Alternative oxidase-mediated respiration prevents lethal mitochondrial cardiomyopathy

Resea ch A icle Al e na i e oxidase-media ed espi a ion p e en s le hal mi ochond ial ca diomyopa hy Jayasimman Rajend an 1,2 , Janne Pu honen 1,2 , Saa a Tegelbe g 1,3,4 , Olli-Pekka Smolande 5 , Ma hias Mö gelin 6 , Jan Rozman 7,8 , Vale ie Gailus-Du ne 7 , Helmu Fuchs 7 , Ma in H abe de Angelis 7,8,9 , Pe i Au inen 5 , Ee o Me aala 10 , Howa d T Jacobs 5,11 , Ma en Szibo 5,11 , Vine a Fellman 1,3,12 & Jukka Kallijä i 1,2,* Abs ac Al e na i e oxidase (AOX) is a non-mammalian enzyme ha can bypass blockade o he complex III-IV segmen o he espi a o y chain (RC). We c ossed a Ciona in es inalis AOX ansgene in o RC complex III (cIII)-de icien Bcs1l p.S78G knock-in mice, displaying mul iple isce al mani es a ions and p ema u e dea h. The homozygo es exp essing AOX we e iable, and hei median su i al was ex ended om 210 o 590 days due o pe manen p e en ion o le hal ca diomyopa hy. AOX also p e en ed enal ubula a ophy and ce eb al as ogliosis, bu no li e disease, g ow h es ic ion, o lipodys ophy, sugges ing dis inc issue- speci ic pa hogene ic mechanisms. Assessmen o eac i e oxygen species (ROS) p oduc ion and damage sugges ed ha ROS we e no ins umen al in he escue. Ca diac mi ochond ial ul as uc- u e, mi ochond ial espi a ion, and pa hological ansc ip ome and me abolome al e a ions we e essen ially no malized by AOX, showing ha he es o ed elec on low ups eam o cIII was su i- cien o p e en ca diac ene ge ic c isis and de imen al decom- pensa ion. These indings demons a e he alue o AOX, bo h as a mechanis ic ool and a po en ial he apeu ic s a egy, o cIII de iciencies. Keywo ds BCS1L; complex III; GRACILE synd ome; mi ochond ial diso de ; espi a o y chain Subjec Ca ego ies Ca dio ascula Sys em; Gene ics, Gene The apy & Gene ic Disease DOI 10.15252/emmm.201809456 | Recei ed 8Augus 2018 | Re ised 8No embe 2018 | Accep ed 12 No embe 2018 | Published online 10 Decembe 2018 EMBO Mol Med (2019)11:e9456 See also: A Saada (Janua y 2019) In oduc ion Mi ochond ial diso de s a e he mos common class o inhe i ed e o s o me abolism. Howe e , e ec i e ea men s a e lacking, and hei clinical managemen emains la gely suppo i e (P e e e al, 2013). In pa ien s wi h RC cIII (ubiquinol:cy och ome c oxido educ ase) de iciency, mu a ions in se e al genes encoding ei he cIII subuni s o assembly ac o s ha e been iden i ied. These comp omise cIII enzyma ic ac i i y and esul in a wide a ie y o clinical mani es a ions (Fe nandez-Viza a & Ze iani, 2015). BCS1L mu a ions a e he mos common cause o cIII de iciency, wi h a i- ous neona al and adul pheno ypes desc ibed wo ldwide (Fe nan- dez-Viza a & Ze iani, 2015), he mos se e e and p e alen o hem being GRACILE synd ome ( e al g ow h es ic ion, aminoacidu ia, choles asis, li e i on o e load, lac ic acidosis, and ea ly dea h du ing in ancy) (Fellman e al, 1998; Visapa ¨a ¨e al, 2002). BCS1L is a mi ochond ial inne memb ane anslocase equi ed o Rieske i on–sul u p o ein (RISP, UQCRFS1) opogenesis and inco po a ion in o cIII (Nob ega e al, 1992; C ucia e al, 1999). Homozygous Bcs1l c.A232G (Bcs1l p.S78G ) knock-in mice bea ing he GRACILE synd ome-analogous mu a ion ecapi ula e many o he clinical mani es a ions, including g ow h ailu e, p og essi e hepa opa hy, kidney ubulopa hy, and, in a C57BL/6JC lBomTac backg ound, sho su i al o 35 days (Le e ´en e al, 2011; Ko a sky e al, 2012; 1Folkhälsan Resea ch Cen e , Helsinki, Finland 2Clinicum, Facul y o Medicine, Uni e si y o Helsinki, Helsinki, Finland 3Depa men o Clinical Sciences, Lund, Pedia ics, Lund Uni e si y, Lund, Sweden 4Molecula Neu ology Resea ch P og am and Neu oscience Cen e , Uni e si y o Helsinki, Helsinki, Finland 5Ins i u e o Bio echnology, Uni e si y o Helsinki, Helsinki, Finland 6Di ision o In ec ion Medicine, Clinical Sciences, Lund Uni e si y, Lund, Sweden 7Ge man Mouse Clinic, Ins i u e o Expe imen al Gene ics, Helmhol z Zen um München, Ge man Resea ch Cen e o En i onmen al Heal h, Neuhe be g, Ge many 8Ge man Cen e o Diabe es Resea ch (DZD), Neuhe be g, Ge many 9Chai o Expe imen al Gene ics, Cen e o Li e and Food Sciences Weihens ephan, TU Munich, F eising-Weihens ephan, Ge many 10 Depa men o Pha macology, Facul y o Medicine, Uni e si y o Helsinki, Helsinki, Finland 11 Facul y o Medicine and Li e Sciences, Uni e si y o Tampe e, Tampe e, Finland 12 Child en’s Hospi al, Helsinki Uni e si y Hospi al, Uni e si y o Helsinki, Helsinki, Finland *Co esponding au ho . Tel: +358 504487006; E-mail: jukka.kallija [email p o ec ed] ª2018 The Au ho s. Published unde he e ms o he CC BY 4.0license EMBO Molecula Medicine 11:e9456 |2019 1o 19 Published online: Decembe 10, 2018 Rajend an e al, 2016; Pu honen e al, 2017). In he sligh ly di e en C57BL/6JC l subs ain, he homozygo es de elop he same ea ly mani es a ions bu do no succumb o he ea ly me abolic c isis. This ex ends hei su i al o o e 150 days (Pu honen e al, 2017) and b ings addi ional la e -onse pheno ypes, such as ce eb al as ogliosis (Tegelbe g e al, 2017). Unde physiological condi ions, quinols ha anspo elec ons in he mi ochond ial inne memb ane a e e icien ly oxidized by cIII, wi h elec on ans e ia cy och ome c and cy och ome c oxidase (complex IV, cIV) o oxygen (B and, 2010; El-Khou y e al, 2014). Howe e , plan s and some lowe o ganisms, bu no mammals, exp ess al e na i e oxidases (AOXs) ha ans e elec ons di ec ly om quinols o oxygen wi hou p o on ansloca ion. Thei main ole is o main ain elec on low when he cIII-cIV segmen o he RC is impai ed, limi ing p oduc ion o ROS and suppo ing edox and me abolic homeos asis (McDonald & Vanle be ghe, 2004; El-Khou y e al, 2014). Ciona in es inalis AOX has been cloned and exp essed in human cul u ed cells (Hakkaa e al, 2006), ui lies and mice (El-Khou y e al, 2013; Szibo e al, 2017). In hese models, AOX is ine unde non-s essed condi ions, mos likely because i accep s elec ons only when he quinone pool is highly educed (Hoe nagel & Wiskich, 1998; Cas o-Gue e o e al, 2004), such as unde inhibi ion o o e load o cIII o cIV (Dassa e al, 2009). Acco dingly, upon inhibi ion o cIII o cIV by mu a ions o chemical inhibi o s, ec opic AOX can main ain espi a ion and p e en cell dea h (Dassa e al, 2009; Fe nandez-Ayala e al, 2009). We se ou o es whe he AOX exp ession could p e en he de i- men al e ec s o cIII de iciency in a mammalian model, by es o ing elec on low ups eam o cIII. To his end, we c ossed mice ca ying a b oadly exp essed AOX ansgene (Szibo e al, 2017) wi h he Bcs1l c.A232G mice and assessed disease p og ession, o gan mani es a- ions, and me abolism in he homozygo es wi h and wi hou AOX exp ession. Resul s B oadly exp essed AOX iples he li e span o cIII-de icien Bcs1l p.S78G mice To assess he e ec o cIII bypass on he su i al and issue mani- es a ions in cIII-de icien mice, we b ed coho s o wild- ype and Bcs1l mu an mice wi h o wi hou a Ciona in es inalis AOX ans- gene. He ea e , we will e e o he Bcs1l p.S78G homozygo es as GRAC (as an abb e ia ion o GRACILE synd ome) mice. The Bcs1l p.S78G homozygo es ca ying AOX ansgene will be e e ed o as GROX mice (GRAC +AOX). Figu e 1A shows a imeline o he appea ance o he p e iously epo ed and no el pheno ypes in GRAC mice, as well as he assessmen s included in his s udy. The GRAC mice eached he c i e ia o eu hanasia be ween pos na al day 180 (P180) and P220, wi h median su i al o P210 (Fig 1B). In con as , he GROX mice showed no signs o e minal de e io a ion o spon aneous dea hs a P200 and su i ed o a median age o 590 days (Fig 1B). To assess whe he he ex ended su i al was due o an o e all imp o emen in ene gy me abolism, we measu ed g ow h, whole-body me abolism, and body composi ion in young adul mice. The GRAC mice we e g ow h es ic ed (Fig 1C and E) and had inc eased lac a e- o-glucose a io (Fig 1D), low a mass (Fig 1F), bone densi y (Fig 1G), and hea p oduc ion (Fig 1H) and, in emales, low espi a o y exchange a io (Fig 1I). Unexpec edly, AOX had no o only small e ec on hese pa ame e s (Fig 1C–I), sugges ing ha he AOX-media ed ex ension o su i al depended on a issue o cell- ype speci ic pa hology a he han whole-body ene gy me abolism. AOX pe manen ly p e en s le hal ca diomyopa hy and alle ia es enal and ce eb al mani es a ions His opa hological analysis o au opsy samples om end s age (P200) mice showed wo no el pheno ypes no p e iously epo ed in s udies o younge GRAC mice: co ical kidney a ophy and ca diomegaly wi h dila ed en icles (Fig 2A and B). Fib osis was p ominen in li e , kidney, and hea (Fig 2C and D). Suspec ing ca diomyopa hy as he cause o dea h, we assessed ca diac unc- ions a se e al ime poin s. Echoca diog aphy showed minimal unc ional changes a P150 (Fig EV1A–F), bu se e e dila ed ca diomyopa hy, ocal ib osis, dec eased ejec ion ac ion, and ac ional sho ening a P200 (Fig 2A and F), indica ing end-s age ca diomyopa hy. S ikingly, he GROX li e ma es had no mal hea size (Fig 2A and B), no ib osis (Fig 2C and D), and o e ly no mal ca diac unc ion (Fig 2E and F), explaining hei ex ended su i al. mRNA exp ession o key ma ke s o ca diac hype ophy and ib osis was signi ican ly al e ed al eady in he p esymp- oma ic (P150) GRAC hea s, and hese changes we e la gely p e en ed in he GROX mice (Fig 2G). A kidney s ess es wi h sal -en iched (6% w/w) chow s a ing a P150 had no e ec on ca diac unc ion a P200, o on su i al, and blood p essu e was only inc eased a he end s age in he sal - ed mice when compa ed o P150 baseline (Fig EV1A–F), consis en wi h a p ima y ca diomyopa hy. Rema kably, he GROX mice had no mal-sized and non- ib o ic hea h oughou hei li e span, o o e P600 (Fig EV2A–D). Kidneys o GRAC mice showed p oximal ubulopa hy wi h ib osis (Fig 2C and D) and ubula degene a ion (dec eased ubu- la mass) a P200 (Appendix Fig S1). In GROX mice, he kidney mass and appa en ubula olume we e p ese ed (Fig 2A and B). On he basis o he p oli e a ion ma ke Ki67 and apop osis ma ke clea ed caspase-3, his was likely due o dec eased apop- osis a he han inc eased egene a ion (Appendix Fig S1I and J). Howe e , AOX had only mino e ec on o he his ological lesions (Appendix Fig S1A–G) o unc ional pa ame e s: albumin- u ia, hema u ia, and u ina y c ea inine (Fig 2H and I, Appendix Fig S1H). A P600, he kidneys we e se e ely ib o ic bu s ill o no mal size, indica ing long- e m p o ec ion om ubula a ophy (Fig EV2A, B and D). Su p isingly, AOX had no e ec on he li e ib osis (Fig 2C and D) o he ele a ed li e enzymes (Fig 2J and K) a end s age. The cause o dea h o he GROX mice emains unknown, bu e en ual de e io a ion due o he p og essing kidney and li e disease is an ob ious possible explana ion. GRACILE synd ome pa ien s ha e no encephalopa hy, bu he b ains o GRAC mice show peculia ocal as ogliosis in he p ima y ba el ield o he soma osenso y co ex (S1BF) (Tegelbe g e al, 2017). S aining o glial ib illa y acidic p o ein (GFAP) showed ha he as ogliosis was almos ully p e en ed by AOX a P200 (Fig 3A). 2o 19 EMBO Molecula Medicine 11:e9456 |2019 ª2018 The Au ho s EMBO Molecula Medicine AOX escues cIII de iciency in mice Jayasimman Rajend an e al Published online: Decembe 10, 2018 A BC DE F GH I Figu e 1. AOX exp ession p olongs he su i al o cIII-de icien Bcs1l p.S78G mice wi hou a ec ing g ow h o whole-body me abolism. A Schema ic p esen a ion o he mul io gan mani es a ions, desc ibed in his s udy o p e iously, in homozygous Bcs1l c.A232G (GRAC) mice, and he ime poin s o he in es iga ions pe o med in his s udy. B Su i al cu es o homozygous Bcs1lmu an mice wi hou (GRAC) and wi h (GROX) al e na i e oxidase (AOX) exp ession (n=18–21/g oup). The median su i al o GRAC mice was 210 days and o GROX mice was 589 days wi h no gende di e ence. C Weigh o mice om P50 o P200 (n>10/g oup) and a P600 (n=4–6/g oup). D Blood lac a e- o-glucose a io a P200 (n=8/g oup). E–G Dual-ene gy X- ay abso p iome y (DEXA) analysis (n=10/g oup) o (E) lean mass, (F) a mass, and (G) bone mine al densi y a P98. H, I Indi ec calo ime ic measu emen (n=10/g oup) o (H) hea p oduc ion and (I) espi a o y exchange a io a P77. Da a in o ma ion: The su i al da a we e analyzed using log- ank (Man el–Cox) es (P<0.0001). Ba g aphs ep esen mean SD. The da a (D–I) we e analyzed using K uskal–Wallis and Mann–Whi ney U- es s o selec ed compa ison. Signi ican di e ences be ween g oups (P- alue) a e indica ed on g aphs. ª2018 The Au ho s EMBO Molecula Medicine 11:e9456 |2019 3o 19 Jayasimman Rajend an e al AOX escues cIII de iciency in mice EMBO Molecula Medicine Published online: Decembe 10, 2018 AOX p ese es mi ochond ial ul as uc u e in escued issues P e ious s udies ha e shown dis up ed hepa ic mi ochond ial ul as uc u e in younge Bcs1l mu an mice (Le e ´en e al, 2011; Pu honen e al, 2017). In he P150 GRAC mice, elec on mic oscopy showed abno mal mi ochond ial ul as uc u e in hepa ocy es as well as in kidney ubula cells, and, impo an ly, in ca diomyocy es al eady a onse o he ca diomyopa hy (Fig 3B). The mi ochond ia A B C DE GH I J K F Figu e 2. 4o 19 EMBO Molecula Medicine 11:e9456 |2019 ª2018 The Au ho s EMBO Molecula Medicine AOX escues cIII de iciency in mice Jayasimman Rajend an e al Published online: Decembe 10, 2018 we e smalle and con ained ewe c is ae, which we e hicke han in WT mi ochond ia. AOX ully o pa ially p e en ed hese changes in ca diomyocy es and kidney ubula cells, bu no in hepa o- cy es, (Fig 3C–F), co ela ing ai h ully wi h he escue o issue pa hology. AOX elie es ca diac me abolic s ess and p e en s decompensa ion Seeking a mechanism o he ema kably issue-speci ic escue e ec s o AOX, we pe o med ansc ip omics and me abolomics a P150, a he onse o he ca diomyopa hy. The h ee a ec ed issues showed ma ked global ansc ip ional changes, including in he GRAC hea despi e i being unc ionally no mal a his s age (Fig 4A–C). In co ela ion wi h he his ological indings, AOX no malized he exp ession o only ew genes in he li e , bu o abou 25% o dys egula ed genes in he kidney and abou 50% in he hea (Appendix Fig S2A–C). The mos obus ly up egula ed gene se s we e ela ed o ex acellula ma ix o ganiza- ion, a signa u e o issue emodeling and ib osis (Fig 4D–F, Appendix Table S1), and cell cycle in li e and kidney, bu no in he hea (Fig 4D–F) e lec ing di e en ial egene a i e capaci y o hese issues. As expec ed, ene gy me abolism- ela ed gene exp es- sion was al e ed in all h ee issues (Fig 4D–F, Appendix Fig S2D). Al e na i e oxidase almos ully p e en ed hese changes in hea , bu no in kidney o li e (Fig 4D–F). No ably, AOX had a signi i- can e ec on gene exp ession in he Bcs1l wild- ype hea , includ- ing up egula ion o genes ela ed o mi ochond ial unc ion (Fig 4D, G, I and Appendix Fig S2D). Exp ession o he majo ca diac me abolic egula o HIF-1aand he me abolic s ess-indu- cible ansc ip ional egula o s ATF3 and ATF4 (Kal on e al, 2017; Qui os e al, 2017) was ele a ed in GRAC hea s and no malized by AOX (Fig 4J and K), indica ing elie ed me abolic s ess. Up egula ion o PGC-1a, he mas e egula o o mi o- chond ial biogenesis, and he mi ochond ial ansc ip ion ac o TFAM, in he GRAC hea sugges ed an a emp o compen- sa o y mi ochond ial biogenesis (Fig 4G–I). Howe e , Wes e n blo analysis using he abundance o RC complex subuni s as a p oxy showed no signi ican changes in mi ochond ial mass be ween he g oups (Fig EV3I) in any issue. We p e iously iden i ied an up egula ed se o genes, which we designa ed cIII s ess signa u e, in P45 Bcs1l mu an li e s (Pu honen e al, 2017). This gene se was also highly up egula ed in all h ee GRAC issues. Despi e escue o kidney ubula mass, AOX ampli ied he cIII s ess signa u e speci ically in his issue, bu no in he hea o li e (Fig 4L). Al e na i e oxidase is unde he s ong syn he ic CAG p omo e and exp essed in all issues in he Rosa26 AOX mice (Szibo e al, 2017). In ou mice, AOX mRNA exp ession was simila in hea and li e and somewha lowe in kidney (Fig EV3H). In o al issue lysa es om he AOX mice, he amoun o AOX p o ein was conside - ably highe in hea han in li e o kidney (Fig EV3H). Howe e , his di e ence was mainly due o he highe mi ochond ial mass in hea , as shown by he mi ochond ial loading con ol VDAC1 and also by mos espi a o y chain subuni s (Fig EV3I). In e es ingly, he amoun o AOX p o ein was a ec ed by he Bcs1l mu a ion (AOX s. GROX mice) so ha he amoun in he h ee GROX issues was almos iden ical (Fig EV3H), which essen ially ules ou ha he di e ences in escue would be due o di e en le els o AOX exp ession. The me abolomics e ealed only modes ca diac me aboli e changes a he onse (P150) o he ca diomyopa hy (Fig 5A, Appendix Table S2). Ne e heless, se e al h ee-ca bon glycoly ic in e media es we e deple ed, and hese ended o be no malized by AOX (Fig 5A). In line wi h his, he gene exp ession o PPAR-1a, he majo d i e o a y acid u iliza ion, was down egula ed in GRAC bu no mal in GROX hea (Fig 4G). The TCA cycle me aboli es mala e and uma a e, adenyla e ene gy cha ge, and NADH/NAD + a io, which all could be a ec ed by bo h he cIII blockade and AOX, we e no changed in he GRAC o GROX hea issue (Fig 5C–F). In e es - ingly, he amino acid p oline, which has been long known o accu- mula e in condi ions wi h lac ic academia (Kowalo e al,1977),was below de ec ion limi in WT hea bu inc eased o app oxima ely 200 nmol/g in GRAC hea . The p oline accumula ion was pa ially p e en ed by AOX (Fig 5G). Concu en ly, glu ama e, he biosyn- he ic p ecu so o p oline, was dec eased o abou 50% in GRAC mice and no malized o WT le el in GROX hea issue (Fig 5A). The mRNA le els o glu ama e c-semialdehyde syn he ase (P5CS, ALDH18A1) and py oline-5-ca boxyla e educ ase 1 (P5CR, PYCR1), bo h d i ing p oline syn hesis om glu ama e, we e up egula ed 1.8- old and 1.7- old, espec i ely, in he GRAC hea (Fig 5H and I). Ta ge ed me abolomics o li e issue (Fig 5B, Appendix Table S2) showed signi ican me aboli e changes cha ac e is ic o ailing ene gy me abolism and glycogen deple ion, as p e iously shown in ju enile mice (Ko a sky e al, 2012). These included dec eased hexose phos- pha es, se e al o he glycoly ic in e media es, ace yl-CoA and NAD + , and ele a ed TCA cycle in e media es and amino acids, he la e sugges ing p o ein deg ada ion o uel. NADH/NAD + a io was inc eased (Fig 5F), in line wi h ou p e iously published da a show- ing dec eased hepa ic NAD + (Pu honen e al, 2018). P oline le el was no changed in li e (Appendix Table S2). Al e na i e oxidase had only a mino e ec on he hepa ic me aboli e le els (Fig 5B, E, and F). ◀Figu e 2. AOX p e en s le hal ca diomyopa hy and p og ession o enal ubulopa hy o kidney a ophy. A Hema oxylin–eosin-s ained c oss sec ions o hea and kidney a P200. B Weigh o hea and kidney a P200 (n=4/g oup). C, D Si ius Red s aining (C) o ib osis in li e , hea , and kidney a P200, and (D) quan i ica ion o ib osis in myoca dium, kidney co ex, and li e (n=5–7). E, F Echoca diog aphy da a (n=4–6/g oup) showing (E) ejec ion ac ion (EF) and ac ional sho ening (FS). (F) Sys olic (LV Vol;S) and dias olic (LV Vol;D) le en icle olume in mice. G Exp ession o ca diac hype ophy and ib osis-associa ed genes in he p esymp oma ic (P150) hea (n=6/g oup). H, I 24-h exc e ion o (H) albumin and (I) c ea inine in u ine a P200 (n=4/g oup). J, K Li e enzymes (J) alanine amino ans e ase (ALT) and (K) alkaline phospha ase (ALP) in plasma a P200 (n=8/g oup). Da a in o ma ion: Ba g aphs ep esen mean SD. S a is ics: one-way ANOVA ollowed by Tukey’s es ( o g aphs B, D, G, H, J, and K), one-way ANOVA ollowed by unpai ed - es wi h Welch’s co ec ion ( o g aphs E and F), and Mann-Whi ney U- es s ( o g aph I). ª2018 The Au ho s EMBO Molecula Medicine 11:e9456 |2019 5o 19 Jayasimman Rajend an e al AOX escues cIII de iciency in mice EMBO Molecula Medicine Published online: Decembe 10, 2018 A BC D F E Figu e 3. AOX amelio a es ce eb al as ogliosis and main ains mi ochond ial ul as uc u e in escued issues. A GFAP s aining o ce eb al as ocy es. The ba el ield o he p ima y soma osenso y co ex (S1BF) is highligh ed wi h b acke s. B Mi ochond ial ul as uc u e in ca diomyocy es, kidney ubula cells, and hepa ocy es a P200 as isualized by elec on mic oscopy. Scale ba 1lm. C–F (C) A e age c oss-sec ional a ea o mi ochond ion, (D) numbe o c is ae pe mi ochond ion, (E) c is a hickness, and (F) a e age dis ance be ween c is ae in mi ochond ia (n=3mice/g oup). Da a in o ma ion: Ba g aphs ep esen mean SD. S a is ics o g aphs (C–F): one-way ANOVA ollowed by Tukey’s es . 6o 19 EMBO Molecula Medicine 11:e9456 |2019 ª2018 The Au ho s EMBO Molecula Medicine AOX escues cIII de iciency in mice Jayasimman Rajend an e al Published online: Decembe 10, 2018 ABC DEF GIH JL K Figu e 4. AOX mi iga es ca diac, bu no hepa ic o enal, me abolic s ess- ela ed gene exp ession changes. A–C P incipal componen analysis o ansc ip ome da a om (A) hea , (B) kidney, and (C) li e . D–F Hea map isualiza ion o pa hway en ichmen analysis (Reac ome da abase). Full pa hway analysis is p o ided in Appendix Table S1. Benjamin–Hochbe g FDR- co ec ed P- alues a e colo labeled as indica ed o he h ee compa isons. G–K Gene exp ession o majo ansc ip ional egula o s o ene gy me abolism (PPAR-a, HIF-1a), mi ochond ial biogenesis (PGC-1a, TFAM), and s ess esponses (ATF4) in hea a P150 (n=6/g oup). L Exp ession o cIII s ess signa u e genes in hea , kidney, and li e (n=6/g oup). E o ba s ep esen 95%con idence in e al o mean di e ence. Da a in o ma ion: Box plo s (G–K) ep esen qua iles, maximum alue, and minimum alue ( ela i e old change “FC” o WT). S a is ics: one-way ANOVA ollowed by Tukey’s es . ª2018 The Au ho s EMBO Molecula Medicine 11:e9456 |2019 7o 19 Jayasimman Rajend an e al AOX escues cIII de iciency in mice EMBO Molecula Medicine Published online: Decembe 10, 2018 AB CDE F HIG Figu e 5. AOX has li le e ec on ca diac ene gy me aboli es bu a enua es p oline accumula ion a onse o disease. A Signi ican ly al e ed me aboli es (FDR <0.2;P<0.05,n=5/g oup) in p esymp oma ic (P150) hea issue o GRAC ( ed) and GROX (g een) mice. B Signi ican ly al e ed me aboli es (FDR <0.2;P<0.05,n=5/g oup) in li e issue o GRAC ( ed) and GROX (g een) mice. C–F Concen a ions o he TCA cycle in e media es (C) uma a e and (D) mala e, (E) adenyla e ene gy cha ge, and (F) NADH/NAD + a io (n=5/g oup). G–I (G) P oline concen a ion in hea issue a P200 (# below de ec ion limi ) and mRNA exp ession o p oline syn hesis- ela ed genes (n=6/g oup) (H) Aldh18a1and (I) Pyc 1a P150. Da a in o ma ion: Ba g aphs (A–G) ep esen mean SD, and box plo s o mRNA exp essions ep esen qua iles, minimum alue, and maximum alue ( ela i e old change “FC” o WT). S a is ics o g aphs (A, B, F, and H): one-way ANOVA ollowed by Tukey’s es . 8o 19 EMBO Molecula Medicine 11:e9456 |2019 ª2018 The Au ho s EMBO Molecula Medicine AOX escues cIII de iciency in mice Jayasimman Rajend an e al Published online: Decembe 10, 2018 AOX es o es ca diac mi ochond ial espi a ion To in es iga e whe he he AOX escue e ec was linked o imp o ed RC assembly and unc ion, we i s pe o med blue na i e gel elec opho esis (BNGE) analyses. These con i med he loss o RISP om ee cIII dime (cIII 2 ) and om he cI-cIII 2 supe complex (SC1) in GRAC issues (Fig 6A). Su p isingly, he amoun o his ully assembled cIII 2 was signi ican ly inc eased in he GROX hea when compa ed o GRAC (Fig 6B and C). In he GROX li e and kidney, he change was opposi e, wi h lowe amoun o RISP in cIII 2 (Fig 6B and C). Spec opho ome ically measu ed cIII ac i i y in isola ed mi o- chond ia was dec eased o <50% o WT alues in all h ee issues a P150, a h eshold o he appea ance o hepa ic pa hology in ju e- nile mice (Le e ´en e al, 2011). Despi e ha AOX should heo e i- cally no a ec cIII ac i i y, i was pa ially es o ed in GROX hea mi ochond ia (Fig 6F), in line wi h he imp o ed RISP assembly. The ac i i ies o cI, cII, and cIV we e unchanged in hea and li e (Fig 6D, E and G). Respi ome y con i med AOX-media ed espi a ion in all h ee GROX issues a P150, as measu ed using he AOX inhibi o n-p opyl galla e (nPG). Respi a ion did no espond o nPG in AOX mice (wi h wild- ype Bcs1l and no mal cIII ac i i y), which indica es ha AOX was ca aly ically ac i e only in he mu an s (Fig 6H). The cIII dys unc ion did no limi s a e 3 espi a ion o li e and kidney mi o- chond ia (Fig 6I and J) when d i en by cI-linked subs a es (mala e, py u a e, and glu ama e o gene a e NADH), bu clea ly did so in he hea (Fig 6I). Fu he s imula ion o espi a ion h ough cII by addi ion o succina e e ealed cIII de iciency in all h ee issues in GRAC mice, wi h he mos se e e e ec in he hea (Fig 6J). Rema kably, cI- and cII-linked s a e 3 espi a ion (mala e, py u a e, glu ama e, and succina e a ADP sa u a ion) we e escued o wild- ype le el in GROX hea mi ochond ia (Fig 6I and J). A simila e ec was ound in s a e 3 espi a ion o hea and li e a P200 (Fig EV4A and B). AOX does no a ec ROS damage o de ense bu no malizes ca diac NO- ela ed gene exp ession Limi ing excessi e ROS p oduc ion om he RC has been posi ed as a majo mechanism o ac ion o AOX (El-Khou y e al, 2014). To assess ROS p oduc ion in he a ec ed GRAC issues, we measu ed hyd ogen pe oxide (H 2 O 2 ) emission om isola ed mi ochond ia a P150. The Amplex Red-pe oxidase assay showed ha o al H 2 O 2 emission was dec eased in GROX hea and kidney as compa ed o GRAC (Fig 7A). Howe e , analysis o gene exp ession ela ed o ROS de ense and damage showed ha glu a hione syn hesis and conjuga ion we e up egula ed only in he GRAC li e (Fig 7B) and we e no a ec ed by AOX. GRAC hea s showed no up egula ion o ROS de ense a all (Fig 7B) despi e he mos se e e cIII dys unc ion. To al issue glu a hione was no changed in GRAC hea o li e bu was inc eased in bo h GROX issues (Fig 7C). Mi ochond ial aconi- ase ac i i y, a sensi i e ma ke o ROS damage o i on–sul u clus- e s (Yan e al, 1997), and p o ein ca bonyla ion (Fig EV4C and D) we e no signi ican ly di e en be ween he geno ypes. U ina y isop os anes, a ma ke o sys emic oxida i e s ess (Fig 7D), was ele a ed in bo h GRAC and GROX. Immunohis ochemical s aining o he lipid pe oxida ion p oduc 4-hyd oxynonenal was inc eased in all h ee GRAC issues, bu he inc ease was no p e en ed by AOX (Fig 7E and F). Finally, we ed he GRAC mice he mi ochon- d ia- a ge ed ROS sca enge mi oQ, which has shown bene icial e ec s in mouse models o ca diac ischemia– epe usion inju y (Adlam e al, 2005). Mi oQ eeding s a ed a P150 had no e ec on he su i al o he mice by P200 o on hei ca diac unc ion (Fig EV1A–F). Since ROS damage was clea ly no ins umen al in he ca diac escue by AOX, we looked a ni ic oxide (NO) me abolism ha is pa icula ly impo an in he ca dio ascula sys em and known o in e ac chemically wi h ROS and he espi a o y chain (Ca nice e al, 2013). We obse ed a signi ican inc ease in NO signaling- ela ed gene exp ession speci ically in GRAC hea s, and hese changes we e a enua ed by AOX (Fig 8A). Dec eased mRNA exp es- sion o cen al a ge s o NO signaling, yanodine ecep o (RyR2), sa co(endo)plasmic e iculum Ca 2+ -ATPase Se ca2 (A p2a2), and phospholamban (Pln) in GRAC hea sugges ed dec eased ca diac con ac ion- ela ed Ca 2+ channel unc ion (Fig 8B–D). AOX up egu- la ed hese Ca 2+ channel genes in bo h Bcs1l mu an and wild- ype mice. Exp ession o endo helial NO syn hase (Nos3) was inc eased in GRAC hea (Fig 8E), and, mos s ikingly, exp ession o neu onal NO syn hase (nNos,Nos1) was inc eased o e 10- old in he GRAC hea wi h a signi ican a enua ion in GROX mice. Wes e n blo anal- ysis con i med simila up egula ion o NOS1 p o ein (Fig 8F and G). Inc eased gene exp ession o dime hyla ginine dime hylaminohyd o- lase 1 (Ddah1), which hyd olyzes he endogenous NOS inhibi o s dime hyla ginine and monome hyla ginine (Fig 8H), and inc eased p oduc /subs a e (ci ulline/a ginine) a io (Fig 8I) in GRAC hea sugges ed inc eased NO p oduc ion. Howe e , nei he p o ein ni o y- osine no o al ni i es we e inc eased in he GRAC hea s (Fig 8J and K). Discussion He e, we show ha AOX, a non-mammalian enzyme ha can bypass cIII blockade by shun ing elec ons di ec ly om he quinone pool o oxygen, is able o pe manen ly p e en le hal ca diomyopa- hy and alle ia e mul iple o he pa hologies in cIII-de icien GRAC mice. The GRAC mice, ini ially b ed in C57BL/6JBomTac back- g ound (Le e ´en e al, 2011), ha e u ned ou pa icula ly use ul in mechanis ic and in e en ional s udies (Rajend an e al, 2016; Pu honen e al, 2017, 2018) due o hei pos na al symp om- ee pe iod and sho su i al. In he cu en s udy, we ound ha , in he C57BL/6JC l backg ound, he homozygo es su i e o median P210 and de elop le hal ca diomyopa hy a e P150, a mani es a- ions no seen in he GRACILE synd ome pa ien s wi h ea ly neona- al le hali y, bu a common mani es a ion in o he mi ochond ial diso de s. AOX p o ided a ull unc ional escue o he ca diomy- opa hy, es o ed cI- and cII-linked espi a ion o wild- ype le els, and ab oga ed he signa u e o me abolic s ess esponses. Hea muscle is highly dependen on mi ochond ial espi a ion and ope - a es a cons an ATP and phosphoc ea ine concen a ions (Ven u a- Clapie e al, 2011; Guzun e al, 2015). Ou esul s clea ly indica e ha he pa hogenesis mechanism leading o ca diomyopa hy in GRAC mice is insu iciency o elec on low, esul ing in dis u bed edox and me abolic homeos asis. Ca diac hype ophy and a me a- bolic swi ch om a y acid o glucose u iliza ion a e hallma ks o ª2018 The Au ho s EMBO Molecula Medicine 11:e9456 |2019 9o 19 Jayasimman Rajend an e al AOX escues cIII de iciency in mice EMBO Molecula Medicine Published online: Decembe 10, 2018 Isop os anes we e quan i ied om u ine using a comme cial ELISA ki (ab175819, Abcam). T ansc ip omics Fo RNA sequencing, RNA was emo ed and he sequencing lib a ies we e p epa ed using Illumina T uSeq S anded To al RNA Lib a y P ep Ki wi h Ribo-Ze o Human/Mouse/Ra . Lib a ies we e sequenced on Illumina Nex Seq 500 ins umen using 75 bp ki . Adap e sequences and low-quali y eads we e emo ed om he da a using cu adap (Ma in, 2011), and da a we e u he sc eened o emaining RNA eads using So MeRNA (Kopylo a e al, 2012). The da a we e mapped o M. musculus genome GRCm38.p4 using STAR (Dobin e al, 2013). The o iginal genome sequence and anno- a ion we e augmen ed wi h sequence and anno a ion o AOX. Coun da a we e p ocessed in R using GenomicFea u es and GenomicAlignmen s (Law ence e al, 2013), and he di e en ial exp ession analysis was ca ied ou using DESeq2 (Lo e e al, 2014). PCA plo s we e gene a ed using Clus Vis online ool (www.b ii .cs.u .ee/clus is). Pa hway en ichmen analysis En ichmen analyses we e pe o med using il e ed gene se s (|FC|>1.5, P<0.05) agains Reac ome da abase using www.mouse mine.o g. To isualize pa hway changes, hea maps we e gene a ed by Gi ool.2.3.1 wi h WT mean-cen e ed Z-sco e alues and gene mapping o selec ed pa hways as desc ibed (Gundem & Lopez- Bigas, 2014; Pe ez-llamas & Lopez-Bigas, 2011). Pa hway hea maps we e gene a ed based on Z-sco e, showing up egula ion (b own colo ) and down egula ion (blue colo ) o each pa hway. P- alues o g oup compa ison (labeled 1, 2, and 3) we e calcula ed using one-way ANOVA ollowed by selec ed compa isons and he Benjamin–Hochbe g FDR co ec ion. Me abolomics Ta ge ed quan i a i e me abolomics co e ing 116 me aboli es was pe o med using capilla y elec opho esis–mass spec ome y (CE- TOMFS and CE-QqQMS) a Human Me abolome Technologies Inc., Japan (h p://humanme abolome.com/en/). De ails o he p o ocol a e a ailable upon eques . We omi ed he AOX g oup in he me a- bolomics since ec opic AOX was shown o be ine in heal hy wild- ype mice (Szibo e al, 2017). S a is ics (excluding ansc ip omics) Fo no mally dis ibu ed da a, one-way ANOVA ollowed by Tukey’s es o by selec ed compa isons using - es s wi h Welch’s co ec ion was used. No mali y o da a was assessed by he d’Agos ino–Pea son omnibus no mali y es , and equali y o a i- ances by Ba le ’s es . Da a no compa ible wi h pa ame ic es s we e assessed by K uskal–Wallis and Mann–Whi ney U- es s. Su i al cu es we e analyzed using log- ank Man el–Cox es . G aphPad P ism 7 so wa e (G aphPad So wa e Inc., La Jolla, CA, USA) was used o he s a is ical analyses. G oup sizes (n) and he s a is ical es s used a e desc ibed in igu e legends, and exac P- alues a e shown in he g aphs. Da a a ailabili y Addi ional aw da a and de ailed p o ocols a e a ailable om he au ho s upon eques . T ansc ip omics da a ha e been deposi ed o A ayExp ess da abase a EMBL-EBI unde he accession ID E-MTAB-7416 (h ps://www.ebi.ac.uk/a ayexp ess/expe imen s/E- MTAB-7416). Expanded View o his a icle is a ailable online. The pape explained P oblem Mi ochond ial diseases a e gene ic diso de s o ene gy me abolism ha can a ec any one o se e al o gans o he body, including skele- al muscle, hea , b ain, and isce al o gans. Nowadays, ea ly diagno- sis o mi ochond ial disease is o en possible using mode n molecula gene ics, bu ea men op ions emain sca ce. Mu a ions in he BCS1L gene a e he mos common cause o mi ochond ial diseases a ec ing he espi a o y chain complex III (cIII). GRACILE synd ome (G ow h Res ic ion, Aminoacidu ia, Choles asis, li e I on o e load, Lac ic acidosis, and Ea ly dea h), wi h neona al le hali y, is he mos se e e o hem. Due o he s ikingly simila disease in all GRACILE synd ome pa ien s, P o . Fellman’s g oup p e iously in oduced he unde lying missense mu a ion (Bcs1l c.A232G ,Bcs1l p.S78G ) in o mice. This mouse model ai h ully ecapi ula es mos o he symp oms o he pa ien s. In sea ch o no el s a egies o alle ia e cIII de iciency- ela ed pa hol- ogy, we c ossed he Bcs1l c.A232G mice wi h ansgenic mice exp essing al e na i e oxidase (AOX), a mi ochond ial inne memb ane enzyme ha can bypass espi a o y elec on low blockade when he quinol oxida ion capaci y o cIII is comp omised. Resul s Mice homozygous o he Bcs1l c.A232G mu a ion su i ed app oxi- ma ely 200 days and died o dila ing ca diomyopa hy, a no el la e- onse pheno ype in he C57BL/6JC l backg ound. In con as , he homozygo es ca ying he AOX ansgene li ed app oxima ely 600 days and ne e de eloped he ca diomyopa hy. AOX also amelio- a ed he se e e kidney disease and ocal as ogliosis o he b ain in he homozygo es. Su p isingly, AOX did no co ec he li e disease, poo g ow h, and loss o whi e a , sugges ing di e en disease mechanism in di e en issues. AOX co ec ed he abno mal ul a- s uc u e o mi ochond ia and mi ochond ial espi a ion in hose issues, in which he issue pa hology was alle ia ed. Hea and kidney mi ochond ia om he homozygo es showed ele a ed eac i e oxygen species (ROS) p oduc ion, bu analyses o ROS damage sugges ed ha he bene icial e ec s o AOX we e no ROS- ela ed. Ins ead, we ound ha AOX no malized ca diac gene exp ession ela ed o ni ic oxide me abolism and signaling, a majo modula o o ca dio ascula unc ions. We conclude ha AOX e icien ly p e en ed issue pa hology in he Bcs1lmu an mouse model o cIII de iciency by es o ing espi a ion, p e e ably in he mos highly oxida i e issues such as hea . Impac Ou s udy in he pa ien mu a ion knock-in model o cIII de iciency is he i s p oo -o -concep ha bypass o he cIII-cIV segmen o he espi a o y elec on ans e can alle ia e pa hological mani es a ions in a physiologically ele an gene ic mouse model o a human mi o- chond ial diso de . These indings highligh he po en ial o AOX as a ool o un a el disease mechanisms, and also u ge u he s udies o AOX in p eclinical models, e.g., using i al deli e y o he a ec ed issues in mouse models, as well as s udies o sea ch o no el pha - macological bypass s a egies po en ially ansla able o pa ien s. 16 o 19 EMBO Molecula Medicine 11:e9456 |2019 ª2018 The Au ho s EMBO Molecula Medicine AOX escues cIII de iciency in mice Jayasimman Rajend an e al Published online: Decembe 10, 2018 Acknowledgemen s We hank P o . Hannu Sa iola o expe assis ance wi h issue his opa hol- ogy; D . E ic Du ou o ad ice on espi ome y; P a een Dhandapani o he AOX geno yping p o ocol; Elisa Al ay, Päi i Leinikka, and Nada Becha a- Hi onen o expe echnical assis ance; he s a o he DNA Sequencing and Genomics Labo a o y, Uni e si y o Helsinki, o unning he RNAseq; he s a o he BioEM Lab (a C-Cina), Biozen um, Uni e si y o Basel, he Co e Facili y o In eg a ed Mic oscopy, Panum Ins i u e, Uni e si y o Copenhagen, and Ola Gus a sson o Mic oscopy Facili y a he Depa men o Biology, Lund Uni e si y, o p o iding he elec on mic oscopy acili ies and assis ance; and inally D . Sanna Ma ja aa a o help ul ini ial discus- sions. This s udy was suppo ed by g an s om Academy o Finland (g an 259296 o VF, 256615 and 272376 o HTJ), Swedish Resea ch Council (g an 521-2011-3877 o VF), Eu opean Resea ch Council (Ad anced G an 232738 o HTJ), Finska Läka esällskape ( o VF), Founda ion o Pedia ic Resea ch in Finland ( o VF), Folkhälsan Resea ch Cen e ( o VF, JK), and Ge man Fede al Minis y o Educa ion and Resea ch (In a on ie g an 01KX1012 o MHdA). Au ho con ibu ions HTJ, MS, VF, and JK in en ed he concep o combine he mouse models; HTJ and MS p o ided he mouse s ain wi h b oad AOX exp ession; VF and JK g oup combined he s ains; JRa, JP, ST, EM, VF, and JK concei ed and designed he expe imen s; HF, VG-D, and MHA concei ed and supe ised pheno yping expe imen s a he GMC; JRa pe o med he animal expe imen s (echoca diog- aphy, me abolic sample collec ion, sickness sco ing, issue sampling); JRa pe o med he labo a o y analyses (his ology, p o ein analyses, BNGE, espi ome y, cIII ac i i y, me abolomics and ansc ip ome pa hway en ich- men analysis, ROS ma ke s); JP pe o med espi ome y, his ology, immuno- his ochemis y quan i a ion, and Amplex Red assay analyses; ST pe o med b ain his ology and immunohis ochemis y analyses; O-PS and PA pe o med ansc ip omics analysis; JK pe o med immunohis ochemis y and ROS ma ke analyses; MM pe o med elec on mic oscopy analyses; JRo and HF pe o med indi ec calo ime y and DEXA analysis; JRo and HF analyzed he da a om GMC; JRa, JP, ST, VF, and JK analyzed he da a om Helsinki; EM, HTJ, and MS pa icipa ed in in e p e a ion o he esul s; JRa, JP, VF and JK w o e he manusc ip d a ; all au ho s e ised he manusc ip and ha e con ibu ed subs an ially o he wo k epo ed. Con lic o in e es MS is a sha eholde o a comme cial en e p ise ha is dedica ed o de eloping he apeu ics based on AOX. The au ho s decla e ha hey ha e no con lic o in e es . Fo mo e in o ma ion (i) GRACILE synd ome: h ps://www.omim.o g/en y/603358. (ii) BCS1Lgene: h ps://www.omim.o g/en y/603647. (iii) Uni ed Mi ochond ial Disease Founda ion (UMDF): h ps://www.umd .o g/. Re e ences Adlam VJ, Ha ison JC, Po eous CM, James AM, Smi h RA, Mu phy MP, Sammu IA (2005) Ta ge ing an an ioxidan o mi ochond ia dec eases ca diac ischemia- epe usion inju y. FASEB J 19:1088 –1095 Al adhel M, Lillquis YP, Wa e s PJ, Sinclai G, S uys E, McFadden D, Hendson G, Hyams L, Sho ne J, Vallance HD (2011) In an ile ca dioencephalopa hy due o a COX15 gene de ec : epo and e iew. Am J Med Gene A 155a: 840 –844 A amna H, Nguyen A, Schul z C, Boyle K, Newbe y J, Ka o H, Ames BN (2008) Me hylene blue delays cellula senescence and enhances key mi ochond ial biochemical pa hways. FASEB J 22:703 –712 Bible E, Gup a P, Ho mann SL, Coope JD (2004) Regional and cellula neu opa hology in he palmi oyl p o ein hioes e ase-1null mu an mouse model o in an ile neu onal ce oid lipo uscinosis. Neu obiol Dis 16: 346 –359 B and MD (2010) The si es and opology o mi ochond ial supe oxide p oduc ion. Exp Ge on ol 45:466 –472 Bu well LS, B ookes PS (2008) Mi ochond ia as a a ge o he ca diop o ec i e e ec s o ni ic oxide in ischemia- epe usion inju y. An ioxid Redox Signal 10:579 –599 Ca nice R, C ab ee MJ, Si akuma an V, Casadei B, Kass DA (2013) Ni ic oxide syn hases in hea ailu e. An ioxid Redox Signal 18:1078 –1099 Cas o-Gue e o NA, K ab K, Mo eno-Sanchez R (2004) The al e na i e espi a o y pa hway o Euglena mi ochond ia. J Bioene g Biomemb 36: 459 –469 Chouchani ET, Me hne C, Nad ochiy SM, Logan A, Pell VR, Ding SJ, James AM, Cocheme HM, Reinhold J, Lilley KS e al (2013) Ca diop o ec ion by S- ni osa ion o a cys eine swi ch on mi ochond ial complex I. Na Med 19: 753 –759 Colombo G, Cle ici M, Ga a aglia ME, Gius a ini D, Rossi R, Milzani A, Dalle- Donne I (2016)A s ep-by-s ep p o ocol o assaying p o ein ca bonyla ion in biological samples. J Ch oma og B 1019:178 –190 C ucia CM, Hell K, Folsch H, Neupe W, S ua RA (1999) Bcs1p, an AAA- amily membe , is a chape one o he assembly o he cy och ome bc(1) complex. EMBO J 18:5226 –5233 Dassa EP, Du ou E, Goncal es S, Paupe V, Hakkaa GA, Jacobs HT, Rus in P (2009) Exp ession o he al e na i e oxidase complemen s cy och ome c oxidase de iciency in human cells. EMBO Mol Med 1:30 –36 Da oudi M, Ko a sky H, Hansson E, Fellman V (2014) Complex I unc ion and supe complex o ma ion a e p ese ed in li e mi ochond ia despi e p og essi e complex III de iciency. PLoS One 9:e86767 Dobin A, Da is CA, Schlesinge F, D enkow J, Zaleski C, Jha S, Ba u P, Chaisson M, Ginge as TR (2013) STAR: ul a as uni e sal RNA-seq aligne . Bioin o ma ics 29:15 –21 Dogan SA, Ce u i R, Beninca C, B ea-Cal o G, Jacobs HT, Ze iani M, Szibo M, Viscomi C (2018) Pe u bed edox signaling exace ba es a mi ochond ial myopa hy. Cell Me ab 28:764 –775 El-Khou y R, Du ou E, Rak M, Ramanan soa N, G andchamp N, Csaba Z, Du illie B, Beni P, Gallego J, G essens P e al (2013) Al e na i e oxidase exp ession in he mouse enables bypassing cy och ome c oxidase blockade and limi s mi ochond ial ROS o e p oduc ion. PLoS Gene 9: e1003182 El-Khou y R, Kemppainen KK, Du ou E, Szibo M, Jacobs HT, Rus in P (2014) Enginee ing he al e na i e oxidase gene o be e unde s and and coun e ac mi ochond ial de ec s: s a e o he a and pe spec i es. B i J Pha macol 171:2243 –2249 Emma F, Be ini E, Sal ia i L, Mon ini G (2012) Renal in ol emen in mi ochond ial cy opa hies. Pedia Neph ol 27:539 –550 Fellman V, Rapola J, Pihko H, Va ilo T, Rai io KO (1998) I on-o e load disease in in an s in ol ing e al g ow h e a da ion, lac ic acidosis, li e haemoside osis, and aminoacidu ia. Lance 351:490 –493 Fe nandez-Ayala DJ, Sanz A, Va iainen S, Kemppainen KK, Babusiak M, Mus alah i E, Cos a R, Tuomela T, Ze iani M, Chung J e al (2009) Exp ession o he Ciona in es inalis al e na i e oxidase (AOX) in D osophila ª2018 The Au ho s EMBO Molecula Medicine 11:e9456 |2019 17 o 19 Jayasimman Rajend an e al AOX escues cIII de iciency in mice EMBO Molecula Medicine Published online: Decembe 10, 2018 complemen s de ec s in mi ochond ial oxida i e phospho yla ion. Cell Me ab 9:449 –460 Fe nandez-Viza a E, Ze iani M (2015) Nuclea gene mu a ions as he cause o mi ochond ial complex III de iciency. F on Gene 6:134 Gundem G, Lopez-Bigas N (2014) Sample-le el en ichmen analysis un a els sha ed s ess pheno ypes among mul iple cance ypes. Genome Med 4: 28 Guzun R, Kaamb e T, Bagu R, G ichine A, Usson Y, Va ikmaa M, Anmann T, Tepp K, Timohhina N, She chuk I e al (2015) Modula o ganiza ion o ca diac ene gy me abolism: ene gy con e sion, ans e and eedback egula ion. Ac a Physiol 213:84 –106 Hakkaa GA, Dassa EP, Jacobs HT, Rus in P (2006) Allo opic exp ession o a mi ochond ial al e na i e oxidase con e s cyanide esis ance o human cell espi a ion. EMBO Rep 7:341 –345 Hansson A, Hance N, Du ou E, Ran anen A, Hul enby K, Clay on DA, Wibom R, La sson NG (2004) A swi ch in me abolism p ecedes inc eased mi ochond ial biogenesis in espi a o y chain-de icien mouse hea s. P oc Na l Acad Sci USA 101:3136 –3141 Hoe nagel MHN, Wiskich JT (1998) Ac i a ion o he plan al e na i e oxidase by high educ ion le els o he Q-Pool and py u a e. A ch Biochem Biophys 355:262 –270 Kal on R, Ko en L, A i am S, Schwa z O, Hai T, A onheim A (2017) ATF3 exp ession in ca diomyocy es p ese es homeos asis in he hea and con ols pe iphe al glucose ole ance. Ca dio asc Res 113:134 –146 Kopylo a E, Noe L, Touze H (2012) So MeRNA: as and accu a e il e ing o ibosomal RNAs in me a ansc ip omic da a. Bioin o ma ics 28: 3211 –3217 Ko de AS, Yada VR, Zheng YM, Wang YX (2011) P ima y ole o mi ochond ial Rieske i on-sul u p o ein in hypoxic ROS p oduc ion in pulmona y a e y myocy es. F ee Radic Biol Med 50:945 –952 Ko a sky H, Kelle M, Da oudi M, Le éen P, Ka ikoski R, Eno DP, Fellman V (2012)Me aboli e p o iles e eal ene gy ailu e and impai ed be a- oxida ion in li e o mice wi h complex III de iciency due o a Bcs1l mu a ion. PLoS One 7:e41156 Kowalo EM, Phang JM, G ange AS, Downing SJ (1977) Regula ion o p oline oxidase ac i i y by lac a e. P oc Na l Acad Sci USA 74:5368 –5371 Law ence M, Hube W, Pages H, Aboyoun P, Ca lson M, Gen leman R, Mo gan MT, Ca ey VJ (2013) So wa e o compu ing and anno a ing genomic anges. PLoS Compu Biol 9:e1003118 Le éen P, Ko a sky H, Mo gelin M, Ka ikoski R, Elme E, Fellman V (2011) The GRACILE mu a ion in oduced in o Bcs1lcauses pos na al complex III de iciency: a iable mouse model o mi ochond ial hepa opa hy. Hepa ology 53:437 –447 Logan GJ, de Alencas o G, Alexande IE, Yeoh GC (2014) Exploi ing he unique egene a i e capaci y o he li e o unde pin cell and gene he apy s a egies o gene ic and acqui ed li e disease. In J Biochem Cell Biol 56:141 –152 Lo e MI, Hube W, Ande s S (2014) Mode a ed es ima ion o old change and dispe sion o RNA-seq da a wi h DESeq2.Genome Biol 15:550 Mak ecka-Kuka M, K umschnabel G, Gnaige E (2015) High-Resolu ion espi ome y o simul aneous measu emen o oxygen and hyd ogen pe oxide luxes in pe meabilized cells, issue homogena e and isola ed mi ochond ia. Biomolecules 5:1319 –1338 Ma in M (2011) Cu adap emo es adap e sequences om high- h oughpu sequencing eads. EMBne J 17:10 –12 McDonald A, Vanle be ghe G (2004) B anched mi ochond ial elec on anspo in he Animalia: p esence o al e na i e oxidase in se e al animal phyla. IUBMB Li e 56:333 –341 Miwa S, T eumann A, Bell A, Vis oli G, Nelson G, Hay S, on Zglinicki T (2016) Ca boxyles e ase con e s Amplex ed o eso u in: implica ions o mi ochond ial H 2 O 2 elease assays. F ee Radic Biol Med 90: 173 –183 Mu phy MP (2009) How mi ochond ia p oduce eac i e oxygen species. Biochem J 417:1–13 Nob ega FG, Nob ega MP, Tzagolo A (1992)Bcs1, a no el gene equi ed o he exp ession o unc ional Rieske i on sul u p o ein in Saccha omyces ce e isiae.EMBO J 11:3821 –3829 Ohmo i T, Nagao Y, Mizukami H, Saka a A, Mu ama su SI, Ozawa K, Tominaga SI, Hanazono Y, Nishimu a S, Nu eki O e al (2017) CRISPR/ Cas9-media ed genome edi ing ia pos na al adminis a ion o AAV ec o cu es haemophilia B mice. Sci Rep 7:4159 Pechano a O, Va ga ZV, Cebo a M, Gi icz Z, Pache P, Fe dinandy P (2015) Ca diac NO signalling in he me abolic synd ome. B i J Pha macol 172: 1415 –1433 Pe ez-llamas C, Lopez-Bigas N (2011) Gi ools: analysis and isualisa ion o genomic da a using in e ac i e hea -maps. PLoS One 6: e19541 P e e G, Ho a h R, Klops ock T, Moo ha VK, Suomalainen A, Koene S, Hi ano M, Ze iani M, Bindo LA, Yu-Wai-Man P e al (2013) New ea men s o mi ochond ial disease-no ime o d op ou s anda ds. Na Re Neu ol 9: 474 –481 Pu honen J, Rajend an J, Mö gelin M, Uusi-Rau a K, Ka ayama S, K ju sko K, Eina sdo i E, Velagapudi V, Ke e J, Jauhiainen M e al (2017) Ke ogenic die a enua es hepa opa hy in mouse model o espi a o y chain complex III de iciency caused by a Bcs1lmu a ion. Sci Rep 7: 957 Pu honen J, Rajend an J, Tegelbe g S, Smolande OP, Pi inen E, Kallijä i J, Fellman V (2018) NAD + eple ion p oduces no he apeu ic e ec in mice wi h espi a o y chain complex III de iciency and ch onic ene gy dep i a ion. FASEB J 32:5913 –5926 Qui os PM, P ado MA, Zamboni N, D’Amico D, Williams RW, Finley D, Gygi SP, Auwe x J (2017) Mul i-omics analysis iden i ies ATF4as a key egula o o he mi ochond ial s ess esponse in mammals. JCellBiol 216:2027 –2045 Rajend an J, Toma si c N, Ko a sky H, Hansson E, Velagapudi V, Kallijä i J, Fellman V (2016) E ec o high-ca bohyd a e die on plasma me abolome in mice wi h mi ochond ial espi a o y chain complex III de iciency. In J Mol Sci 17:E1824 Sena LA, Li S, Jai aman A, P ak iya M, Ezponda T, Hildeman DA, Wang CR, Schumacke PT, Lich JD, Pe lman H e al (2013) Mi ochond ia a e equi ed o an igen-speci ic T cell ac i a ion h ough eac i e oxygen species signaling. Immuni y 38:225 –236 Szibo M, Dhandapani PK, Du ou E, Holms om KM, Zhuang Y, Salwig I, Wi ig I, Heidle J, Giza ullina Z, Gainu dino T e al (2017) B oad AOX exp ession in a gene ically ac able mouse model does no dis u b no mal physiology. Dis Model Mech 10:163 –171 Tegelbe g S, Toma si c N, Kallijä i J, Pu honen J, Elme E, Lindbe g E, No d DG, Solle M, Lesko N, Wedell A e al (2017) Respi a o y chain complex III de iciency due o mu a ed BCS1L: a no el pheno ype wi h encephalomyopa hy, pa ially phenocopied in a Bcs1lmu an mouse model. O phane J Ra e Dis 12:73 Tuomainen T, Ta i P (2017) The ole o ca diac ene gy me abolism in ca diac hype ophy and ailu e. Exp Cell Res 360:12 –18 Ven u a-Clapie R, Ga nie A, Veksle V, Joube F (2011) Bioene ge ics o he ailing hea . Biochim Biophys Ac a Mol Cell Res 1813:1360 – 1372 18 o 19 EMBO Molecula Medicine 11:e9456 |2019 ª2018 The Au ho s EMBO Molecula Medicine AOX escues cIII de iciency in mice Jayasimman Rajend an e al Published online: Decembe 10, 2018 Visapää I, Fellman V, Vesa J, Das a ma A, Hu on JL, Kuma V, Payne GS, Maka ow M, Van Cos e R, Taylo RW e al (2002) GRACILE synd ome, a le hal me abolic diso de wi h i on o e load, is caused by a poin mu a ion in BCS1L.Am J Hum Gene 71:863 –876 Vos M, Esposi o G, Edi isinghe JN, Vilain S, Haddad DM, Slabbae JR, Van Meensel S, Schaap O, De S oope B, Megana han R e al (2012) Vi amin K 2 Is a mi ochond ial elec on ca ie ha escues Pink1de iciency. Science 336:1306 –1310 Yan LJ, Le ine RL, Sohal RS (1997) Oxida i e damage du ing aging a ge s mi ochond ial aconi ase. P oc Na l Acad Sci USA 94:11168 –11172 License: This is an open access a icle unde he e ms o he C ea i e Commons A ibu ion 4.0 License, which pe mi s use, dis ibu ion and ep oduc- ion in any medium, p o ided he o iginal wo k is p ope ly ci ed. Jayasimman Rajend an e al AOX escues cIII de iciency in mice EMBO Molecula Medicine ª2018 The Au ho s EMBO Molecula Medicine 11:e9456 |2019 19 o 19 Published online: Decembe 10, 2018