agi-lent Technologies), we identified 29 specific genes in-volved in carbohydrate metabolism and 16 genes withknown roles in lipid biosynthesis (Fig. 3C and supplemen-tal Table S2). Correlation-based hierarchical clustering ofgenes involved in carbohydrate metabolism and lipid bio-synthesis is shown in Fig. 3, C and D. Additionally, wefound suites of genes involved in regulating electron trans-port, adipocyte secretory products, and intracellulartransport (supplemental Table S2). We further used Gene-Spring Gx 7.3 (Agilent Technologies) and pathway anal-ysis software (IPA) to identify common regulators of thealtered genes. SREBP-1, ChREBP, and leptin were iden-tified as critical nodes of regulation by HC diets, consistent

FIG. 2. A, Representative photomicrographs of liver and adipose tissues from rats fed diets with LC or HC via TEN at 187 kcal/kg3/4 � d (n � 7 and9 in LC and HC, respectively) or 220 kcal/kg3/4 � d (n � 9 and 9 in LC and HC, respectively). a-d, H&E-stained liver sections; e-h, Oil-Red-O-stainedlivers; i-l, H&E-stained retroperitoneal adipose tissues. B, Estimation of adipocyte size in the retroperitoneal adipose tissue of rats fed 187kcal/kg3/4 � d (top panel) or 220 kcal/kg3/4 � d (bottom panel) (n � 7–9/group). Diameter of a minimum of 300 adipocytes at random wasestimated, magnification, �100. Data are expressed as the average (�SEM) percentage of adipocytes in a given size range. Differing superscriptsindicate significant differences (P � 0.05). Serum insulin (C and D) and glucose (E and F) after an oral glucose challenge (2.5 g/kg) to fasted ratsfed LC or HC diets via TEN for 4 wk. Statistical differences were determined using a Student’s t test at each time point. *, P � 0.05 compared withlean rats at the same time point. G, Linear regression analyses between percent fat mass derived via NMR analyses and serum insulin at 15 minafter oral glucose challenge in rats fed LC or HC diets via TEN for 4 wk. Immunoblot analyses of fatty acid synthase and ACC protein expression inhepatic (H) and adipose (I) tissue total lysates from rats fed LC or HC diets via TEN for 4 wk. J, Immunoblot analyses of ChREBP and lamin A proteinexpression adipose tissue nuclear extracts from rats fed LC or HC diets for 4 wk.


FIG. 3. A, Venn diagram of differentially expressed adipose tissue genes at 187 kcal/kg3/4 � d or 220 kcal/kg3/4 � d after HC diet. Genes werefiltered based on minimum �1.8-fold change (HC vs. LC) and P value �0.05 using Student’s t test. B, Pie chart of GO molecular function ofdifferentially expressed transcripts by HC diets at either caloric intake. C and D, Two-way hierarchical clustering of genes with specific functions incarbohydrate or lipid metabolism derived from the list of HC altered genes. The heat map was generated using GeneSpring Gx version7.3 (AgilentTechnologies). Colors orange, yellow, and blue represent up-regulation, no relative effect, and down-regulation of adipose tissue genes,respectively. E, IPA gene network of highest significance identified using IPA software from the list of HC altered genes. Interactions betweenSREBP-1, ChREBP (MLXIPL), Glut4 (SLC2A4), and leptin signaling with several lipogenic targets is evident. Colors green and red represent down-regulation and up-regulation, respectively.


with increased adipose hypertrophy and expression of li-pogenic genes (Fig. 3E).

We performed independent verification of 25 genes us-ing real-time RT-PCR (Fig. 4). To further understand thephysiological changes in the adipose after HC diets, weincluded important candidates whose mRNA expressionwas not changed in the microarray analyses (at the setthresholds of �1.8-fold) (Fig. 4). These included adipo-genic genes, inflammatory cytokines, and other genes in-volved in fatty acid metabolism (Fig. 4). Linear regressionof mRNA expression for selected genes with percent fatmass (derived from whole body NMR) was carried out(supplemental Fig. S4). A summary of significant HC-in-duced gene expression changes of selected biological func-tions is described below.

Carbohydrate metabolismExpression of genes involved in carbohydrate transport

(Glut4), glycolysis (hexokinase, Aldo2), malonyl CoA,and precursor biosynthesis (Me1, ATP-citrate lyase,OCD, acetoacetyl CoA synthase, acetyl CoA carboxylase;supplemental Table S2) and mitochondrial decarboxylateimport were significantly induced by carbohydrates. In

addition, expression of genes regulating the pentose phos-phate pathway (G6PD, transketolase) was transcription-ally increased in the adipose tissue after HC. Previousstudies implicated activation of ChREBP, via increasedglucose flux through the pentose phosphate pathwaythrough xylulose-5-phosphate (8). Our results show thatHC diets also transcriptionally activate ChREBP mRNAin adipose tissues (Fig. 4 and supplemental Table S2), con-sistent with recent studies in primary hepatocytes (11).

Fatty acid, triglyceride, and cholesterol biosynthesisConsistent with increased glycolytic flux into fatty acid

biosynthesis, several genes regulating fatty acid biosyn-thesis, elongation, desaturation, and transport (FASN,Theddc1, Adiponutrin, Scd2, Elolv4, Elovl5 and Elvol6,Fads2, FABP1 and -5) were induced in rats fed HC diets.Adiponutrin, which has been shown to be regulated byglucose in liver and adipose tissues, was induced about 3-to 5-fold at both adequate and overfed caloric intakes (11,24). Several genes involved in triglyceride biosynthesis(G3PD, AGPAT2, AGPAT3, DGAT1) were also increasedafter HC diets. In addition, we observed a coordinatedinduction of 13 genes involved cholesterol biosynthesis

FIG. 4. A and B, mRNA expression of genes in retroperitoneal adipose tissues of rats fed LC or HC via TEN at 187 kcal/kg3/4 � d (n � 7 and 9 in LCand HC, respectively) or 220 kcal/kg3/4 � d (n � 9 and 9 in LC and HC, respectively). Gene expression was assessed via real-time RT-PCR. Expressionof each gene was quantitated on a standard curve and normalized to the expression of cyclophilin A. Data are expressed as means � SEM.Statistical differences were examined using a two-way ANOVA for the effects of HC and overfeeding, followed by Student-Neuman-Keuls post hocanalyses. Overall P values for each effect are given below the respective histogram.


and transport, including key genes in this pathway (Mvd,Lss, Sc5d, CYP51, DHCR7, ApoAI and -AII) (supplemen-tal Fig. S3).

Transcription factorsHC diets potently induced mRNA expression of several

transcription factors including CRE modulator (CREM),Forkhead A1, ChREBP, SREBP-1, and the basic helix-loop-helix B2 (Bhlhb2) proteins (Fig. 4 and supplemental TableS2). Real-time RT-PCR revealed that the expression of adi-pogenic transcription factors peroxisome proliferator-acti-vated receptor (PPAR)-�2 and CCAAT/enhancer-bindingprotein-� was significantly induced by HC diets (P �0.05), suggesting that compensatory adipogenesis was in-duced after HC diets. However, gene expression ofPPAR-� and -� was not altered by HC diets (Fig. 4).

Inflammatory cytokines and other adipocyte-secretedfactors

Consistent with increased fat mass and serum leptinlevels, mRNA of leptin was increased with both HC dietsand overfeeding (P � 0.01, Fig. 4). However, we alsoidentified several novel factors that were altered by HCdiet and overfeeding. These included four serine protein-ase inhibitors (serpin) isoforms (A1, A7, A12, C1) thatwere induced by HC and SerpinI2 that was robustly down-regulated by HC diets. In addition, we observed down-regulation of Wnt genes (Wnt2a and -4) in the adiposetissue (supplemental Table S2). Finally, we examinedexpression of inflammatory cytokines using real-timeRT-PCR. Expression of TNF-�, IL-6 or monocyte che-motactic protein-1 was not altered by either HC diet oroverfeeding (Fig. 4).

Discussion

Both the total caloric intake and percentage of caloriesderived from carbohydrates in the American diet havesteadily increased over the last 30 yr (25, 26). Studies inhuman subjects as well as animal studies (in vivo and invitro) clearly indicate that overconsumption of HC dietsmodulate cellular pathways controlling de novo lipogen-esis (13, 27–31). In an isocaloric diet, studying macronu-trient caloric sources necessitates alterations in relativeratios of macronutrients. The present study is the first tocomprehensively address adipose tissue gene expression inrats fed different dietary C/F and maintaining constantprotein calories. For ease of discussion, we use HC and LCas abbreviations for high and low C/F, respectively. Sev-eral novel findings are evident from these studies: 1) HCdiets led to greater adiposity than LC diets at two levels ofcaloric intake, 2) overconsumption of calories irrespective

of diet composition results in body fat accretion, hyper-insulinemia, and insulin resistance, albeit leading to mark-edly varied adipose gene expression signatures, and 3) de-spite the low total amount of dietary fat in the HC diets,adiposity after overfeeding of HC diets was associatedwith metabolic derangements akin to models of either ge-netic or high-fat diet-induced obesity.

HC diets induce expression of genes involved in glucoseuptake, glycolysis, and lipogenesis in a ChREBP-depen-dent fashion (8, 11). Upon activation by high glucose lev-els, the ChREBP.Mlx heterodimer mediates transcriptionof target genes through carbohydrate response elementelements. Mice lacking ChREBP are intolerant to simplecarbohydrates, present decreased expression of the gly-colytic enzyme L-pyruvate kinase, and have excess he-patic glycogen due to inability to metabolize glucose(32). Most importantly, ChREBP is required for thenormal lipogenic response after a carbohydrate load(32). Whereas ChREBP is highly expressed in the liver,it is also expressed in adipose tissues, intestine, and pan-creatic islets. Expression of ChREBP is induced in 3T3-L1cells during differentiation, and ChREBP knockout micehave smaller adipose tissue mass on a high carbohydratediet, suggesting that ChREBP may also control lipogenesisin the adipocyte (32, 33). Similar to the response in the liver,insulin-responsive induction of lipogenic genes in adipocytesis greatly enhanced by higher ambient concentrations of glu-cose (12). However, it remains unknown whether ChREBPmediates induction of lipogenic genes in the adipocyte. Ourresults from the present study establish that a HC dietpositively regulates adipose tissue ChREBP mRNA andprotein in vivo and other enzymes involved in the pentosephosphate pathway (transketolase, glucose-6-phosphatedehydrogenase). These enzymes collectively provide thereduced form of nicotinamide adenine dinucleotide phos-phate (NADP�) required for fatty acid biosynthesis andgenerate xylulose-5-phosphate, leading to activation ofChREBP.

Recently elegant studies by Ma et al. (11) used microar-rays and identified the profile of glucose-responsive genesin primary hepatocytes. These studies confirmed thatChREBP.Mlx is critical in mediating glucose-regulatedgene expression. Results from our gene expression anal-yses are remarkably similar to findings from these exper-iments. Expression of GLUT4, Aldolase B, G6PD, tran-sketolase, Me1, glycogen synthase 2, ATP-citrate lyase,FASN, ACC-1, AACS, G3PD, Dhcr7, Adiponutrin, Adi-poR2, ChREBP were some of the key genes commonlyinduced by HC diets in the adipose tissue, reported to beincreased in a ChREBP-dependent manner. Our resultssuggest that the adipose response is similar to liver, a find-ing that is not altogether surprising, considering that in-


sulin and glucose enhance lipogenesis in both tissues.However, it appears that the carbohydrate-responsivegene expression profile in the adipose tissue is muchgreater than previously appreciated. Genes in the entirepathway of glycolysis, pentose phosphate pathway, fattyacid and triglyceride biosynthesis, and desaturation/elon-gation and lipolysis are affected by dietary carbohydratesin vivo (Fig. 5).

The present data also identified several novel carbohy-drate-responsive targets in vivo unique to adipose tissue.Expression of several SREBP-1c target genes such asELOVL6, SCD1 and -2, FABP5, and suite of genes con-trolling cholesterol biosynthesis and transport were ele-vated by overfeeding of HC diets. mRNA expression ofseveral lipogenic genes correlated significantly with per-cent fat mass (supplemental Fig. S4). Expression of adi-pokines leptin, SerpinA12 (regulating insulin sensitivity),haptoglobin (acute phase response protein), and IGF bind-ing protein-3 among others was also altered. Pathwayanalyses revealed that SREBP-1, ChREBP, and leptin weremajor regulators of the observed gene expression profile.These results differ from in vitro studies because overfeed-ing of LC or HC diets in vivo concomitantly increasesserum insulin. Previous studies have shown that overfeed-

ing of mice also causes an increase in SREBP-1 in the ad-ipose tissue (34).

Changes in several metabolic and endocrine parame-ters were diet composition dependent. Serum triglycerideswere elevated by overfeeding carbohydrates, whereasNEFA levels increased more robustly as a function of over-all caloric excess. These findings are consistent with in-creased hepatic de novo lipogenesis, steatosis, and in-creased secretion of hepatic very low-density lipoprotein,resulting in higher serum triglyceride after HC feeding(35). Higher NEFA levels, on the other hand, presumablyresult from increased lipolysis of dietary fat in the LC diets.Adiponectin levels were decreased specifically by overcon-sumption of LC diets, consistent with several lines of ev-idence showing high-fat diets, and more specifically freefatty acids, may be the primary culprits in reducing adi-ponectin levels (36). Despite these differences, significantinsulin resistance that correlated with increased fat massdeveloped with obesity.

Previous studies highlighted a role for adipose tissueinflammatory gene expression in mediating insulin resis-tance (36, 37). In addition, in adipocytes activation oftoll-like receptor and c-Jun N-terminal kinase signalinghave been implicated in this process (38, 39). However, we

FIG. 5. Schematic summarizing the effect of HC diets on targets involved in cellular glucose disposal and lipid biosynthesis in adipose tissues.Genes represented in black and gray boxes are transcriptionally up-regulated and down-regulated, respectively, by HC diets. Overall, genesinvolved in glycolysis, fatty acid, lipid, and triglyceride biosynthesis are transcriptionally induced after HC diets.


did not find increases in gene expression of inflammatorycytokines, TNF-�, IL-6, and monocyte chemotactic pro-tein-1, presumably because of the relatively short durationof obesity (4 wk), compared with 8–24 wk in standarddiet-induced obesity models. Hence, it appears that insulinresistance after HC-induced obesity develops via path-ways independent of inflammation. Other literature sug-gests that hypertrophic adipocytes are per se less insulinsensitive (40, 41). Because HC diets lead to dramatic hy-pertrophy of adipocytes, mechanisms related to this maybe important. Furthermore, because skeletal muscle insu-lin action is by far the most important determinant ofwhole-body insulin sensitivity, changes in endocrine pro-files such as increases in serum resistin after HC diets mayalso play an important role. Additionally, we observed in-creases in expression of adipogenic genes, PPAR-�2 andCCAAT/enhancer-binding protein-�, suggesting increasedadipogenesis in HC-fed rats. Consistent with this, compo-nents of the Wnt pathway (Wnt2b, Wnt4) that normallyantagonize adipogenesis were down-regulated in HC-fedrats (supplemental Table S2). Because TNF-� expressionalso antagonizes adipogenesis (42), it is likely that the lack ofincreases in TNF-� along with lower Wnt expression, allowsfor differentiation of preadipocytes to occur.

In conclusion, we have demonstrated that consumptionof nutritionally complete diets with high C/F in a con-trolled manner results in extensive gene expressionchanges in the adipose tissues. Increased expression of anumber of genes regulating glycolysis and lipid biosyn-thesis appears to be coordinated via the lipogenic tran-scription factors SREBP-1 and ChREBP. The increasedlipogenic drive is reflected in and correlated with greateraccretion adipose mass and the concomitant developmentof insulin resistance and metabolic dysfunction. These re-sults suggest that targeting carbohydrate-regulated lipo-genic pathways may be an effective strategy in mitigatingincreased adiposity in at least a subset of the population.

Acknowledgments

We thank Matt Ferguson and the members of the Arkansas Chil-dren’s Nutrition Center-Animal Research Core Facility for theirassistance with TEN. We thank Dr. John C. Marecki, MichaelBlackburn, Jamie Badeaux, Renee Till, Crystal Combs, andMichele Perry for their technical assistance. We also thank Dr.Ying Su for helpful discussions regarding microarray data anal-ysis and presentation.

Address all correspondence and requests for reprints to:Kartik Shankar, Ph.D., Arkansas Children’s Nutrition Center,15 Children’s Way, Slot 512-20B, Little Rock, Arkansas72202. E-mail: [email protected].

This work was supported by the U.S. Department of Agri-culture-Agricultural Research Service (Current Research Infor-mation System Grant 6251-51000-005-03S).

Disclosure Summary: The authors have nothing to disclose.

References

1. Ludwig DS 2002 The glycemic index: physiological mechanismsrelating to obesity, diabetes, and cardiovascular disease. JAMA 287:2414–2423

2. Flatt JP 1996 Carbohydrate balance and body-weight regulation.Proc Nutr Soc 55:449–465

3. Schwarz JM, Linfoot P, Dare D, Aghajanian K 2003 Hepatic de novolipogenesis in normoinsulinemic and hyperinsulinemic subjects con-suming high-fat, low-carbohydrate and low-fat, high-carbohydrateisoenergetic diets. Am J Clin Nutr 77:43–50

4. Kim TS, Freake HC 1996 High carbohydrate diet and starvationregulate lipogenic mRNA in rats in a tissue-specific manner. J Nutr126:611–617

5. Taniguchi CM, Emanuelli B, Kahn CR 2006 Critical nodes in sig-nalling pathways: insights into insulin action. Nat Rev Mol Cell Biol7:85–96

6. Towle HC 2001 Glucose and cAMP: adversaries in the regulation ofhepatic gene expression. Proc Natl Acad Sci USA 98:13476–13478

7. Ishii S, Iizuka K, Miller BC, Uyeda K 2004 Carbohydrate responseelement binding protein directly promotes lipogenic enzyme genetranscription. Proc Natl Acad Sci USA 101:15597–15602

8. Yamashita H, Takenoshita M, Sakurai M, Bruick RK, Henzel WJ,Shillinglaw W, Arnot D, Uyeda K 2001 A glucose-responsive tran-scription factor that regulates carbohydrate metabolism in the liver.Proc Natl Acad Sci USA 98:9116–9121

9. Towle HC 2005 Glucose as a regulator of eukaryotic gene tran-scription. Trends Endocrinol Metab 16:489–494

10. Dentin R, Benhamed F, Hainault I, Fauveau V, Foufelle F, Dyck JR,Girard J, Postic C 2006 Liver-specific inhibition of ChREBP im-proves hepatic steatosis and insulin resistance in ob/ob mice. Dia-betes 55:2159–2170

11. Ma L, Robinson LN, Towle HC 2006 ChREBP*Mlx is the prin-cipal mediator of glucose-induced gene expression in the liver. J BiolChem 281:28721–28730

12. Jones BH, Standridge MK, Claycombe KJ, Smith PJ, Moustaid-Moussa N 1998 Glucose induces expression of stearoyl-CoA de-saturase in 3T3-L1 adipocytes. Biochem J 335(Pt 2):405–408

13. Minehira K, Bettschart V, Vidal H, Vega N, Di Vetta V, Rey V,Schneiter P, Tappy L 2003 Effect of carbohydrate overfeeding onwhole body and adipose tissue metabolism in humans. Obes Res11:1096–1103

14. Badger TM, Crouch J, Irby D, Hakkak R, Shahare M 1993 Episodicexcretion of ethanol during chronic intragastric ethanol infusion inthe male rat: continuous vs. cyclic ethanol and nutrient infusions.J Pharmacol Exp Ther 264:938–943

15. Badger TM, Ronis MJ, Ingelman-Sundberg M, Hakkak R 1993Pulsatile blood alcohol and CYP2E1 induction during chronic al-cohol infusions in rats. Alcohol 10:453–457

16. Badger TM, Hidestrand M, Shankar K, McGuinn WD, Ronis MJ2005 The effects of pregnancy on ethanol clearance. Life Sci 77:2111–2126

17. Shankar K, Harrell A, Liu X, Gilchrist JM, Ronis MJ, Badger TM2008 Maternal obesity at conception programs obesity in the off-spring. Am J Physiol Regul Integr Comp Physiol 294:R528–R538

18. Shankar K, Hidestrand M, Liu X, Xiao R, Skinner CM, Simmen FA,Badger TM, Ronis MJ 2006 Physiologic and genomic analyses ofnutrition-ethanol interactions during gestation: Implications for fe-tal ethanol toxicity. Exp Biol Med (Maywood) 231:1379–1397

19. Singhal R, Shankar K, Badger TM, Ronis MJ 2008 Estrogenic status


modulates aryl hydrocarbon receptor-mediated hepatic gene ex-pression and carcinogenicity. Carcinogenesis 29:227–236

20. Su Y, Shankar K, Simmen RC 2009 Early soy exposure via maternaldiet regulates rat mammary epithelial differentiation by paracrinesignaling from stromal adipocytes. J Nutr 139:945–951

21. Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ,Scherf U, Speed TP 2003 Exploration, normalization, and summa-ries of high density oligonucleotide array probe level data. Biosta-tistics 4:249–264

22. Hochberg Y, Benjamini Y 1990 More powerful procedures for mul-tiple significance testing. Stat Med 9:811–818

23. Shankar K, Liu X, Singhal R, Chen JR, Nagarajan S, Badger TM,Ronis MJ 2008 Chronic ethanol consumption leads to disruption ofvitamin D3 homeostasis associated with induction of renal 1,25dihydroxyvitamin D3–24-hydroxylase (CYP24A1). Endocrinology149:1748–1756

24. Baulande S, Lasnier F, Lucas M, Pairault J 2001 Adiponutrin, atransmembrane protein corresponding to a novel dietary- and obe-sity-linked mRNA specifically expressed in the adipose lineage.J Biol Chem 276:33336–33344

25. Wright JD, Kennedy-Stephenson J, Wang CY, McDowell MA,Johnson CL 2004 Trends in intake of energy and macronutrients,United States 1971–2000. MMWR Morb Mortal Wkly Rep 53:80–82

26. Hu FB, Stampfer MJ, Manson JE, Grodstein F, Colditz GA, SpeizerFE, Willett WC 2000 Trends in the incidence of coronary heartdisease and changes in diet and lifestyle in women. N Engl J Med343:530–537

27. Hellerstein MK, Schwarz JM, Neese RA 1996 Regulation of hepaticde novo lipogenesis in humans. Annu Rev Nutr 16:523–557

28. McDevitt RM, Bott SJ, Harding M, Coward WA, Bluck LJ, PrenticeAM 2001 De novo lipogenesis during controlled overfeeding withsucrose or glucose in lean and obese women. Am J Clin Nutr 74:737–746

29. Aarsland A, Chinkes D, Wolfe RR 1997 Hepatic and whole-body fatsynthesis in humans during carbohydrate overfeeding. Am J ClinNutr 65:1774–1782

30. Girard J, Perdereau D, Foufelle F, Prip-Buus C, Ferre P 1994 Reg-ulation of lipogenic enzyme gene expression by nutrients and hor-mones. FASEB J 8:36–42

31. Pawlak DB, Kushner JA, Ludwig DS 2004 Effects of dietary gly-

caemic index on adiposity, glucose homoeostasis, and plasma lipidsin animals. Lancet 364:778–785

32. Iizuka K, Bruick RK, Liang G, Horton JD, Uyeda K 2004 Deficiencyof carbohydrate response element-binding protein (ChREBP) re-duces lipogenesis as well as glycolysis. Proc Natl Acad Sci USA 101:7281–7286

33. He Z, Jiang T, Wang Z, Levi M, Li J 2004 Modulation of carbo-hydrate response element-binding protein gene expression in3T3–L1 adipocytes and rat adipose tissue. Am J Physiol EndocrinolMetab 287:E424–E430

34. DengQG,SheH,ChengJH,FrenchSW,KoopDR,XiongS,TsukamotoH 2005 Steatohepatitis induced by intragastric overfeeding in mice.Hepatology 42:905–914

35. Parks EJ, Parks EJ 2002 Changes in fat synthesis influenced by di-etary macronutrient content. Proc Nutr Soc 61:281–286

36. Nguyen MT, Satoh H, Favelyukis S, Babendure JL, Imamura T,Sbodio JI, Zalevsky J, Dahiyat BI, Chi NW, Olefsky JM 2005 JNKand tumor necrosis factor-� mediate free fatty acid-induced insulinresistance in 3T3-L1 adipocytes. J Biol Chem 280:35361–35371

37. Neels JG, Olefsky JM 2006 Inflamed fat: what starts the fire? J ClinInvest 116:33–35

38. Hirosumi J, Tuncman G, Chang L, Gorgun CZ, Uysal KT, MaedaK, Karin M, Hotamisligil GS 2002 A central role for JNK in obesityand insulin resistance. Nature 420:333–336

39. Tsukumo DM, Carvalho-Filho MA, Carvalheira JB, Prada PO,Hirabara SM, Schenka AA, Araujo EP, Vassallo J, Curi R, VellosoLA, Saad MJ 2007 Loss-of-function mutation in Toll-like receptor4 prevents diet-induced obesity and insulin resistance. Diabetes 56:1986–1998

40. Crandall DL, Fried SK, Francendese AA, Nickel M, DiGirolamo M1983 Lactate release from isolated rat adipocytes: influence of cellsize, glucose concentration, insulin and epinephrine. Horm MetabRes 15:326–329

41. Jernås M, Palming J, Sjoholm K, Jennische E, Svensson PA, GabrielssonBG, Levin M, Sjogren A, Rudemo M, Lystig TC, Carlsson B, CarlssonLM, Lonn M 2006 Separation of human adipocytes by size: hypertro-phic fat cells display distinct gene expression. FASEB J 20:1540–1542

42. Cawthorn WP, Heyd F, Hegyi K, Sethi JK 2007 Tumour necrosisfactor-� inhibits adipogenesis via a �-catenin/TCF4(TCF7L2)-de-pendent pathway. Cell Death Differ 14:1361–1373


carbohydrate-responsive gene expression in the adipose

Documents