Metformin inhibits mitochondrial complex I of cancer cells to reduce tumorigenesis

eli esciences.o g Whea on e al. eLi e 2014;3:e02242. DOI: 10.7554/eLi e.02242 1 o 18 Me o min inhibi s mi ochond ial complex I o cance cells o educe umo igenesis William W Whea on1†, Samuel E Weinbe g1†, Robe B Hamanaka1, Saul Sobe anes1, Lucas B Sulli an1, Elena Anso1, And ea Glasaue 1, E ic Du ou 2, Gokhan M Mu lu1, GR Sco Budigne 1, Na deep S Chandel1* 1Depa men o Medicine, The Feinbe g School o Medicine, No hwes e n Uni e si y, Chicago, Uni ed S a es; 2Ins i u e o Biomedical Technology, Uni e si y o Tampe e, Tampe e, Finland Abs ac Recen epidemiological and labo a o y-based s udies sugges ha he an i-diabe ic d ug me o min p e en s cance p og ession. How me o min diminishes umo g ow h is no ully unde s ood. In his s udy, we epo ha in human cance cells, me o min inhibi s mi ochond ial complex I (NADH dehyd ogenase) ac i i y and cellula espi a ion. Me o min inhibi ed cellula p oli e a ion in he p esence o glucose, bu induced cell dea h upon glucose dep i a ion, indica ing ha cance cells ely exclusi ely on glycolysis o su i al in he p esence o me o min. Me o min also educed hypoxic ac i a ion o hypoxia-inducible ac o 1 (HIF-1). All o hese e ec s o me o min we e e e sed when he me o min- esis an Saccha omyces ce e isiae NADH dehyd ogenase NDI1 was o e exp essed. In i o, he adminis a ion o me o min o mice inhibi ed he g ow h o con ol human cance cells bu no hose exp essing NDI1. Thus, we ha e demons a ed ha me o min's inhibi o y e ec s on cance p og ession a e cance cell au onomous and depend on i s abili y o inhibi mi ochond ial complex I. DOI: 10.7554/eLi e.02242.001 In oduc ion Me o min is widely used o ea pa ien s wi h ype II diabe es melli us who ha e high le els o ci cu- la ing insulin (Na han e al., 2009). Me o min supp esses li e gluconeogenesis he eby educing glucose elease om he li e (Inzucchi e al., 1998; Violle e al., 2012). In se e al ecen e ospec- i e s udies, in es iga o s ha e obse ed an associa ion be ween me o min use and diminished umo p og ession in pa ien s su e ing om di e en ypes o cance s (E ans e al., 2005; Bowke e al., 2006; Dowling e al., 2012). These da a ha e p omp ed se e al p ospec i e clinical ials o de e - mine he e icacy o me o min as an an i-cance agen . Howe e , he unde lying mechanism by which me o min diminishes umo g ow h is no ully unde s ood. The e a e wo pos ula ed mechanisms by which me o min educes umo g ow h. Me o min may ac a he o ganismal le el, educing le els o ci cula ing insulin, a known mi ogen o cance cells. Al e na i ely, me o min may ac in a cance cell au onomous manne . Me o min is known o inhibi mi ochond ial complex I in i o (O a e al., 2009; El-Mi e al., 2000; Owen e al., 2000) and i is hus possible ha his a ge ing o he elec on anspo chain could inhibi umo cell g ow h (Bi soy e al., 2012). This la e hypo hesis has been ques ioned as cance cells ha e he abili y o su i e on ATP p oduced exclusi ely by glycolysis. Fu he mo e, cance cells ha e been shown o conduc glu- amine-dependen educ i e ca boxyla ion o gene a e he TCA cycle in e media es equi ed o cell p oli e a ion when he elec on anspo chain is inhibi ed (Mullen e al., 2012; Fend e al., 2013). Thus, i is no clea whe he inhibi ion o complex I by me o min would esul in dec easing umo g ow h. In he p esen s udy, we di ec ly es ed whe he inhibi ion o cance cell mi ochond ial com- plex I by me o min was equi ed o dec ease cell p oli e a ion in i o and umo p og ession in i o. *Fo co espondence: na @ no hwes e n.edu †These au ho s con ibu ed equally o his wo k Compe ing in e es s: The au ho s decla e ha no compe ing in e es s exis . Funding: See page 16 Recei ed: 09 Janua y 2014 Accep ed: 15 Ap il 2014 Published: 13 May 2014 Re iewing edi o : Chi Van Dang, Uni e si y o Pennsyl ania, Uni ed S a es Copy igh Whea on e al. This a icle is dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License, which pe mi s un es ic ed use and edis ibu ion p o ided ha he o iginal au ho and sou ce a e c edi ed. RESEARCH ARTICLE Human biology and medicine Whea on e al. eLi e 2014;3:e02242. DOI: 10.7554/eLi e.02242 2 o 18 Resea ch a icle Resul s Human HCT116 p53−/− colon cance cells ha e p e iously shown o be sensi i e o me o min (Buzzai e al., 2007). To de e mine i me o min ea men inhibi ed cellula oxygen consump ion in hese cells, we ea ed HCT116 p53−/− cells wi h inc easing concen a ions o me o min in media con aining he me abolic subs a es glucose, py u a e, and glu amine o 24 h . Subsequen ly, we measu ed cel- lula oxygen consump ion. Me o min inhibi ed cellula oxygen consump ion o HCT 116 p53−/− cells a concen a ions (0.25–1.0 mM) simila o hose epo ed o a ec bioene ge ics and gluconeogene- sis in p ima y hepa ocy es in i o (Figu e 1A; Fo e z e al., 2010; Mille e al., 2013). To de e mine whe he me o min's inhibi ion o cellula oxygen consump ion depended on mi o- chond ial complex I, we s ably o e exp essed he Saccha omyces ce e isiae p o ein NDI1 in HCT 116 p53−/− cells (he eon e e ed o as NDI1-HCT 116 p53−/− cells). NDI1 is a single-subuni NADH dehy- d ogenase, which oxidizes NADH in a p ocess simila o he mul i-subuni mammalian complex I; howe e wi hou p o on pumping o ROS gene a ion (Seo e al., 1998). By con as , mammalian complex I con ains 45 subuni s ha pumps p o ons and gene a es ROS. NDI1-HCT 116 p53−/− cells demons a ed a sligh , non-signi ican ele a ion in basal cellula oxygen consump ion compa ed o con ol cells and we e comple ely esis an o he e ec s o me o min on cellula oxygen consump ion (Figu e 1— igu e supplemen 1, Figu e 1B). To ensu e ha he inhibi ion o cellula oxygen consump ion by me o min was a di ec e ec o me o min on complex I, we examined mi ochond ial espi a o y unc ion in saponin-pe meabilized cells. Saponin emo es choles e ol om plasma memb anes, allowing he en y o me abolic sub- s a es di ec ly o mi ochond ia (Jamu and Oli e , 2010). In he p esence o ADP and he complex I subs a es py u a e and mala e, me o min ully inhibi ed oxygen consump ion in pe meabilized Con ol-HCT 116 p53−/− cells (Figu e 1C). By con as , me o min had no e ec on py u a e/mala e-d i en oxygen consump ion in NDI1-HCT 116 p53−/− cells (Figu e 1D). Me o min also had no e ec on oxygen consump ion in saponin-pe meabilized cells espi ing on he complex II subs a e succina e in he p esence o ADP (Figu e 1E). In e es ingly, in saponin-pe meabilized cells, me o min signi ican ly inhibi ed complex I-dependen espi a ion a a much lowe concen a ion han ha equi ed o inhibi oxygen consump ion o in ac cells, sugges ing ha anspo ac oss he plasma memb ane is a ba ie o me o min's inhibi ion o complex I. Me o min is known o slowly accumula e in cells in which i s eLi e diges Me o min is widely used o educe he high blood suga le els caused by diabe es. Recen ly, se e al s udies ha e sugges ed ha pa ien s aking me o min who also de elop cance ha e umo s ha g ow mo e slowly han a e age. As clinical ials ha e al eady s a ed o in es iga e i me o min is an e ec i e an i-cance ea men , i is impo an o unde s and how i migh es ic umo g ow h. Resea che s ha e p oposed wo ways ha me o min could a ec umo s. Fi s , insulin is known o p omp cance cells o di ide, so he slowe a e o umo g ow h could jus be a side-e ec o he me o min educing he amoun o insulin in he blood. Al e na i ely, me o min could a ge cance cells mo e di ec ly by cu ing he ene gy supply p oduced by hei mi ochond ia. Me o min has been shown o dis up complex I o he elec on anspo chain ha is used by cells o gene a e ene gy. Howe e , i is no known i dis up ing complex I would ac ually s op cells di iding because hey can gene a e ene gy in o he ways. Whea on, Weinbe g e al. ha e now demons a ed ha me o min does a ge complex I in cance cells, and ha i s e ec s depend on he amoun o glucose a ailable o cells o con e , wi hou in ol ing mi ochond ia, in o ene gy. When he e is plen y o glucose, me o min slows down he a e a which cance cells di ide, which slows down umo g ow h. When he cells a e dep i ed o glucose, me o min kills he cells ins ead. Me o min also inhibi s he pa hways ha egula e hypoxia inducible ac o s (HIFs), which a e pa o a sys em ha helps cells o su i e low-oxygen condi ions, a p ominen ea u e o many umo s. This means ha me o min may comba cance mo e e ec i ely i used alongside o he ea men s ha educe he a ailabili y o bo h oxygen and glucose inside cells. Me o min could also po en ially ea condi ions ha a e linked o o e ac i e HIFs, such as pulmona y hype ension. DOI: 10.7554/eLi e.02242.002 Human biology and medicine Whea on e al. eLi e 2014;3:e02242. DOI: 10.7554/eLi e.02242 3 o 18 Resea ch a icle Figu e 1. Me o min inhibi s mi ochond ial complex I unc ion. (A) Rela i e mi ochond ial oxygen consump ion a e (OCR) o in ac Con ol-HCT 116 p53−/− and (B) NDI1-HCT 116 p53−/− cells ea ed wi h me o min in comple e media o 24 h . (C) Rela i e complex I (2 mM mala e, 10 mM py u a e, 10 mM ADP)-d i en oxygen consump ion a e o saponin pe meabilized Con ol-HCT 116 p53−/− cells and (D) NDI1-HCT 116 p53−/− cells ea ed wi h me o min o 20 min in mi ochond ial assay bu e . (E) Rela i e complex II-d i en oxygen consump ion a e o saponin pe meabilized Con ol-HCT 116 p53−/− cells ea ed wi h 10 mM succina e and 10 mM ADP in he p esence o 1 mM me o min o he complex II inhibi o 3-Ni op opionic acid (3-NPA). (F) Rep esen a i e wes e n blo and quan i ica ion o le els o OCT1 p o ein in Con ol BFP-HCT 116 p53−/− and NDI1-HCT 116 p53−/− cells. E o ba s a e SEM (OCR: n = 4; OCT1: n = 4). * indica e signi icance p<0.05. DOI: 10.7554/eLi e.02242.003 The ollowing igu e supplemen s a e a ailable o igu e 1: Figu e supplemen 1. NDI1 exp ession sligh ly inc eases oxygen consump ion. DOI: 10.7554/eLi e.02242.004 Human biology and medicine Whea on e al. eLi e 2014;3:e02242. DOI: 10.7554/eLi e.02242 4 o 18 Resea ch a icle up ake is media ed by o ganic ca ion anspo e s (OCTs) (Emami Riedmaie e al., 2013). To ensu e ha NDI1-HCT 116 p53−/− cells a e no e ac o y o me o min because o a change in me o min up ake, we analyzed he exp ession o OCT 1 in bo h con ol and NDI1-HCT 116 p53−/− cells. Exp ession o OCT1 p o ein did no change wi h he p esence o NDI1 (Figu e 1F). We nex sough o de e mine i me o min-dependen inhibi ion o complex I esul ed in changes in p oli e a ion and su i al o HCT116 p53−/− cells. Me o min did no induce cell dea h in Con ol- HCT 116 p53−/− o NDI1-HCT 116 p53−/− cells in he p esence o glucose (Figu e 2A,B), howe e , in he absence o glucose, me o min induced cell dea h in Con ol-HCT 116 p53−/− bu no in NDI1-HCT 116 p53−/− cells (Figu e 2C,D). Me o min diminished cell p oli e a ion in Con ol-HCT 116 p53−/− cells bu no in NDI1-HCT 116 p53−/− cells in media con aining glucose (Figu e 2E,F). These esul s indica e ha he me o min-dependen inhibi ion o complex I dec eases cell p oli e - a ion in he p esence o glucose and inc eases cell dea h unde glucose dep i a ion. These inhibi o y e ec s o me o min we e no speci ic o HCT116 p53−/− cells as me o min inhibi ed oxygen consump- ion and cellula p oli e a ion o Con ol-HCT 116 p53+/+ cells and Con ol-A549 human lung cance cells bu no NDI1-HCT 116 p53+/+ o NDI1-A549 cells (Figu e 2— igu e supplemen 1 and 2). Taken oge he , hese esul s indica e ha he an i-p oli e a i e and cell dea h p omo ing e ec s o me - o min equi e mi ochond ial complex I inhibi ion. We also examined whe he phen o min, a mo e lipophilic biguanide, also exe i s an i-p oli e a i e e ec s on cance cells h ough inhibi ion o complex I. Phen o min inhibi ed oxygen consump ion in Con ol-HCT 116 p53−/− cells and saponin-pe meabilized Con ol HCT 116 p53−/− cells a 100- old lowe concen a ion compa ed o me o min (Figu e 3A,C). Exp ession o NDI1 escued he phen- o min-media ed dec ease in oxygen consump ion (Figu e 3B,D). Phen o min diminished cell p oli e - a ion in he con ol bu no NDI1 exp essing HCT116 p53−/− cells (Figu e 3E,F), and did no induce cell dea h in media con aining glucose, simila o me o min (Figu e 3G,H). Collec i ely, hese esul s indica e ha phen o min also exe s i s biological e ec s h ough inhibi ion o mi ochond ial complex I. To de e mine whe he me o min and phen o min diminish p oli e a ion and su i al o cells lacking endogenous complex I ac i i y, we u ilized a a ian o CCL16 hams e ib oblas s ha ha bo s a mu a ion in complex I (B2-CCL16) (Seo e al., 1998). Me o min inhibi ed p oli e a ion o wild- ype Con ol-CCL16 hams e ib oblas s bu no ha o B2-CCL16 cells o o B2-CCL16 cells econs i u ed wi h NDI1 (NDI1-CCL16). Me o min and phen o min inhibi ed cellula oxygen consump ion in wild- ype CCL16 bu no in CCL16-NDI1 cells (Figu e 3— igu e supplemen 1 and 2). When hese cells we e cul u ed in galac ose-subs i u ed media, bo h me o min and phen o min induced cell dea h only in wild- ype CCL16 cells. Su i al o NDI1-CCL16 cells was no a ec ed by me o min o phen o min (Figu e 3— igu e supplemen s 1 and 2). The B2-CCL16 cells die in galac ose in he absence o me - o min o phen o min since hey ha bo a mu a ion in complex I. Taken oge he , hese esul s con- i m ha he an i-p oli e a i e e ec s o me o min and phen o min equi e mi ochond ial complex I inhibi ion. The posi i e cha ge o me o min has been p oposed o accoun o i is accumula ion wi hin he ma ix o mi ochond ia ha exhibi a obus inne mi ochond ia memb ane po en ial (Owen e al., 2000). Al e na i ely, he non-pola hyd oca bon-side chain o he d ug could p omo e binding o com- plexes wi hin mi ochond ial memb anes. We es ed whe he he mi ochond ial memb ane po en ial is necessa y o me o min-dependen inhibi ion o complex I. Saponin-pe meabilized Con ol-HCT 116 p53−/− cells we e induced o espi e on py u a e/mala e in he p esence o ei he ADP o CCCP. Al hough bo h ADP and CCCP induce mi ochond ial espi a ion, only CCCP depola izes mi ochond ial inne memb ane po en ial. Me o min inhibi ed ADP bu no CCCP s imula ed oxygen consump ion indica ing ha he me o min-media ed inhibi ion o mi ochond ial complex I equi ed pola ized mi o- chond ia (Figu e 4A,B). Ro enone, an i e e sible inhibi o o complex I, does no equi e pola ized mi ochond ia o inhibi mi ochond ial oxygen consump ion (Figu e 4C,D). Ou esul s sugges ha me o min would no be e ec i e in supp essing complex I ac i i y o in ac cells i mi ochond ial inne memb ane po en ial was dis up ed. Me o min-media ed inhibi ion o he elec on anspo chain diminishes p o on pumping, which migh depola ize he mi ochond ial memb ane, hus limi ing accu- mula ion o he d ug. Howe e , we did no obse e a educ ion in he mi ochond ial inne memb ane po en ial measu ed using TMRE luo escen dye in Con ol and NDI1 HCT116 p53−/− cells a e me - o min ea men (Figu e 4E,F). When elec on anspo unc ion is inhibi ed, he ATP syn hase can unc ion in e e se such ha i uses ATP gene a ed by glycolysis o pump p o ons ac oss he inne mi ochond ial memb ane, main aining memb ane po en ial (Appleby e al., 1999). The ATP syn hase Human biology and medicine Whea on e al. eLi e 2014;3:e02242. DOI: 10.7554/eLi e.02242 5 o 18 Resea ch a icle Figu e 2. Me o min dec eases cell p oli e a ion by inhibi ing mi ochond ial complex I. (A) Pe cen age o li e Con ol-HCT 116 p53−/− o (B) NDI1-HCT 116 p53−/− ea ed wi h me o min o 72 h in media con aining 10 mM glucose. (C) Pe cen age o li e Con ol-HCT116 p53−/− o (D) NDI1-HCT 116 p53−/− ea ed wi h me o min o 24 h ollowed by glucose wi hd awal o 16 h . (E) Cell numbe o Con ol-HCT 116 p53−/− cells and (F) NDI1-HCT 116 p53−/− cells 24, 48, and 72 h pos ea men wi h 0.5 mM o 1 mM me o min in comple e media. E o ba s a e SEM (n = 4). * indica es signi icance p<0.05. DOI: 10.7554/eLi e.02242.005 The ollowing igu e supplemen s a e a ailable o igu e 2: Figu e supplemen 1. Me o min dec eases cellula p oli e a ion h ough inhibi ion o mi ochond ial complex I unc ion in HCT 116 p53+/+ cells. DOI: 10.7554/eLi e.02242.006 Figu e supplemen 2. Me o min dec eases cellula p oli e a ion h ough inhibi ion o mi ochond ial complex I unc ion in A549 cells. DOI: 10.7554/eLi e.02242.007 Human biology and medicine Whea on e al. eLi e 2014;3:e02242. DOI: 10.7554/eLi e.02242 6 o 18 Resea ch a icle Figu e 3. Phen o min dec eases cell p oli e a ion by inhibi ing mi ochond ial complex I. (A) Rela i e mi ochond ial oxygen consump ion a e (OCR) o in ac Con ol-HCT 116 p53−/− and (B) NDI1-HCT 116 p53−/− cells ea ed wi h phen o min in comple e media o 24 h . (C) Rela i e complex I (2 mM mala e, 10 mM py u a e, 10 mM ADP)-d i en oxygen consump ion a e o saponin pe meabilized Con ol-HCT 116 p53−/− cells and (D) NDI1-HCT 116 p53−/− cells ea ed wi h phen o min o 20 min in mi ochond ial assay bu e . (E) Cell numbe o Con ol-HCT 116 p53−/− cells and (F) NDI1-HCT 116 p53−/− cells 24, 48, and 72 h pos ea men wi h 0 o 5 µM phen o min in comple e media. (G) Pe cen age o li e Con ol-HCT 116 p53−/− o (H) NDI1-HCT 116 p53−/− ea ed wi h me o min o 72 h ollowed in comple e media. E o ba s a e SEM (Rela i e OCR n = 5; Cell numbe n = 4). * indica es signi icance p<0.05. DOI: 10.7554/eLi e.02242.008 Figu e 3. Con inued on nex page Human biology and medicine Whea on e al. eLi e 2014;3:e02242. DOI: 10.7554/eLi e.02242 7 o 18 Resea ch a icle inhibi o , Oligomycin A, diminished TMRE luo escence in Con ol-HCT 116 p53−/− cells ea ed wi h me o min sugges ing ha in he p esence o me o min, in ac cells main ain hei mi ochond ial memb ane po en ial by e e sal o he ATP syn hase (Figu e 4E). Ro enone i e e sibly inhibi s complex I, which con ibu es o i s high oxici y in i o. Because me - o min is well ole a ed, we sough o de e mine i me o min migh e e sibly bind o complex I. Saponin-pe meabilized Con ol-HCT 116 p53−/− cells we e ea ed wi h py u a e and mala e o main- ain he mi ochond ial inne memb ane po en ial. Me o min was hen added, ollowed by injec ion o ei he ADP o CCCP. Me o min inhibi ed ADP, bu no CCCP-s imula ed oxygen consump ion (Figu e 5A,B). As me o min accumula ion equi es mi ochond ial memb ane pola iza ion (Figu e 4A,B), hese esul s indica e ha me o min e e sibly inhibi s mi ochond ial complex I. I he me o min ha accumula ed in he mi ochond ial ma ix i e e sibly inhibi ed complex I, hen oxygen consump ion would ha e emained a enua ed in CCCP- ea ed cells a e me o min ea men . An eme ging unc ion o mi ochond ia dis inc om hei abili y o pe o m biosyn he ic and bio- ene ge ic eac ions is he gene a ion o H2O2, which p omo es signaling in no mal and cance cells (Hamanaka and Chandel, 2010; Sena and Chandel, 2012). Mi ochond ial complexes I and III p oduce supe oxide in o he mi ochond ial ma ix, whe e i is con e ed quickly o H2O2 by SOD2 (B and, 2010). Mi ochond ial complex III also gene a es supe oxide in o he mi ochond ial in e memb ane space whe e i escapes h ough VDACs o cy osol and is con e ed in o H2O2 by SOD1 (Han e al., 2003; Mulle e al., 2004). We measu ed p oduc ion and subsequen elease o H2O2 om isola ed mi ochond ia in he p esence o me o min, o enone, o an imycin A (complex III inhibi o ) using py - u a e and mala e as subs a es. Consis en wi h p e ious epo s, o enone and an imycin inc eased he elease o H2O2 om mi o- chond ia isola ed om Con ol-HCT 116 p53−/− cells (Figu e 6A,B; S -Pie e e al., 2002; Mulle e al., 2004). In con as , me o min did no subs an ially inc ease H2O2 elease, sugges ing ha me o min and o enone ac on di e en si es o complex I (Figu e 6A). When mi ochond ia we e isola ed om NDI1–HCT 116 p53−/− cells, only an imycin lead o a signi ican inc ease in H2O2 elease (Figu e 6B). P e ious epo s ha e shown ha me o min does no subs an ially inc ease H2O2 p oduc ion in iso- la ed li e mi ochond ia and ha me o min diminishes mi ochond ial H2O2 p oduc ion in esponse o pa aqua , which induces mi ochond ial ROS p oduc ion (Ba andie e al., 2006; Algi e e al., 2012). One biological consequence o mi ochond ial-gene a ed H2O2 is hypoxic s abiliza ion o he hypoxia-inducible ac o s (HIFs) (Chandel e al., 2000). HIFs a e in ol ed in me abolic adap a ion o umo cells o hypoxia (Semenza, 2012). Me o min educed hypoxic s abiliza ion o HIF-1α in Con ol- HCT 116 p53−/− bu no in NDI1-HCT 116 p53−/− (Figu e 6C,D). Me o min did no dec ease de e ox- amine (DFO) s abiliza ion o HIF-1α p o ein. DFO is an i on chela o known o di ec ly s abilize HIF-1α p o ein independen o ups eam signaling e en s. Me o min also signi ican ly diminished hypoxic ac i a ion o HIF-dependen a ge genes, ascula endo helial g ow h ac o (VEGF), and ca bonic anhyd ase 9 (CA9) in Con ol-HCT 116 p53−/− bu no in NDI1-HCT 116 p53−/− (Figu e 6E). Thus, me - o min is an e ec i e agen o educe hypoxic ac i a ion o HIF-1. Finally, we di ec ly es ed whe he umo cell au onomous inhibi ion o mi ochond ial complex I by me o min was equi ed o dec ease umo p og ession in i o. As ou NDI1-HCT 116 p53−/− cells a e e ac o y o mul iple e ec s o me o min in i o, we easoned ha i me o min ac ed di ec ly on mi ochond ial complex I wi hin he umo cells o educe umo igenesis hen NDI1-HCT 116 p53−/− xenog a umo s would no be inhibi ed in hei g ow h. Howe e , i me o min ac s a he o ganismal le el o diminish umo igenesis hen NDI1-HCT 116 p53−/− xenog a umo g ow h would be sup- p essed simila o con ol umo s. Con ol-HCT 116 p53−/− cells subcu aneously injec ed in o he le lank o nude mice apidly g ew in i o, while umo s om mice ed me o min h ough d inking wa e ad libi um s a ing 4 days pos -implan a ion exhibi ed a ma ked educ ion in g ow h (Figu e 7A,B). NDI1-HCT 116 p53−/− xenog a g ow h was esis an o me o min he apy (Figu e 7A,B), sugges ing The ollowing igu e supplemen s a e a ailable o igu e 3: Figu e supplemen 1. Me o min inhibi s mi ochond ial complex I o CCL16 cells. DOI: 10.7554/eLi e.02242.009 Figu e supplemen 2. Phen o min inhibi s mi ochond ial complex I o CCL16 cells. DOI: 10.7554/eLi e.02242.010 Figu e 3. Con inued Human biology and medicine Whea on e al. eLi e 2014;3:e02242. DOI: 10.7554/eLi e.02242 8 o 18 Resea ch a icle ha he me o min ca ies ou i s umo inhibi o y e ec s in a cance cell au onomous manne h ough inhibi ion o mi ochond ial complex I. Impo an ly, he consump ion o wa e con aining me o min was simila be ween con ol and NDI1 umo ba ing mice (Figu e 7C). T ansc ip s o he HIF a ge Figu e 4. Me o min inhibi ion o complex I equi es an in ac mi ochond ial inne memb ane po en ial. (A) Complex I (2 mM mala e, 10 mM py u a e)-d i en oxygen consump ion a e o saponin pe meabilized Con ol-HCT 116 p53−/− cells o e ime. A = 5 min pe meabilized cells we e ea ed wi h ei he 10 mM ADP o induce espi a- ion wi h an in ac mi ochond ial memb ane po en ial o (B) 10 µM CCCP o induce espi a ion in absence o mi ochond ial memb ane po en ial. A = 12 min 1 mM me o min was added o cells. A = 48 min an imycin A was added. (C) Complex I (2 mM mala e, 10 mM py u a e)-d i en oxygen consump ion a e o saponin-pe meabilized Con ol-HCT 116 p53−/− cells. A = 5 min pe meabilized cells we e ea ed wi h ei he 10 mM ADP o induce espi a ion wi h an in ac mi ochond ial memb ane po en ial o (D) 10 µM CCCP o induce espi a ion in absence o mi ochond ial memb ane po en ial. A = 15 min, 1 μM o enone was added o cells. A = 25 min an imycin A was added. (E) Mi ochond ial memb ane po en ial measu ed by TMRE s aining o Con ol-HCT116 p53−/− cells o (F) NDI1-HCT 116 p53−/− in he p esence o 1 mM Me o min, 10 µM CCCP o 2.5 µM Oligomycin A. E o ba s a e SEM (n = 4). * indica es signi icance p<0.05. DOI: 10.7554/eLi e.02242.011 Human biology and medicine Whea on e al. eLi e 2014;3:e02242. DOI: 10.7554/eLi e.02242 9 o 18 Resea ch a icle genes CA9 and VEGF we e diminished in con ol umo s ea ed wi h me o min bu no in NDI1 exp essing umo s (Figu e 7D,E). The blood glucose, plasma lac a e, insulin, and IGF-1 le el displayed no di e ences be ween he me o min- ea ed animals and con ol animals a he end o he s udy (Figu e 7— igu e supplemen 1), consis en wi h p e ious epo s (Tomimo o e al., 2008). To u he bols e ou conclusions, we examined umo g ow h o A549 cells exp essing NDI1 and shRNA a ge ing mammalian NDUFS3, a subuni o he human mi ochond ial complex I (Vogel e al., 2007). A549 cells a e null o he umo supp esso LKB1 and a e known o be esponsi e o me o min he apy (Rocha e al., 2011). Fu he mo e, LKB1-de icien umo s a e mo e suscep ible o he ela ed biguanide phen o min (Shackel o d e al., 2013). We eplaced he endogenous mi ochond ial complex I by exp essing he NDI1 p o ein in A549 cells s ably exp essing shRNA agains NDUFS3 (Figu e 7— igu e supplemen 2A), e e ed o as NDI1-shNDUFS3-A549 cells. The Con ol-A549 cells con ain he emp y ec o s wi h selec ion ma ke s BFP and pu omycin o NDI1 and shRNA, espec- i ely. NDI1-NDUFS3-A549 cells we e esis an o he me o min-media ed educ ion in cellula and mi ochond ial oxygen consump ion and cell p oli e a ion (Figu e 7— igu e supplemen 2B–G). Me o min also dec eased HIF-1a p o ein le els in con ol bu no in NDI1-NDUFS3-A549 cells (Figu e 7— igu e supplemen 3A). Con ol-A549 cells subcu aneously injec ed in o he le lank o nude mice apidly g ew in i o, while umo s om mice ed me o min h ough d inking wa e ad libi um s a ing 2 weeks p io o umo induc ion exhibi ed a ma ked educ ion in g ow h o e 45 days (Figu e 7— igu e supplemen 3B). By con as , NDI1-NDUFS3-A549 xenog a s we e comple ely esis an o me o min he apy, sugges ing ha me o min ca ies ou i s umo inhibi o y e ec s in a cell au onomous manne h ough inhibi ion o mi ochond ial complex I in hese cells (Figu e 7— igu e supplemen 3C). NDI1-NDUFS3-A549 xenog a s g ew slowe han con ol xenog a s in un ea ed mice. An al e na i e explana ion o he esis ance o NDI1-NDUFS3-A549 cells o me o min could be ha e ec s o me o min a e blun ed in slowe -g owing cells. Howe e , based on ou esul s om HCT116 cells in i o and ex ensi e analysis o A549 cells in i o, we ind ha i is likely ha NDI1 exp essing A549 cells a e also esis an o me o min in i o due o escue o complex I ac i i y by he NDI1 p o ein. Discussion The mechanisms by which me o min inhibi s cance g ow h is no ully unde s ood. Me o min has been p e iously shown o inhibi mi ochond ial complex I, ye i is no known whe he me o min exhibi s i s an i- umo e ec s h ough inhibi ion o complex I. I is impo an o no e ha many Figu e 5. Me o min e e sibly inhibi s mi ochond ial complex I. (A) Complex I (2 mM mala e, 10 mM py u a e)-d i en oxygen consump ion a e o saponin pe meabilized Con ol-HCT 116 p53−/− cells o e ime. A = 5 min pe meabilized cells we e exposed o 1 mM me o min. A = 25 min espi a ion was s imula ed wi h ei he 10 mM ADP o induce espi a ion wi h an in ac mi ochond ial memb ane po en ial o (B) 10 µM CCCP o induce espi a- ion lacking memb ane po en ial wi h 10 mM ADP. A = 42 min an imycin A was added. Fo mi ochond ial memb ane po en ial e o ba s a e SEM (n = 4). Fo oxygen consump ion a es, e o ba s a e s anda d de ia ion (n = 6). * indica es signi icance p<0.05. DOI: 10.7554/eLi e.02242.012 Human biology and medicine Whea on e al. eLi e 2014;3:e02242. DOI: 10.7554/eLi e.02242 16 o 18 Resea ch a icle be ween 0.5 µg/ml and 2 µg/ml (G aham e al., 2011). This ea men c ea ed ou expe imen al g oups: Con ol-HCT 116 p53−/− wi h H2O (n = 8), Con ol-HCT 116 p53−/− wi h Me o min (n = 8), NDI1-HCT 116 p53−/− wi h H2O (n = 8), NDI1-HCT 116 p53−/− wi h Me o min (n = 8). Expe imen s a e om wo independen coho s o ou mice each. Tumo s we e measu ed h ee imes pe week using calipe s and umo olume was de e mined using he equa ion (3.14/6 × L × W2). A he comple ion o he s udy, mice we e eu hanized and he umo s we e ex ac ed and weighed. Fo A549s, 3 × 106 Con ol-A549 o NDI1-NDUFS3-A549 cells we e injec ed in o he le lank o emale nude mice (Nu:Nu) adminis e ed me o min (250 mg/kg) in hei d inking wa e o 2 weeks p io o cell injec ion and con inuously adminis e ed h oughou he expe imen . All mouse wo k was done in acco dance wi h No hwes e n Uni e si y Ins i u ional Animal Ca e and Use Commi ee. S a is ical analysis Da a a e p esen ed as he mean ± SEM. S a is ical signi icance was de e mined using 1-way ANOVA wi h a Bon e oni pos es co ec ion, 2-way ANOVA when wo a iables we e p esen , o he s uden s es compa ing con ol o expe imen al condi ions o p<0.05. Fo all di e ences unco e ed, a s u- den es was pe o med o e i y di e ences be ween con ol and expe imen al g oups. Acknowledgemen s We a e g a e ul o he ollowing people o p o iding eagen s: E Sche le (CCL16-B2 cells), D Takao Yagi (CCL16-NDI1 cells). This wo k is suppo ed by g an s om he NIH (R01CA123067, 5P01HL071643) o NSC, NIH (ES015024) o GMM and Ve e ans Adminis a ion Me i Awa d (GRSB). The wo k was also suppo ed by NIH aining g an T32 GM08061 o LBS and 5T32HL076139-10 o SW. Addi ional in o ma ion Funding Funde G an e e ence numbe Au ho Na ional Ins i u es o Heal h RO1 CA123067 Na deep S Chandel Na ional Ins i u es o Heal h T32GM08061 Lucas B Sulli an Na ional Ins i u es o Heal h T32HL076139 Samuel E Weinbe g The unde s had no ole in s udy design, da a collec ion and in e p e a ion, o he decision o submi he wo k o publica ion. Au ho con ibu ions WWW, SEW, RBH, Concep ion and design, Acquisi ion o da a, Analysis and in e p e a ion o da a, D a ing o e ising he a icle; SS, LBS, EA, AG, Concep ion and design, Acquisi ion o da a, Analysis and in e p e a ion o da a; ED, Concep ion and design, Con ibu ed unpublished essen ial da a o eagen s; GMM, GRSB, NSC, Concep ion and design, Analysis and in e p e a ion o da a, D a ing o e ising he a icle E hics Animal expe imen a ion: Ins i u ional animal app o al: all mouse wo k was done in acco dance wi h No hwes e n Uni e si y Ins i u ional Animal Ca e and Use Commi ee app o ed p o ocol #2012-2840. Re e ences Algi e C, Moisee a O, Deschenes-Sima d X, Am ein L, Pe uccelli L, Bi man E, Violle B, Fe bey e G, Pollak MN. 2012. Me o min educes endogenous eac i e oxygen species and associa ed DNA damage. Cance P e en ion Resea ch 5:536–543. doi: 10.1158/1940-6207.CAPR-11-0536. Appleby RD, Po eous WK, Hughes G, James AM, Shannon D, Wei YH, Mu phy MP. 1999. Quan i a ion and o igin o he mi ochond ial memb ane po en ial in human cells lacking mi ochond ial DNA. Eu opean Jou nal o Biochemis y 262:108–116. doi: 10.1046/j.1432-1327.1999.00350.x. Bailey CJ, Tu ne RC. 1996. Me o min. The New England Jou nal o Medicine 334:574–579. doi: 10.1056/ NEJM199602293340906. Bailey CJ, Wilcock C, Sca pello JH. 2008. Me o min and he in es ine. Diabe ologia 51:1552–1553. doi: 10.1007/ s00125-008-1053-5. Human biology and medicine Whea on e al. eLi e 2014;3:e02242. DOI: 10.7554/eLi e.02242 17 o 18 Resea ch a icle Ba andie C, Guigas B, De aille D, El-Mi MY, Fon aine E, Rigoule M, Le e e XM. 2006. The ROS p oduc ion induced by a e e se-elec on lux a espi a o y-chain complex 1 is hampe ed by me o min. Jou nal o Bioene ge ics and Biomemb anes 38:33–42. doi: 10.1007/s10863-006-9003-8. Bell EL, Klimo a TA, Eisenba J, Mo aes CT, Mu phy MP, Budinge GR, Chandel NS. 2007. The Qo si e o he mi ochond ial complex III is equi ed o he ansduc ion o hypoxic signaling ia eac i e oxygen species p oduc ion. The Jou nal o Cell Biology 177:1029–1036. doi: 10.1083/jcb.200609074. Bi soy K, Possema o R, Lo bee FK, Bay ak a EC, Thi u P, Yucel B, Wang T, Chen WW, Clish CB, Saba ini DM. 2014. Me abolic de e minan s o cance cell sensi i i y o glucose limi a ion and biguanides. Na u e 508: 108–112. doi: 10.1038/na u e13110. Bi soy K, Saba ini DM, Possema o R. 2012. Un uning he umo me abolic machine: a ge ing cance me abo- lism: a bedside lesson. Na u e Medicine 18:1022–1023. doi: 10.1038/nm.2870. Bowke SL, Majumda SR, Veugele s P, Johnson JA. 2006. Inc eased cance - ela ed mo ali y o pa ien s wi h ype 2 diabe es who use sul onylu eas o insulin. Diabe es Ca e 29:254–258. doi: 10.2337/diaca e.29.02.06. dc05-1558. B and M. 2010. The si es and opology o mi ochond ial supe oxide p oduc ion. Expe imen al Ge on ology 45:466–472. doi: 10.1016/j.exge .2010.01.003. Buzzai M, Jones RG, Ama a adi RK, Lum JJ, DeBe a dinis RJ, Zhao F, Violle B, Thompson CB. 2007. Sys emic ea men wi h he an idiabe ic d ug me o min selec i ely impai s p53-de icien umo cell g ow h. Cance Resea ch 67:6745–6752. doi: 10.1158/0008-5472.CAN-06-4447. Chandel N, McClin ock D, Feliciano C, Wood T, Melendez J, Rod iguez A, Schumacke P. 2000. Reac i e oxygen species gene a ed a mi ochond ial complex III s abilize hypoxia-inducible ac o -1alpha du ing hypoxia: a mechanism o O2 sensing. The Jou nal o Biological Chemis y 275:25130–25138. doi: 10.1074/jbc.M001914200. Dowling RJ, Ni aula S, S ambolic V, Goodwin PJ. 2012. Me o min in cance : ansla ional challenges. Jou nal o Molecula Endoc inology 48:R31–R43. doi: 10.1530/JME-12-0007. El-Mi MY, Noguei a V, Fon aine E, A e e N, Rigoule M, Le e e X. 2000. Dime hylbiguanide inhibi s cell espi a ion ia an indi ec e ec a ge ed on he espi a o y chain complex I. The Jou nal o Biological Chemis y 275:223–228. doi: 10.1074/jbc.275.1.223. Emami Riedmaie A, Fisel P, Nies AT, Schae ele E, Schwab M. 2013. Me o min and cance : om he old medicine cabine o pha macological pi alls and p ospec s. T ends in Pha macological Sciences 34:126–135. doi: 10.1016/j. ips.2012.11.005. E ans JM, Donnelly LA, Emslie-Smi h AM, Alessi DR, Mo is AD. 2005. Me o min and educed isk o cance in diabe ic pa ien s. B i ish Medical Jou nal 330:1304–1305. doi: 10.1136/bmj.38415.708634.F7. Fend SM, Bell EL, Keible MA, Da idson SM, Wi h GJ, Fiske B, Maye s JR, Schwab M, Bellinge G, Csibi A, Pa naik A, Blouin MJ, Can ley LC, Gua en e L, Blenis J, Pollak MN, Olumi AF, Vande Heiden MG, S ephanopoulos G. 2013. Me o min dec eases glucose oxida ion and inc eases he dependency o p os a e cance cells on educ i e glu amine me abolism. Cance Resea ch 73:4429–4438. doi: 10.1158/0008- 5472.CAN-13-0080. Fo e z M, Heb a d S, Lecle c J, Za inpashneh E, So y M, Mi hieux G, Sakamo o K, And eelli F, Violle B. 2010. Me o min inhibi s hepa ic gluconeogenesis in mice independen ly o he LKB1/AMPK pa hway ia a dec ease in hepa ic ene gy s a e. The Jou nal o Clinical In es iga ion 120:2355–2369. doi: 10.1172/JCI40671. Gohil VM, She h SA, Nilsson R, Woj o ich AP, Lee JH, Pe occhi F, Chen W, Clish CB, Aya a C, B ookes PS, Moo ha VK. 2010. Nu ien -sensi ized sc eening o d ugs ha shi ene gy me abolism om mi ochond ial espi a ion o glycolysis. Na u e Bio echnology 28:249–255. doi: 10.1038/nb .1606. Goodwin PJ, Thompson AM, S ambolic V. 2012. Diabe es, me o min, and b eas cance : lilac ime? Jou nal o Clinical Oncology 30:2812–2814. doi: 10.1200/JCO.2012.42.3319. G aham GG, Pun J, A o a M, Day RO, Doogue MP, Duong JK, Fu long TJ, G een ield JR, G eenup LC, Ki kpa ick CM, Ray JE, Timmins P, Williams KM. 2011. Clinical pha macokine ics o me o min. Clinical Pha macokine ics 50:81–98. doi: 10.2165/11534750-000000000-00000. Hamanaka R, Chandel N. 2010. Mi ochond ial eac i e oxygen species egula e cellula signaling and dic a e biological ou comes. T ends in Biochemical Sciences 35:505–513. doi: 10.1016/j. ibs.2010.04.002. Han D, An unes F, Canali R, Re o i D, Cadenas E. 2003. Vol age- dependen anion channels con ol he elease o he supe oxide anion om mi ochond ia o cy osol. The Jou nal o Biological Chemis y 278:5557–5563. doi: 10.1074/jbc.M210269200. Inzucchi SE, Maggs DG, Spolle GR, Page SL, Ri e FS, Wal on V, Shulman GI. 1998. E icacy and me abolic e ec s o me o min and ogli azone in ype II diabe es melli us. The New England Jou nal o Medicine 338:867–872. doi: 10.1056/NEJM199803263381303. Jamu MC, Oli e C. 2010. Pe meabiliza ion o cell memb anes. Me hods in Molecula Biology 588:63–66. doi: 10.1007/978-1-59745-324-0_9. Lin X, Da id CA, Donnelly JB, Michaelides M, Chandel NS, Huang X, Wa io U, Weinbe g F, To mos KV, Fesik SW, Shen Y. 2008. A chemical genomics sc een highligh s he essen ial ole o mi ochond ia in HIF-1 egula ion. P oceedings o he Na ional Academy o Sciences o he Uni ed S a es o Ame ica 105:174–179. doi: 10.1073/pnas.0706585104. Memmo RM, Me cado JR, Maie CR, Kawaba a S, Fox SD, Dennis PA. 2010. Me o min p e en s obacco ca cinogen–induced lung umo igenesis. Cance P e en ion Resea ch 3:1066–1076. doi: 10.1158/1940-6207. CAPR-10-0055. Mille RA, Chu Q, Xie J, Fo e z M, Violle B, Bi nbaum MJ. 2013. Biguanides supp ess hepa ic glucagon signalling by dec easing p oduc ion o cyclic AMP. Na u e 494:256–260. doi: 10.1038/na u e11808. Human biology and medicine Whea on e al. eLi e 2014;3:e02242. DOI: 10.7554/eLi e.02242 18 o 18 Resea ch a icle Mullen AR, Whea on WW, Jin ES, Chen PH, Sulli an LB, Cheng T, Yang Y, Linehan WM, Chandel NS, DeBe a dinis RJ. 2012. Reduc i e ca boxyla ion suppo s g ow h in umou cells wi h de ec i e mi ochond ia. Na u e 481:385–388. doi: 10.1038/na u e10642. Mulle F, Liu Y, Van Remmen H. 2004. Complex III eleases supe oxide o bo h sides o he inne mi ochond ial memb ane. The Jou nal o Biological Chemis y 279:49064–49073. doi: 10.1074/jbc.M407715200. Na han DM, Buse JB, Da idson MB, Fe annini E, Holman RR, She win R, Zinman B, Ame ican Diabe es Associa ion, and Eu opean Associa ion o S udy o Diabe es. 2009. Medical managemen o hype glycemia in ype 2 diabe es: a consensus algo i hm o he ini ia ion and adjus men o he apy: a consensus s a emen o he Ame ican Diabe es Associa ion and he Eu opean Associa ion o he S udy o Diabe es. Diabe es Ca e 32:193–203. doi: 10.2337/dc08-9025. O a S, Ho igome K, Ishii T, Nakai M, Hayashi K, Kawamu a T, Kishino A, Taiji M, Kimu a T. 2009. Me o min supp esses glucose-6-phospha ase exp ession by a complex I inhibi ion and AMPK ac i a ion-independen mechanism. Biochemical and Biophysical Resea ch Communica ions 388:311–316. doi: 10.1016/j. bb c.2009.07.164. Owen MR, Do an E, Hales ap AP. 2000. E idence ha me o min exe s i s an i-diabe ic e ec s h ough inhibi ion o complex 1 o he mi ochond ial espi a o y chain. The Biochemical Jou nal 348:607–614. doi: 10.1042/0264-6021:3480607. P oc o WR, Bou de DL, Thakke DR. 2008. Mechanisms unde lying sa u able in es inal abso p ion o me o m- in. D ug Me abolism and Disposi ion: he Biological Fa e o Chemicals 36:1650–1658. doi: 10.1124/ dmd.107.020180. P yde KR, Hi s J. 2011. Supe oxide is p oduced by he educed la in in mi ochond ial complex I: a single, uni ied mechanism ha applies du ing bo h o wa d and e e se elec on ans e . The Jou nal o Biological Chemis y 286:18056–18065. doi: 10.1074/jbc.M110.186841. Reagan-Shaw S, Nihal M, Ahmad N. 2008. Dose ansla ion om animal o human s udies e isi ed. FASEB Jou nal 22:659–661. doi: 10.1096/ j.07-9574LSF. Rocha GZ, Dias MM, Ropelle ER, Oso io-Cos a F, Rossa o FA, Ve cesi AE, Saad MJ, Ca alhei a JB. 2011. Me o min ampli ies chemo he apy-induced AMPK ac i a ion and an i umo al g ow h. Clinical Cance Resea ch 17:3993–4005. doi: 10.1158/1078-0432.CCR-10-2243. Sca pello JH, Howle HC. 2008. Me o min he apy and clinical uses. Diabe es & Vascula Disease Resea ch 5:157–167. doi: 10.3132/d d .2008.027. Semenza GL. 2012. Hypoxia-inducible ac o s in physiology and medicine. Cell 148:399–408. doi: 10.1016/j. cell.2012.01.021. Sena LA, Chandel NS. 2012. Physiological oles o mi ochond ial eac i e oxygen species. Molecula Cell 48:158–167. doi: 10.1016/j.molcel.2012.09.025. Seo BB, Ki ajima-Iha a T, Chan EK, Sche le IE, Ma suno-Yagi A, Yagi T. 1998. Molecula emedy o complex I de ec s: o enone-insensi i e in e nal NADH- quinone oxido educ ase o Saccha omyces ce e isiae mi ochond ia es o es he NADH oxidase ac i i y o complex I-de icien mammalian cells. P oceedings o he Na ional Academy o Sciences o he Uni ed S a es o Ame ica 95:9167–9171. doi: 10.1073/ pnas.95.16.9167. Shackel o d DB, Ab E, Ge ken L, Vasquez DS, Seki A, Leblanc M, Wei L, Fishbein MC, Cze nin J, Mischel PS, Shaw RJ. 2013. LKB1 inac i a ion dic a es he apeu ic esponse o non-small cell lung cance o he me abolism d ug phen o min. Cance Cell 23:143–158. doi: 10.1016/j.cc .2012.12.008. Shimoda LA, Semenza GL. 2011. HIF and he lung: ole o hypoxia-inducible ac o s in pulmona y de elopmen and disease. Ame ican Jou nal o Respi a o y and C i ical Ca e Medicine 183:152–156. doi: 10.1164/ ccm.201009-1393PP. S -Pie e J, Buckingham JA, Roebuck SJ, B and MD. 2002. Topology o supe oxide p oduc ion om di e en si es in he mi ochond ial elec on anspo chain. The Jou nal o Biological Chemis y 277:44784–44790. doi: 10.1074/jbc.M207217200. S ephenne X, Fo e z M, Taleux N, an de Zon GC, Sokal E, Hue L, Violle B, Guigas B. 2011. Me o min ac i a es AMP-ac i a ed p o ein kinase in p ima y human hepa ocy es by dec easing cellula ene gy s a us. Diabe ologia 54:3101–3110. doi: 10.1007/s00125-011-2311-5. Sulli an LB, Ma inez-Ga cia E, Nguyen H, Mullen AR, Du ou E, Suda shan S, Lich JD, Debe a dinis RJ, Chandel NS. 2013. The p o o-oncome aboli e uma a e binds glu a hione o ampli y ROS-dependen signaling. Molecula Cell 51:236–248. doi: 10.1016/j.molcel.2013.05.003. Tomimo o A, Endo H, Sugiyama M, Fujisawa T, Hosono K, Takahashi H, Nakajima N, Nagashima Y, Wada K, Nakagama H, Nakajima A. 2008. Me o min supp esses in es inal polyp g ow h in ApcMin/+ mice. Cance Science 99:2136–2141. doi: 10.1111/j.1349-7006.2008.00933.x. Violle B, Guigas B, Sanz Ga cia N, Lecle c J, Fo e z M, And eelli F. 2012. Cellula and molecula mechanisms o me o min: an o e iew. Clinical Science 122:253–270. doi: 10.1042/CS20110386. Vogel RO, Die e en CE, an den Heu el LP, Willems PH, Smei ink JA, Koopman WJ, Nij mans LG. 2007. Iden i ica ion o mi ochond ial complex I assembly in e media es by acing agged NDUFS3 demons a es he en y poin o mi ochond ial subuni s. The Jou nal o Biological Chemis y 282:7582–7590. doi: 10.1074/jbc.M609410200. Weinbe g F, Hamanaka R, Whea on W, Weinbe g S, Joseph J, Lopez M, Kalyana aman B, Mu lu G, Budinge G, Chandel N. 2010. Mi ochond ial me abolism and ROS gene a ion a e essen ial o K as-media ed umo igenici y. P oceedings o he Na ional Academy o Sciences o he Uni ed S a es o Ame ica 107:8788–8793. doi: 10.1073/pnas.1003428107.