Annika Mehlem, Isolde Palombo, Xun Wang, Carolina E. Hagberg, Ulf Eriksson,and Annelie Falkevall

PGC-1a Coordinates MitochondrialRespiratory Capacity and MuscularFatty Acid Uptake via Regulationof VEGF-BDiabetes 2016;65:861–873 | DOI: 10.2337/db15-1231

Vascular endothelial growth factor (VEGF) B belongs tothe VEGF family, but in contrast to VEGF-A, VEGF-Bdoes not regulate blood vessel growth. Instead, VEGF-Bcontrols endothelial fatty acid (FA) uptake and wasidentified as a target for the treatment of type 2 diabetes.The regulatory mechanisms controlling Vegfb expressionhave remained unidentified. We show that peroxisomeproliferator–activated receptor g coactivator 1a (PGC-1a)together with estrogen-related receptor a (ERR-a) regu-lates expression of Vegfb. Mice overexpressing PGC-1aunder the muscle creatine kinase promoter (MPGC-1aTGmice) displayed increased Vegfb expression, and thiswas accompanied by increased muscular lipid accumu-lation. Ablation of Vegfb in MPGC-1aTG mice fed a high-fat diet (HFD) normalized glucose intolerance, insulinresistance, and dyslipidemia. We suggest that VEGF-Bis the missing link between PGC-1a overexpressionand the development of the diabetes-like phenotypein HFD-fed MPGC-1aTG mice. The findings identifyVegfb as a novel gene regulated by the PGC-1a/ERR-asignaling pathway. Furthermore, the study highlightsthe role of PGC-1a as a master metabolic sensor thatby regulating the expression levels of Vegfa and Vegfbcoordinates blood vessel growth and FA uptake withmitochondrial FA oxidation.

Vascular endothelial growth factor B (VEGF-B) has beenshown to control endothelial fatty acid (FA) transcytosisand tissue accumulation of lipids through regulation ofthe FA transport proteins (FATPs) FATP3 and FATP4 (1).

Reduction of VEGF-B levels improves the development oftype 2 diabetes in diabetic rodent models by decreasingmuscular lipid accumulation (2). However, the regulationof Vegfb expression is not well understood. Vegfb ex-pression is high in mitochondria-dense tissues (3),and bioinformatic analysis showed that Vegfb is coex-pressed with a set of nuclear-encoded mitochondrialgenes, the so-called OXPHOS genes (1). Peroxisomeproliferator–activated receptor g coactivator 1a (Ppargc1a/PGC-1a) is a major regulator of mitochondrial energy me-tabolism, and in response to extrinsic factors, PGC-1abinds to and coactivates several transcription factors, in-cluding estrogen-related receptor a (Esrra/ERR-a), peroxi-some proliferator–activated receptor g (PPARg), and nuclearrespiratory factor 1 (NRF1) (4).

A common pathology in type 2 diabetes is dysfunc-tional and insufficient muscular mitochondria content(5), and therefore, elevated PGC-1a expression wasexpected to be beneficial by inducing mitochondrial bio-genesis. Of note, mice overexpressing PGC-1a in skele-tal muscle (muscle creatine kinase PGC-1a transgenic[MPGC-1aTG] mice) displayed reduced insulin sensitivityupon high-fat diet (HFD) feeding despite increased mito-chondrial density and higher respiratory capacity (6,7). Giventhe coexpression of Vegfb with OXPHOS genes and the roleof VEGF-B in ectopic lipid accumulation and insulin resis-tance, we stipulated that Vegfb expression may be regulatedby PGC-1a as well. In the current study, we show thatPGC-1a and ERR-a control the expression of Vegfb. This

Department of Medical Biochemistry and Biophysics, Division of Vascular Biology,Karolinska Institutet, Stockholm, Sweden

Corresponding authors: Annelie Falkevall, annelie.falkevall@ki.se, and UlfEriksson, ulf.pe.eriksson@ki.se.

Received 2 September 2015 and accepted 4 January 2016.

This article contains Supplementary Data online at http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db15-1231/-/DC1.

U.E. and A.F. were equal senior authors.

C.E.H. is currently affiliated with the Department of Cell and Molecular Biology,Karolinska Institutet, Stockholm, Sweden.

© 2016 by the American Diabetes Association. Readers may use this article aslong as the work is properly cited, the use is educational and not for profit, andthe work is not altered.

Diabetes Volume 65, April 2016 861

METABOLISM

previously undescribed regulatory pathway allows PGC-1a tocoordinate FA uptake with mitochondrial FA oxidationthrough regulation of VEGF-B levels.

RESEARCH DESIGN AND METHODS

Animal HandlingMPGC-1aTG mice (6) were crossed with Vegfb-deficientmice (3) to ultimately create MPGC-1aTG//Vegfb2/2

mice, and only male age-matched mice from the F3 gen-eration were used. For HFD studies, wild-type (WT),MPGC-1aTG, MPGC-1aTG//Vegfb2/2, and Vegfb2/2

mice where fed 60% HFD (Research Diets) for 12weeks starting from 5 weeks of age. The study wasrepeated twice. All animals had ad libitum access tofood and water and were housed in standard cages inan environment with 12-h light/dark cycles. All mousework was conducted in accordance to the SwedishAnimal Welfare Board at Karolinska Institutet, Stock-holm, Sweden.

Cell CultureCell lines were maintained in DMEM supplemented with10% FBS at 37°C in 5% CO2. C2C12 was maintained asdescribed (8). For nutrient deprivation studies, cells werestarved in DMEM.

Bioinformatic AnalysisPutative consensus binding sites for ERR-a (ERR1_Q2)within the first kilobase (kb) promoter and first intron ofVegfb were analyzed using Evolutionary Conserved Region(ECR) Browser, Mulan, and multiTF databases (9) (http://ecrbrowser.dcode.org). The sequences within the ECRswere fed to Mulan and then continued with multiTF anal-ysis with a predefined factor of 0.80.

Cloning and MutagenesisThe plasmids pCMX-ERR-a, pcDNA3.1/myc-HisA-PGC-1a-FLAG, pTK-LUC, and pCMX-b-Gal and correspondingempty vectors were used in the experiments. PCR fragmentsof the first kb Vegfb promoter and first Vegfb intron were

Table 1—Primer sequencesa

Fwd primer Rev primer

Vegfb TCTGAGCATGGAACTCATGG TCTGCATTCACATTGGCTGT

Ndufa5 ATCACCTTCGAGAAGCTGGA ACTTCACCACCCTGAAGCAA

Cycs CCAAATCTCCACGGTCTGTT CCAGGTGATGCCTTTGTTCT

Vegfa CAGGCTGCTGTAACGATGAA TATGTGCTGGCTTTGGTGAT

L19 GGTGACCTGGATGAGAAGGA TTCAGCTTGTGGATGTGCTC

B2m CTGACCGGCCTGTATGCTAT CCGTTCTTCAGCATTTGGAT

Slc2a4 ACTCTTGCCACACAGGCTCT CCTTGCCCTGTCAGGTATGT

Fatp1 TCAATGTACCAGGAATTACAGAAGG GAGTGAGAAGTCGCCTGCAC

Fatp3 CGCAGGCTCTGAACCTGG TCGAAGGTCTCCAGACAGGAG

Fatp4 GCAAGTCCCATCAGCAACTG GGGGGAAATCACAGCTTCTC

Cd36 GATGAGCATAGGACATACTTAGATGTG CACCACTCCAATCCCAAGTAAG

Tfam CCTGAGGAAAAGCAGGCATA ATGTCTCCGGATCGTTTCAC

Esrra TGGCCTCTGGCTACCACTAC CGCTTGGTGATCTCACACTC

Ppargc1a GACATGTGCAGCCAAGACTC TCAGGAAGATCTGGGCAAAG

Ndufb10 TGGAGCAGTTCACCAAAGTG TTCCAGCATTCTCTGCTTCT

Slc2a1 GCTGTGCTTATGGGCTTCTC CACATACATGGGCACAAAGC

Acox1 CACTGCCACATATGACCCCAA AGAGGCTTGTGGGTCCAA

Acadm TGCCTGTGATTCTTGCTGGAA CTTTCCCCCGTTGGTTATCCA

Acadvl CGACACTTTGCAGGGACTCA GGCCTTTGTGCCATAGAGCA

Hk2 CTTGCTGAAGGAAGCCATTCG CCGTCCACCAGTTCCACATTA

Pdk4 GGCTTGCCAATTTCTCGTCTC CACCAGTCATCAGCTTCGGA

Cs AGCCCTCAACAGTGAAAGCA TCAATGGCTCCGATACTGCTG

Plpp1 CAGTCCTTGACTGACATCGCTAA GAGAATGAAGAGTGTCCCGAGT

Sptlc1 ATCAGCGGCTCTCCGGTCAA AAGCGCCGGAGAAAGGGACT

Cers2 GGGCGCTAGAAGTGGGAAA GCTTTGGCATAGACACGTCC

Atgl GCCAACGCCACTCACATCTA CGGATGGTCTTCACCAGGTT

Acadm, acyl-CoA dehydrogenase, medium chain; Acadvl, acyl-CoA dehydrogenase, very long chain; Acox1, acyl-CoA oxidase 1;B2m, b-2-microglobulin; Cers2, ceramide synthase 2; Cs, citrate synthase; Hk2, hexokinase 2; Slc2a1, solute carrier family 2(facilitated glucose transporter), member 1; Slc2a4, solute carrier family 2 (facilitated glucose transporter), member 4; Sptlc1, serinepalmitoyltransferase, long chain base subunit 1. aAll sequences are written 59→ 39.

862 PGC-1a Regulates Vegfb Expression Diabetes Volume 65, April 2016

generated with overhangs of Hind III and Sbf I restrictioncleavage sites and 12 more random base pairs (bps). Thefragments were digested with Hind III and Sbf I and thensubcloned into pTK-LUC. Putative ERR-a binding siteswere mutated into Nco I restriction cleavage site (CCATGG)by site-directed mutagenesis.

Transient Transfection and Luciferase AssayCOS-1 cells were transfected in 24-well plates usingLipofectamine LTX according to the manufacturer’sinstructions. pCMX-b-Gal was cotransfected to allownormalization of transfection efficiency, and 24 or 48 hposttransfection, the cells were harvested and lysed.OD405 for b-galactosidase activity and chemilumines-cence signal for luciferase activity were read by using aPOLARstar Omega plate reader.

RNA Extraction and Real-Time Quantitative PCRPGC-1a overexpression in the MPGC-1aTG mice is highestin glycolytic type II fibers (6); therefore, quadriceps femoriswas used for all downstream analyses. Total RNA wasextracted by using the RNeasy Mini Kit (QIAGEN); there-after, cDNA was synthesized using iScript cDNA SynthesisKit according to the manufacturer’s instructions. Real-timequantitative PCR (qPCR) was performed as previously de-scribed (2). Primer sequences are listed in Table 1.

Oil Red O StainingMuscles were dissected, sectioned, and stained with OilRed O (ORO), and the staining was quantified as previouslyreported (10).

Glucose Measurements and Metabolic TestsBody weights and postprandial blood glucose levels weremeasured by intraperitoneal glucose tolerance tests (IPGTTs)and intraperitoneal insulin tolerance tests (IPITTs) per-formed in mice as previously described (2). Pooled datafrom at least two independent experiments, totaling 10–14 animals per genotype, were included in the analysis.Metabolic analyses of mouse plasma were performed aspreviously described (2). HOMA of insulin resistance(HOMA-IR) was calculated as (fasting glucose [mmol/L] 3fasting insulin [mU/L])/22.5.

Immunohistological Analyses of Muscle SectionsMuscles were flash frozen in liquid nitrogen, cryosec-tioned into 12-mm sections, and stained using standard

Figure 1—PGC-1a activates the Vegfb promoter through coacti-vation of ERR-a. A: Relative luciferase activity in cells transfectedwith the first kb of a Vegfb promoter/intron reporter construct, plusvectors expressing PGC-1a, ERR-a, NRF1, or PPARG (n = 4/construct). B: Relative luciferase activity in COS-1 cells trans-fected with the Vegfb promoter/intron reporter construct, withdeletion constructs or with constructs where the indicated

putative Esrra binding sites were mutated plus vectors expressingPGC-1a and ERR-a. C–E: Relative muscular mRNA expression ofPpargc1a, Vegfb, and Esrra (C); Vegfa, Tfam, Ndufa5, Ndufb10, andCycs (D); and Cd36, Fatp1, Fatp3, and Fatp4 (E) in chow-fed WT,MPGC-1aTG, and MPGC-1aTG//Vegfb2/2 mice (n = 6–7 for eachgenotype). Data are mean 6 SEM. *P < 0.05, **P < 0.01, ***P <0.001, MPGC-1aTG or MPGC-1aTG//Vegfb2/2 compared with WT;##P < 0.01, ###P < 0.001, MPGC-1aTG//Vegfb2/2 comparedwith MPGC-1aTG; †††P < 0.001, control construct compared withPGC-1a + ERR-a reporter construct or mutated construct. a.u.,arbitrary unit; D, mutated bp; Luc, transcription start site of theluciferase reporter gene.

diabetes.diabetesjournals.org Mehlem and Associates 863

procedures. The primary antibodies were goat anti-FATP3(Santa Cruz Biotechnology, Santa Cruz, TX), rat anti-CD31 (BD Biosciences, San Jose, CA), rabbit anti-FATP4 (Santa Cruz Biotechnology), and goat anti-CD31(R&D Systems, Minneapolis, MN). At least 10 frames peranimal within each section were photographed with anAxioVision microscope (Carl Zeiss, Oberkochen, Germany)at 340 magnification. For quantification of the stainings,the number of pixels per square micrometer was calculatedusing the AxioVision quantification program.

Ex Vivo b-Oxidation Assayb-Oxidation activity was measured in isolated skeletalmuscle preparations as previously described (1).

Liquid Chromatography–Mass Spectrometry–BasedLipidomic AnalysisTen milligrams of muscle tissue were analyzed by ultra-fast liquid chromatography–mass spectrometry (LC-MS)–based lipidomic analysis by the Swedish MetabolomicsCentre at the Swedish University of Agricultural Sci-ences. The analyses were carried out as previouslydescribed (11).

StatisticsIn all figures, data are presented as mean 6 SEM, withn referring to the number of independent data points.P values were calculated with two-tailed Student t test,and P , 0.05 was considered significant.

Figure 2—Muscular lipid accumulation is decreased in MPGC-1aTG mice lacking Vegfb expression. Analysis of chow-fed WT,MPGC-1aTG, MPGC-1aTG//Vegfb2/2 (n = 6–10 for each genotype), and Vegfb2/2 mice (n = 3–7). A: Representative imagesshowing ORO staining of muscle sections. B: Relative muscular mRNA expression of Acox1, Acadm, Acadvl, and Fasn. C: Post-prandial blood glucose levels. D: IPGTT. E and F: Relative mRNA expression of Glut1 (gene name Slc2a1) and insulin-sensitiveglucose transporter Glut4 (gene name Slc2a4) (E ) and Hk2, Pdk4, and Cs (F ). Data are mean 6 SEM. Scale bar = 100 mm. Insetmagnification 36. *P < 0.05, **P < 0.01, ***P < 0.001, MPGC-1aTG or MPGC-1aTG//Vegfb2/2 compared with WT; #P < 0.05, ##P <0.01, MPGC-1aTG//Vegfb2/2 compared with MPGC-1aTG. Acadm, acyl-CoA dehydrogenase, medium chain; Acadvl, acyl-CoAdehydrogenase, very long chain; Acox1, acyl-CoA oxidase 1; a.u., arbitrary unit; Fasn; fatty acid synthase; Hk2, hexokinase 2; ns, notsignificant; Slc2a1, solute carrier family 2 (facilitated glucose transporter), member 1; Slc2a4, solute carrier family 2 (facilitatedglucose transporter), member 4.

864 PGC-1a Regulates Vegfb Expression Diabetes Volume 65, April 2016

RESULTS

PGC-1a and ERR-a Regulate Vegfb Promoter ActivityTo investigate whether Vegfb can be induced by starva-tion, we monitored the expression levels of Ppargc1a andVegfb in serum-deprived C2C12 myotubes (SupplementaryFig. 1). Eight hours of nutrient deprivation inducedthe expression of both Ppargc1a and Vegfb by 2.5- and1.5-fold, respectively, and the induction was increasedto 3.5- and 1.8-fold after 16 h of starvation (Supplemen-tary Fig. 1A). The induced expression of Ppargc1a andVegfb upon nutrient deprivation was confirmed in fibro-blast (COS-1) and pericyte precursor (10T1/2) cell lines(Supplementary Fig. 1B and C).

We characterized the Vegfb promoter by cloning thefirst kb of the 59-untranslated region and the first exonand intron of the Vegfb gene in front of a luciferase re-porter construct. Plasmids encoding PGC-1a and/or thetranscription factors ERR-a, NRF1, or PPARg were thencotransfected with the luciferase reporter construct inCOS-1 cells (Fig. 1A). The results show that only coex-pression of PGC-1a and ERR-a induced significant trans-activation of the reporter construct. To identify putativeERR-a binding sites in the Vegfb promoter, the murinegenomic sequence was analyzed by the ECR Browser,Mulan, and multiTF algorithms (9,12). The analysesidentified two ERR-a consensus DNA binding sites (13)

located in the 59promoter region (2315 to 2310) and inthe first intron of the Vegfb gene (+340 to +345) andone putative binding site (2571 to 2566) (Fig. 1B). Toidentify the ERR-a binding sites responsible for thetransactivation of the Vegfb promoter, new reporter con-structs were generated where each ERR-a binding sitewas mutated by a single bp substitution. No luciferase ac-tivity was detected in the absence of PGC-1a and ERR-a,showing that the Vegfb construct per se did not induceluciferase activity (Fig. 1B, upper bar). Cotransfection ofPGC-1a and ERR-a with WT or modified luciferase con-structs revealed that mutating the 2571 to2566 bindingsite significantly decreased luciferase activity comparedwith control reporter constructs (Fig. 1B). Mutations inthe two other ERR-a DNA binding sites did not decreaseluciferase activity. However, additional cryptic ERR-abinding sites most likely exist between 20.5 and 0 kbin the Vegfb promoter and in the first intron of theVegfb gene because reporter constructs of both thesegenomic regions alone also displayed high luciferaseactivity (Fig. 1B).

We crossed MPGC-1aTG mice (6) with Vegfb-deficientanimals (3) to ultimately create MPGC-1aTG//Vegfb2/2

mice. qPCR analysis demonstrated a sevenfold increase inmuscular expression of Ppargc1a transcripts in MPGC-1aTG mice compared with WT controls (Fig. 1C). Similar

Figure 3—HFD feeding induces the expression of Vegfb and genes for FA handling proteins, whereas expression of mitochondrialOXPHOS genes are decreased. A and B: Relative muscular mRNA expression of Vegfb, Esrra, Fatp1, Cd36, Fatp3, and Fatp4 (A) andNdufa5, Ndufb10, Tfam, and Cycs (B) in chow-fed and HFD-fed WT, MPGC-1aTG, and MPGC-1aTG//Vegfb2/2 mice (n = 4–7 for eachgenotype). Data are mean 6 SEM. For clarity, only significance between chow-fed and HFD-fed animals within genotypes and HFD-fedMPGC-1aTG compared with MPGC-1aTG//Vegfb2/2 mice are indicated. †P< 0.05, ††P< 0.01, †††P< 0.001, HFD compared with chow;#P < 0.05, ##P < 0.01, ###P < 0.001, HFD-fed MPGC-1aTG//Vegfb2/2 compared with HFD-fed MPGC-1aTG. ns., not significant.

diabetes.diabetesjournals.org Mehlem and Associates 865

analyses showed six- and threefold upregulation of theexpression of Vegfb and Esrra, respectively (Fig. 1C).MPGC-1aTG mice also showed an approximate twofoldupregulation of the previously reported PGC-1a target

genes Vegfa; transcription factor A, mitochondrial(Tfam); NADH dehydrogenase (ubiquione) 1 a subcomplex5 (Ndufa5); NADH dehydrogenase (ubiquione) 1 b sub-complex 10 (Ndufb10); and cytochrome C, somatic (Cycs)

Figure 4—Muscular lipid accumulation and expression of FATPs are reduced in HFD-fed MPGC-1aTG mice lacking Vegfb. Analysis ofHFD-fed WT, MPGC-1aTG, MPGC-1aTG//Vegfb2/2 mice (n = 6–10 for each genotype), and Vegfb2/2 mice (n = 3–7). A: Representativeimages showing ORO staining of muscle sections; quantification of the ORO staining is shown on the right. B: Relative muscular mRNAexpression of Acox1, Acadm, Acadvl, and Fasn. C: Ex vivo b-oxidation measured as levels of formed 14CO2 in skeletal muscles from HFD-fed MPGC-1aTG mice and MPGC-1aTG//Vegfb2/2 mice. Etomoxir was used to inhibit mitochondrial b-oxidation and served as an internalcontrol (n = 2–3). D: Representative images of muscle sections stained with anti-FATP3 or anti-FATP4 antibodies together with CD31antibodies. Quantifications are shown on the right (n = 3–6). Data are mean 6 SEM. *P < 0.05, **P < 0.01, ***P < 0.001, MPGC-1aTG orMPGC-1aTG//Vegfb2/2 compared with WT; ##P < 0.01, ###P < 0.001, MPGC-1aTG//Vegfb2/2 compared with MPGC-1aTG. In A, scalebar = 100 mm and inset magnification 36, whereas in D, scale bars = 50 mm and 100 mm in top vs. bottom panels. Acadm, acyl-CoAdehydrogenase, medium chain; Acadvl, acyl-CoA dehydrogenase, very long chain; Acox1, acyl-CoA oxidase 1; a.u., arbitrary unit; Cd31,platelet endothelial cell adhesion molecule; Fasn; fatty acid synthase.

866 PGC-1a Regulates Vegfb Expression Diabetes Volume 65, April 2016

Figure 5—Ablation of Vegfb in HFD-fed mice reduced muscular levels of TGs, DAGs, and ceramides. LC-MS analysis of muscle biopsyspecimens from chow-fed WT and HFD-fed WT, MPGC-1aTG, MPGC-1aTG//Vegfb2/2, and Vegfb2/2 mice. A: Total ion chromatogram(TIC) of TGs and quantification of total TG content by AUC. B: Quantification of the most abundant TG species as fold change comparedwith chow-fed WT. C: TIC of DAGs and quantification of total DAG content by AUC. D: Quantification of the most abundant DAG species asfold change compared with chow-fed WT. E: TIC of ceramides and quantification of the total ceramide content by AUC. F: Quantification ofthe most abundant ceramide species as fold change compared with chow-fed WT. G and H: Relative muscular mRNA expression of Atgland Plpp1 (G) and Cers2 and Sptlc1 (H). Data are mean 6 SEM. Statistical analyses between chow-fed and HFD-fed WT are omitted for clarity.

diabetes.diabetesjournals.org Mehlem and Associates 867

(Fig. 1C and D). Ablation of Vegfb in MPGC-1aTG micedid not affect the expression of any of these genes.

We analyzed muscular expression of FATPs and sev-eral other FA handling proteins previously reported to beregulated by PGC-1a (14). The expression of fatty acidtranslocase (Cd36), Fatp1, and Fatp4 was upregulated inMPGC-1aTG mice compared with WT mice (Fig. 1E).Fatp3, the vascular-specific target of VEGF-B signaling,was downregulated in the MPGC-1aTG//Vegfb2/2 micecompared with the WT and MPGC-1aTG animals (Fig.1E). Taken together, PGC-1a and ERR-a regulate Vegfbexpression, which requires at least one intact ERR-a bind-ing site located in the promoter region of Vegfb. In vivo,PGC-1a controls Vegfb expression in parallel with theexpression of OXPHOS genes and Vegfa. The expressionof the OXPHOS genes was not altered by Vegfb deletion,whereas Fatp3 was downregulated only in the MPGC-1aTG//Vegfb2/2 mice.

Muscular Lipid Accumulation Is Reduced inMPGC-1aTG//Vegfb2/2 MiceMuscle sections from WT, MPGC-1aTG, MPGC-1aTG//Vegfb2/2, and Vegfb2/2 mice were stained with ORO tovisualize accumulation of neutral lipids. Quantificationsshowed a fivefold increase in lipid accumulation inMPGC-1aTG compared with WT mice (Fig. 2A). In con-trast, MPGC-1aTG//Vegfb2/2 mice had almost normal-ized levels of lipid accumulation (Fig. 2A). The lowestlevel of lipid accumulation was detected in Vegfb2/2

mice, in line with previous results (1).Intracellular lipids can accumulate either through in-

creased FA uptake, decreased b-oxidation, or increasedde novo FA synthesis. Therefore, muscular expression ofgenes encoding critical enzymes in b-oxidation and denovo FA synthesis were measured by qPCR analysis(Fig. 2B). The transcript levels of regulatory genes inb-oxidation and de novo FA synthesis were upregulatedin MPGC-1aTG compared with WT mice (Fig. 2B), whichis in line with previous reports (7). No further increasein the expression of any of the genes was detected inMPGC-1aTG//Vegfb2/2 mice (Fig. 2B), and in line withprevious studies, no difference in expression levels ofthese genes was detected between WT and Vegfb2/2

mice (1).Vegfb2/2 mice have a shift in nutrient utilization to-

ward increased use of glucose and decreased use of FAscompared with WT animals (1). However, on chow diet,no differences in postprandial blood glucose levels orglucose tolerance were detected among WT, MPGC-1aTG, MPGC-1aTG//Vegfb2/2, and Vegfb2/2 mice (Fig.2C and D). A minor increase in the muscular expression

of Glut4 (gene name Slc2a4) but not of Glut1 (gene nameSlc2a1) was detected in MPGC-1aTG//Vegfb2/2 micecompared with MPGC-1aTG mice (Fig. 2E). PGC-1a/ERR-a has been reported to regulate glucose utiliza-tion by controlling the expression of several intracel-lular glucose handling enzymes (15,16); therefore, weassessed muscular expression of these transcripts in WT,MPGC-1aTG, MPGC-1aTG//Vegfb2/2, and Vegfb2/2

mice (Fig. 2F). Muscular pyruvate dehydrogenase kinase4 (Pdk4) expression was downregulated, with ;60% inMPGC-1aTG//Vegfb2/2 and Vegfb2/2 mice comparedwith MPGC-1aTG mice. A minor increase in citrate syn-thase (Cs) levels (25%) were detected in MPGC-1aTG//Vegfb2/2 compared with MPGC-1aTG mice. In sum-mary, no major differences in glucose tolerance, bloodglucose levels, or expression of glucose transporterswere found between WT and MPGC-1aTG, MPGC-1aTG//Vegfb2/2, and Vegfb2/2 mice on chow diet.

VEGF-B Is Regulated by PGC-1a in HFD Mice andControls Muscular Lipid Accumulation Through FATPs

To characterize the impact of VEGF-B–driven muscularlipid accumulation of the diabetic phenotype in HFD-fedMPGC-1aTG mice, transcriptional profiling of Vegfb andgenes involved in lipid uptake and oxidation were ana-lyzed in chow- and HFD-fed animals. Vegfb expressionwas induced fourfold by HFD feeding in WT mice com-pared with chow-fed mice and twofold in HFD-fed MPGC-1aTG mice compared with chow-fed MPGC-1aTG mice(Fig. 3A). HFD feeding did not increase Esrra levels inany of the mouse strains compared with the respectivegenotype fed the chow diet (Fig. 3A). However, HFD feed-ing increased muscular expression of genes encoding FAhandling proteins in both WT and MPGC-1aTG mice (Fig.3A). Deletion of Vegfb in the HFD-fed MPGC-1aTG micesignificantly prevented the upregulation of Fatp4 (Fig.3A). On the contrary, the expression of mitochondrialgenes was reduced upon HFD feeding in WT, as previouslyreported (2), and also in MPGC-1aTG mice (Fig. 3B). De-letion of Vegfb in the HFD-fed MPGC-1aTG preservedthe expression of several OXPHOS genes compared withMPGC-1aTG mice (Fig. 3B). Thus, transcriptional analysissuggested that HFD feeding induced an imbalance be-tween lipid uptake and lipid oxidation, and reducingVEGF-B levels may prevent this, leading to metabolic nor-malization. To analyze whether ablation of Vegfb reducedmuscular FA uptake during HFD, muscular tissue sectionsfrom the various genotypes were stained with ORO.Muscular lipid accumulation was increased eightfold inHFD-fed MPGC-1aTG mice compared with WT HFD-fedmice but only fourfold in MPGC-1aTG//Vegfb2/2 mice

*P < 0.05, **P < 0.01, ***P < 0.001, MPGC-1aTG or MPGC-1aTG//Vegfb2/2 compared with WT; #P < 0.05, ##P < 0.01, MPGC-1aTG//Vegfb2/2 compared with MPGC-1aTG (n = 3–6). a.u., arbitrary unit; Cers2, ceramide synthase 2; Sptlc1, serine palmitoyltransferase, longchain base subunit 1.

868 PGC-1a Regulates Vegfb Expression Diabetes Volume 65, April 2016

(Fig. 4A). Hence, deletion of Vegfb alone in MPGC-1aTGmice lowered muscular lipid accumulation by 80%. In ac-cordance with our previous findings (2), the lowest levelsof lipid accumulation was seen in the HFD-fed Vegfb2/2

mice. The decreased muscular lipid accumulation in HFD-fed MPGC-1aTG//Vegfb2/2 mice was not due to a com-pensatory increase in b-oxidation or decrease in de novoFA synthesis because transcript levels of the regulatorygenes in these pathways were not altered (Fig. 4B). Wealso assessed whether genetic ablation of Vegfb in HFD-fed MPGC-1aTG mice directly affected the capacity ofb-oxidation by incubating isolated skeletal musclepreparations with 14C-oleic acid and measured formed14CO2 (Fig. 4C). However, no differences in the levelsof released 14CO2 were observed between genotypes.

The transcriptional alterations of the FATPs weremodest in relation to the drastic reduction in lipidaccumulation in the HFD-fed MPGC-1aTG//Vegfb2/2

mice, and we used immunohistochemistry to analyze pro-tein expression of FATP3 and FATP4 in chow-fed WT,HFD-fed WT, MPGC-1aTG, MPGC-1aTG//Vegfb2/2, andVegfb2/2 mice (Fig. 4D). CD31, a vessel-specific markerprotein, was used to identify vascular localization of theFATPs. The expression of both FATP3 and FATP4 wasstrongly upregulated in HFD-fed MPGC-1aTG mice com-pared with HFD-fed WT mice. Strikingly, the protein ex-pression of both FATP3 and FATP4 were downregulatedby ;70% and 90% in HFD-fed MPGC-1aTG//Vegfb2/2

and HFD-fed MPGC-1aTG mice, respectively. In sum-mary, HFD feeding induced the expression of Vegfb andFA handling protein transcripts in both WT and MPGC-1aTG mice, and muscular lipid deposition was increasedaccordingly. Of note, muscular lipid accumulation wasdramatically reduced in the MPGC-1aTG//Vegfb2/2 mice,which was caused by lowered FA uptake through decreasedexpression of FATP3 and FATP4.

Ablation of Vegfb in HFD-Fed Mice Targets MuscularLipid Accumulation by Reducing Triglycerides,Diacylglycerols, and CeramidesHigh intramyocellular lipid content in both humans andexperimental animal models have been shown to beassociated with insulin resistance (17,18). Specifically,both diacylglycerols (DAGs) derived from lipid dropletsand ceramides have been shown to affect insulin sig-naling and/or insulin-mediated muscular glucose up-take (19). Therefore, we used ultrafast LC-MS–based

Figure 6—Genetic deletion of Vegfb in HFD-fed MPGC-1aTG micereduces insulin resistance and type 2 diabetes. Analysis of chow-fed WT (n = 4–5) and HFD-fed WT, MPGC-1aTG, MPGC-1aTG//Vegfb2/2, and Vegfb2/2 mice (n = 7–12 for each genotype). A:Postprandial blood glucose levels and body weights. B: Insulin lev-els and HOMA-IR. C and D: Relative mRNA expression of Glut1(gene name Slc2a1) and insulin sensitive glucose transporter

Glut4 (gene name Slc2a4) (C ) and Hk2, Pdk4, and Cs (D). E and F:IPGTT (E) and IPITT (F ) and AUC analyses. Data are mean 6 SEM.Statistical analyses between chow-fed and HFD-fed WT are omit-ted for clarity. *P < 0.05, **P < 0.01, ***P < 0.001, MPGC-1aTG//Vegfb2/2 or Vegfb2/2 compared with WT; #P < 0.05, ##P < 0.01,###P < 0.001, MPGC-1aTG//Vegfb2/2 or Vegfb2/2 compared withMPGC-1aTG. a.u., arbitrary unit; Hk2, hexokinase 2; Slc2a1, solutecarrier family 2 (facilitated glucose transporter), member 1; Slc2a4,solute carrier family 2 (facilitated glucose transporter), member 4.

diabetes.diabetesjournals.org Mehlem and Associates 869

lipidomics to quantify the levels of triglycerides (TGs),DAGs, and ceramides in muscle biopsy specimens fromchow-fed WT mice and HFD-fed WT, MPGC-1aTG,MPGC-1aTG//Vegfb2/2, and Vegfb2/2 mice (Fig. 5).Area under the curve (AUC) analysis of all lipid speciesshowed that MPGC-1aTG//Vegfb2/2 and Vegfb2/2mice had;35% and 60% lower muscular TG content, respectively,compared with MPGC-1aTG mice (Fig. 5A and B). Similaranalysis showed that muscular levels of DAGs andceramides were reduced by ;35% and 25% in MPGC-1aTG//Vegfb2/2 mice and by ;70% and 35% in Vegfb2/2

mice compared with MPGC-1aTG mice (Fig. 5C–F). Toanalyze whether VEGF-B controls the degradation ofTGs to DAGs or de novo DAG synthesis, we measuredmuscular expression levels of adipose triglyceride lipase(Atgl) and phospholipid phosphatase 1 (Plpp1) (Fig. 5G).Expression of Atgl was downregulated in both HFD-fedMPGC-1aTG//Vegfb2/2 and Vegfb2/2 mice, suggestingthat lower amounts of TGs are available for degrada-tion. No differences were found between genotypes inexpression of Plpp1 or in the expression of transcriptscoding for key enzymes for ceramide formation (Fig. 5Gand H).

MPGC-1aTG//Vegfb2/2 Mice Fed an HFD AreProtected Against the Development of Type 2 DiabetesWe analyzed whether deletion of Vegfb could reverse in-sulin resistance in HFD-fed MPGC-1aTG mice (Fig. 6).Postprandial blood glucose levels were normalized inHFD-fed MPGC-1aTG//Vegfb2/2 and Vegfb2/2 animals(Fig. 6A). As previously reported, Vegfb2/2 mice had anincreased body weight compared with WT littermatesupon HFD feeding (Fig. 6A) (2). However, no differencesin body weights were detected among WT, MPGC-1aTG,and MPGC-1aTG//Vegfb2/2 mice (Fig. 6A). HFD-fed WTand MPGC-1aTG mice but not MPGC-1aTG//Vegfb2/2

mice developed hyperinsulinemia, whereas HFD-fedVegfb2/2 mice displayed moderately increased plasma in-sulin levels (Fig. 6B). The HOMA-IR index mirrored theplasma insulin levels (Fig. 6B). qPCR analysis showed thatMPGC-1aTG//Vegfb2/2 mice had higher Glut4 expression(Fig. 6C), lower Pdk4 expression (Fig. 6D), and higher Csexpression (Fig. 6D) compared with MPGC-1aTG mice,suggesting higher glucose utilization (20). In line withreduced muscular lipid accumulation and increased ex-pression of glucose handling enzymes, the HFD-fedMPGC-1aTG//Vegfb2/2 and Vegfb2/2 mice showed im-proved glucose tolerance and insulin sensitivity comparedwith HFD-fed WT and MPGC-1aTG mice (Fig. 6E and F).

To visualize how insulin resistance correlated to VEGF-Bsignaling and muscular lipid accumulation, we plottedmuscular lipid content and measurements of insulinresistance in individual HFD-fed mice (Fig. 7A). HFD-fedMPGC-1aTG were characterized by high muscular lipidcontent and insulin resistance. Reducing Vegfb2/2 levelsin HFD-fed MPGC-1aTG mice strongly affected lipid ac-cumulation and increased insulin sensitivity. In summary,

deletion of Vegfb can completely prevent insulin resis-tance in HFD-fed MPGC-1aTG mice.

In addition to insulin resistance, dyslipidemia is a keyfeature of type 2 diabetes. Hence, the levels of severalplasma lipid species were analyzed in HFD-fed mice (Fig.7B and C). The results showed that HFD feeding leads toincreased levels of plasma TGs in WT and MPGC-1aTGmice. In contrast, TG levels were decreased in HFD-fedMPGC-1aTG//Vegfb2/2 and Vegfb2/2 mice comparedwith both WT and MPGC-1aTG animals (Fig. 7B). LDLand VLDL cholesterol were also decreased in MPGC-1aTG//Vegfb2/2 mice compared with MPGC-1aTG mice,whereas no difference in HDL cholesterol was detected(Fig. 7B). Similar levels of nonesterified fatty acids(NEFAs) and ketone bodies (KBs) were also detected inMPGC-1aTG and MPGC-1aTG//Vegfb2/2 mice (Fig. 7C).In summary, the metabolic tolerance tests showed thatgenetic deletion of Vegfb in HFD-fed MPGC-1aTG mice issufficient to prevent hyperglycemia and hyperinsulinemia,reduce insulin resistance, and improve dyslipidemia, thusrescuing the aberrant phenotype of this previously enig-matic mouse model.

DISCUSSION

In this study, we demonstrate that Vegfb is a downstreamtarget of the PGC-1a/ERR-a signaling pathway. PGC-1ais induced by physiological stimuli, including exercise,fasting, and cold temperature (4). Of note, mRNA levelsof Vegfb are also increased by nutrient deprivation (Sup-plementary Fig. 1) and during exercise, which is shown inmuscles biopsy specimens from mice and human subjects(21,22). The coregulation of PGC-1a and VEGF-B ensuresthat lipid uptake from the circulation is coordinated withtissue lipid oxidation (Fig. 7D). PGC-1a/ERR-a have alsobeen shown to induce VEGF-A levels and angiogenesis inskeletal muscle in vivo during starvation (8,23–25). Thus,PGC-1a and ERR-a coordinate VEGF-A and VEGF-B ex-pression levels, mitochondrial biogenesis, and lipid oxida-tion not only by controlling vessel growth but also byregulating lipid uptake.

Given the impact of PGC-1a as the main activator ofmitochondrial biogenesis and energy metabolism, effortshave been made to elucidate the role of PGC-1a in met-abolic conditions such as insulin resistance and type 2diabetes. The muscular expression of Ppargc1a andOXPHOS genes is reduced in patients with type 2 diabe-tes, suggesting that low PGC-1a activity could be the un-derlying mechanism of type 2 diabetes (26,27). Therefore,it was surprising and paradoxical that HFD-fed MPGC-1aTG mice developed insulin resistance and type 2 diabe-tes (7). In this study, we have unraveled the mechanismunderlying the diabetic phenotype reported for theHFD-fed MPGC-1aTG mice (7). We show that muscularoverexpression of PGC-1a upregulates Vegfb levels,driving FA uptake and tissue accumulation and, as a con-sequence, insulin resistance. Inactivation of Vegfb in HFD-fed MPGC-1aTG mice reduced ectopic lipid accumulation

870 PGC-1a Regulates Vegfb Expression Diabetes Volume 65, April 2016

and the development of type 2 diabetes by reduced proteinlevels of endothelial FATPs. Hence, the HFD-induced in-sulin resistance in MPGC-1aTG mice first described byChoi et al. (7) is most likely caused by a PGC-1a–dependentupregulation of Vegfb expression.

Ablation of Vegfb in MPGC-1aTG mice did not affectplasma levels of NEFAs or KBs, despite reduction of mus-cular lipid accumulation. The current data agree with thatof Choi et al. (7), which showed that HFD-fed MPGC-1aTG mice had lower plasma levels of KB, despite in-creased muscular FA inflow. We have previously shownthat reducing VEGF-B levels in db/db mice reduced bothmuscular lipid accumulation and plasma levels of NEFAsand KBs (2). Furthermore, lipid infusions in unchallengedVegfb2/2 mice redirected lipids from the peripheral tis-sues to the adipose tissue (AT) (1). PGC-1a is described asa regulator of several myokines shown to induce alter-ations in the AT (22). Therefore, the lack of differencesin plasma NEFA levels between MPGC-1aTG and MPGC-1aTG//Vegfb2/2 could be caused either by the effect ofPGC-1a on AT tissue or by VEGF-B–dependent shuntingof lipids to the AT.

Efforts have been made to try to link the VEGF-Bsignaling pathway to insulin resistance in humans, but itwas reported that serum VEGF-B levels did not differbetween patients with type 2 diabetes and healthy controlsubjects (28). However, a major confounding factor wasthat the trial enrolled patients taking oral antidiabeticdrugs, such as metformin and thiazolidinediones. Theauthors showed that thiazolidinediones reduced cir-culating levels of VEGF-B. Similarly, in a more recentclinical trial of metformin treatment in women withpolycystic ovary syndrome and insulin resistance, met-formin was shown to reduce plasma VEGF-B levels (29).This trial enrolled only patients who had not receivedantidiabetic drugs within the past 3 months before themetformin treatment was initiated. Measurementsof serum VEGF-B levels before metformin treatmentshowed that VEGF-B levels were increased in the sub-jects with insulin resistance and correlated positivelywith insulin resistance. In addition, gene associationstudies have linked the VEGF-B signaling pathway toinsulin resistance in humans because a sequence variantin Fatp4 was identified to be associated with the meta-bolic syndrome and insulin resistance (30). Moreover,serum VEGF-B levels as well as AT Vegfb expression have

Figure 7—Reducing VEGF-B levels in HFD-fed mice targets intra-myocellular lipid storage, insulin resistance, and dyslipidemia. Aand B: Analysis of chow-fed WT (n = 4–5) and HFD-fed WT,MPGC-1aTG, MPGC-1aTG//Vegfb2/2, and Vegfb2/2 mice (n =7–12 for each genotype). A: Neutral lipid accumulation quantifiedby ORO analysis plotted against the levels of insulin resistance asmeasured by AUC analysis of IPITTs. B and C: Plasma levels ofTGs, HDL cholesterol, and LDL/VLDL cholesterol (B) and NEFAs,and ketones (C). D: Schematic illustration of PGC-1a–dependentregulation of VEGF-B and VEGF-A and how b-oxidation is corre-lated to FA uptake and angiogenesis. In the normal state, PGC-1acoordinates mitochondrial biogenesis and b-oxidation with FA up-take and angiogenesis by coexpression of mitochondrial OXPHOSgenes Vegfb and Vegfa. During nutrient deprivation and/or exercise,the expression of PGC-1a induces mitochondrial biogenesis thatsubsequently upregulates expression of VEGF-B and induces FAuptake in the tissue cells and induces expression of VEGF-A with

subsequent angiogenesis. This type of regulation ensures that in-creased blood vessel growth and nutrient supply are correlated withthe higher oxidative capacity in the tissue to preserve the metabolichomeostasis. Data in B and C are mean6 SEM. Statistical analysesbetween chow-fed and HFD-fed WT are omitted for clarity. *P <0.05, **P < 0.01, HFD-fed MPGC-1aTG//Vegfb2/2 or Vegfb2/2

compared with HFD-fed WT; #P < 0.05, ##P < 0.01, HFD-fedMPGC-1aTG//Vegfb2/2 or Vegfb2/2 compared with HFD-fedMPGC-1aTG. a.u., arbitrary unit.

diabetes.diabetesjournals.org Mehlem and Associates 871

been shown to be upregulated in patients who are obese(31,32). Taken together, several human studies haveshown interesting links among VEGF-B levels, obesity,and insulin resistance.

In summary, the current study identifies Vegfb as adownstream target of the PGC-1a/ERR-a signaling path-way and a mediator of the PGC-1a–induced adverse ef-fects upon HFD feeding. We provide evidence that tissuemetabolism is intimately linked to the functional featuresof the vasculature and highlights the importance of PGC-1a as a major metabolic sensor/regulator to coordinaterespiratory capacity with vessel growth through VEGF-Aand lipid uptake through VEGF-B.

Acknowledgments. The authors thank Janet Rossant (Departments ofMolecular Genetics and Obstetrics and Gynaecology, University of Toronto) andGuo-Hua Fong (Department of Cell Biology and Center for Vascular Biology,University of Connecticut Health Center) for the gift of the MPGC-1aTG mice;Thomas Perlmann (Department of Cell and Molecular Biology, KarolinskaInstitutet) for the gifts of pCMX-ERR-a, pcDNA3.1/myc-HisA-PGC-1a-FLAG,pTK-LUC, pCMX-b-Gal plasmids; and the Swedish Metabolomics Centre(Umeå University; www.swedishmetabolomicscentre.se) for help with lipidanalysis. The authors also thank Sofia Wittgren and Karin Pettersson (Depart-ment of Medical Biochemistry and Biophysics, Karolinska Institutet) for mousecare and mouse genotyping.Funding. This study was supported by the Wilhelm och Else StockmannsStiftelse (C.E.H.), Swedish Heart-Lung Foundation (U.E.), Novo Nordisk Foundation(U.E.), Swedish Cancer Foundation (U.E.), Swedish Research Council (U.E.),Torsten Söderbergs Stiftelse (U.E.), Ragnar Söderbergs Stiftelse (U.E.), LudwigInstitute for Cancer Research (U.E.), CSL Ltd., Melbourne, Australia (U.E.), andKarolinska Institutet.Duality of Interest. A.M., U.E., and A.F. are shareholders in a companywithin the diabetes field. This study was supported by CSL Ltd., Melbourne,Australia (U.E.). U.E. is a consultant for CSL Ltd., Melbourne, Australia. This doesnot alter the author’s adherence to all policies of Diabetes. No other potentialconflicts of interest relevant to this article were reported.Author Contributions. A.M. and A.F. contributed to the study design andresearch, data analysis, and writing and review of the manuscript. I.P. performedexpression, immunohistochemistry, and lipid analyses and contributed to thereview of the manuscript. X.W. performed the in vitro analysis and contributed tothe in vivo experiments and review of the manuscript. C.E.H. initiated the projectand contributed to the review of the manuscript. U.E. designed the research andcontributed to the review of the manuscript. U.E. and A.F. are the guarantors ofthis work and, as such, had full access to all data in the study and takeresponsibility for the integrity of the data and the accuracy of the data analysis.Prior Presentation. Parts of this study were presented in abstract format 2013 Keystone Symposia on Molecular and Cellular Biology Diabetes—NewInsights into Mechanism of Disease and its Treatment, Keystone, CO, 27January–1 February 2013.

References1. Hagberg CE, Falkevall A, Wang X, et al. Vascular endothelial growth factor Bcontrols endothelial fatty acid uptake. Nature 2010;464:917–9212. Hagberg CE, Mehlem A, Falkevall A, et al. Targeting VEGF-B as a noveltreatment for insulin resistance and type 2 diabetes. Nature 2012;490:426–4303. Aase K, von Euler G, Li X, et al. Vascular endothelial growth factor-B-deficient mice display an atrial conduction defect. Circulation 2001;104:358–3644. Chan MC, Arany Z. The many roles of PGC-1a in muscle–recent devel-opments. Metabolism 2014;63:441–451

5. Lowell BB, Shulman GI. Mitochondrial dysfunction and type 2 diabetes.

Science 2005;307:384–3876. Lin J, Wu H, Tarr PT, et al. Transcriptional co-activator PGC-1 alpha drives

the formation of slow-twitch muscle fibres. Nature 2002;418:797–8017. Choi CS, Befroy DE, Codella R, et al. Paradoxical effects of increased

expression of PGC-1alpha on muscle mitochondrial function and insulin-

stimulated muscle glucose metabolism. Proc Natl Acad Sci U S A 2008;105:

19926–199318. Arany Z, Foo SY, Ma Y, et al. HIF-independent regulation of VEGF and

angiogenesis by the transcriptional coactivator PGC-1alpha. Nature 2008;451:

1008–10129. Ovcharenko I, Loots GG, Giardine BM, et al. Mulan: multiple-sequence local

alignment and visualization for studying function and evolution. Genome Res

2005;15:184–19410. Mehlem A, Hagberg CE, Muhl L, Eriksson U, Falkevall A. Imaging of neutral

lipids by oil red O for analyzing the metabolic status in health and disease. Nat

Protoc 2013;8:1149–115411. Nygren H, Seppänen-Laakso T, Castillo S, Hyötyläinen T, Oreši�c M. Liquid

chromatography-mass spectrometry (LC-MS)-based lipidomics for studies of

body fluids and tissues. Methods Mol Biol 2011;708:247–25712. Ovcharenko I, Nobrega MA, Loots GG, Stubbs L. ECR Browser: a tool for

visualizing and accessing data from comparisons of multiple vertebrate genomes.

Nucleic Acids Res 2004;32:W280–W28613. Sladek R, Bader JA, Giguère V. The orphan nuclear receptor estrogen-

related receptor alpha is a transcriptional regulator of the human medium-chain

acyl coenzyme A dehydrogenase gene. Mol Cell Biol 1997;17:5400–540914. Summermatter S, Baum O, Santos G, Hoppeler H, Handschin C. Peroxisome

proliferator-activated receptor gamma coactivator 1alpha (PGC-1alpha) promotes

skeletal muscle lipid refueling in vivo by activating de novo lipogenesis and the

pentose phosphate pathway. J Biol Chem 2010;285:32793–3280015. Pilegaard H, Neufer PD. Transcriptional regulation of pyruvate dehydroge-

nase kinase 4 in skeletal muscle during and after exercise. Proc Nutr Soc 2004;

63:221–22616. Wende AR, Huss JM, Schaeffer PJ, Giguère V, Kelly DP. PGC-1alpha co-

activates PDK4 gene expression via the orphan nuclear receptor ERRalpha: a

mechanism for transcriptional control of muscle glucose metabolism. Mol Cell

Biol 2005;25:10684–1069417. Krssak M, Petersen KF, Bergeron R, et al. Intramuscular glycogen and in-

tramyocellular lipid utilization during prolonged exercise and recovery in man: a13C and 1H nuclear magnetic resonance spectroscopy study. J Clin Endocrinol

Metab 2000;85:748–75418. Clerk LH, Rattigan S, Clark MG. Lipid infusion impairs physiologic insulin-

mediated capillary recruitment and muscle glucose uptake in vivo. Diabetes

2002;51:1138–114519. Samuel VT, Petersen KF, Shulman GI. Lipid-induced insulin resistance:

unravelling the mechanism. Lancet 2010;375:2267–227720. Majer M, Popov KM, Harris RA, Bogardus C, Prochazka M. Insulin down-

regulates pyruvate dehydrogenase kinase (PDK) mRNA: potential mechanism

contributing to increased lipid oxidation in insulin-resistant subjects. Mol Genet

Metab 1998;65:181–18621. Vind BF, Pehmøller C, Treebak JT, et al. Impaired insulin-induced site-

specific phosphorylation of TBC1 domain family, member 4 (TBC1D4) in skeletal

muscle of type 2 diabetes patients is restored by endurance exercise-training.

Diabetologia 2011;54:157–16722. Boström P, Wu J, Jedrychowski MP, et al. A PGC1-a-dependent myokine

that drives brown-fat-like development of white fat and thermogenesis. Nature

2012;481:463–46823. Chinsomboon J, Ruas J, Gupta RK, et al. The transcriptional coactivator

PGC-1alpha mediates exercise-induced angiogenesis in skeletal muscle. Proc

Natl Acad Sci U S A 2009;106:21401–21406

872 PGC-1a Regulates Vegfb Expression Diabetes Volume 65, April 2016

24. Zhang K, Lu J, Mori T, et al. Baicalin increases VEGF expression and an-giogenesis by activating the ERRalpha/PGC-1alpha pathway. Cardiovasc Res2011;89:426–43525. Wallace MA, Hock MB, Hazen BC, Kralli A, Snow RJ, Russell AP. Striatedmuscle activator of Rho signalling (STARS) is a PGC-1a/oestrogen-relatedreceptor-a target gene and is upregulated in human skeletal muscle afterendurance exercise. J Physiol 2011;589:2027–203926. Mootha VK, Lindgren CM, Eriksson KF, et al. PGC-1alpha-responsive genesinvolved in oxidative phosphorylation are coordinately downregulated in humandiabetes. Nat Genet 2003;34:267–27327. Patti ME, Butte AJ, Crunkhorn S, et al. Coordinated reduction of genes ofoxidative metabolism in humans with insulin resistance and diabetes: potentialrole of PGC1 and NRF1. Proc Natl Acad Sci U S A 2003;100:8466–847128. Sun CY, Lee CC, Hsieh MF, Chen CH, Chou KM. Clinical associ-ation of circulating VEGF-B levels with hyperlipidemia and target organ

damage in type 2 diabetic patients. J Biol Regul Homeost Agents 2014;28:225–23629. Cheng F, Zhao L, Wu Y, et al. Serum vascular endothelial growthfactor B is elevated in women with polycystic ovary syndrome and canbe decreased with metformin treatment. Clin Endocrinol (Oxf) 2016;84:386–39330. Gertow K, Bellanda M, Eriksson P, et al. Genetic and structural evaluation offatty acid transport protein-4 in relation to markers of the insulin resistancesyndrome. J Clin Endocrinol Metab 2004;89:392–39931. Gómez-Ambrosi J, Catalán V, Rodríguez A, et al. Involvement of serumvascular endothelial growth factor family members in the development of obesityin mice and humans. J Nutr Biochem 2010;21:774–78032. Gómez-Ambrosi J, Catalán V, Diez-Caballero A, et al. Gene expressionprofile of omental adipose tissue in human obesity. FASEB J 2004;18:215–217

diabetes.diabetesjournals.org Mehlem and Associates 873