Branched-Chain Amino Acid Catabolism Fuels Adipocyte Differentiation and Lipogenesis

Home , Glutamine (data page), Isoleucine (data page), Protein catabolism

UC San Diego UC San Diego Previously Published Works

Title Branched-chain amino acid catabolism fuels adipocyte differentiation and lipogenesis

Permalink https://escholarship.org/uc/item/5hd6t56w

Journal Nature Chemical Biology, 12(1)

ISSN 1552-4450

Authors Green, CR Wallace, M Divakaruni, AS et al.

Publication Date 2016

DOI 10.1038/nchembio.1961

Peer reviewed

eScholarship.org Powered by the California Digital Library University of California © 2015 Nature America, Inc. All rights reserved. A San 1 of modulation differentiation genetic 3T3-L1 Whereas during significantly increases enzymes adipocytes to thatdifferentiation duringprecipitously increases suggest consumption BCAA metabolism 3T3-L1 of profiling systems-based and oxidation substrateactivity, Enzyme defined. well not is nutrients other of those to relative metabolism TCA to acids amino these of quantitativecontribution the though homeostasis, BCAA in role a syndrome are not understood. fully changes and their ultimate consequences in the context of metabolic resistance insulin of nes) and other essential amino acids become elevated in the context downstream their acylcarniti catabolitesas BCAAs, (suchof levels plasma that suggested recently have studies metabolomics clinical genic and/or anaplerotic substrates in a number of tissues keto important are valine and isoleucine, leucine, (BCAAs) acids amino branched-chain The cells. in metabolism AcCoA to cantly contributeandnonessentialessential also aminoboth acids signifi sites these in generation AcCoA anaplerosisand fueling sources carbon primary the consideredare acids fatty and glucose and tissue, adipose and liver the in (DNL) metabolic regulation and activity inadipocytes. understandbetter to need the highlights mellitus(T2DM) diabetes functions, the increased adiposity associated with obesity and type 2 resistance insulin to contribute can ways appetite,sensitivity insulin and inflammation liver, the in metabolism acid fatty influence adipokines signaling paracrine and endocrine through functions regulatory important Adipocytes, the major cellular constituent of adipose tissue, execute role inadipocyte differentiation. results These adipogenesis. quantitatively highlight compromised the contribution catabolism of BCAAs to BCAA adipocyte metabolism and of suggest that inhibition BCAA Finally,catabolism has a mitochondria. functional entering substrates of ance accumulation and odd-chain fatty acid synthesis. Vitamin B12 supplementation reduced these metabolites and altered the bal catabolism accounted for as much as 30% of lipogenic AcCoA pools. Medium cobalamin deficiency caused methylmalonic acid increased branched-chain aminoacid(BCAA) catabolicfluxsuchthatleucineandisoleucinefrom mediumand/or from protein showed (AcCoA)adipocytes differentiatedA generation,acetyl–coenzyme for glutamine and glucose use which cells, erating to contributionssubstratesthe different adipocytesquantify of totricarboxylic (TCA)acid differentiated lipogenesis.and metabolism contrast In toprolif and pre-adipocytes to tracing isotope applied we Here adipocytes. in metabolism acid amino of regulation the about known is less but homeostasis, carbohydratelipid regulating and in roles playsimportant tissue Adipose T Courtney adipocyte differentiation andlipogenesis Branched-chain aminoacidcatabolismfuels nature CH *e-mail: of D heodore epartment of epartment C Several past studies provide evidence that adipose tissue plays tissue adipose that evidence provide studies past Several and utilization glucose stimulates Insulin alifornia, San D iego, tss i te trg o ecs ntins n ii droplets lipolysis. through lipidsubstrates bioenergetic of in release the nutrientsand excess of storage the via stasis homeo lipid and glucose in role major a has tissue dipose [email protected] 1 E Fr xml, h snhss n rlae f iis and lipids of release and synthesis the example, For . L MIC a Jolla, R 12– A P B L BIOLOGY Green ioengineering, 1 D Ciaraldi 4 . In addition, the transcription of BCAA catabolic BCAA oftranscription the addition, In . iego, C alifornia, 9– 1 L 1 a Jolla, . However, the mechanisms leading to these these to However,leading mechanisms. the

| Adv 1 , Martina Wallace, Martina 3 U , 4 U SA. a C &Christian MMetallo niversity of nce online public alifornia, 3 V 6 eterans Affairs San Byn croyrts n fat, and carbohydrates Beyond . Bcat2 2– 4 . Dysfunction in these path these in Dysfunction . U n ie les circulating alters mice in C SA. alifornia, San 5 . Beyond these signaling these Beyond . 5 I nstitute of a e novo de tion | 1 , D A www.nature.com/naturechemicalbiology iego Healthcare System, San D jit E iego, lipogenesis ngineering inMedicine, S 7 , 8

. In fact, D L 1 a Jolla, , 5 ivakaruni * 1 5 - - - - - .

C pu alifornia, bli Fig. 1b Fig. [U- with cultured cells 3T3-L1 ated differenti and proliferating in extracts) nonpolar saponified from differentiation of ( induction after d 7 accumulation lipid stantial sub displayed adipocytes 3T3-L1 differentiation. after and before changes duringadipogenesis, wequantified sources the of lipogenic utilization AcCoA substrate mitochondrial how understand To Adipogenesis reprograms amino acidmetabolism RESULTS cellular between interplay complex differentiation and metabolic pathway biology. inadipocyte flux the highlight results These adipogenesis. 3T3-L1 influenced negatively catabolism BCAA of (OCFAs).Finally,acids inhibitionfatty odd-chain of synthesis (MMA) and methylmalonate of to accumulation led non-physiological and metabolism the acid fatty and BCAA perturbed cultures 3T3-L1 in availability cobalamin inadequate Furthermore, tissues. isolatedadipocytes from human subcutaneous and adipose visceral in as well as differentiatedadipocytes terminally 3T3-L1 in AcCoA mitochondrial the of one-third as much as for accounted BCAAs substrates, these catabolizeappreciably not did pre-adipocytes ing proliferat Whereas adipocytes. differentiated and pre-adipocytes analysisspectral (ISA) to quantify BCAA utilization in proliferating not has employed13 DNL we Here adipocytes. and in determined quantitatively anaplerosis been to contributes substrates, other context inthe effects ofbeneficial T2DM. tissue adipose receptor- proliferator-activated peroxisome(PPAR of clinically activators (TZDs), used thiazolidinediones with animals and subjects BCAAs plasma normalizes mice levels, transplantationBCAA tissue adipose of to wild-type Supplementary Results Supplementary -aee iooe rcr, as pcrmty n isotopomer and spectrometry mass tracers, isotope C-labeled s o ae te xet o hc oiain f CA, eaie to relative BCAAs, of oxidation which to extent the date, To h 2 ed , U S γ U ) had the expected pattern of labeling arising via pyruvate via arising labeling of pattern expected the had ) , nrae te rncito o BA ctblc ee in genes catabolic BCAA of transcription the increases ), SA. o D usan niversity of n iego, li 2 D ne: 16

epartment of epartment 19 C , A 2 alifornia, 0 , suggesting that this metabolic activity may have may activity metabolic this that suggesting ,

P C No alifornia, San hillips ve U mb , SA. P SupplementaryFig.1a harmacology, er 2015 | 3 4 , D 4 , epartment ofMedicine, epartment D A 16– iego, nne 1 8

d Fnly tetet f human of treatment Finally, . 13 L o C a Jolla, U i 6 N : niversity of guoe ( ]glucose

1 0 Murphy . 1 article 0 C 3 alifornia, 8 / ). Palmitate). (total n C Supplementary Supplementary c alifornia, h U e niversity 2 m U , SA. b i o Bcat2 .

1 9

6 1 −/− γ - - - - - 1

© 2015 Nature America, Inc. All rights reserved. article two-tailed asterisks represent significant differences between groups by Student’s representative of three biological replicates (independent experiments); from three technical replicates (separate wells in cell culture) interval ( acid tracers. an M1labelindicatingtransamination from theindicated [ pre-adipocytes andadipocytes. ( and adipocytes. ( newly synthesized (labeled)from [ ( ( A549, HuH-7 cancer cells and 3 [ adipocyte differentiation. Figure 1 | Characterization of metabolic reprogramming during nonessential amino acids initially present at high levels in the culture proliferation,of lack Despite their several type. cell each in thesized measure the percentage of each amino acid pool that was newly syn cells proliferatingrapidly of hallmark up—a taken glucose of mole per ( stage pre-adipocyte the at those to compared as cells 3T3-L1 tiated tion and glutamine uptake were all significantly reduced in differen aftercally differentiation ( dramati increased palmitate to glutamine and glucose from flux lipogenic absolute expected, As glutamine. and glucose than other substrates from arose a carbon lipid of cells fraction differentiatedgreater substantially in cycle, cell active an with cells to contrast for ( origins tissue various of lines cell cancer in results to compared and adipocytes to [U- of contributions relative the quantify to employed was ISA glucose. labeled with observed [U- two (i.e., containing AcCoAAcCoA M2 of generation (PDH)-mediated dehydrogenase 2 3T3-L1 adipocytes, including glutamine, glycine differentiatedand serine in ( rates high strikingly at synthesized were medium c b Fig.1 U a d

) ) - AA uptake and secretion Contribution to 13 When normalized to protein levels, glucose uptake, lactate secre P e novo de U 13 C ≥ ercentage of intracellular glutamine, serine and glycine pools that were 2 ﬂuxes (nmol/h/mg protein) lipogenic AcCoA (%) ptake andsecretion fluxes in3 1 C 100 200 250 5 0 o te ioei ACA n l poieaig el. In cells. proliferating all in AcCoA lipogenic the of 80% 40 100 80 60 . Importantly, the use of stable isotope tracers allowed us to us allowed tracers isotope stable of use Importantly, the . 20 150 ]glutamine to lipogenic Ac 50 0 b 5 0 guaie ( ]glutamine CI ). Proliferating). pre-adipocytesmoresecreted3T3-L1 lactate Leucine 143B t *** ), andthosein 3T3-L1 pre-adipocytes [U- [U- Other -test: * Isoleucine D amtt snhss n T-1 r-dpcts versus pre-adipocytes 3T3-L1 in synthesis palmitate A549 13 13 Pre-adipocytes Adipocytes Uptake ata presented in C C *** HuH-7 3T3-L1 adipocytes 6 5 ]Glutamine ]Glucose Fig. 1 Fig. d P

) Valine *** <0.05, ** N et amino acid uptake and secretion in 3

upeetr Fg 1c Fig. Supplementary Serine a ). Notably, glucose and glutamine accounted glutamine andNotably, glucose ). b

( Alanine Secretion – b a *** e 13 Uptake and secretion ﬂuxes ) Supplementary Fig. 1d are mean P Glucose uptake iooe) Iooi ercmn from enrichment Isotopic isotopes). C

C ( mol h mg protein) <0.01, *** µ / / a Glycine T 0 4 2 3 ontribution of[ 1 are modeloutput *** e 3-

C Lactate secretion ) 13 U oA for palmitate synthesis in 143 L P T Adipocytes Pre-adipocytes C * 1 pre-adipocytes and adipocytes. - ercentage of amino acids that contain 3-

13 Glutamine uptake 6 guoe n [U- and ]glucose C L ** ± e 6 1 pre-adipocytes andadipocytes. Glutamate secretion ]glucose in 3

s.d. M1 label on AAs from P

<0.001. speciﬁed tracers (%) ** 40 60 20 0 D U ata shown in Serine - 13 ws oe ta that than lower was ) ± C 95% confidence

6 Glycine c ]glucose and T Total enrichment 3- 13 ). Aspartate from [U- C6]glucose (%) L 40 80 60 20 15 1 pre-adipocytes 0 N 13 Glutamine a C T ]amino nature

– Glutamate [ [ [ 3- *** Adipocytes Pre-adipocytes 5 15 15 15 e ]glutamine N]BCAAs N]Valine N]Leucine are L Glycine 1 *** Fig. 1

Glutamine B

, ch

Serine *** e c mic N ), ), - - - -

ature a l

nitd dpcts ( adipocytes entiated and isoleucine were readily incorporated into citrate pools in differ subsequently is leucine fromarising molecules AcCoA succinyl-CoA. convertedto which (PropCoA), propionyl-CoA M3 one and [U- and lism, [U- from produced are molecules AcCoA M1 catabolism, BCAA see of map detailed more a For intermediates. mitochondrial in labeling resultant and catabolism BCAA with associated sitions cultures. 3T3-L1 to adipocytes, we applied cultured [U- in lipogenesis and metabolism mitochondrial to tribute con isoleucine or leucine which to extent the quantify better To BCAAs fuelTCA metabolismandlipogenesis inadipocytes acid synthesis indifferentiated ( adipocytes amino tocontributed substrates these from nitrogen that indicated vations ( pre-adipocytes to pared com as adipocytes 3T3-L1 differentiated in dramatically increased quantitation of amino acid fluxes indicated thatIndeed, BCAA consumptionreactions. these for nitrogen necessary the provide to tion differentia upon increased acids amino other consumptionof that atoccurring significant rates in differentiated cells, we hypothesized that uh ht ltmn ws ertd no h clue eim at medium 1e Fig. culture ( consumed) the being still into (while rate considerable secreted a was glutamine that such synthesized, newly was pool the of glu 20% approximatelyglutamine, synthesized tested cells proliferating tamine the of none whereas [ substrate in3 isoleucine. ( ( dehydrogenase; Ac each ( Figure 2|BCAA catabolismisinitiated uponadipocyte differentiation. A asterisks represent significantdifferences between groups by two-way representative ofthree biologicalreplicates andrepresent mean tissue ( and adipocytes isolated from subcutaneous(SA biology b a U c a NOV ) Summaryof ) - CO Citrate MPE from CO 13 C L C -Valine 2 [U- indicated tracer substrate (%) 2 BC itrate labelingin3 e novo de 40 80 60 13 20 6 A: * 0 ]leucine and[

O C ), in line with results from studies in adipose tissue adipose in studies from results with line in ), 1 Bckdh AA tracer. Abbreviations: PropCoA

de novo de 6

3 ]Glucose | Adv A chemical . Furthermore, metabolic tracing using tracing Furthermore,metabolic . P

T [U-

< 0.05, ** 13 ) depots. c

) Molepercent enrichment(M C T a 5 ]Glutamine CO ytei o guaie n ohr mn ais was acids amino other and glutamine of synthesis 3- L nce online public -Isoleucine BC 13 2 L . After 24 h of culture in medium containing 4 mM mM 4 containingmedium in culture of h 24 After . Supplementary Figure 2a C 1 pre-adipocytes andadipocytes. ( [U- C AA catabolismandcarbon-atom transitions from 13 Bckdh 6 oA, acetyl- PropCoA AcCoA ioecn ctbls yed oe 2 AcCoA M2 one yields catabolism ]isoleucine U C D P 6 ]Leucine - < 0.01, *** Figure 2 Figure Adipocytes Pre-adipocytes ata in T 13 [U- 3- C i. 2 Fig. 13

6 C L ]isoleucine inprimaryhumanpre-adipocytes CO

6 1 adipocytes from [ ]Isoleucine L bi -Leucine Fig. 1 Fig. b 2 13 – C C d o Bckdh b are from three technical replicates a AcCoA CO oA; 6 a P . o opr cags eoe and before changes compare To ). B ]leucine or [U- summarizes the carbon atom tran atom carbon the summarizes l tion | < 0.001. ckdh, branched-chain ketoacid 2 d ogy P ), consistent with previous obser previous with consistent ), rop b d

www.nature.com/naturechemicalbiology Citrate labeling from

C Citrate MPE from indicated tracer substrate (%) oA, propionyl- 100 PE indicated tracer substrate (%) 80 90 d 10 10 0 5 0 4 8 6 2 oi ) ofcitrate from eachtracer . Two M2 AcCoA and one

U [U- T

Fig. 1 13 : 10.1038/n - M0 Number ofisotopespermolecule ) oromentaladipose

13 C OAT adipocytes OAT pre-adipocytes SAT adipocytes SA 13 C 6 ]Leucine 6 * C T d

13 M1 ]leucine and[ 15 pre-adipocyte ) M 6 C e ]isoleucine tracers N-labeled BCAAs BCAAs N-labeled ** ).

6 M2 ]leucine catabo ]leucine Supplementary Supplementary PE C oA. [U- [U- ofcitrate from [U-

M3 ch 13

C 13 13 s

C C 6 ]Isoleucine e

± M4 6 6 mbio ]Isoleucine ]Leucine s.d; 2 * 2 U . Given Given . M5 - *** 13 .196 C M6 6

] ------ 1

© 2015 Nature America, Inc. All rights reserved. N of three biologicalreplicates. * D labeling over thecourse oftheexperiment. to tracer dilution. palmitate from [ to therespirometry medium.( malate; K permeabilized 3 Figure 3|BCAA catabolismfuelsmitochondrial metabolismandlipogenesis inadipocytes. infcnl uo dfeetain ( differentiation upon significantly The contributions of both glucose and glutamine to AcCoA decreased [U- from arising AcCoA labeled partially foraccount to model ISA our [U- and leucine [U- of contributions tive ( lipogenicAcCoApool ment the in facilitatesquantitation which enrich isotopeDNL,of of levels high ( BCKAs oxidize to capacity increased the explain entirely not does biogenesis mitochondrial differentiation-induced substrates, other that for suggesting observed was than increase greater much a differentiation,after 25-fold to 14- by increased BCKAs these dize ( evident was (KIV) ketoisovalerate and (KMV) ketomethylvalerate (KIC), ketoisocaproate products reaction BCAT2 the of oxidation the in increase a disproportionate but tested, substrates all from consumption oxygen in increase respiration of adipocytes and pre-adipocytes, we observed a marked adipocytes meabilized per of analysis respirometry via obtained was (BCKAs) ketoacids tive capacity of cellsthese for each of the respective branched-chain than human pre-adipocytes ( rates higher at BCAAs metabolize also adipocytes human that ing suggest cells, 3T3-L1 in observed those to similar were adipocytes differentiated to compared as pre-adipocytes human in isoleucine [U- both from MPE citrate comparing adipogenesis also human catabolism during BCAA of occurs induction omental whether observe or to subcutaneous depots from isolated pre-adipocytes human aboutbeginning 2dinto differentiation ( levels of 3.0%, leucine and isoleucine enrichment ( and 3.3% to ( respectively differentiation adipocyte increased upon contribution considerably their pre-adipocytes, in detectable barely was citrateinto isoleucine and leucine fromatoms carbon of ration ( measured was citratepool the in (MPE)enrichmentpercent steady-statemole [U- of presence [U- the in cultured were d 6–7 for ated differenti those and pre-adipocytes 3T3-L1 differentiation, after nature CH respectively,adipocytes. 9%, in and 23% to pre-adipocytes in levels undetectable from increased contributions isoleucine and leucine Supplementary Fig. 3a ata in ature In contrast to cultured human adipocytes, 3T3-L1 cells exhibit cells 3T3-L1 adipocytes, human cultured to contrast In oxida the of evidence enrichment, atom carbon to addition In in catabolism BCAA of extent relative the quantified also We 13 13 C C 5 6 a guaie [U- ]glutamine, ]leucine catabolism ( catabolism ]leucine represent atleastfive biologicalreplicates eachwithfour technical replicates, whilethosein IV E , ketoisovalerate; MIC Fig. 2 Fig.

Fig. 2 chemical A T U a L BIOLOGY 3- BC 13 - Substrate-speciﬁc OCR 13 c C L C and and (pmol O /min/µg total AA contributions were adjusted to account for dilutionofintracellular aminoacidsfrom protein turnover usingtheaverage 1 pre-adipocytes andadipocytes. Substrate key: M,malate; K 2 c 6 5 ]isoleucine to this metabolic pool, we modified modified we pool, metabolic this to ]isoleucine ). Other TCA). Other intermediates showed comparable cell protein) ]valine. ( 20 25 10 15 0 5 26 Supplementary Fig. 2b Fig. Supplementary 13

, 2 | Adv ). 7 C M . When comparing uncoupler-stimulatedcomparing When . 13 P KIC-M 6 Adipocytes Pre-adipocytes -M, pyruvate andmalate; S-R,succinate androtenone; G-M,glutamate andmalate. 250nMAdo b P C luie r [U- or ]leucine * c ) <0.05, ** Fig. 2 Fig. a 6 )

Fig. 2 KMV-M

guoe [U- ]glucose, P nce online public bi C almitate labelingin3 * ontribution ofeachtracer substrate to lipogenic Ac o KIV * a d l Fig. 3 Fig. and and i. 3 Fig. ). i. 3 Fig. ogy P 23–

P-M ** <0.01 by Student’s two-tailed D 2 ata presented in 5 Supplementary Fig. 2d . Trends observed when when observed Trends . ** 13

Supplementary Table 1 Table Supplementary S-R b C c a 13 ). To). rela the determine Supplementary Fig. 2c 13 . n h ohr hand, other the On ). . h cpct t oxi to capacity The ).

6 C G-M d C luie n [U- and ]leucine 5 oi a ). Although incorpo Although ). 6 guaie [U- ]glutamine, ]isoleucine and the the and ]isoleucine tion | : 10.1038/n T b

3- Palmitate labeling from

L indicated tracer substrate (%) www.nature.com/naturechemicalbiology 10 1 adipocytes from [ 70 85 10 0 0 5 13 a are mean C M0 Number ofisotopespermolecule 6 ]glucose, ch M2 e 13 13 mbio [U- [U- ). C C M4 ± 13 13 6 6 ). ). ), ), t ------] ] s.e.m., datain C -test. C

6 6 ]Isoleucine ]Leucine M6 .196 U increased 3 d after induction and continued until 6 d after induc after d 6 until continued and induction after d 3 increased dramatically utilization BCAA toward reprogramming metabolic Finally,this differentiation. adipocyte upon observed we oxidation These results are consistent with the pronounced induction of BCAA aeig f ltmn fo guoe ( glucose from glutamine of labeling increased by for accounted entirely almost be could contribution BCAAs ( of decreased contribution acids fatty relative to the while glucose increased, of AcCoA contribution to the that indicated tracers specific using ( unchanged were rates synthesis and uptake acid ( secretion amino and glycolysis in observed reduction we significant media, a gluc+AA Low versus control in adipocytes for complete formulation). In comparing the metabolism of 3T3-L1 200 and tamine glu mM full 1 glucose, mM the 6 gluc+AA’—forexample, ‘Low levels and (termed physiological more glucose at present with acids amino medium of complement in adipocytes cultured we catabolism, BCAA ofextent the influences cultures 3T3-L1 in cose glu and acids amino of concentrations high the whether test To over are DMEM) is, four-fold higher (thatthan those measured in medium human and mouse adipocyte plasma standard in BCAAs levels and glucose of case the in particularly pathways, signaling cell of regulation the or action mass through activity metabolic influence can concentrations nutrient Elevated Protein BCAA metabolism catabolismsupports palmitate and other fatty acids ( ( tion active cells and tissues. metabolically in recycling acid amino of importance the highlights and catabolism synthesisprotein of regulationcoordinate the given surprising not Though differentiatedby adipocytes. utilized arecatabolismhighly proteinand sources extracellular both from particular) in (BCAAs acids amino that evidence provide data However, these culture. of stronglywouldendprotein induce which the atcatabolism valine), fasting from plasma in subjects human observed those below well were valine) (25 BCAA each of tions ( unlabeled became pool acids amino present): within 24 h, approximately essential 50% of the intracellular BCAA of source other only (the turnover protein unlabeled from dilution by affected strongly label was pools BCAA-derivedAcCoA in in decrease the importantly, more Perhaps - 1

IC M8 13 C -M, ketoisocaproate andmalate; KM 6 ]leucine and[

Supplementary Fig. 3b Fig.Supplementary M1

C 0 oA in3 b are mean Fig. 4a Fig. Contribution to c (

a lipogenic AcCoA (%) T 100 ) Substrate-specific oxygen consumption rate ( 40 80 60 20 3 3- 0 2 0 , the marked dilution of [ 9

μ Pre-adipocytes (125 L U , b M for each BCAA; see see BCAA; each for M 1 pre-adipocytes correction andadipocytes after due - ± ). On the other hand, total fatty acid pools and pools acid fatty total hand, other the On ). 13 b s.d.anddatain C and 6 ]isoleucine. Minimallabelwas detected in μ M leucine, 75 leucine, M μ c i. 4 Fig. M leucine, 40 leucine, M i. 4 Fig. are three technical replicates representative

), coinciding with the accumulation coincidingthe of with), Adipocytes Supplementary Fig. 3c c d ). Notably, the increased glucose glucose increased the Notably, ). . oal, h fnl concentra final the Notably, ). upeetr Fg 4a,b Fig. Supplementary c are modeloutput V * * * * -M, ketomethylvalerate and μ upeetr Fg 4c Fig. Supplementary M isoleucine and 200 and isoleucine M [U- [U- [U- Other [U- μ 13 M isoleucine and 75 and isoleucine M C Supplementary Table 2 Table Supplementary C]BCAAs observed C]BCAAs hereobserved 13 13 13 13 bl andbiotinwere added C C C C 6 5 6 6 ]Glutamine ]Glucose ]Leucine ]Isoleucine 2 article 8 . Indeed, BCAA BCAA Indeed, . ± 95% ). OC BC CI R) in AA . . ISA ).

μ μ M M ). 9 3 - - - - .

© 2015 Nature America, Inc. All rights reserved. article by Student’s two-tailed representative ofthree biologicalreplicates; ** output 24 h. cultured inindicated tracer substrate incontrol and glucose-derived synthesis. ( the course oftheexperiment. Glutaminedilutionoccurred primarilyvia amino acidsfrom protein turnover usingtheaverage BC Ac and and ( Figure 4|BCAA by utilization issupported protein catabolism. after culture with these BCAA tracers ( tracers afterBCAA culture these with peak M3 prominent a contained OCFAs also C17:0 and C15:0 the architecture.pathwaySimilarly,known of basis the on expected as with [U- remained unclear has origin their but cells, 3T3-L1 using reports other in observed ( species C17:0 and C15:0 including pre-adipocytes, not but adipocytes 3T3-L1 5a Fig. present surprisingly ( BCAAs of was those to similar levels intracellular which at cells, 3T3-L1 differentiated (MMA) in acid methylmalonic of accumulation substantial sured metabolite from [U- ( cycle TCA 2a the Fig. enter not did isoleucine from derived (MMA-CoA) CoA methylmalonyl and PropCoA that ing turing differentiated 3T3-L1 cells with [U- ( Intriguingly,prominent ofcitrateobserve labelingto M3 failed we BCAA andlipidmetabolism Medium B12deficiencyaffects 4 via than rather DNL derived Fig. 5c a Fig. 2 Fig. c a

) C AA contributions were adjusted to account for dilutionofintracellular Contribution to Uptake and secretion ﬂuxes Distinct M3 labeling was observed inMMAafter 24hofDistinct M3labeling culture was observed U L L oA in3 ptake andsecretion fluxes in3 ow gluc+AAmedia.( ow gluc+AAmedia.(

D lipogenic AcCoA (%) (nmol/h/well) 400 200 300 100 100 40 ata presented in ± 80 60 20 Glucose uptake ), providing evidence that these fatty acids arose from BCAA- b 0 0 ). We also failed to detect detect to failed also We ). 95% . utemr, ih ees f CA acmltd in accumulated OCFAs of levels high Furthermore, ). ) or other TCA intermediates (data not shown) when cul when shown)not intermediates (dataTCA other or ) 13 C T 3- 6 Control *** ]isoleucine or [U- [U- [U- [U- [U- Other CI

L Lactate secretion 1 adipocytes cultured incontrol and 13 13 13 13 . C C C C D i. 5 Fig. 5 6 6 6 ]Glutamine ata shown in ]Glucose ]Leucine ]Isoleucine 3 1 Low gluc+AA Control . ** 13 t

a Low gluc+AA Glutamine uptake -test. C , a b 5 ). These fatty acids have previously been been previously have acids fatty These ). c b , ]valine ]valine (data not shown). Notably, we mea d ) ) Aminoaciduptake andsecretion incontrol d are mean C ) ontribution ofeachtracer to lipogenic P *** ercentage of poolwithoutlabelwhen a α 13 – d C -oxidation. These data suggest that suggest data These -oxidation. T are from three technical replicates 3- 5 ]valine ]valine ( L ± d b 13 1 adipocytes cultured incontrol s.d.;datain AA metabolism icroain n n TCA any in incorporation C Dilution of amino acid (nmol/h/well) Fig.5 40 20 50 30 tracer from protein (%) 10 0 40 60 20 Low gluc+AA Control P

0 Leucine Supplementary Fig. 5b <0.01, *** [U- 13 13 ***

C Isoleucine

b C 6 Uptake L and and 6 ]Leucine L ]isoleucine, suggest BC *** ow gluc+AAfor c ow gluc+AAmedia. are model ***

AA labelingover [U- Valine Supplementary Supplementary Supplementary Supplementary Supplementary 13 Low gluc+AA Control ** P C <0.001 6 ]Isoleucine Serine nature **

Secretion Alanine

***

Glycine ch e mic N ), - - -

ature a l

mulation of these metabolites ( tion in human plasma) to 500 nM (DMEM/F12) prevented the accu approximate (the concentranM 2 from ranging levels at media to we OCFAs.andcobalamin ofadditionMMA Indeed,accumulationof metabolism, the this to led availability support cobalamin in to deficiencies that hypothesized sufficient be not might serum untreated or dialyzed 10% with supplementation and cobalamin, lacking DMEM employconditions culture 3T3-L1 typical Because cultures. these in undetectable were MMA and OCFAs both and cobalamin, nM 500 contains which DMEM/F12, in cultured were (AdoCbl) to MMA-CoA 5 of coenzyme conversion the requires the succinyl-CoA as metabolism, cobalamin in MMA-CoA( mutaseencoding gene the in mutations by caused be may phenotype This aciduria. ylmalonic presumablycytes, catabolism. due to inBCAA defects adipo 3T3-L1 differentiated in accumulate PropCoA and MMA TCA metabolism. For example, addition of cobalamin from day 0 day from cobalamin of addition example, For metabolism. TCA to BCAAs contributionof the as well as species acid fatty of dance 3T3-L1 culture medium ( We obtained similar results when we added cobalamin or AdoCbl to three technical replicates representative ofthree biologicalreplicates. + cobalamin condition. condition by Student’s two-tailed acids in control and + cobalamin conditions. *** multiple-comparison test. ( compared to control condition by two-way A cultureafter with500nMcobalamin or100nMAdo beginning onday 0ofdifferentiation. ( cells cultured in isoleucine and[ 3 3T3-L1 adipocytes. Figure 5 | BCAAs contribute to MMA, biology e c a T

3- Relative abundance MMA abundance Lipid accumulation in MMA accumulates in patients with the genetic disorder meth disorder genetic the with patients in accumulates MMA upeetto o cblmn nlecd h rltv abun relative the influenced cobalamin of Supplementation

L of each fatty acid (%) (nmol/well) di erentiating 3T3-L1 1 adipocytes 0–7 induction.( dafter 0.4 4.0 0.8 0.6 0.2 5.0 4.5 40 1.0 80 60 20 adipocytes (nmol/well) 0 0 40 60 20

| Adv 0 01 C14:0 chemical 3 Cobalamin concentration(nM) 2 . Notably, the human adipocytes analyzed in this study this in analyzed adipocytes human Notably, the . 01 a C15:0 00 *** nce online public U D Days postinduction - M 13

C16:0 200 23 *** C ( E a 5 M +10%F ]valine in3 )

D C17:0 *** O C17:0 C15:0 300 ata presented in dd-chain fatty acid( 45 e + cobalamin C18:1 Control

) Relative abundance of the most abundant fatty 400 bi Supplementary Fig. 5e

B C18:0 67 o T S and0, 2,8,32,125or500nMcobalamin t -test. ( 3- a 500 Fig. 5 l tion | L ogy 1 adipocytes. ( d O ) CFAs and BCFAs in differentiated f OC f d b b a ) www.nature.com/naturechemicalbiology c ) – and and U BCAA uptake OCFA abundance C17:0 labeling from f C FA levels in3 NOV are mean ptake of

(nmol/well) OC (nmol/h/well) indicated tracer substrate (%) d 17:0 labelingfrom [ 100 100 ′ 40 80 60 80 60 20 30 20 -deoxyadenosylcobalamin 10 10 P 15 oi 0 0 0 5 < 0.001 compared to control Supplementary Fig. 5d FA) accumulation in MUT A andHolm-Sidak’s : 10.1038/n Leucine Number ofisotopespermolecule M0 c [U- [U- C ) MMA abundance in PropCoA BC M1 bl. *** 13 C15:0 13 ± ) or genes involvedgenesor ) *** C C AAs incontrol and and s.d.andrepresent M2 6 5 T ]Valine Isoleucine ]Isoleucine *** 3- Control + cobalamin M3 P L <0.001 ch Fig. 5 1 adipocytes 1 adipocytes M4 U M5 e Control + AdoCbl + cobalamin mbio - C17:0 13

Valine *** M6 d C ). 6 **

] M7 .196 * M8 ). - - - - - 1 © 2015 Nature America, Inc. All rights reserved. N ( [U- of ( trols uptakeBCAA were significantly also lower in KD cells than in con ( cells control than levels protein Bckdha differentiation, After markers. differentiation and characterized morphology metabolism, each their to respect and with rosiglitazone and cobalamin of the absence in differentiation to subjected were cells these confluence At (control). sequences nonspecific or sequences) unique two of one RNAshairpin (shRNAs) targeting polyclonal populations of 3T3-L1 cells that stably express generated we either question, this Toshort addressadipogenesis. in role tional func a plays pathway this whether examined next we adipocytes, differentiated in observed BCAAs of flux catabolic high the Given BCAA metabolismcontributes to adipogenesis in metabolism cultured adipocytes. mitochondrial and acid fatty BCAA, perturbs ity availabil cobalamin inadequate to due PropCoA,precursor, their presumablyOCFAsand MMA, of accumulation that indicate data ( cobalamin nM 500 with cultured [U- adipocytes 3T3-L1 in intermediates fromTCA all across decreased enrichment cose hand, other of the regulation On broader pathway. in the results OCFAs and MMA of mulation ( medium cobalamin-supplemented in cultured when includ leucine, BCAAs, ing more up took adipocytes 3T3-L1 Furthermore, [U- from intermediates denced by increased enrichment of fumarate, malate and other TCA supplementationcobalamin promoted evi anaplerosisas valine, of relative the ringin amount of palmitate ( the total intracellular fatty acids, with a concomitant increase occur of percentage a as levelsOCFA reduced differentiation of 7 day to A replicates (independentlentiviral knockdown anddifferentiation suchthatK * From control left: K acids. ( control K hairpins) normalized to nuclearquantitation. Full gel isshown in Figure 6|InhibitionofBCAA catabolismimpairs adipocyte differentiation. nature CH sized synthe acids fatty including cells, KD differentiated in decreased Fig.6 5 Fig. P ad Bckdha β NOV f <0.05, ** -actin Relative expression ature 0.5 1.0 1.5 0 de novo de A withHolm-Sidak’s multiple-comparison test. i. 6b Fig.

f Control KD 13 d f ) Quantitative , ugsig ht oaai dfcec ado te accu the and/or deficiency cobalamin that suggesting ), C ). Furthermore,). abundance the ofnumerous fatty was acids D Bckdha E and 6 luie o at ais a mrel lwr n D cells KD in lower markedly was acids fatty to ]leucine MIC

† Bckdha P <0.01, †

chemical as well as those taken up from the medium ( medium the fromup taken those as well as A , Bckdha

c KD1

L BIOLOGY Pparg ). Consistent with this observation, lipogenic flux flux lipogenic observation, this with Consistent ). Bckdha † D Supplementary Figure 6 , † Bckdha P

K KD2 PC

<0.001. Glut4 D R analysis ofadipocyte-specific gene expression. ( † adipocytes. (

| Adv * b 13 K

C Mitochondrial respiration D Plin4 5 a 1, † (pmol O /min) vln ( ]valine

U 2

** nce online public bi 200 100 Bckdha 150 nless otherwisespecified,dataare from three technical replicates representative ofthree ( 50 0

Bckdha Adipoq Control KD o Bckdha † d Supplementary Fig. 5i Fig. Supplementary Bckdha l ) Absolute lipogenic fluxofleucineto palmitate synthesis. ( K ogy upeetr Fg 5f Fig. Supplementary D

. ( Bckdha KD1 D el ehbtd lower exhibited cells KD 2.

Lep † D Control KD Bckdha Bckdha Fig.5 † b ata are from four technical replicates representative ofthree biologicalreplicates. ( Fig. 6 Fig. (knockdown (KD) using (knockdown(KD) D ) † KD2 ** B ata presented in

d asal respiration incontrol and KD2 KD1 oi a e tion | ). At). same the time, a ce : 10.1038/n

). Respiration and Respiration ). BCAA uptake g (nmol/h/well) 40 80 60 www.nature.com/naturechemicalbiology 20 0

Leucine b 13 † Fig. 6 Fig. ch ). These These ). – C † f are mean 6

Isoleucine e ]glu mbio – h e † Bckdha Bckdha Control KD ). ). ------†

g D .196 ( ) Representative images ofadipocyte differentiation (scale bars, 200 1 andK

a Valine eedn idcin f CA aaoim n T-1 el and cells 3T3-L1 in catabolism BCAA of induction dependent sensitive to leucine availability in particular (mTOR) rapamycin activity and associated of protein translation required for proliferation target is Mammalian biosynthesis. for nutrients protein these of utilization dedicated the with consistent is cultured inlow-glutamine medium adipo differentiated in pancreaticobtainedwith cellscancer those result to a similar cytes, DNL) (and metabolism mitochondrial for source carbon significant a as serve acids aminoproteinogenic that indicate data These adipocytes. differentiated turnover in observed was protein from arising dilution significant as protein catabolism, via sustained was catabolism this BCAAs, of centrations con physiologicalmore of presence the in Even AcCoA. lipogenic of one-third approximately for accounted isoleucine and leucine that such BCAAs catabolized adipocytes differentiated time, same significant while version of glucose and glutamine carbon to fatty acids was reduced, con relative the differentiation, Upon proliferating. when glu tamine and glucose from DNL for AcCoA obtained predominantly pre-adipocytes, subcutaneous and visceral human including gins, differentially ori tissue diverse nutrientsof lines cell mouse and Human specific differentiation. how explored adipocyte of function a as contributemetabolismmitochondrial to have we Here DISCUSSI of upstream flux pathway or MMA-CoA mutase contributes to adipogenesis. metabolite some that suggesting shown), not (data differentiation adipogenic affect observably not ( formation droplet lipid decreased by differentiation, adipogenicevidenced asaffected that suggested also cultures differentiated of images Representative adipogenesis. to contributes catabolism ( cells control the in than Plin4 Finally, the transcription of selected adipogenic genes ( ± KD2 KD1 ) Western blotof * Bckdha s.d.Asterisks in 1 The suppression of BCAA catabolism during cell proliferation cell during catabolism BCAA of suppression The D , Adipoq 2 were eachderived two orthree times) analyzed by two-way K Absolute lipogenic D ﬂux to palmitate (nmol adipocytes (two cell linesseparately derived usingunique

O of leucine/h/well) 20 30

10 and and N Control KD 0 B b Bckdha ckdha and – de novo de Lep e e

represent significance compared to control K Bckdha ) Relative abundance ofthemostabundantfatty KD1 i. 6 Fig. † ) was significantly lower in the KD conditionKD the lowerin significantly was ) KD2 ** glutamine synthesis was initiated. At the the At initiated. was synthesis glutamine i. 6 Fig. β g -actin indifferentiated 3 . upeetto o cblmn did cobalamin of Supplementation ). Relative FA abundance 100 150 50 0 f ), further suggesting that BCAA BCAA that suggesting further ),

3 C14:0 3 . † † Control KD Bckdha Bckdha

a C15:0 – † e 28 † KD1 KD2 ) ortwo ( Bckdha , c 3 4 ) C16:0 . The differentiation- BC † † article T AA uptake in

3- C17:0 D negatively KD f † L ) biological 1 adipocytes. † Pparg C18:0 * μ D , Glut4 m). : C18:1

† † 5 - - - - - ,

© 2015 Nature America, Inc. All rights reserved. article xii dcesd dpst aog te mtblc changes metabolic other among adiposity decreased exhibit in observations previous with consistent is adipogenesis and catabolism BCAA between link a of onstration dem Our effect. regulatory a such identify to additionalrequired are studies though adipogenesis, on effects observed the mediate catabolic pathway that are upstream of PropCoA and AcCoA could BCAA the along metabolite(s) some Presumably, differentiation. adipogenic for important is oxidation BCAA that evidence direct first provide the data differentiation. accumulationandlipid These impaired pre-adipocytes in pathway the of knockdown functional cofactor can human affect health. gut the microbiome of composition the influence can itself availability lamin coba and transporters cobalamin of expression the Furthermore, causes other or metformin to due malabsorption of result a as cate that patients with T2DM are commonly deficient in cobalamin indi studies Notably, several serum). dialyzed employing systems arising intermediates culturenotablyin limited also is (carnitinecatabolism BCAA from downstream of accumulation the induces as obese in compared to downregulatedlean patients is that pathway a as succinyl-CoA to conversion MMA-CoA identifies data transcriptional and teomic thatproincorporatedmetabolism clinical adipocyte of model flux recent a Indeed, biotin. to manner similar a in catabolism BCAA pathways these in perturbations of cause unifying potential a be to hypothesized resistance with insulin associated and been obesity all have species, C5-acylcarnitine and C3- these metabolites with associated pathology demonstrated the such given from systems obtained results affect may accumulation OCFA and cells glial in observed been has as catabolism, BCAA for cobalamin sufficient lack may that media employ adipocytes) and these tissues (such as representrat cardiomyocytes, toprimary murine myocytes used cofactor routinely are this that models in biological Numerous deficiencies by affected particularly be may and BCAAs metabolize to knownare kidney andmuscle heart, the Indeed,oxidation. BCAA completesupport to inadequateare cells 3T3-L1 differentiatedfor conditions culture standard thatonstrate in other that cell types catabolize dem findings These BCAAs. also availability cobalamin with correlates it as function mitochondrial lonic aciduria methylma untreated acute, with associated pathologies severe the to activatedthe AdoCbl cofactor for necessary MUT activity. Given the appropriate mitochondrial machinery for converting cobalamin and MMA abundances in culture, suggesting that 3T3-L1 cells have α (FASN)than rathersynthase acid fatty by extension and PropCoA of accumulation the from arose metabolites these that confirmed OCFAs of levels significant accumulated cells 3T3-L1 differenti ated studies, published other model in and various hands our affect In systems. could that deficiencies potential highlight and processes metabolic importantmediating in cofactors enzyme through identified more be systematic analyses. may process this driving coactivators and models cell and catabolic expressionenzyme withBCAA humanin tissues, animals associated been recently has and adipogenesis early during cantly PPARofActivity metabolism. model system for identifying key activators or suppressors of BCAA a as thereforeserve maydemonstrated here pre-adipocyteshuman 6 adipose- disease urine potential syrup maple with into patients in pathogenesis insights mediated provide also may results These

-oxidation u rsls lo eosrt te rtcl oe f iais as vitamins of role critical the demonstrate also results Our While BCAA catabolism clearlyis inducedduringWhileBCAA adipogenesis, as well as OCFAs, and methionine BCAAs, plasma Elevated 3 3 4 9 5 5 . Splmnain f oaai sprse OCFA suppressed cobalamin of Supplementation . 3 , highlighting the complexity of the ways in which this 6 , our results highlight the importance of investigating 1 13 0 cblmndfcec mgt lo compromise also might cobalamin-deficiency , , 15 , 19 , 2 0 Hwvr adtoa tasrpin factors transcription additional However, . 4 2 . We thatobserved cobalamin deficiency γ 40 , a target of TZDs, increases signifi increases TZDs, of target a , , 4 1 Atog boi aalblt is availability biotin Although . Bcat2 −/− mice, which which mice, nature 3 8 . MMA . 3 ch 1 We . 16– 7 43 e , 8 mic , , 1 4 3 N 8 4 7 ------. . . ature a l

6. 5. 4. 3. 2. 1. R version of pape the the in available are references associated any and Methods M published online16November 2015 R differentiation, lipogenesis and bioenergetics. rolescal for catabolism BCAA and vitamin availability in adipocyte criti highlight findings quantitativeourCollectively, MSUD. with individuals of (relative) health the andhere noted effects moderate a prerequisite for adipogenic differentiation, as demonstrated by the (MSUD). However, cell-autonomous catabolismBCAA is likely not 22. 21. 20. 19. 18. 17. 16. 15. 14. 13. 12. 11. 10. 9. 8. 7. biology eceived 10March 2015; accepted 7October 2015;

eferences eth

473–481 (2000). 473–481 metabolism. glucose systemic implications. resistance. insulin of 409 Cell E151–E154 (1994). E151–E154 vivo. in tissue adipose subcutaneous rat by glutamate (2009). 1029–1033 proliferation. cell of requirements metabolic the Warburgeffect: (2011). E164–E174 rats. obese in pathways metabolic and 106 sensitization. insulin thiazolidinedione-mediated Metab.Genet. mice. in metabolism acid amino chain branched of errors inborn for levels. BCAA circulating modulates metabolism (BCAA) acid amino chain branched Metab.Cell cycle. turnoverprotein futile a of activation the with associated expenditure obesity. human in flux acid amino cross-adipose and expression enzyme catabolism differentiation. during adipocytes 3T3-L1 in complexdehydrogenase acid 2-oxo Metab.Endocrinol. formation. adipocyte novo de of changes metabolic dependent Biophys. consequences. metabolic and levels, proteinactivity,mRNA, of analysis adipogenesis: during enzymes catabolic valine of expression Med.Nat. ONE PLoS women. African-American obese diabetic 2 type and non-diabetic in resistance. insulin to contributes and humans lean and obese differentiates that signature (1974). 411–418 muscle. and tissue adipose in precursor sterol (1972). 8085–8096 respiration.pyruvate and glucose, acids, fatty of effect The rat. the of diaphragms and hearts isolated by acids amino chain Kahn, B.B. & Flier, J.S. Obesity and insulin resistance. insulin and Obesity J.S.Flier, & B.B. Kahn, M.A. Herman, clinical Turer,and insightsScherer,P.E. mechanisticA.T. & Adiponectin: etiology the in signaling lipid and Inflammation Olefsky,J.M. & C.K. Glass, C.M. Steppan, fat. about talk we when about talk we What B.M. Spiegelman, & E.D. Rosen, Kowalski, T.J. & Watford, M. Production of glutamine and utilization of utilization and glutamine of T.J.Watford,ProductionKowalski, & M. the Understanding C.B. Thompson, & Cantley,L.C. VanderM.G., Heiden, G. Hsiao, D.D. Sears, transplantAdipose C.J.Lynch, & G. Chen, K.C., Olson, H.A., Zimmerman, tissue Adipose B.B. Kahn, P.,& She,C.J.O.D., Peroni,Lynch, M.A., Herman, P.She, D.E.Lackey, branched-chain the of InductionD.T., Chuang,M.S. C.W.Patel, Hu, & time- of analysis modularity and profile Flux K. Y., Lee, Yoon,Si, & J. Coordinated R.A. Harris, & J.A.N.Y.,Kedishvili, Jaskiewicz,Popov, K.M., Wang,T.J. O. Fiehn, C.B. Newgard, significant a leucine: of fate MetabolicJ. Farkas, & A. Angel,J., Rosenthal, branchedJ.F. of Buse, Oxidation & J.F.,K.H. Biggers, Friderici, M.G., Buse, , 307–312 (2001). 307–312 , , 18745–18750 (2009). 18745–18750 ,

156 o

| Adv chemical J. Biol. Chem. Biol. J. et al. et ds

Am. J. Physiol. Endocrinol. Metab.Endocrinol.Physiol. J. Am. , 20–44 (2014). 20–44 , 315

et al. et et al. et 17

et al. et

5 a et al. et 6 Disruption of BCATm in mice leads to increased energy energy increased to leads mice BCATmofin Disruption , 317–322 (1994). 317–322 , , e15234 (2010). e15234 , et al. et , 448–453 (2011). 448–453 , nce online public , 181–194 (2007). 181–194 , Diabetologia

109 Plasma metabolomic profiles reflective of glucose homeostasis glucose of reflective profiles metabolomic Plasma Multi-tissue, selective PPARy modulation of insulin sensitivity insulin PPARyof modulation selective Multi-tissue, et al. et et al. et et al. et Biochem. J.Biochem. Metabolite profiles and the risk of developing diabetes. developing of risk the and profiles Metabolite Mechanisms of human insulin resistance and resistance insulin human of Mechanisms Regulation of adipose branched-chain amino acid amino branched-chain adipose of Regulation , 345–353 (2013). 345–353 , Cell Metab.Cell

292 r The hormone resistin links obesity to diabetes. to obesity links resistin hormone The A branched-chain amino acid-related metabolic acid-related amino branched-chain A A novel ChREBP isoform in adipose tissue regulates tissue adipose in isoform ChREBP novel A

. 285 Cell Metab.Cell , E1637–E1646 (2007). E1637–E1646 , , 11348–11356 (2010). 11348–11356 ,

55 214 bi , 2319–2326 (2012). 2319–2326 ,

Nature 9 , 177–181 (1983). 177–181 , o , 311–326 (2009). 311–326 , a 15 l tion | , 635–645 (2012). 635–645 , ogy

484 Am. J. Physiol. Endocrinol. Metab.Endocrinol.Physiol. J. Am. www.nature.com/naturechemicalbiology , 333–338 (2012). 333–338 ,

304

d , E1175–E1187 (2013). E1175–E1187 , Am. J. Physiol.J. Am. oi J. Biol. Chem. Biol. J. : 10.1038/n Proc. Natl. Acad. Sci. USA Sci. Acad.Natl. Proc. Am. J. Physiol.J. Am. J. Clin. Invest.Clin. J.

Arch. Biochem.Arch. 226 ch

247 Am. J. Physiol.J. Am. Science , e ,

mbio Nature

106 266 Mol.

online

324 , , 300 .196

J. Lipid Res. LipidJ. metabolism. B12 vitamin deranged with patient a of system nervous the in aldehydes and acids fatty odd-numbered and Branched-chain culture. tissue in Physiol.RenalPhysiol. proteins. carrier and vitamins of uptake tubular for receptors of (2013). 472–476 follow-up.42-year a aciduria: methylmalonic (cblA) adipocytes. differentiating in utilization and storage lipid of determinants major (2003). mTOR.involvingpathway cells. Ras-transformed in route transport. and absorption, function, of Genomics differentiation. adipocyte of model a of phenotypingMetabolic mTORC1.by (2002). significance. metabolic and regulation, characteristics, autophagy.and (2013). carrier.pyruvate mitochondrial electrode. oxygen Clark-type a or analyzer XF seahorse the using cells permeabilized in function tissue. adipose human cultured in dependent depot is pioglitazone to response and secretion Med. complications. obesity for tissue adipose in differences depot depots. fat among varies apoptosis and adipogenesis, replication, for capacities distinct Kishimoto, Y., Williams, M., Moser, H.W., Hignite, C. & Biemann, K. Biemann, & Moser,H.W.,Y.,Kishimoto,C. M., Hignite,Williams, deficiency B12 vitamin of effect An R.H. Barley,F.W.,Abeles, & G.H. Sato, characterization homeostasis: folate and B12 vitamin in kidney The H. Birn, A. Haarmann, X. Su, C.J.Lynch, C. Commisso, disorders I—inherited part disease: and health in B12 Vitamin C.R. Kapadia, J.L. A.W.Griffin, Nicholls, & A., Vidal-Puig, S., Virtue,L.D., Roberts, Y.Zhang, transport: membrane on note a with levels acid amino Plasma Cynober,L.A. P.Nicklin, A.S. Divakaruni, mitochondrial MeasuringG.W.Murphy, Rogers,A.N. & A.S., Divakaruni, Adiponectin Henry,R.R. & T.P.,M.K. Ciaraldi, Savu, D.K.,S.A., Oh, Phillips, of implication heterogeneity: tissue Adipose S.K. M.-J.J.,Wu,Fried, Y. Lee, & Tchkonia,T.

34 et al. et E , 1–11 (2013). 1–11 , MIC

39 et al. et et al. et Am. J. Physiol. Endocrinol. Metab.Endocrinol.Physiol. J. Am. sequential ordered fatty acid fatty ordered sequential

et al. et

Biochemistry 14 chemical , 109–119 (2009). 109–119 , et al. et A Nature L BIOLOGY , 69–77 (1973). 69–77 , et al. et et al. et Coordinated regulation of protein synthesis and degradation and synthesis protein of regulation Coordinated Am. J. Physiol. Endocrinol. Metab.Endocrinol.Physiol. J. Am. Bidirectional transport of amino acids regulates mTORregulates acids amino of transport Bidirectional Cell Potential role of leucine metabolism in the leucine-signaling the in metabolism leucine of rolePotential et al. et J. Biol. Chem. Biol. J. Abundance of two human preadipocyte subtypes with subtypes preadipocyte human two of Abundance

Macropinocytosis of protein is an amino acid supply acid amino an is protein of Macropinocytosis Renal involvement in a patient with cobalamin A type type A cobalamin with patient a involvementin Renal 136

513

Thiazolidinediones are acute, specific inhibitors of the of inhibitors specific acute, are Thiazolidinediones 291 , 521–534 (2009). 521–534 ,

, 440–443 (2014). 440–443 , 43

| Adv , F22–F36 (2006). F22–F36 , Am. J. Physiol. Endocrinol. Metab.Endocrinol.Physiol. J. Am. , 5033–5044 (2004). 5033–5044 , Curr. Protoc. Toxicol.Curr.Protoc. a 247 Nature Proc. Natl. Acad. Sci. USA Sci. Acad.Natl. Proc. nce online public bi , 4270–4276 (1972). 4270–4276 , o

497 Gastroenterologist l α ogy , 633–637 (2013). 633–637 , -oxidation and -oxidation

288

, E267–E277 (2005). E267–E277 , 295

d 60 oi a Mol. Genet. Metab.Genet. Mol. , E842–E850 (2008). E842–E850 , , 25.2.1–25.2.16 (2014). 25.2.1–25.2.16 , tion | Nutrition : 10.1038/n

Δ 110 , 329–344 (1995). 329–344 , 9 desaturation aredesaturation 9 www.nature.com/naturechemicalbiology

285 Mol. Aspects Mol. , 5422–5427 ,

Physiol. 18 Am. J. Am. , E854–E863 , , 761–766 , ch

e 110

mbio

, .196 45. 44. 43. 42. 41. 40. 1R01NS087611 (A.N.M.)and agrant from Bioscience. Seahorse of Veterans Affairs and VA the San Diego Healthcare NIHgrant System (R.R.H.), USDepartment (1I010X00635-01A1),the Medical the Research(7-05-DCS-04), Service Scholar Award (C.M.M.),as well as grants from American the Diabetes Association US Department of grant Defense (DOD) W81XWH-13-1-0105 (C.M.M.)and aSearle InstituteCalifornia of Regenerative Medicine (CIRM)Award RB5-07356(C.M.M.), byin part, USNational Institutes of Health (NIH)grant R01CA188652(C.M.M.), San Diego, USA)for California, human material. adipocyte This work was supported, We Veterans (R.R.H.; Henry thank R.R. Affairs San Diego Healthcare System, A to C.M.M. com/reprints/index.htm papethe Any supplementary information, and source data are available inthe A The authors declare no competing interests. financial Competing financialinterests authors.all research;C.M.M. designed M.W. C.R.G., and C.M.M.wrote paper the with help from M.W. other M.W., experiments. all performed C.R.G., A.S.D., A.N.M.,T.P.C. and tissue; A.S.D. and oxygen A.N.M.performed consumption and experiments; C.R.G. S.A.P. and T.P.C. obtained biopsies and isolated human pre-adipocytes from adipose A 1 cknowledgments dditional information uthor contributions

compete in the gut. the in compete B vitamin distinguish to transporters multiple use microbes gut (2014). e109878 study.cross-sectional a diabetes: 2 type with patients in combination insulin with compared combination sulfonylurea in deficiency 1999–2006. Survey, ExaminationNutrition and HealthNational the supplements: B12 vitamin and therapy metformin with deficiency B12 biochemical of Association adipocyte. human the of model metabolic state.insulin-resistant the and obesity resistance. insulin of development Degnan, P.H., Barry, N.A., Mok, K.C., Taga, M.E. & Goodman, A.L. Human A.L. Goodman, & Taga,M.E. K.C., Mok,P.H., N.A., Degnan, Barry, D. Kang, Y.P.,J.V.G.P.Qi, Oakley,Garn, Jr.Reinstatler,L., & R.S., Williamson, A. Mardinoglu, in metabolism acid amino essential on perspectives Emerging S.H. Adams, in acids amino branched-chain and lipids between InterplayC.B. Newgard, r . Reprints and permissions information is available online at et al. et

Higher prevalence of metformin-induced vitamin B12 vitamin metformin-induced of prevalence Higher et al. et l . Correspondence and requests for material should addressed be Diabetes Care Diabetes Cell Host MicrobeHost Cell Integration of clinical data with a genomescale a with data clinical ofIntegration

35 Cell Metab.Cell , 327–333 (2012). 327–333 ,

15 Adv.Nutr. , 47–57 (2014). 47–57 , Mol. Syst. Biol.Syst.Mol.

, 606–614 (2012). 606–614 , 2 , 445–456 (2011). 445–456 , article PLoS ONE PLoS

9 , 649 (2013). 649 , online version of http://ww 12 analogs and analogs

, w.nature.

7 © 2015 Nature America, Inc. All rights reserved. glucose glucose Dulbecco’s mediumEagle modified (DMEM) (Life Technologies) sup Life high in from cultured and Collection CultureType purchased American the were from purchased were pre-adipocytes 3T3-L1 sera Murine stated. and otherwise media unless Technologies All noted. otherwise unless reagents. and culture Cell ( stearate and oleate palmitate, myristate, of production for accounted that sis was conducted using INCA analysis ISA algorithms. in-house using abundance natural for corrected and by integrating metabolite ion fragments Isotopomer analysis.spectral and spent with internal media standards labeled for quantification. freshderivatization of GC-MS using calculated was secretion anduptake acid glutamateAminomeasurements. and glutamine lactate, glucose, for analyzed andremoveimpurities to min 15 for RPM 15,000 at centrifuged natantswere 300 at centrifuged was media spentBriefly, measurements. flux extracellular obtain to used was measurements. flux Extracellular choninic acid assay (Pierce) with matched plates of whole cells. bicin the by measured as protein cell total to normalized were data All A. maximum respiratory rate in response to FCCP that was sensitive to antimycin previously Uncoupler-stimulated respiration described was as calculated the difference the between as are substrates for respiratory KIV, used commonly (except for malate mM 0.5 FADH with only which ketoacids chain branched mM 5 of 2 Bioscience) recombinantPerfringolysin O (purchased as XF nM Plasma Membrane Permeabilizer, 3 Seahorse and ADP mM 4 BSA, (w/v) 0.2% with supplemented mediumMAS in measured initially initiated.Respirationentiationwas was Measurements of differentiated cells were conducted 10–14 days after differ 1 × 10 (Life Technologies) appropriately. and scaled CyQuant using number cell for normalized were rates experiment, the after sumption rate sensitive to 1 and pyruvate con oxygen mM the as calculated 1 is respirationMitochondrial carnitine. mM glutamine, 0.5 mM 3 glucose, mM 8 with supplemented initiated, growth medium was replaced with unbuffered DMEM (Sigma #5030) differentiationwas after days 10 above. described asdifferentiated and tained 10 × 5 atplated were cells cells, intact respirationForin Respirometry. culture was as performed previously described 800 at centrifuged 250 through centrifuged and aspirated was SVF the 50 at centrifuged 450 a through filtered waterFollowingshakingbath. °C suspension digestion,was37 the a in h for1 (Worthington collagenase units 333 Biochemical Corp.) per ml buffer (~1,000 units per gram containing of AT) and incubated buffer 4% a salts into placed BSA/HEPES and minced was SAT Fresh techniques. sterile using out isolation. Pre-adipocyte were cells and cells water.washed 4×with milli-Q from removed was solution Stain min. 30 for solution (stock:water) 3:2 with stained finally isopropanol,and 60% with once washed paraformaldehyde for 30 min at room temperature, washed 2× in milli-Q water, 0.22 a through filtered Imaging. O nature differentiation, For contamination. mycoplasma for screened regularly Cells were confluence. 70% below (BCS) serum calf bovine 10% with plemented Supplementary Table 1 NLINE METH For respiration in permeabilized cells, For cells cells, were onto respiration seeded in XF96 permeabilized plates at μ M FCCP, and 2 4 cells/well and were maintained and differentiated as described above. above. described as differentiated and maintained were and cells/well ch 03% wv Ol e O tc ws rprd n spoao and isopropanol in prepared was stock O Red Oil (w/v) 0.35% A e 2 6 mic and sequentially offered 2 Respirometry was conducted using a Seahorse XF96 Analyzer.XF96 Seahorse a using conducted was Respirometry a g l for 10 min at 25 °C. The pellet and infranatant containinginfranatant and pellet The °C. 25 at min 10 for g

g biology for 10 min at 20 °C. Further pre-adipocyte isolation and isolation pre-adipocyte Further °C. 20 at min 10 for 2 for 5 min to remove cell debris. Culture medium super medium Culture debris. cell remove to min 5 for lne rsiain a maue) n concentrations and measured) was respiration -linked O μ M antimycin A. Permeabilized adipocytes were offered DS μ μ All procedures for AT explant culture were carried carried were culture explant AT for procedures All ). m filter. Cells were washed with PBS, fixed with 4% with fixed PBS, with washed were Cells filter. m nln eh Cmoet upy opn) and Company) Supply (Component mesh nylon m All reagents were purchased form Sigma-Aldrich form purchased were reagents All 4 μ 6 and a more complete model for fatty acid synthe M rotenone and 2 Mass isotopomer distributions were determined A Yellow Springs Instruments (YSI) 2950 (YSI) Instruments YellowSprings A 4 summarized in μ g/ml g/ml oligomycin, successive additions 2 3 μ . M antimycin A. Immediately Supplementary Table 3 μ 3 m nylon mesh, then then mesh, nylon m cells/well and main and cells/well 2 7 ------.

33 10 T3, nM 2 dexamethasone, nM 100 insulin, nM 100 IBMX, mM 0.5 FBS, 3% + DMEM/F12differentiated werein cells reached), was confluence after days (2 0 AtDay FBS. 10% + DMEM/F12 Human differentiation. adipocyte previously described greater omentum the superficial abdominal SAT depot and visceral (V) AT was obtained from the Adipose tissue (AT) biopsy. San Diego, approved studies. the subjects gave All informed consent. MemorialofScripps Boards Hospital UniversityReview and the of California, Institutionalisolation.The pre-adipocyteprovidefor materialtostudy the for scopic gastric bypass via Roux-en-Y for the treatment of obesity were recruited subjects. Human Cobalamin cultures used. noted. when was FBS 10% + 5 or nM) (500 DMEM thereafter and 4, Day 1 insulin. with FBS 10% + DMEM to 2 Day on changed was Medium (FBS). 1 0.25 (IBMX), 3-isobutyl-1-methylxanthine mM 0.5 with induced was differentiation 0 Day On –2). Day (termed confluence reach to allowed and (ThermoFisher)plates 12-well into seeded were cells/well 10,000 dryness dryness before redissolving in 50–100 500 two with solution NaCl saturated 100 of addition after extracted FAMEswere min. 90–120 for °C 50 at tion (FAMEs) through addition of 500 lites were evaporated, saponified and toesterified form fatty acid methyl esters metabo nonpolar Extracted analysis. for vials sample GC to transferred was 30–45 min. Samples were centrifuged at 15,000 RPM for 5 min and supernatant tert N and 30 with silylated then pyridine were Samples min. 60–90 for in °C 37 at incubated Scientific) (Thermo hydrochloride methoxyamine (w/v) 20 in derivatized were buffers metabolites accordingly.Polaradjusted extraction were of volumes the and plate 12-well a in cultured were cells 3T3-L1 differentiated while plates, 6-well in cultured were adipocytes human described and previously methanol/water/chloroform as using analyzed extracted were acids fatty and lites analysis. (GC/MS) spectrometry chromatography/mass Gas abundance of the where arelabeled: that pool metabolite the atomswithin all of percent the as calculated was isotopes foundin be can media gluc+AA Low CompleteformulationandDMEM more of labeled). or one (with experiments all in cells to available were nutrients all that such FBS dialyzed 10% and components, chemical other of versions unlabeled tracer, specified the with formulated was Inc.) Laboratories, (Hyclone DMEM/F12 10,000 at seeded cells/cm were cells 143B and stated. HuH-7 otherwise A549, unless 3T3-L1, h Proliferating 24 for medium tracer in post-induction days 7 cells by culturing accomplished was cells 3T3-L1 differentiated in metabolites intracellular of labeling isotope Stable Inc. Isotopes Cambridge from chased experiments. Tracing tracing experiments were conducted on Day 14. metabolic until dexamethasone nM 10 and insulin nM 10 with DMEM/F12 in maintainedwere cells induction, After days. 2 every changesmedium with - μ tert μ btliehlhooiae tDS (ei Tcnlge) t 7 C for °C 37 at Technologies) (Regis (tBDMS) -butyldimethylchlorosilane /l nui, n 10 M oiltzn i DE cnann 1% FBS 10% containing DMEM in rosiglitazone nM 100 and insulin, g/ml boi ad 17 and biotin M -butyldimethylsilyl- n is the number of carbon atoms in the metabolite and 2 n clue i tae mdu fr 8 . utm MM or DMEM Custom h. 48 for medium tracer in cultured and ′ -adenosylcobalamin (AdoCbl; 100 nM) were supplemented to supplementedwere nM) 100 (AdoCbl; -adenosylcobalamin Supplementary Table 3 Supplementary 2 egt tbe bs sbet udron eetv laparo elective undergoing subjects obese stable Weight 5 i . Biopsy tissue was placed in a sterile HEPES salts solution as th th mass isotopomer. 4 7 and transported to lab the for immediate processing. All [ μ pnohnc acid pantothenic M N mtytilooctmd (TSF) ih 1% with (MTBSTFA) -methyltrifluoroacetamide Subcutaneous (S) AT biopsies were obtained from 13 ] lcs ad mn ai taes ee pur were tracers acid amino and glucose C] μ 4 Human pre-adipocytes were maintained in i 9 ∑ l 2% (w/v) H . Proliferating 3T3-L1 cells and primary primary and cells 3T3-L1 Proliferating . = n 1 μ Mi l of hexane and transfer to glass GC vials μ n μ i g/ml transferrin, 1 transferrin, g/ml . Mole percent enrichment (MPE) of (MPE) enrichment percent Mole . i hxn wse ad vprtd to evaporated and washes hexane l

4 8 nuto ccti fr days 7 for cocktail induction 2 SO doi:10.1038/nchembio.1961 4 in methanol and incuba μ M dexamethasone, M μ M M rosiglitazone, M oa metabo Polar i is the relative μ l of 2% of l μ μ g/ml g/ml l of l μ - - - - - l

© 2015 Nature America, Inc. All rights reserved. a three-stage program provided by the manufacturer: 95 °C for 3 min, 40 cycles ried out on 96-well plates on a CFX Connect Real time System (Bio-Rad), using car simultaneously.was amplified PCR was The quantification,18S of zation 1 and (Bio-Rad), 2 of consisted reactions to the manufacturer’s instructions. Individual 20 Reverse Transcription Supermix for RT-PCR (Bio-Rad Laboratories) according tions. First-strand cDNA was synthesized from 1 tured cells using Trizol Reagent (Life Technologies) per manufacturer’s instruc RT-PCR.quantitative and isolation RNA removed from medium the on beginning Day 0. was Puromycin rosiglitazone. without but above described as differentiation 2 with selected tion of 2 ml of virus-free medium. After 24 h of recovery, transduced cells were addi before h 4 forplate 6-well a in medium virus-containing of ml 0.5 with added to a final concentration of 6 (0.45 filtered and later days two collected was constructs ing plasmids VSV-G, gag/pol, and rev. HEK293T medium containing lentiviral requiredpackag andagenttransfection Fugene 6 the using cells HEK293T in non-targeting control construct were purchased from Sigma-Aldrich, packaged a and GAAGGTCATCTCGAGATGACCTTCTTTCGTGACTGCTTTTTG), AAGGATTTTTG; KD2: NM_007533.2-1188s1c1: CCGGGCAGTCACGAAA CCGGTCCTTCTACATGACCAACTATCTCGAGATAGTTGGTCATGTAG mouse pLKO.1-purotargeting shRNA of KD shRNA and production Lentiviral nal standard mixture was prepared and added during extraction. mass spectrometer (MS). 5975C Agilent an with interfaced chromatograph(GC) gas 7890A Agilent an 0.25 × i.d. mm 0.25 × m (30 column DB-35MS a using GC-MS by analyzed were samples Derivatized analysis. for doi:10.1038/nchembio.1961 For quantitation of amino acids in plasma samples, an isotope-labeled inter μ g/ml puromycin. Cells were then plated to 12-well plates for plates 12-well to plated then were puromycin. Cells g/ml μ l of each 5 each of l μ l of diluted cDNA, 10 cDNA, diluted of l μ M forward and reverse primers.Forreverseandforward standardi M μ g/ml. 3T3-L1 pre-adipocytes were infected μ m, Agilent J&W Scientific) installed in installed Scientific) J&W Agilent m, Bckdha Total RNA was purified from cul from purified was TotalRNA Bckdha (KD1: NM_007533.2-421s1c1: (KD1: μ μ μ l SYBR Green real-time PCR g of total RNA using iScript l of SYBR Green Supermix Green SYBR of l . Glycerol stocks of TRC2- of stocks Glycerol μ m). Polybrenewas m). ------

*, P mean as presented and experiments independent multiple of averages experiment.Respirometryareresults an within plate)culture tissue a of wells (i.e., replicates spatial separate as defined are replicates technical temporally; (biologicalreplicates). Biologicalreplicates separatearedefinedasexperiments mean as Statistical analysis. Scientific). (Genesee film Substrate (Thermo Scientific) and developed using Blue Devil Autoradiography WestChemiluminescent SuperSignalPico using (1:10,000) anti-mouserabbit radish peroxidase-conjugated antibody goat secondary anti-rabbit (1:2,500)or horse with detected was signal Specific (1:5,000). 8H10D10) Signaling (Cell anti-Bckdha (Novus Biologicals rabbitNBP1-79616) (1:1,000 dilution) and with mouse anti-Beta-Actin immunoblotted and membrane nitrocellulose a to transferred were proteins The SDS-PAGEgel. 12% on separated was protein RIPAice-cold protease1× withbufferinhibitor 30 (Sigma-Aldrich. Western blots. Supplementary Table 4 in listed are used primers Gene-specific s. 30 for °C 60 and s 10 for °C 95 of intervals determined via parameter continuation/sensitivity analysis. <0.001.Errors associated withMFA andofISAlipogenesis areconfidence95% 46. 49. 48. 47. values were calculated using Student’s two-tailed P

value between 0.01 and 0.05; **, lipogenesis under hypoxia. under lipogenesis Obesity phenotypes. improve metabolic and preadipocytes human differentiation of Metab. humans. in therapies antidiabetic by adiponectin analysis. flux metabolic Young, J.D. INCA: a computational platform for isotopically non-stationary isotopically for computationalplatform Young,a J.D.INCA: Metallo, C.M. Metallo, maximize to protocol modified A S.K. M.-J.J.,Wu,Fried, Y. Lee, & S.A. Phillips, ±

297 .. n ae ersnaie w o mr idpnet experiments independent more or representative two are and s.d. 20 , 2334–2340 (2012). 2334–2340 , , E767–E773 (2009). E767–E773 , 3T3-L1 adipocytes with Control KD or et al. et et al. et Results shown as averages of technical replicates are presented Selective regulation of cellular and secreted multimeric secreted and cellular of regulation Selective Reductive glutamine metabolism by IDH1 mediates IDH1 by metabolism glutamine Reductive . Bioinformatics Nature P

481 value between 0.001 and 0.01; ***,

, 380–384 (2012). 380–384 , 30 , 1333–1335 (2014). 1333–1335 , t test or ANOVA, as required; nature CH Am. J. Physiol. Endocrinol.Physiol. J. Am. Bckdha KD were lysed in E MIC A μ L BIOLOGY g oftotal g ± P s.e.m. value -