Con en s lis s a ailable a ScienceDi ec
BBA - Molecula Basis o Disease
jou nal homepage: www.else ie .com/loca e/bbadis
Al e na i e espi a o y chain enzymes: The apeu ic po en ial and possible
pi alls
☆
Sina Saa i
a,1
, Geo ana S. Ga cia
b,1
, Ka ha ina B eme
a,1
, Ma ina M. Chioda
b
, Ana Andjelko ić
a
,
Paul V. Debes
c,d
, Mikko Nikinmaa
c
, Ma en Szibo
a
, E ic Du ou
a
, Pie e Rus in
e,
,
Ma cos T. Oli ei a
b,2
, Howa d T. Jacobs
a,g,⁎,2
a
Facul y o Medicine and Li e Sciences, BioMediTech Ins i u e and Tampe e Uni e si y Hospi al, FI-33014, Uni e si y o Tampe e, Finland
b
Depa amen o de Tecnologia, Faculdade de Ciências Ag á ias e Ve e iná ias, Uni e sidade Es adual Paulis a “Júlio de Mesqui a Filho”, 14884-900 Jabo icabal, SP, B azil
c
Depa men o Biology, FI-20014, Uni e si y o Tu ku, Finland
d
Facul y o Biological and En i onmen al Sciences, FI-00014, Uni e si y o Helsinki, Finland
e
INSERM UMR1141, Hôpi al Robe Deb é, 48, Boule a d Sé u ie , 75019 Pa is, F ance
Uni e si é Pa is 7, Facul é de Médecine Denis Dide o , Pa is, F ance
g
Ins i u e o Bio echnology, FI-00014, Uni e si y o Helsinki, Finland
ARTICLE INFO
Keywo ds:
Mi ochond ia
Mi ochond ial disease
The mogenesis
AOX
Reac i e oxygen species
ABSTRACT
The al e na i e espi a o y chain (aRC), comp ising he al e na i e NADH dehyd ogenases (NDX) and quinone
oxidases (AOX), is ound in mic obes, ungi and plan s, whe e i buffe s s esses a ising om es ic ions on
elec on flow in he oxida i e phospho yla ion sys em. The aRC enzymes a e also ound in species belonging o
mos me azoan phyla, including some cho da es and a h opods species, al hough no in e eb a es o in
D osophila. We pos ula ed ha he aRC enzymes migh be deployed o alle ia e pa hological s esses a ising om
mi ochond ial dys unc ion in a wide a ie y o disease s a es. Howe e , be o e such he apies can be con-
empla ed, i is essen ial o unde s and he effec s o aRC enzymes on cell me abolism and o ganismal phy-
siology. He e we epo and discuss new findings ha shed ligh on he unc ions o he aRC enzymes in animals,
and he unexpec ed benefi s and de imen s ha hey con e on model o ganisms. In Ciona in es inalis, he aRC is
induced by hypoxia and by sulfide, bu is un esponsi e o o he en i onmen al s esso s. When exp essed in
D osophila, AOX esul s in impai ed su i al unde es ic ed nu i ion, in addi ion o he p e iously epo ed
male ep oduc i e anomalies. In con as , i con e s cold esis ance o de eloping and adul flies, and coun e ac s
cell signaling de ec s ha unde lie de elopmen al dysmo phologies. The aRC enzymes may also influence li e-
span and s ess esis ance mo e gene ally, by elici ing o in e e ing wi h ho me ic mechanisms. In sum, hei
judicious use may lead o majo benefi s in medicine, bu his will equi e a ho ough cha ac e iza ion o hei
p ope ies and physiological effec s.
1. In oduc ion
P o is s, ungi and plan s possess an auxilia y espi a o y sys em in
hei mi ochond ia, which buffe s me abolic s esses ha a ise om
limi a ions on elec on flow in he sys em o oxida i e phospho yla ion
(OXPHOS). This al e na i e espi a o y chain (aRC) comp ises e-
p esen a i es o jus wo classes o enzyme: one o mo e al e na i e
NADH dehyd ogenases (NDX), ha ans e elec ons om NADH o an
in e media e elec on ca ie , ubiquinone [1], and al e na i e oxidases
(AOX), ha comple e elec on ans e om ubiquinol di ec ly o
h ps://doi.o g/10.1016/j.bbadis.2018.10.012
Recei ed 20 May 2018; Recei ed in e ised o m 3 Oc obe 2018; Accep ed 5 Oc obe 2018
☆
This a icle is pa o a Special Issue en i led: The powe o me abolism: Linking ene gy supply and demand o con ac ile unc ion edi ed by To s en Doens ,
Michael Schwa ze and Ch is ine Des Rosie s.
⁎
Co esponding au ho a : Facul y o Medicine and Li e Sciences, FI-33014, Uni e si y o Tampe e, Finland.
E-mail add esses: sina.saa i@u a.fi(S. Saa i), [email p o ec ed] (G.S. Ga cia), [email p o ec ed] (K. B eme ),
[email p o ec ed] (M.M. Chioda), ana.andjelko ic@u a.fi(A. Andjelko ić), [email p o ec ed] (P.V. Debes), miknik@u u.fi(M. Nikinmaa),
ma en.szibo @u a.fi(M. Szibo ), e ic.du ou @u a.fi(E. Du ou ), [email p o ec ed] (P. Rus in), [email p o ec ed] (M.T. Oli ei a),
howa d. .jacobs@u a.fi(H.T. Jacobs).
1
Who con ibu ed equally o he wo k.
2
Who con ibu ed equally o he supe ision o he wo k.
BBA - Molecula Basis o Disease 1865 (2019) 854–866
A ailable online 17 Oc obe 2018
0925-4439/ © 2018 The Au ho (s). Published by Else ie B.V. This is an open access a icle unde he CC BY-NC-ND license
(h p://c ea i ecommons.o g/licenses/BY-NC-ND/4.0/).
T
oxygen [2]. In e ms o hei ne edox chemis y, NDX and AOX can
unc ionally eplace espi a o y complex I (cI, NADH:ubiquinone oxi-
do educ ase) and complexes III (cIII, ubiquinol:cy och ome coxido -
educ ase) plus IV (cIV, cy och ome coxidase) o he OXPHOS sys em,
espec i ely.
In con as o he s anda d OXPHOS complexes o he mi ochond ial
espi a o y chain (RC), he al e na i e enzymes ha e fi e dis inc
p ope ies. Fi s , hey a e each composed o a single polypep ide.
Second, hei eac ion chemis y is non p o on-mo i e, ins ead dis-
sipa ing he eleased ee ene gy as hea . Thi d, hey a e uni e sally
coded only by he nuclea genome. Fou h, hey a e e ac o y o he
commonly used OXPHOS inhibi o s. Finally, hei biochemical p op-
e ies limi hei ac i i y o me abolic condi ions whe e hey a e unc-
ionally equi ed, a leas in he case o AOX. As a esul , AOX may be
conside ed a sel - egula ing enzyme ha does no ‘sho -ci cui ’ he
OXPHOS sys em, excep when he la e is al eady dys unc ional. In
plan s, he basis o his es ic ion is ela i ely well unde s ood, in ha
he enzyme only becomes ac i e when he subs a e pool o educed
quinones accumula es o ele a ed le els, ha would be conside ed
abno mally high in non-pho osyn he ic o ganisms [3,4]. In effec , he
enzyme displays a much highe K
m
o ubiquinol han does cIII. This
p ope y has been assumed o apply o AOX om o he axa, bu has no
ye been o mally demons a ed.
In a ious species, AOX has been epo ed o become ac i a ed by
me aboli es such as py u a e [3,5–7] and o he o ganic acids [8,9], ha
may accumula e unde condi ions o RC o e load o inhibi ion, o in
esponse o specific me abolic s esses. In some axa he aRC enzymes
ha e also been shown o be egula ed by hea s ess [10], calcium
[10,11] and pu ine nucleo ides [7,9,12]. The molecula basis o hese
phenomena, o e en whe he egula ion is di ec o indi ec , is mos ly
no known, al hough ac i a ion o AOX in plan s is ela i ely well un-
de s ood. In his case, AOX is belie ed o exis no mally as a dime ,
which is ac i a ed (mos likely by hio edoxin) when he in e -subuni
disulfide b idge is b oken. Fu he ac i a ion is b ough abou by he
allos e ic binding o py u a e and/o o he o ganic acids [13]. Plan
AOX is also egula ed a he le el o gene exp ession in many con ex s
whe e i s p esence con e s p edic able esis ance agains a ious me-
abolic s esses, including d ough , hea , cold, sal and e en pa hogen
in asion [14,15].
Much less is known abou he me abolic egula ion o NDX, which is
ound in diffe en sub-mi ochond ial compa men s, and is likely o
ope a e subjec o he a ailabili y o subs a e in hese loca ions. In he
yeas Saccha omyces ce e isiae, which lacks cI, he NDX a ian s Nde1
and Ndi1, which espec i ely ca alyze NADH oxida ion a he ex e io -
and in e io - acing sides o he inne mi ochond ial memb ane, may be
co- egula ed wi h he es o he OXPHOS machine y.
In plan s, he aRC enzymes a e also belie ed o main ain mi-
ochond ial me abolic p ocesses and edox homeos asis du ing day-
ligh , a leas unde s ess condi ions [16], when excess ATP p oduc ion
by pho osyn hesis may es ic flux h ough he mi ochond ial RC. Two
o he physiological ai s con e ed by aRCs a e he mogenesis and
egula ion o aging. aRC enzymes do no conse e he ee ene gy o
biological oxida ions by cha ge sepa a ion ac oss he inne mi ochon-
d ial memb ane. Ins ead, hey dissipa e he ene gy as hea . Bu his hea
can also be ha nessed. The mos spec acula examples o aRC-based
he mogenesis a e in plan s, whe e he eleased hea is used o ola ilize
insec a ac an s, as in he flowe s o a um lilies [17,18]. In ungi, a
well documen ed example o a li e-cycle shi om OXPHOS o al e -
na i e espi a ion unde lies he swi ch om ege a i e g ow h o long-
e m main enance in Podospo a anse ina. In senescen cul u es he aRC
ac s o eplenish p ima y elec on ca ie s such as NAD in hei oxidized
o m bu does no suppo con inuous g ow h. Howe e , i also gen-
e a es a less eac i e oxygen species (ROS) han he OXPHOS sys em,
hus limi ing oxida i e damage and acili a ing long- e m su i al [19].
De ence agains excess ROS may be a common unc ion a leas o AOX,
since he accumula ion o educed quinols ha igge s i s ac i a ion
po en ially a ou s he passage o single elec ons o oxygen bo h a cIII
and, ia e e se elec on flow, a cI. Con e sely, since ROS has also
been posi ed o se e a signaling unc ion, especially in ega d o e-
e se elec on anspo a cI [20], AOX ac i a ion migh in e e e wi h
o modi y cell signaling pa hways in some con ex s, wi h impo an
physiological consequences.
The aRC enzymes a e also ound in bac e ia, and in ep esen a i e
o ganisms om many me azoan phyla [21], including some a h opods
and some cho da es [22,23]. In iguingly, hey appea o ha e been los
du ing he cou se o e olu ion in wha a e commonly ega ded as he
c own g oups o me azoan e olu ion: he e eb a es and he mo e
ad anced insec s, such as honeybees o flies [21,22] as well as, mos
likely, he cephalopod molluscs. Whils a leas some physiological
unc ions o aRC enzymes in lowe euka yo es and plan s a e ela i ely
well es ablished, he oles hey play in animals a e almos comple ely
unknown. They ha e been p oposed o acili a e adap a ion o en-
i onmen al s esses and ansi ions [22], bu expe imen al da a sup-
po ing his concep emains o be ga he ed. In plan s, some al e na i e
NADH dehyd ogenases can use NADPH as a subs a e, bu his issue is
one o many ha ha e no been explo ed in ega d o he enzyme in
animals.
We easoned ha , since many o he same s esses agains which
AOX p o ides a de ence in plan s and lowe euka yo es also occu pa-
hologically in humans, as a esul o mi ochond ial OXPHOS dys-
unc ion, deploying aRC enzymes xeno opically migh help alle ia e he
associa ed disease s a es [24]. To es and de elop his concep , we se
ou o examine he p ope ies con e ed upon cells and model o gan-
isms by exp ession o me azoan aRC enzymes.
In pionee ing expe imen s, Yagi and colleagues had al eady de-
mons a ed ha yeas Ndi1 could be exp essed in human cells, and
could unc ionally eplace he edox ac i i y o cI [25–29]. By de el-
oping his concep u he , hey showed i s po en ial in pa hological
models o cI- ela ed diseases in oden s [30–32], in o ming and in-
spi ing ou wo k using he me azoan enzymes. Fo ou own expe i-
men s, we selec ed, as a sou ce, he g oup o animals closes o humans,
bu which s ill e ains genes o bo h AOX and o NDX; namely he
unica es, a sis e -g oup o he e eb a es.
In ini ial ials we we e able o exp ess in human cells he AOX om
he unica e Ciona in es inalis [33]. The Ciona AOX p o ein was ou ed o
mi ochond ia ia i s own, in insic N- e minal mi ochond ial a ge ing
pep ide. I con e ed esis ance o OXPHOS oxins such as an imycin A
o cyanide in isola ed mi ochond ia o in whole cells. Mo e ema kably,
when in oduced ansgenically in o D osophila o in o he mouse, AOX
could be exp essed ubiqui ously, wi h almos no de ec able effec on
pheno ype [34]. T ansgenic exp ession in D osophila o he Ciona NDX
[35], o he single-subuni NADH dehyd ogenase Ndi1 om yeas [36],
was also achie ed wi h ha dly any de ec able effec on no mal de el-
opmen o physiology, a leas unde non-s essed condi ions.
Gi en hese findings, we and o he s ha e p oceeded o es how a
hese ansgenes p o ide p o ec ion agains physiological s esses ha
a ise h ough, o a e media ed, by mi ochond ial dys unc ion, com-
mencing wi h s udies in D osophila. Al hough AOX was unable o escue
a null mu a ion in a subuni o cIV [34], i was able o compensa e, a
leas pa ially, o he effec s o cIV knockdown di ec ed a specific
subuni s and issues [37,38], as well as o he effec s o a he e oplasmic
m DNA mu a ion affec ing cy och ome oxidase [6]. Mo e in iguingly,
AOX exp ession was able o a enua e he pa hological pheno ypes o
se e al diffe en neu ological disease models, including wo D osophila
models o Pa kinson's disease [34,39] and one o Alzheime 's disease
[40]. The exac mechanisms behind hese pheno ypic ans o ma ions
emain o be elucida ed, bu alle ia ion o excess ROS p oduc ion due
o blocked elec on flow has been pu o wa d as one plausible,
common explana ion. Con e sely, AOX was unable o imp o e he
pheno ype o a fly model o mi ochond ial ansla ional disease ( ko
25
),
which exhibi s de elopmen al delay and mechanically induced sei-
zu es, associa ed wi h a dec eased ac i i y o all ou OXPHOS
S. Saa i e al. BBA - Molecula Basis o Disease 1865 (2019) 854–866
855
complexes ha a e dependen on mi ochond ial ansla ion p oduc s
[41]. Co-exp ession o Ndi1 ac ually wo sened he pheno ype, sug-
ges ing ha he unde lying de ec is insufficien ATP p oduc ion a he
han dis u bed edox o me abolic homeos asis. AOX was also unable o
escue le hal mu a ions in he m DNA helicase (Twinkle) o DNA
polyme ase γ, as well as he le hali y p oduced by global knockdown o
hese genes [42].
Flies exp essing Ndi1 showed e idence o inc eased li espan, which
was o iginally sugges ed o be due o a compensa ion o oxida i e s ess
[36]. Howe e , his effec appea s o esul om a pa adoxical ho me ic
p ocess, in which he cons i u i e exp ession o Ndi1 ac ually inc eases
mi ochond ial ROS p oduc ion due o e e se elec on flow h ough cI,
which hen ac i a es s ess esponses ha p e en oxida i e damage
h oughou li e [43].
Whils ou findings in D osophila a e in some ways ema kable,
implemen ing aRC enzymes as ac ual he apy emains only a dis an
goal. The ange o medical condi ions in which mi ochond ial RC o
OXPHOS dys unc ion is a con ibu o y o essen ial ac o is e y b oad,
spanning om p ima y mi ochond iopa hies caused by mu a ions in
m DNA o in he appa a us o m DNA main enance and exp ession
[44,45], h ough o common disease en i ies whe e mi ochond ial dis-
up ion is due o ischemia/ epe usion inju y, oxida i e o p o eo oxic
s ess, oxic damage o o he ex e nal causes. We need o elucida e in
much g ea e de ail which o hese condi ions is alle ia ed by he ex-
p ession o aRC enzymes, using he mos app op ia e animal models.
Thus, a majo h us o cu en wo k is o es , using mouse models now
a ailable, how a AOX (o NDX/Ndi1) can nega e he pa hological
consequences o mi ochond ial dys unc ion in all o hese con ex s. The
ocus o hese s udies will inc easingly be on ca dio ascula and neu-
ological diseases, whe e he impo ance o mi ochond ial dys unc ion
is now widely ecognized.
Despi e he encou aging findings o da e in animal models, he
po en ial p oblems in he use o aRC enzymes in he apy need o be
conside ed in much g ea e de ail. We need o be su e ha hei acu e
o ch onic adminis a ion does no ha e ha m ul long- e m con-
sequences; in pa icula when he body is s essed in ways ha may o
may no be ela ed o he o iginal o unde lying condi ion ha
p omp ed hei use. No all ypes o mi ochond ial dys unc ion a e
equi alen , and no all issues espond in he same manne o a gi en
biochemical de ec o ex e nal s ess.
In o de o add ess he many issues a ising om he po en ial use o
me azoan aRC enzymes in u u e he apy, i is necessa y o es ablish in
fine de ail hei inhe en p ope ies, in pa icula when and how hey
become enzyma ically ac i e and how hey affec model o ganisms
unde s ess condi ions whe e hey may become ac i a ed. The wide
ange o medical condi ions associa ed wi h mi ochond ial dys unc ion,
and agains which aRC enzyme-based he apies migh be effec i e,
combined wi h he di e si y and complexi y o hei egula ion al eady
known om he plan and mic obial wo ld, mean ha his ask is as
and complica ed.
As a fi s s ep owa ds his goal, we he e se ou o in es iga e as-
pec s o he biology o aRC enzymes in he animal sou ce ha we in-
i ially selec ed, C. in es inalis. In pa icula , in o de o ob ain clues on
hei p obable physiological unc ions, we es ed how hei exp ession
is modified by ex e nal condi ions, ocusing on s esso s ound in hei
na u al en i onmen . In a pa allel se o s udies, we in es iga ed how
he exp ession o aRC enzymes, specifically AOX, can modi y he phy-
siological esponses o model o ganisms unde s ess ul en i onmen al
condi ions, ocusing ini ially on empe a u e and nu i ion, and on he
al eady well cha ac e ized D osophila models which a e also much mo e
easily manipula ed han hei mammalian coun e pa s.
2. Ma e ials and me hods
2.1. Ciona specimens, main enance and s ess ea men s
Ciona adul s we e supplied as li ing specimens by he S a ion
Biologique de Roscoff. P io o he expe imen s, hey we e acclima ed
o labo a o y condi ions o 24–48 h wi hou eeding, a 18 °C in glass
aqua ia a a densi y o one animal pe 2.5 l a ificial sea wa e (Ree
C ys als Aqua ium Sys ems, Sa ebou g, F ance). To es he effec s o
sulfide, sodium sulfide nonahyd a e (Sigma Ald ich) was added so as o
each combined concen a ions o dissol ed ee sulfides o 100 and
300 μM. To es he effec s o al e ed oxygen le els, no moxia was de-
fined as he equilib ium s a e wi h a mosphe ic oxygen. Hypoxic con-
di ions we e es ablished by displacemen o ai by ni ogen gas, and
hype oxic condi ions by o e sa u a ion using gaseous oxygen. Oxygen
le els we e moni o ed using an op ical fib e p obe and associa ed
so wa e (OxyView - PST3-V6.02, P ecision Sensing GmbH, Ge many)
and adjus men s we e made so o main ain s able condi ions
h oughou each expe imen . Fo de ails o o he s esses applied, see
Supplemen a y Da a File.
2.2. RNA ex ac ion and analysis
To al RNA was ex ac ed om ozen dissec ed Ciona o gans (o a y,
hea and s omach) by homogeniza ion in 1 ml TRI Reagen (Molecula
Resea ch Cen e , USA) using ce amic beads, o 3 × 20 s a 4000 pm
(Powe Lyze 24 Bench Top Bead-Based Homogenize ; MO BIO
Labo a o ies, USA). A e chlo o o m ex ac ion, ollowed by cen-
i uga ion a 12,000g
max
o 15 min a 4 °C, RNA was p ecipi a ed om
he aqueous phase by isop opanol a −20 °C o 1 h, ollowed by cen-
i uga ion a 12,000g
max
o 8 min a 4 °C. Pelle s we e washed in 75%
e hanol, ai d ied, and esuspended in RNase- ee wa e a 60 °C o
10 min. RNA was p epa ed om ozen neu al complex samples using
Single Cell RNA Pu ifica ion Ki (No gen Bio ek, Canada). Samples we e
manually homogenized in 100 μl lysis buffe using plas ic pes les, ol-
lowed by incuba ion a 42 °C o 5 min, subsequen o exing o 15 s,
and he ea e ea ed acco ding o manu ac u e 's ins uc ions. RNA
ex ac s we e quan ified spec opho ome ically, p e- ea ed wi h
RNase- ee DNase I (The mo Fishe Scien ific Inc., Wal ham, MA, USA)
acco ding o manu ac u e 's ins uc ions, o emo e esidual genomic
DNA, hen e e se ansc ibed using High-Capaci y cDNA Re e se
T ansc ip ion Ki (Applied Biosys ems), also acco ding o manu ac-
u e 's ins uc ions. Quan i a i e PCR o cDNA used S epOne Plus PCR
ins umen (Applied Biosys ems), wi h ini ial dena u a ion o 20 s a
95 °C, 40 cycles o 3 s a 95 °C, 30 s a 60 °C, and a final mel -cu e s ep
o 15 s a 95 °C, 60 s a 62 °C, and 15 s a 95 °C. Reac ions using cus o-
mized gene-specific p ime s (Supplemen a y Table S1) we e un in
duplica e, wi h 3.125–50 ng cDNA (depending on he a ge gene),
0.5 μM o each p ime , and Fas SYBR G een Mas e Mix (Applied
Biosys ems). To minimize echnical bias, samples om all ea men s
o a gi en gene and issue we e analyzed on he same pla e, whe e
possible. Measu emen s ha exhibi ed di e gen mel ing empe a u es
o amplifica ion cu es we e excluded and analyses epea ed.
Quan ifica ion cycles (C
q
) and amplifica ion efficiencies (E) we e cal-
cula ed using he online ool Real- ime PCR Mine and C
q
alues we e
co ec ed o among-pla e bias and s anda dized agains he mean o
wo e e ence genes, β-ac in and RPL5.
2.3. D osophila s ains and cul u e
D osophila s ains used in he s udy we e 7 ansgenic lines o C.
in es inalis AOX cons uc ed in-house and desc ibed p e iously: UAS-
AOX
F6
, inse ed on ch omosome 2 [34], UAS-AOX
F24
, inse ed on
ch omosome 3 [34] and UAS-AOX
8.1
, inse ed by ΦC31 ecombina ion
a a lowe -exp ession si e on ch omosome 2 [38], each con aining he
AOX coding sequence placed unde he con ol o he exogenously
S. Saa i e al. BBA - Molecula Basis o Disease 1865 (2019) 854–866
856
supplied GAL4 ansc ip ion ac o ; ubAOX
35
, ubAOX
112
and ubAOX
7
,
con aining inse ions on ch omosomes X, 2 and 3, espec i ely, o he
AOX coding sequence unde he con ol o he α- ubulin p omo e ; and
UAS-mu AOX
2nd
, bea ing a mu a ed, ca aly ically inac i e AOX a ian
also unde GAL4 con ol, on ch omosome 2. We also gene a ed lines
2x ubAOX, homozygous o bo h he ubAOX
112
and ubAOX
7
inse ions,
and 3x ubAOX, homozygous addi ionally o ubAOX
35
. Recipien line
w
1118
was used as a con ol o all ansgenic lines excep hose c ea ed
by ΦC31 ecombina ion, o which a line wi h he UAS-con aining
ec o inse ed a he same si e as UAS-AOX
8.1
and UAS-mu AOX
2nd
,
UAS-emp y, was used. An addi ional con ol, UAS-GFP (S inge ), ex-
p essing nuclea -localized GFP unde GAL4 con ol, was used in some
expe imen s. Flies we e main ained and cul u ed on s anda d high-
suga medium [34] a 18 o 25 °C, wi h 12 h cycles o ligh and da k-
ness, excep whe e indica ed in specific expe imen s.
2.4. De elopmen al assays
To es effec s o empe a u e, de elopmen al assays we e con-
duc ed a 12, 15, 18, 25, and 29 °C, as ollows. 20 p e-ma ed emales, in
he p esence o 10 males o he same line, we e allowed o lay eggs o
≤24 h a 25 °C in a o al o ou ials pe expe imen . The eggs we e
coun ed, ans e ed o he indica ed empe a u es, and he ials
moni o ed daily o he appea ance o pupae and adul s. Egg- o-pupa
iabili y was calcula ed o each ial as he a io be ween he o al
numbe o pupae and he o al numbe o eggs laid. Egg- o-pupa and
egg- o-adul de elopmen al ime was de e mined by eco ding he
numbe o pupae o adul s pe ial on successi e days ela i e o he
day o egg laying. Expe imen s we e pe o med in duplica e a 12 and
15 °C, and in iplica e a o he empe a u es. To es effec s o he
composi ion o he cul u e medium, flies o a gi en geno ype o c osses
as indica ed in figu e legends we e g own on s anda d high-suga
medium o on a low-nu ien medium comp ising, excep whe e s a ed,
3.5% yeas and 5% glucose in s anda d aga wi h an imic obials (nia-
pigin and p opionic acid). Pupae pe ial and he numbe o eclosed
adul s we e eco ded.
2.5. Ligh mic oscopy o pupae
Uneclosed pupae, dissec ed om he pupal case i sufficien ly ad-
anced de elopmen ally, we e isualized using a Nikon SMZ 745T
zoom s e eomic oscope.
2.6. P o ein ex ac ion and analysis
Ba ches o 10 la ae o flies we e snap ozen a −80 °C and c ushed
in an Eppendo ube in 100 μl o lysis buffe , comp ising 0.3% SDS in
PBS plus p o ease inhibi o cock ail (Roche). Following incuba ion o
15 min a oom empe a u e and cen i uga ion o 10 min a
15,000g
max
a oom empe a u e, supe na an s we e decan ed and
p o ein concen a ions measu ed by NanoD op spec opho ome y
(The moFishe Scien ific). A e dilu ion wi h wa e and 5 × SDS-PAGE
sample buffe , 63 μg p o ein aliquo s in 20 μl we e esol ed on 12%
polyac ylamide gels, we -blo ed o ni ocellulose memb ane, washed
in PBS-Tween, blocked o 3 h wi h 5% non a milk in PBS-Tween a
oom empe a u e, and eac ed wi h cus om-made abbi an i-AOX
an ibody [34], 1:10,000 o e nigh a 4 °C. A e 5 × 5 min washes in
PBS-Tween, he memb ane was eac ed wi h seconda y an ibody (goa
an i- abbi , Vec o Labo a o ies, 1:10,000) o 1 h a oom empe a u e,
e-washed, p ocessed o imaging using Lumina a™C escendo (Milli-
po e) o 5 min, hen imaged o chemiluminescence using BioRad
ChemiDoc MP. Equal loading was confi med by s aining he memb ane
wi h Ponceau S o 5 min, washing wi h wa e and plain imaging.
Images we e op imized o b igh ness and con as and c opped, o a ed
and masked o cla i y, bu no manipula ed in any o he way.
Fig. 1. AOX and NDX induc ion by en i onmen al s esso s in Ciona in es inalis.
Re e se ansc ip ion quan i a i e eal- ime PCR (RT-qPCR) analysis o he le els o AOX and NDX ansc ip s in RNA ex ac ed om he indica ed issues o C.
in es inalis adul s, ea ed as indica ed (see Ma e ials and Me hods): (A) o 6 h a oxygen le els o 1.62 (hypoxia), 7.92 (no moxia) o 15.94 (hype oxia) mg/l o o
8 h in 0 (con ol), 100 o 300 μM sulfide, n = 6 o each g oup, in 2 aqua ia; (B) o 6 h wi h an oxygen le el o 1.96 mg/l (hypoxia), o 100 μM sulfide o he
combina ion o bo h (1.91 mg/l O
2
, 100 μM sulfide), n = 12 o each g oup, in 4 aqua ia. The co ec ed C
q
s a is ic, ep esen ing ela i e ansc ip le els no malized
agains wo e e ence genes, was calcula ed as desc ibed in Ma e ials and Me hods, wi h da a plo ed as ma ginal model means ± 95% CI, showing da a poin s o
indi idual animals in g ey. Panel (A) is a subse o he da a shown in Fig. S1A and S1B, omi ing NDX and indica o genes. Fo s a is ical analyses, see Supplemen a y
Tables S2 and S3, ela ing o he da a shown in panels (A) and (B), espec i ely. No e ha he animals used in he expe imen o panel (B) we e a mix u e o C.
in es inalis ypes A and B, cu en ly p oposed as dis inc species [77]. The p ime s used ecognize RNA om bo h C. in es inalis ypes, and he da a shown in he figu e
a e no malized o ype B (see Supplemen a y Table S3) o compa abili y wi h o he expe imen s, whe e only ype B was used.
S. Saa i e al. BBA - Molecula Basis o Disease 1865 (2019) 854–866
857
3. Resul s & discussion
3.1. AOX and NDX a e induced in Ciona by hypoxia and sulfide
To add ess he ques ion o he unc ional oles o NDX and AOX in
animals, we ocused on C. in es inalis, he o ganism ha has been used
as he sou ce o he aRC ansgenes ha we ha e exp essed in flies,
mice and human cells. We subjec ed Ciona adul s o a se o s esso s
expe ienced in he na u al en i onmen , hen analyzed NDX and AOX
induc ion a he RNA le el in diffe en issues (Figs. 1, S1). Bo h AOX
and NDX we e induced by hypoxia (Figs. 1A, S1A) o exposu e o sul-
fide (Figs. 1A, S1B) in hea , neu al complex and, o a lesse ex en , in
s omach. Physiologically s ess ul empe a u e o hea y-me al exposu e
did no induce he exp ession o aRC mRNAs (Fig. S1C, S1D), whils he
effec s o hypoxia and sulfide we e addi i e, a leas o AOX in he
issues showing he g ea es induc ion (Fig. 1B). These esponses
he e o e esul om independen sensing and signal ansduc ion
p ocesses. Sulfide is a na u ally occu ing inhibi o o cy och ome
oxidase, agains which Ciona AOX is able o p o ec mammalian cells
(Fig. S1E), p o iding a a ionale o i s induc ion in sulfide-con aining
seawa e . AOX induc ion by hypoxia is mo e unexpec ed, since he
enzyme equi es oxygen as a subs a e jus as does cy och ome oxidase.
Howe e , AOX main enance does no equi e he elabo a e biogene ic
p og am o he cy och ome-con aining complexes cIII and cIV. In pa -
icula , he pa hway o haem biosyn hesis in all highe euka yo es
in ol es wo oxygen-dependen s eps: cop opo phy inogen oxidase
consumes wo molecules o oxygen, eleasing wo each o wa e and
CO
2
[46], whils oxygen is used as he e minal elec on accep o o he
nex s ep in he pa hway, ca alyzed by p o opo phy inogen oxidase
[47]. AOX ca alysis depends on a dii on cen e a he han a haem
co ac o . The e o e, AOX biosyn hesis should no be affec ed by dis-
u bed haem biosyn hesis du ing p olonged hypoxia. Ano he possible
a ionale o AOX exp ession being esponsi e o hypoxia is o minimize
ischemia/ epe usion inju y by ROS, when he sys em becomes eox-
ygena ed, bu whils he RC ca ie s a e s ill in he educed s a e. In a
pa allel s udy in molluscs, AOX was ecen ly shown o be induced unde
B
A55
50
45
40
35
30
25
20
15
10
5
w1118 ubAOX72x ubAOX
ubAOX35 ubAOX112 3x ubAOX
)apupo gge(syad
35
30
25
20
15
10
5
w1118 3x ubAOX
2x ubAOX
days (egg o adul )
a b c a b c a b c a b c
% (egg o pupa)
100
80
60
40
20
0
w1118
1x ubAOX
a = ubAOX35
b = ubAOX7
c = ubAOX112
2x ubAOX
3x ubAOX
C
12 °C 15 °C 18 °C 25 °C
12 °C
15 °C
18 °C
25 °C
29 °C
Fig. 2. AOX exp ession in D osophila accele a es de elopmen a low empe a u es.
(A) Egg- o-pupa de elopmen ime o flies o he indica ed geno ypes and cul u e empe a u es (means ± SD o ≥100 flies in each case, cul u ed in 8–12 ials in
2–4 biological eplica es). Fo abula ed da a and s a is ical analysis see Supplemen a y Table S4. (B) Egg- o-adul de elopmen ime o flies o he indica ed
geno ypes and cul u e empe a u es (means ± SD o a o al o 8–16 ials in each case, om 2 o 4 biological eplica e expe imen s). Fo cla i y, only he effec s o 2
and 3 copies o he ubAOX ansgene e sus he w
1118
backg ound s ain a e shown in he figu e. Comple e abula ed da a and s a is ical analysis a e p esen ed in
Supplemen a y Table S5. (C) Egg- o-pupa iabili y o flies o he indica ed geno ypes, cul u ed a diffe en empe a u es, as shown: means ± SD o 8–12 ials om
2 o 4 biological eplica es in each case. Fo comple e abula ed da a and s a is ical analysis see Supplemen a y Table S6. No e ha , o cla i y, da a o he 29 °C
s ess condi ion is omi ed om he figu e. Al hough AOX was also beneficial a his empe a u e, he ela ionship wi h gene dosage was less clea han a low
empe a u es, compa ed wi h s anda d g ow h condi ions o 18–25 °C whe e AOX was neu al.
S. Saa i e al. BBA - Molecula Basis o Disease 1865 (2019) 854–866
858

hypoxia in a eshwa e bi al e [48].
Possible medical implica ions: Wha a e he po en ial implica ions o
his finding o he use o AOX in he apy? Toxic le els o sulfide as well
as oxygen-dep i a ion a e condi ions expe ienced by animals in he
na u al en i onmen bu a e also ound in humans in cases o disease.
Al hough low le els o hyd ogen sulfide ac in endoc ine signaling and
aso elaxa ion [49–52], na u al o e -p oduc ion o lack o de oxifica-
ion o H
2
S can also become pa hological, wi h cIV as a majo a ge . A
equen cause o dis u bed H
2
S me abolism is he al e a ion o he gu
mic obio a by an ibio ic use, which has been implica ed in ulce a i e
coli is and o he in es inal diseases [51]. Sulfide can also accumula e in
he body as a seconda y effec o inhe i ed diso de s, no ably e hyl-
malonyl encephalopa hy, now ecognized as a mi ochond ial disease
[53]. The abili y o AOX o p o ec agains sulfide oxici y hus has he
same double-edged aspec as H
2
S i sel . Whe e sulfide is p esen a
pa hological le els, causing impai ed mi ochond ial espi a ion, AOX
can po en ially o e come he p oblem, i applied he apeu ically.
Con e sely, i cy och ome oxidase inhibi ion is di ec ly ope a ing as a
physiological senso o H
2
S, o example in aso elaxa ion [52], AOX
could dis u b homeos a ic esponses. In ega d o hypoxia, p e ious
da a showed compa able effec s o low oxygen on he ac i i ies o AOX
and cy och ome oxidase [54]. Howe e , because he subuni -iso o m
composi ion o cIV a ies be ween issues [55], wi h some combina-
ions exhibi ing al e ed kine ic p ope ies in ega d o oxygen, i canno
be concluded ha AOX will pe o m compa ably o cIV in all issues and
condi ions. Since solid umou s ep esen a pa hologically impo an
low-oxygen en i onmen , he exp ession o AOX o manage o he dis-
eases may a ou (o dis a ou ) he g ow h o specific umou ypes,
leading o unin ended consequences.
3.2. AOX exp ession p omo es empe a u e-dependen g ow h accele a ion
P e ious s udies o D osophila ansgenic o Ciona AOX showed ha
ubiqui ous AOX exp ession esul ed in a sligh bu s a is ically sig-
nifican de elopmen al delay (egg- o-adul iming), and exagge a ed
weigh loss as young adul s [34]. Bo h o hese obse a ions a e con-
sis en wi h he idea ha AOX migh become a leas pa ially ac i e
du ing de elopmen , esul ing is less efficien use o s o ed nu i ional
esou ces. In such a case, by ca alyzing he same edox chemis y as he
mi ochond ial cy och ome chain, bu wi hou ene gy conse a ion ia
p o on-pumping, AOX should simply con e a g ea e p opo ion o
eleased ee ene gy o hea , as al eady demons a ed in i o [56]. One
co olla y o his is ha ac i a ion o AOX could po en ially be beneficial
a low empe a u es ha a e o he wise sub-op imal o he comple ion
o de elopmen , ia a he mogenic effec .
To es his we made use o a se o ansgenic fly lines exp essing
AOX unde he con ol o he α- ubulin p omo e in one, wo o h ee
diploid copies. These we e cul u ed unde diffe en empe a u es. The
pe iod o la al de elopmen , cha ac e ized by biomass accumula ion,
was essen ially unaffec ed by AOX exp ession a he s anda d g ow h
empe a u e o 25 °C, o a he ele a ed empe a u e o 29 °C (Fig. 2A).
Howe e , a low empe a u e, when he g ow h pe iod was g ea ly
ex ended, AOX exp ession esul ed in a ma kedly inc eased a e o
de elopmen , which was mos p onounced in flies bea ing h ee copies
o he ansgene, and a he lowes empe a u e es ed, 12 °C (Fig. 2A).
Pupal de elopmen was less affec ed by AOX (Fig. 2B): ins ead, he
2–3 d accele a ion p oduced du ing la al de elopmen was simply
main ained du ing me amo phosis (Fig. 2B). A low empe a u e we
also obse ed a significan ly highe p opo ion o eggs able o each he
pupal s age, also co ela ing wi h inc eased AOX gene dosage (Fig. 2C).
As in o he con ex s whe e apid g ow h p e ails (cance cells, yeas
in exponen ial g ow h phase in glucose- ich medium) ATP p oduc ion
in D osophila la ae depends la gely on glycolysis, conside ed as he
main high-capaci y pa hway o gene a ing bo h ene gy and o ganic
in e media es o biosyn hesis. Unde such condi ions, mi ochond ial
espi a ion ne e heless emains indispensable, since he p ocessing o
ca bon skele ons o biosyn hesis depends on he TCA cycle which, in
u n, equi es he efficien eoxida ion o p ima y elec on ca ie s ia
he espi a o y chain, a condi ion ecognized by Wa bu g as ‘ae obic
glycolysis’[57]. The capaci y o he sys em o ulfil his ole does no
appea o be limi ing in la ae g own unde s anda d condi ions, since
AOX exp ession has no effec on la al g ow h a e a 25 °C (Fig. 2A). I
he TCA cycle we e being es ained by he coupling o he RC o ATP
p oduc ion, enzyma ically ac i e AOX should accele a e he cycle, since
i p oduces less ATP pe molecule o ubiquinol oxidized. Howe e , a
low empe a u e, whe e all chemical eac ions a e slowed, g ow h
p oceeds much mo e slowly, e.g. la ae cul u ed a 12 °C g ow
a < 20% o he a e exhibi ed by hose g own a 25 o 29 °C (Fig. 2A).
The clea g ow h accele a ion p o ided by AOX exp ession unde hese
condi ions can hus be a ibu ed ei he o i s abili y o alle ia e lim-
i a ions on elec on flow, which migh cons ain he TCA cycle a low
empe a u e, o else o a di ec he mogenic effec aising he em-
pe a u e o he mi ochond ia [58] and o he whole o ganism. In whole
o pe meabilized mammalian cells g own a 37 °C [78], o in mi-
ochond ial homogena es om flies g own and assayed a diffe en
empe a u es be ween 18 and 29 °C [34], AOX exp ession did no
suppo 100% o he espi a o y capaci y o cIII unde uninhibi ed
condi ions. Thus, we s ongly a ou he second mechanis ic hypo h-
esis, whe eby a di ec he mogenic effec ac ually wa ms la ae suffi-
cien ly o accele a e de elopmen al p ocesses which a e sub-op imal a
low empe a u e. Tes ing his will no be s aigh o wa d, howe e ,
since D osophila la ae a e opaque o he dyes and epo e s hus a
de eloped as in acellula empe a u e epo e s.
Possible medical implica ions: based on hese findings, he use o AOX
o compensa e RC de ec s po en ially ca ies he unexpec ed isk ha i
could c ea e me abolic condi ions ha p omo e he g ow h o some
umou s. The ole o me abolism in cance has ecen ly a ac ed a lo o
in e es , al hough i is misleading o asse ha his ole is always in he
same di ec ion. Ra he , he ela i e dependence on glycolysis e sus
OXPHOS a ies g ea ly be ween umou ypes, o e en be ween a single
cance a diffe en s ages o he disease [59]. Howe e , apid g ow h is
gene ally associa ed wi h he high- h oughpu glycoly ic pa hway as
he majo sou ce o ATP, bu accompanied by a epu posing o he TCA
cycle (and OXPHOS) o anabolism [59], which applies bo h o de el-
oping D osophila la ae and o mos umou cells. Al hough he g ow h
accele a ion con e ed upon fly la ae by AOX was seen only a low
empe a u e, he mechanism could po en ially ope a e unde o he
s ess condi ions in cance cells, whe e he p ocessing o ca bon ske-
le ons o biosyn hesis could po en ially be limi ed by insufficiency o
he RC. An ob ious cause o such insufficiency would be he clonal
amplifica ion o m DNA mu a ions du ing he es ablishmen o he u-
mou , which could hen limi i s g ow h in la e s ages o he disease.
Many umou s a e indeed ound o ha bou dele e ious m DNA mu a-
ions. In such cases, AOX could alle ia e g ow h cons ain s a ising
om mu a ions in genes o subuni s o cIII o cIV o om mu a ions in
he mi ochond ial p o ein syn he ic appa a us whe e cIII o cIV we e
he mos affec ed p oduc s, such as he A8344 MERRF mu a ion [60].
No e, howe e , ha i is also possible o cons ue an opposi e a gumen ,
in cases whe e a cIII de ec p omo es umou g ow h o me as asis
h ough inc eased ROS p oduc ion [61]. In such cases, AOX may elie e
his effec and es ic cance p og ession. The in e ed he mogenic
effec o AOX aises addi ional po en ial issues, as discussed u he in
he ollowing sec ion.
3.3. AOX-exp essing flies show empe a u e dependen le hali y on nu ien -
poo media
The ea lie obse a ions o a sligh de elopmen al delay and mild
weigh loss in adul flies exp essing AOX sugges ha AOX-exp essing
flies use nu i ional esou ces less efficien ly han hei wild- ype
coun e pa s. To in es iga e his issue u he , we cul u ed flies ubi-
qui ously exp essing AOX, as well as a wide panel o con ols, on
S. Saa i e al. BBA - Molecula Basis o Disease 1865 (2019) 854–866
859
nu ien -poo media. On s anda d high-suga medium [62], AOX-ex-
p essing flies eclosed a he same equency as con ols (Fig. 3A),
whe eas on low-nu ien medium, con aining only 3.5% yeas and 5%
glucose plus aga and an imic obials, mos (~80%) AOX-exp essing
flies died as pupae (Fig. 3A). Con ols, including AOX ansgenic bu
non-exp essing flies, as well as flies exp essing GFP, eclosed almos
no mally on his medium (Fig. 3A). A second AOX ansgenic line be-
ha ed simila ly (Fig. 3B), whe eas a hi d, in which AOX exp ession is
much lowe [38], as well as a ansgenic line exp essing a ca aly ically
inac i e a ian o AOX [38] did no (Fig. 3C). AOX-exp essing pupae
cul u ed on low-nu ien medium died a a ious diffe en s ages o
me amo phosis (Fig. 3D), sugges ing ha hey had exhaus ed a gene al
componen equi ed o he comple ion o de elopmen , a he han
becoming blocked in a s age-specific p ocess. When he flies we e cul-
u ed in igh ly empe a u e-con olled incuba o s, he pheno ype was
ound o be ex emely empe a u e sensi i e (Fig. 4A). Specifically, a
22 °C AOX-exp essing flies eclosed a a nea -no mal equency on low-
nu ien medium, whe eas a 25 °C, and mo e acu ely a 26 °C, hey
mos ly died as pupae, whils a 27 °C, ha dly any AOX-exp essing flies
eclosed e en when cul u ed on s anda d medium. A 29 °C con ol flies
also showed a dec eased eclosion equency, especially on he low-nu-
ien medium. The pheno ype, including i s modula ion by empe a-
u e, was main ained e en when he glucose le el was aised o 10%
(Fig. 4A). Al hough his empe a u e-dependence may pa ly be ex-
plained by he inc eased exp ession o he ansgene a high empe a-
u e, unde he influence o he daGAL4 d i e (Fig. 4B), he simila i ies
wi h he beha iou o con ol flies a high empe a u e sugges s a
con ibu ion also om he he mogenic effec o he enzyme. Inc easing
he yeas concen a ion in he low-nu ien medium o 10% also ailed
o e e se he de elopmen al pheno ype, whils dec easing i o 1% led
o de elopmen al a es o a majo i y o wild- ype pupae as well
(Fig. 4C). E en when cul u ed on medium con aining only yeas and
aga , con ol flies eclosed no mally, whils > 90% o AOX-exp essing
flies died as pupae.
These findings imply ha componen s o ou s anda d high-suga
medium mus be c ucial in enabling AOX-exp essing flies o comple e
de elopmen . The s anda d medium is a complex mix u e con aining
se e al highly he e ogeneous componen s (including yeas , as well as
eacle, soya flou and maize flou ). Fu he analysis will be needed o
es ablish which specific nu ien (s) a e essen ial o AOX flies.
As al eady indica ed, D osophila de elopmen in ol es dis inc
phases. Du ing me amo phosis, he pupa does no eed, ins ead elying
upon biomass accumula ed du ing la al de elopmen . Me amo phosis
in ol es d as ic issue eo ganiza ion, uelled la gely by s o ed igly-
ce ides [63]. I AOX we e o become ac i a ed unde such condi ions,
ATP gene a ion could be se e ely cu ailed, since he pa hways o i-
glyce ide b eakdown ( ia glyce ol‑3‑phospha e dehyd ogenase and
a y-acid β-oxida ion) gi e ise o educing equi alen s ha mos ly
en e he espi a o y chain ia ubiquinol a cIII. This may accoun o
de elopmen al ailu e, i insufficien such esou ces had been accu-
mula ed p io o me amo phosis. Al e na i ely, i AOX we e ac i a ed
in la ae unde low nu ien condi ions, impai ing he efficiency o
biosyn hesis, his may accoun o insufficien amoun s o s o ed i-
glyce ides being laid down, and leading o de elopmen al ailu e a a
A
w1118
UAS-AOXF6
daGAL4
UAS-GFP
UAS-AOXF6 x daGAL4
UAS-GFP x daGAL4
UAS-AOXF6 x w1118
UAS-GFP x w1118
daGAL4 x w1118
nois
ol
c
e %
s anda d medium low-nu ien medium
100
80
60
40
20
0
% eclosion
100
80
60
40
20
0
B
low-nu ien medium
D
CUAS-emp y x daGAL4
UAS-mu AOX2nd x daGAL4
UAS-AOX8.1 x daGAL4
noisolce %
100
80
60
40
20
0
s anda d low-nu ien
medium medium
w1118
daGAL4
UAS-GFP x daGAL4
UAS-AOXF24 x daGAL4
UAS-AOXF24 x w1118
daGAL4 x w1118
Fig. 3. AOX-exp essing flies show de elopmen al a es on low-nu ien medium.
(A–C) P opo ion o pupae eclosing on he indica ed media, o he geno ypes o c osses ( emale × male) as shown (means ± SD). (D) Rep esen a i e images a
compa able zoom, o dying/de elopmen ally a es ed AOX-exp essing pupae, cul u ed on low-nu ien medium.
S. Saa i e al. BBA - Molecula Basis o Disease 1865 (2019) 854–866
860
la e s age.
Possible medical implica ions: ou findings can be conside ed as a
u he example o he inc easingly documen ed link be ween nu i ion,
mi ochond ial unc ion, and he accumula ion and use o a ese es.
Dys egula ion o he ele an p ocesses is inc easingly conside ed o
unde lie obesi y and o he me abolic diso de s [56,64,65]. He e we
ha e shown ha al e ing he balance be ween mi ochond ial subs a e
u iliza ion and ene gy p oduc ion can ha e p o ound implica ions o
biological p ocesses, leading o de elopmen al a es . T ying o d aw
exac pa allels be ween insec and mammalian physiology may be
misleading, especially gi en ha mammals show p ecise he mo-
egula ion and possess a issue dedica ed o his ole, b own a , al-
hough i s impo ance declines wi h age [66]. I AOX exp ession leads
mo e globally o inc eased a bu ning, o dec eased a accumula ion,
i s use in pa hologies wi h a me abolic dimension may ha e impo an ,
and in some cases undesi ed consequences. An analogy may be d awn
wi h human CPT2 deficiency, whe e he inabili y o use long-chain a y
acids as a me abolic uel leads ei he o ea ly le hali y o o sensi i i y
o s a a ion, depending on he se e i y o he mu a ion [67–69]. A
simila pheno ype is p oduced in flies homozygous o a null mu a ion
in D osophila CPT2. Once mo e, he e is also a po en ial ele ance in
cance , whe e me abolic inflexibili y, i.e. loss o he abili y o su i e
on diffe en subs a es, is a common obse a ion [70]. Con e sely, AOX
may be o po en ial use o co ec me abolic imbalances, o a leas o
in es iga e he ole o mi ochond ial ene ge ics in pa hology. Simila
ideas ha e been p oposed elsewhe e o ea men s ha induce he -
mogenic o ene ge ically ‘ u ile’pa hways [71]. Indeed, he mogenesis
is clea ly a ea u e o he immune esponse, and AOX may offe one
ou e o in ensi y i and a ge i mo e effec i ely.
3.4. AOX accele a es eco e y om cold-induced pa alysis in D osophila
adul s
The abo e findings sugges ha AOX may be unc ionally he mo-
genic unde specific physiological condi ions. E en hough he an-
sc ip ion o he gene was no induced in Ciona by cold empe a u e (Fig.
S1), ac i a ion in he cold may be an inhe en p ope y o he enzyme.
We he e o e es ed i s effec s on cold-exposed flies. In ials, 12 days a
4 °C was sufficien o kill 80–90% o con ol flies, whe eas flies homo-
zygous o 3 copies o he ubAOX ansgene in he same gene ic
backg ound all su i ed his ea men . A sho e pe iod o cold ex-
posu e (15 h a 4 °C in ood-con aining ials) was non-le hal o all flies,
bu induced pa alysis. When ials we e ans e ed o 24.5 °C, he AOX-
exp essing flies s a ed o become mobile a e 25 min, whe eas con ol
flies only began o eme ge om pa alysis a e 45 min (Fig. 5).
Possible medical implica ions: he simples in e p e a ion o hese
esul s is ha AOX becomes a leas pa ially ac i e a low empe a u e,
physically wa ming he o ganism o mi iga e le hali y a low
A+ daGAL4, 5% glc
+ daGAL4, 10% glc
+ daGAL4, s anda d
- daGAL4, 5% glc
- daGAL4, 10% glc
- daGAL4, s anda d
daGAL4 only, 5% glc
daGAL4 only, 10% glc
daGAL4 only, s anda d
noisolce%
B
w1118
UAS-AOXF6
daGAL4
UAS-AOXF6 x daGAL4
UAS-AOXF6 x w1118
daGAL4 x w1118
% eclosion
10% yeas 1% yeas
5% glc 5% glc
100
80
60
40
20
0
C
22 °C 25 °C 26 °C 27 °C 29 °C
100
80
60
40
20
0
UAS-AOXF6 x daGAL4
UAS-GFP x daGAL4
% eclosion
s anda d 3.5% yeas
medium 0% glc
100
80
60
40
20
0
D
18 25 26 29 °C c 3x ubAOX
AOX (37 kDa)
Ponceau S
UAS-AOXF6
Fig. 4. De elopmen al a es o AOX-exp essing flies on low-nu ien medium is empe a u e-dependen .
(A, C, D) P opo ion o pupae eclosing on diffe en media and a indica ed cul u e empe a u es, o he geno ypes o c osses ( emale x male) as shown (means ± SD).
Whe e no shown, expe imen s we e conduc ed a 25 °C and on media con aining 3.5% yeas as well as he indica ed amoun s o glucose (glc). (B) Wes e n blo s o
p o ein ex ac s om AOX-exp essing L3 la ae (UAS-AOX
F6
x daGAL4 c oss), cul u ed a he indica ed empe a u es, nega i e con ol (daGAL4) la ae g own a
25 °C (c) and posi i e con ol 3x ubAOX adul s, p obed o AOX, alongside Ponceau S-s ained memb ane o indica e ela i e loading. Molecula weigh o AOX,
37 kDa, ex apola ed om PageRule ™Plus P es ained P o ein Ladde (The moFishe Scien ific).
S. Saa i e al. BBA - Molecula Basis o Disease 1865 (2019) 854–866
861
empe a u e and accele a ing eco e y om cold-induced pa alysis.
Since a he mogenic effec o AOX in plan s is well documen ed
[17,18], ou findings ha i may ac simila ly in some animals is no
implausible. Howe e , i AOX we e deployed he apeu ically in
homeo he ms such as humans, excess hea p oduc ion should ac i a e
he mo egula o y mechanisms and, in an ex eme si ua ion, migh
o e whelm hem. Mo eo e , i he excess hea is no efficien ly con-
duc ed away om he mi ochond ia, as sugges ed by ecen s udies
[58], i may lead o unp edic able changes in enzyme ac i i ies and
dis u b, a he han es o e, me abolic homeos asis. On he o he hand,
i p ope ly egula ed, AOX may p o ide a possible ea men o some
o ms o obesi y, by inc easing he amoun o a o ca bohyd a e ha
mus be bu ned o gene a e a gi en amoun o ATP, bu wi hou he
po en ially ha m ul effec s a ising om loss o mi ochond ial mem-
b ane po en ial p oduced by chemical uncouple s o uncoupling p o-
eins such as UCP1.
3.5. AOX has o he unexpec ed effec s on de elopmen and physiology
When unde aking he exp ession o Ciona AOX in me azoan models
such as D osophila o he mouse, we ini ially conside ed he possibili y
ha widesp ead exp ession o he enzyme would be le hal, due o i s
abili y o sho -ci cui mi ochond ial ATP p oduc ion when unc ionally
ac i e. We ini ially made he su p ising obse a ion ha ubiqui ous
AOX exp ession was ully compa ible wi h de elopmen and an os en-
sibly no mal physiology in bo h flies and mice [34,72]. This, plus he
known p ope ies o he plan and algal enzyme, which is ac i a ed only
unde condi ions o o e - educ ion o he quinone pool [3,4], hen led
us in he opposi e di ec ion. I AOX is only ac i e unde condi ions o
me abolic imbalance and oxida i e s ess, i can be p oposed as a po-
en ial wide-spec um he apeu ic ool ha is benign in he uns essed
o ganism. The findings p esen ed abo e now a gue o an in e media e
conclusion, namely ha AOX, whils displaying clea beneficial effec s,
does impac physiological p ocesses in ways ha we e no expec ed,
e en i hey can be a ionalized om he known p ope ies o mi-
ochond ia and cells. Ac i a ion o AOX by specific me aboli es o en-
e ge ic s a es can po en ially in e e e wi h specific de elopmen al
p ocesses. The causes and consequences a e no easy o p edic me-
chanis ically, since we do no ha e a ull desc ip ion o he me abolic
changes ha occu in animal de elopmen , no do we ye ha e a clea
idea o how he ac i i y o AOX om Ciona is egula ed. Some o he
examples om ou ecen wo k illus a e hese poin s.
AOX-exp essing flies o bo h sexes a e e ile and p oduce no mal
numbe s o offsp ing [34]. Howe e , when es ed in a di ec compe i-
ion assay, as illus a ed in Fig. 6, AOX-exp essing males we e sys em-
a ically ou -compe ed by con ol males [73]. We aced he eason o
his abno mali y o a ela i e lack o spe m p oduc ion by AOX-ex-
p essing males [73]. Su p isingly, his did no appea o be due o AOX
exp ession in he ge mline, which was seen only a low le els com-
pa able wi h he de ec ion limi , bu o i s exp ession in he pigmen
cells o he es is shea h. Mo eo e , i was associa ed no wi h a block
on diffe en ia ion as such, bu wi h i s spa ial diso ganiza ion and
consequen ailu e o accumula e ma u e spe m in o he seminal e-
sicle. The p ecise mechanis ic basis o he pheno ype emains o be
elucida ed, al hough i logically in ol es de anged signaling om he
mi ochond ia- ich pigmen cells o he unde lying smoo h muscle, im-
peding he pe is al ic mo emen o ma u ing spe m cys s. Why AOX
would be ac i a ed in hose cells, and whe he he effec is a ibu able
o dec eased ATP p oduc ion, impai ed ROS signaling, inc eased em-
pe a u e o some o he p ocess, emain open ques ions.
I is o be expec ed ha me abolic s ess can also lead o ansien
ac i a ion o AOX, blun ing he signals ha enable o ganisms o e-
spond o and nega e such s esses. Thus, he e may be ins ances whe e
AOX impai s egene a i e esponses and exace ba es s ess-induced
damage. Con e sely, we ha e iden ified se e al con ex s whe e AOX
exp ession appea s able o po en ia e no mal signaling o o quench
abno mal signaling, in ei he case p e en ing a pa hological ou come.
Pe haps he mos ema kable example is he abili y o AOX o p e en a
ange o de elopmen al dysmo phologies induced by abno mal nuclea
ecep o ac i i y. These a e p o oked by he combined ac ion o a
modified GAL4 d i e (‘GeneSwi ch’), which includes he ligand-
binding domain o he p oges e one ecep o , plus an excess o he a -
ificial s e oid ha induces i , bu is b ough abou in he absence o any
ansgene. Exp ession o AOX, e en om a GAL4-independen p o-
mo e , was able o la gely p e en he appea ance o hese dysmo -
phological pheno ypes, including cle ho ax o abdomen, apop o ic
wing segmen s, mal o med legs o senso y b is les [74]. AOX exp es-
sion was also able o co ec he mos equen ly obse ed o hese
abno mali ies, cle ho ax, when induced by a comple ely sepa a e
manipula ion, h ough he impai men o Jun N- e minal kinase (JNK)
signaling in he do sal ho acic midline du ing me amo phosis [75].
AOX was also ound o p omo e cell mig a ion in immo alized mouse
emb yonic fib oblas s, and o coun e ac he nega i e effec s on his
p ocess o a leas one p o ein kinase inhibi o [75].
Ou s udies o he p ope ies o mouse models is much less ad anced
a his poin , bu one in iguing obse a ion is he ac ha AOX is able
o block he le hal effec s o lipopolysaccha ide (LPS), in a mouse model
o sepsis [76]. This finding was pa o a wide s udy o unde s and he
ole o mi ochond ia in cy okine elease by ac i a ed mac ophages. I
no p ope ly con olled, his is belie ed o lead o sep ic shock and
o gan ailu e. AOX was in e ed o p o ide a shun o blocked elec-
ons, no malizing mi ochond ial me abolism in mac ophages and
p e en ing excess ROS p oduc ion a ising om e e se elec on flow
h ough cI. This s udy p o ides an impo an benchma k o how AOX
can be used o p obe he mi ochond ial ole in di e se pa hologies,
physiological and de elopmen al p ocesses. The lesson om s udies in
D osophila is ha a mi ochond ial ole should always be conside ed in
any complex o unexplained pa hological pheno ype, and he a ail-
abili y o ansgenic lines o AOX and NDX p o ides a way o es ing
his whene e a c edible animal o cell-cul u e model exis s.
Possible medical implica ions: i is a guably anci ul o ansla e
knowledge di ec ly om D osophila o e en he mouse o human dis-
ease, wi hou exhaus i e alida ion s udies. The e o e, he ac ha
AOX impai s male ep oduc i e compe i i eness o co ec s cle ho ax
in flies does no mean ha male in e ili y mus au oma ically be con-
side ed a mi ochond ial disease o ha AOX could somehow be de-
ployed o ea midline closu e de ec s in human de elopmen , such as
spina bifida o cle pala e. Ne e heless, he use o AOX and NDX o
p obe mi ochond ial in ol emen in disease- ela ed p ocesses should
p o ide powe ul clues ha could ha e un o eseen applica ions in
medicine.
5
1/s
eil el
i
bom
m
mi
o.
o
N
20 25 30 35 40 45
mins
16
14
12
10
8
6
4
2
0
w1118
3x ubAOX
Fig. 5. AOX exp ession accele a es eco e y om cold-induced pa alysis.
Numbe o flies emaining immobile ( om ba ches o 15 i gin emales) a he
indica ed imes a e shi ing om 4 °C o 24.5 °C: flies exp essing AOX
(3x ubAOX) e sus backg ound con ol line w
1118
.
S. Saa i e al. BBA - Molecula Basis o Disease 1865 (2019) 854–866
862