RESEARCH ARTICLE Open Access
Exp ession o Ciona in es inalis AOX causes
male ep oduc i e de ec s in D osophila
melanogas e
Sina Saa i
1†
, Ana Andjelko ić
1†
, Geo ana S. Ga cia
2
, Howa d T. Jacobs
1,3*
and Ma cos T. Oli ei a
1,2
Abs ac
Backg ound: Mi ochond ial al e na i e espi a o y-chain enzymes a e phylogene ically widesp ead, and bu e
s esses a ec ing oxida i e phospho yla ion in species ha possess hem. Howe e , hey ha e been los in he
e olu iona y lineages leading o e eb a es and a h opods, aising he ques ion as o wha su i al o ep oduc i e
disad an ages hey con e . Recen in e es in using hem in he apy lends a biomedical dimension o his ques ion.
Me hods: He e, we examined he impac o he exp ession o Ciona in es inalis al e na i e oxidase, AOX, on he
ep oduc i e success o D osophila melanogas e males. Spe m-compe i ion assays we e pe o med be ween lies
ca ying h ee copies o a ubiqui ously exp essed AOX cons uc , d i en by he α- ubulin p omo e , and wild- ype
males o he same gene ic backg ound.
Resul s: In spe m-compe i ion assays, AOX con e ed a subs an ial disad an age, associa ed wi h dec eased
p oduc ion o ma u e spe m. Spe m di e en ia ion appea ed o p oceed un il he las s ages, bu was spa ially
de anged, wi h spe ma ozoids e ained in he es is ins ead o being eleased o he seminal esicle. High AOX
exp ession was de ec ed in he ou e mos cell-laye o he es is shea h, which we hypo hesize may dis up a signal
equi ed o spe m ma u a ion.
Conclusions: AOX exp ession in D osophila hus has e ec s ha a e dele e ious o male ep oduc i e unc ion. Ou
esul s imply ha AOX he apy mus be de eloped wi h cau ion.
Keywo ds: Mi ochond ia, Respi a o y chain, Spe ma ogenesis, Spe m compe i ion
Backg ound
In he animal kingdom, he al e na i e mi ochond ial e-
spi a o y chain is widely ep esen ed, as illus a ed by
he b oad phylogene ic dis ibu ion o he gene o he
al e na i e oxidase (AOX) [1, 2]. AOX is a single-
subuni enzyme able o bypass espi a o y complexes III
and IV, ans e ing elec ons om ubiquinone o oxy-
gen wi hou p o on pumping and ATP p oduc ion. I is
belie ed o bu e s esses a ec ing oxida i e phospho -
yla ion, whe he due o o e load, en i onmen al insul s
o gene ic damage. Howe e , AOX, as well as o he al-
e na i e espi a o y enzymes, was los independen ly
du ing he ea ly e olu ion o bo h e eb a es and a -
h opods. This aises he ques ion o wha disad an ages
AOX migh ha e con e ed on he su i al and/o e-
p oduc i e i ness o he ances o s o hese axa.
In ac , only ma ine in e eb a es appea o s ill pos-
sess genes o al e na i e enzymes [1–3]. None heless,
when he AOX gene om he unica e Ciona in es inalis
(Ascidiacea) was exp essed in cul u ed human cells, i
p omo ed esis ance o inhibi o s o complexes III and
IV [4] and compensa ed o he g ow h de ec s and
oxidan -sensi i i y o complex IV-de icien cell-lines [5].
In he ui ly D osophila melanogas e , ansgenic ex-
p ession o C. in es inalis AOX had no appa en e ec
on iabili y, ep oduc ion and heal h o he o ganism [6].
I con e ed esis ance o o he wise le hal le els o com-
plexes III and IV inhibi o s [6], mi iga ed locomo o de-
ec s in he dj-1βPa kinson’s disease model [6] and
* Co espondence: [email p o ec ed]
†
Equal con ibu o s
1
Ins i u e o Biosciences and Medical Technology, Tampe e Uni e si y
Hospi al, Uni e si y o Tampe e, FI-33014 Tampe e, Finland
3
Ins i u e o Bio echnology, Uni e si y o Helsinki, FI-00014 Helsinki, Finland
Full lis o au ho in o ma ion is a ailable a he end o he a icle
© The Au ho (s). 2017 Open Access This a icle is dis ibu ed unde he e ms o he C ea i e Commons A ibu ion 4.0
In e na ional License (h p://c ea i ecommons.o g/licenses/by/4.0/), which pe mi s un es ic ed use, dis ibu ion, and
ep oduc ion in any medium, p o ided you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o
he C ea i e Commons license, and indica e i changes we e made. The C ea i e Commons Public Domain Dedica ion wai e
(h p://c ea i ecommons.o g/publicdomain/ze o/1.0/) applies o he da a made a ailable in his a icle, unless o he wise s a ed.
Saa i e al. BMC De elopmen al Biology (2017) 17:9
DOI 10.1186/s12861-017-0151-3
amelio a ed di e se pheno ypes associa ed wi h complex
IV de iciency [7], de ec i e m DNA eplica ion [8], o
exp ession o human β-amyloid [9]. C. in es inalis AOX
has also been success ully exp essed in he mouse wi h-
ou any signi ican e ec on majo physiological pa ame-
e s [10, 11]. I led o dec eased p oduc ion o eac i e
oxygen species (ROS) when he espi a o y chain was
blocked, was able o suppo cyanide- esis an espi -
a ion by in ac o gans, and con e ed p olonged p o ec-
ion agains le hal concen a ions o hyd ogen cyanide in
whole animals [10, 11].
These bene i s o AOX exp ession aise he possibili y
o i s e en ual use o ea men o human pa ien s wi h
complexes III and/o IV de iciencies, as a by-pass he -
apy [12]. Howe e , he ad an ages o AOX exp ession in
‘highe ’animal sys ems unde pa hological condi ions
con as s wi h i s e olu iona y loss om he genome o
e eb a es and a h opods. A de ailed s udy o he biol-
ogy o he al e na i e pa hways is he e o e equi ed be-
o e any use o his enzyme in gene he apy o
mi ochond ial o neu odegene a i e diso de s can be
con empla ed [13].
We se ou o in es iga e he pa adox be ween he
bene icial by-pass and he possible maladap i e conse-
quences o AOX exp ession in highe me azoans, by di -
ec ly es ing i s e ec s on male ep oduc i e unc ion.
Spe m a e highly dependen on biological ene gy. The e-
o e, we easoned ha i may be a ele an a ge in
which o es o sub le bu unc ionally meaning ul de -
imen s o AOX exp ession. We challenged D. melano-
gas e males exp essing AOX in compe i ion assays wi h
AOX-nonexp esso s. AOX p oduced a clea impai men
o ep oduc i e unc ion ha was associa ed wi h de-
c eased p oduc ion o ma u e spe m cells. This sugges s
ha possible gene he apy applica ions o AOX mus be
de eloped wi h cau ion.
Resul s
Males exp essing AOX a e de ec i e in spe m-compe i ion
assays
In D osophila, as in many species, emales ypically ma e
wi h a succession o males, s o ing spe m inside he e-
male body o e ilize oocy es as hey ma u e [14]. Spe m
compe i ion occu s by means o a subs i u ion mechan-
ism, in which a second male dislodges he spe m o he
i s male om he emale’s ep oduc i e ac and e-
places i wi h i s own game es [14, 15]; his is appa en ly
he mo i e o ce o he co ela i e e olu ion be ween
he male’s gian spe m cells and he emale’s spacious
spe m-s o age o gans [16]. Newly acqui ed spe m com-
pe es wi h and usually displaces ha om p e ious
males, unless i is unc ionally comp omised. These
mechanisms can d i e e olu ion by maximizing male e-
p oduc i e success, such as by he de elopmen o la ge
es es, p oduc ion o mo e abundan , mo e iable o
longe -las ing game es, o by modi ying he esponse o ele-
a ed compe i ion isk, among o he cha ac e is ics [17].
To es he ep oduc i e success o D. melanogas e
males exp essing AOX, we pe o med spe m-
compe i ion assays be ween lies ca ying h ee copies o
he α- ubulin-AOX cons uc , as desc ibed p e iously
(3X ub-AOX line: geno ype ub-AOX
35
/Y; ub-AOX
112
/
ub-AOX
112
; ub-AOX
7
/ ub-AOX
7
[7]) and wild- ype
males o he same gene ic backg ound (w
1118
). In he
‘de ensi e’app oach, 3X ub-AOX males we e i s
allowed o ma e wi h i gin w
1118
emales, which we e
hen ma ed wi h w
1118
males in he absence o 3X ub-
AOX males. P ogeny om he compe ing males can be
dis inguished by eye-colou : hose om ansgenic males
ha e ed eyes, whils hose om w
1118
males ha e whi e
eyes. Con ols consis en ly show ha he eye-colou
ma ke , as such, has no in luence o e he ou come
(Addi ional ile 1: Fig. S1).
The numbe o p ogeny o igina ing om he spe m o
3X ub-AOX males was dec eased a e he emales we e
ma ed subsequen ly wi h w
1118
males (Fig. 1a, uppe
panel). This esul is consis en wi h p e iously pub-
lished da a using di e se gene ic backg ounds [14, 18–
21] and wi h ou own da a using se e al o he ans-
genes (Fig. 1b, Addi ional ile 1: Fig. S1–S3): spe m o
he i s male is eplaced (pa ially o comple ely) by he
spe m o he second male. To implemen a igo ous s a -
is ical analysis o he indings, we de i ed he pa ame e
P1’which measu es he p opo ion o p ogeny si ed by
he i s male o ma e in spe m-compe i ion assays [22].
This was simila o all lines analyzed he e, al hough
3X ub-AOX males si ed s a is ically ewe o sp ing han
some o he AOX-nonexp esso con ols (Addi ional ile
1: Table S1).
In con as , when 3X ub-AOX males we e challenged
in an ‘o ensi e’app oach, i.e., when he i gin w
1118
e-
males we e i s c ossed wi h w
1118
males, hen wi h
AOX-exp essing males, he numbe o 3X ub-AOX p o-
geny ne e o e came he numbe o w
1118
p ogeny (Fig.
1a, lowe panel). In ac , e en a e 10 days since w
1118
males we e emo ed om he ma ing ials, eggs e il-
ized wi h he o iginal male spe m we e s ill p e e en ially
laid o e he ones e ilized wi h 3X ub-AOX spe m (Fig.
1a, lowe panel, ial IV). Con ol expe imen s pe o med
wi h ed-eyed daugh e less-GAL4 (daGAL4) (Fig. 1b,
lowe panel) and ubulin-GeneSwi ch ( ubGS) and also
UAS-emp y
2nd
and UAS-emp y
3 d
(Addi ional ile 1: Fig.
S1) males imply ha he ailu e o AOX-exp essing
males o compe e success ully in he o ensi e pa adigm
is speci ic o AOX, and is no a p ope y o ansgenic
lines in gene al. The de ec appea ed o be AOX
ansgene-dose dependen . Males ca ying only wo cop-
ies o he ub-AOX ansgene (Addi ional ile 1: Fig. S2)
Saa i e al. BMC De elopmen al Biology (2017) 17:9 Page 2 o 11
appea ed less impai ed in he ‘o ensi e’pa adigm han
hose ca ying h ee (Fig. 1a), whils hose wi h jus a
single copy o ub-AOX we e no signi ican ly di e en
om con ol males (Addi ional ile 1: Fig. S2). These
conclusions e lec he s a is ical analysis o he p opo -
ion o p ogeny si ed by he second male (P2’, [22]),
whe ein 3X and 2X ub-AOX males we e signi ican ly im-
pai ed in he o ensi e pa adigm compa ed o all con ol
classes (Addi ional ile 1: Table S2), al hough he di e -
ences be ween 3X and 2X ub-AOX (Addi ional ile 1:
Table S2) hemsel es we e no signi ican . The obse ed
pheno ypes also co ela e app oxima ely wi h he
amoun o exp ession o he ub-AOX ansgene in he
male ep oduc i e sys em, obse ed by immunoblo ing
(Addi ional ile 1: Fig. S4).
Howe e , exp essing GAL4-dependen UAS-AOX
using he daGAL4 d i e [6], p oduced no loss o
spe m-compe i i eness (Addi ional ile 1: Fig. S3A, and
Table S2), despi e he ac ha he le el o AOX p o ein
in he male ep oduc i e sys em was compa able wi h
ha exp essed om mul iple copies o ubAOX (Add-
i ional ile 1: Fig. S4). These da a sugges ha he e ec
o AOX on spe m compe i i eness depends on he
p ecise cellula con ex o i s exp ession a he han sim-
ply i s o e all amoun .
AOX-exp essing males accumula e a dec eased amoun o
ma u e spe m
Success ul spe m compe i ion in D osophila depends
bo h on he numbe and quali y o spe ma ozoids and
on he compa ibili y be ween he male’s seminal luid
p o eins and hei ecep o s in he emale ep oduc i e
ac [17, 23–25]. We checked how he p oduc ion o
ma u e spe ma ozoids in he 3X ub-AOX males is a -
ec ed by dissec ing he ep oduc i e o gans in adul
males o inc easing age. Spe ma ogenesis in D. melano-
gas e s a s a he dis al ip o he es es, aking place in-
side cys s. As cell di e en ia ion p oceeds, he cys s
mo e owa ds he p oximal end o he es es, whe e
hese o gans connec o he SVs. The ma u e spe ma o-
zoids a e hen deposi ed in he SVs, whe e hey a e
s o ed un il ma ing [26]. The e o e, as adul males age,
he es es end o ge hinne , because di e en ia ion in
he cys s p oceeds and he ully o med spe ma ozoids
mo e o he SVs, which in u n inc ease in size. As
shown in Fig. 2 (panels a and c), he hickness o he
Fig. 1 AOX-exp essing males a e de ec i e in spe m-compe i ion assays. In he de ensi e pa adigm (uppe panels), i gin w
1118
emales (whi e
eyes) we e ini ially c ossed indi idually wi h ansgenic males ( ed eyes), and hen wi h w
1118
males (whi e eyes). In he o ensi e pa adigm (lowe
panels), he i gin w
1118
emales we e i s c ossed wi h w
1118
males, and hen wi h ansgenic males (see Me hods o de ails). The ansgenic
males used we e 3X ub-AOX ( ubAOX
35
/Y; ubAOX
112
/ ubAOX
112
; ubAOX
7
/ ubAOX
7
)ina, and daGAL4 (X/Y; 2/2; daGAL4/daGAL4)inb. Vials I-IV ep e-
sen he mean numbe o whi e- o ed-eyed p ogeny ± SD (e o ba s), eclosed a e 3 days om he ini ial c osses ( ial I), 3 days o egg-laying
du ing second c osses ( ial II), 3 days o o iposi ion only, subsequen o second c osses ( ial III, males disca ded), ollowed by 5 days o u he
o iposi ion ( ial IV)
Saa i e al. BMC De elopmen al Biology (2017) 17:9 Page 3 o 11
es es o con ol lies dec eases >50% in he i s 10 days
o adul li e, whe eas he SVs iple in size. In con as ,
he es es o 3X ub-AOX males exhibi a much less
p onounced dec ease in hickness (~25%), whils a he
same ime hei SVs emain imma u e. In addi ion, we
obse ed an accumula ion o whi ish ma e ial a he
A
B
C
Fig. 2 Ma u e spe m cells do no accumula e in he seminal esicles (SVs) o AOX-exp essing males. Rep esen a i e samples o dissec ed es es and SVs
om adul males o he indica ed age a e shown o con ol and AOX-exp essing lines: aw
1118
(X/Y;2/2;3/3)and3X ub-AOX ( ubAOX
35
/Y; ubAOX
112
/
ubAOX
112
; ubAOX
7
/ ubAOX
7
); bUAS-AOX
F6
/w
1118
(X/Y; UAS-AOX
F6
/2; 3/3) and UAS-AOX
F6
/ ubGS (X/Y; UAS-AOX
F6
/2; ubGS/3) in he absence (uninduced) o
p esence (induced) o 200 μM mi ep is one. AOX p o ein induc ion by mi ep is one was isualized by immunoblo ing (Addi ional ile 1: Fig. S5). Red
ci cles indica e one o he dissec ed SVs o he ep esen a i e samples o he indica ed geno ype. No e he unde de eloped s a e o he SVs in 10 day-
old lies om he lines exp essing AOX ei he cons i u i ely (3X ub-AOX)o inducibly(UAS-AOX
F6
/ ubGS plus mi ep is one). Fo an explana o y isualiza ion
o D. melanogas e ep oduc i e o gans, see Addi ional ile 1: Fig. S6A. In a, heinse highligh s he whi ish ma e ial ha accumula es in he p oximal end
o he es is in 3X ub-AOX males; in b, he ed a ow poin s o he same ma e ial o UAS-AOX
F6
/ ubGS males ea ed wi h mi ep is one. cQuan i ica ion o
he da a shown in aand b, in which he da a poin s ep esen he mean hickness o 10–20 dissec ed o gans ± SD (e o ba s). * indica es s a is ically
signi ican di e ence be ween o gans o AOX-exp essing and nonexp essing males (p≤0.01). Addi ional ile 1: Fig. S8B illus a es schema ically he
hickness measu emen p o ocol used o es ima e he amoun o imma u e (in he es es) and ma u e (in he SVs) spe m cells pe ly
Saa i e al. BMC De elopmen al Biology (2017) 17:9 Page 4 o 11
p oximal end o he es es o 10-day-old 3X ub-AOX
males (Fig. 2a, inse ), which we in es iga ed u he (see
below). The ma u e spe ma ozoids p oduced by hese
males, al hough ew in numbe , did no appea o ha e
any mo ili y de ec s, as judged by isual inspec ion.
We we e able o p oduce a simila pheno ype using
he inducible GeneSwi ch ( ubGS) sys em, in which he
same α- ubulin p omo e as used in he ub-AOX con-
s uc s con ols he exp ession o a modi ied GAL4 ha
is inducible by mi ep is one (RU486). This allows ime-
egula ed exp ession o a GAL4-dependen AOX ans-
gene in exac ly he same issues as in he ubAOX lines.
By ans e ing males o ood- ials con aining he indu-
cing d ug on he day o eclosion, any possible de elop-
men al dis u bance is a oided, whils accu a e con ols
can be implemen ed, no ably lies wi h ansgene, bu
lacking d i e and/o d ug (see Addi ional ile 1: Fig. S5
o immunoblo s indica ing igh egula ion o AOX p o-
ein le els using his sys em). Simila ly as o 3X ub-
AOX males, he es es o UAS-AOX ansgenic males
d i en by ubGS in he p esence o mi ep is one
emained hick du ing he i s 10 days o adul li e,
whils hei SVs emained small (Fig. 2b and c). These
o gans we e as wild- ype in he absence o he d i e o
inducing d ug (Fig. 2b and c). The no mal hinning o
he es is o e he i s 10 days o adul li e was also seen
when bo h he d i e and inducing d ug we e p esen ,
whe he d i ing a con ol ansgene, GFP, a ca aly ically
inac i e o m o AOX, mu AOX [27] o e en wi h no
ansgene a all (Addi ional ile 1: Fig. S6A, S6C). In
hese cases, howe e , he SV emained unde eloped, bu
only when d i e and d ug we e bo h p esen (Add-
i ional ile 1: Fig. S6B, S6D). The whi ish ma e ial nea
he p oximal end o he es is also accumula ed p omin-
en ly when ubGS was used o exp ess AOX (bu no
GFP o mu AOX), and again only unde inducing condi-
ions (Fig. 2b, ed a ow). To con i m ha his is an α-
ubulin p omo e -speci ic pheno ype, we dissec ed he
ep oduc i e o gans o 10-day-old males exp essing
UAS-AOX d i en by daGAL4. The SVs had a no mal
mo phology (Addi ional ile 1: Fig. S7), and appea ed o
ha e accumula ed a simila amoun o ma u e spe m
cells as con ols, in ag eemen wi h he esul s o he
spe m-compe i ion assays (Addi ional ile 1: Fig. S3).
O he ep oduc i e o gans we e also e alua ed mo -
phologically (see Addi ional ile 1: Fig. S8A o sche-
ma ics o he ep oduc i e sys em o D. melanogas e
males), bu no ob ious al e a ions we e obse ed be-
ween AOX-exp essing and con ol males, based on is-
ual inspec ion o >70 dissec ed lies o each geno ype.
Al hough AOX was no exp essed in he accesso y
glands (Addi ional ile 1: Fig. S9A), we also checked he
p esence o Sex Pep ide in hese o gans by immuno-
s aining and con ocal mic oscopy, and again obse ed
no di e ence (Addi ional ile 1: Fig. S9B). Sex Pep ide is
one o he mos impo an componen s o he male’s
seminal luid; i has been implica ed in ep oduc i e suc-
cess in spe m-compe i ion assays [24, 28] and i s ac ion
is dependen on he compa ibili y wi h ecep o s in he
emale ep oduc i e ac [23]. Al oge he , ou da a indi-
ca e ha AOX-exp essing males a e de ec i e in spe m-
compe i ion assays due o a dec eased p oduc ion o ma-
u e spe ma ozoids.
AOX is exp essed in cells o he es is and seminal esicle
shea hs
Al hough he SVs o AOX-exp essing males p esen pe -
haps he mos p onounced pheno ype in ou s udy, hese
a e pu ely s o age o gans o ma u e spe m cells [26],
and hei mo phology was no diagnos ic o AOX ex-
p ession (Addi ional ile 1: Fig. S6B).
The e o e, we hypo hesized ha he de ec in spe m
p oduc ion caused by AOX exp ession mos likely a -
ec s he es es, whe e spe ma ogenesis akes place. The
α- ubulin p omo e has been used p e iously o d i e
ansgene exp ession in he ly es is in bo h soma ic
cells and he ge mline ( om he s em-cell s age o la e
spe ma ocy es, e iewed in [29]). Howe e , using im-
muno luo escen con ocal mic oscopy (Fig. 3 and Add-
i ional ile 1: Fig. S10), AOX exp ession om he α-
ubulin p omo e , whe he di ec ly (Addi ional ile 1:
Fig. S10) o d i en by ubGS plus mi ep is one (Fig. 3),
was below he le el o de ec ion in ge mline cells, and
seen only a e y low le els in some soma ic cys cells.
Ne e heless, i was abundan ly exp essed in he soma ic
cells o he es is and SV shea hs. In addi ion, all o he
imma u e s ages o cell di e en ia ion appea ed o be
p esen (Addi ional ile 1: Fig. S11), judging by he cha -
ac e is ic changes in nuclea mo phology, mi ochond ial
ne wo k a angemen and cellula elonga ion ha he
ge m line cells go h ough owa ds he inal s eps in
spe ma ogenesis [26].
A highe esolu ion, i was clea ha no all cells in
he enshea hing issues o he es es and SVs exp ess
AOX a a high le el when d i en by he α- ubulin p o-
mo e (ei he di ec ly, as 3X ub-AOX o indi ec ly,
h ough ubGS). Some cells highly exp essing ATP5A, a
complex V subuni adi ionally used as a mi ochond ial
ma ke , appea ed nega i e o AOX (Addi ional ile 1:
Fig. S12A). Using UAS-S inge GFP in combina ion wi h
he ubGS d i e , in o de o ma k he posi i e cells wi h
GFP in he nucleus, we obse ed high exp ession in spe-
ci ic cells in he ou e mos shea h-cell laye (Addi ional
ile 1: Fig. S12B). The inne shea h-cell laye , o med o
smoo h-muscle cells wi h mul iple small nuclei, ap-
pea ed nega i e o AOX o GFP, in ag eemen wi h co-
s aining o AOX and ac in (Addi ional ile 1: Fig. S13A)
in he es es o 3X ub-AOX males. Posi i e cells mos
Saa i e al. BMC De elopmen al Biology (2017) 17:9 Page 5 o 11

likely co espond wi h he pigmen cells, which ca y a
single la ge nucleus, and a e abundan a ound he SVs
and he p oximal end o he es is, whe e hese wo o -
gans connec [30]. This is suppo ed (Addi ional ile 1:
Fig. S13B–D) by co-s aining o he emp y spi acles gene
p oduc (ems), which is a ma ke o he pigmen cells
[31, 32]. Pa ial h ee-dimensional econs uc ion o he
mic oscopy images shows ha he AOX and ems signals
a e loca ed in he ou e mos cell laye o he shea h o
he SV (Addi ional ile 1: Fig. S13C) and es is (Add-
i ional ile 1: Fig. S13D), whils he unde lying (smoo h
muscle) cells s ain highly o ATP5A and ac in. No e
ha he pigmen cells a e abundan in he ana omical
egion whe e he mos p onounced mo phological al e -
a ions we e ound in AOX-exp essing males, ma ked by
he accumula ion o whi ish ma e ial a he p oximal
end o he es is combined wi h a dec eased amoun o
ma u e spe m cells in he SVs (Fig. 2a).
Ma u e-looking spe ma ozoids a e lodged in he p oximal
end o he es is in AOX-exp essing males
The p oximal end o he D. melanogas e es is is also
he egion whe e he indi idualiza ion p ocess s a s.
This p ocess is essen ial o he inal s age o elonga ed
spe ma id di e en ia ion [26], and is accomplished by an
indi idualiza ion complex (IC) ha is assembled in he
cys egion con aining he spe ma id nuclei and which
hen mo es owa ds he ip o he ails, collec ing synci-
ial cy oplasm and c ea ing cys ic bulges and was e bags
(WBs) along he way [33, 34]. By s aining he es is sam-
ples o ac in and ac i a ed caspase-3, wo componen s
o he ICs, we obse ed no signi ican di e ences in he
numbe o s a ing and es ablished ICs in males exp ess-
ing AOX d i en by ubGS, bu did obse e an ele a ed
numbe o WBs (Addi ional ile 1: Fig. S14). Mos no-
iceably, we obse ed in hese males a d ama ic change
in he dis ibu ion o ICs h oughou he o gan. S a ing
ICs a e usually p esen in he p oximal egion o he
es is, bu in 10 day-old males exp essing AOX d i en
by ubGS, a signi ican numbe o hem we e ound o-
wa ds he middle egion o he o gan. Es ablished ICs,
which a e usually ound all along he es es, we e con-
cen a ed a he middle and p oximal end o he AOX-
exp essing es es. Finally, WBs appea ed o be dis ib-
u ed in all es is egions in he AOX-exp essing males,
whe eas in con ol es es hey accumula ed a he dis al
end (Addi ional ile 1: Fig. S14C). Ou obse a ions
imply ha AOX exp ession d i en by ubGS in he
Fig. 3 AOX p o ein is localized in mi ochond ia o cells om he es is and seminal esicle (SV) shea hs. Immuno lo escence s aining and con ocal
mic oscopy o he indica ed p o eins and DAPI, om a 10-day old male o geno ype X/Y; UAS-AOX
F6
/2; ubGS/3, in he p esence o mi ep is one. AOX
p o ein is localized o he ou e mos cell laye o hese o gans, whe e i o e laps ha o he mi ochond ial ma ke ATP5A, bu no o ge mline cells a
any s age o di e en ia ion. The ligh mic oscopy image (le ) is om a con ol male (w
1118
), used o illus a ion pu pose only. Spe ma ogenesis akes
place inside cys s, s a ing a he ip o he es es (uppe panels - no e he concen a ion o nuclei ha ep esen he i s s ages o his p ocess) and
mo ing owa ds he p oximal end o he o gan as cell di e en ia ion p oceeds (middle panels - no e he p esence o a cys con aining onion-s age
spe ma ids wi h hei single mi ochond ial mass, called he Nebenke n). A he p oximal end o he es is, ma u e spe ma ozoid cells a e eleased om
he cys and s o ed in he SV (lowe panels - no e he needle-shaped nuclei ha is ypical o hese cells). No e also he low-le el AOX exp ession in he
soma ic cells o he cys s, bu no in he ge mline. See Addi ional ile 1: Fig. S10 o compa able images om 3X ub-AOX males
Saa i e al. BMC De elopmen al Biology (2017) 17:9 Page 6 o 11
soma ic cells o he cys s, and/o in he pigmen cells
o he es is shea h causes in e nal ea angemen s in
his o gan, dis up ing he p oduc ion and deli e y o
ma u e spe m.
To in es iga e he na u e o he whi ish ma e ial ound
a he p oximal end o he es is o AOX-exp essing
males, we used a ansgenic cons uc ha exp esses
GFP- agged Don Juan p o ein (DJ-GFP) in he ail o
elonga ed spe ma ids and ma u e spe ma ozoids [35,
36], enabling us o e alua e he ansi ion be ween hese
wo s ages. The exp ession o DJ-GFP e ealed a high
concen a ion o wha appea ed o be indi idualized ma-
u e spe ma ozoids in he p oximal end o he es is o
AOX-exp essing males (Fig. 4, lowe panels). In con ol
males, ew o hese indi idualized GFP-posi i e cells
we e ound in his egion o he es es (Fig. 4, uppe
panels), as expec ed, gi en ha hey should mo e o he
SVs a e indi idualiza ion. The ailu e o he indi idual-
ized, elonga ed spe ma ids o mo e in o he SV and he
associa ed dis u bance in he spa ial dis ibu ion o ICs
could accoun o he accumula ion o he whi ish ma-
e ial seen by ligh mic oscopy, and is he simples ex-
plana ion o he de ec seen in spe m-compe i ion
assays, al hough he eason why hey a e e ained in he
es is is no ob ious mo phologically.
Discussion
In p e ious s udies [6, 7] we epo ed ha cons i u i e
exp ession o C. in es inalis AOX in D osophila did no
a ec iabili y o e ili y unde non-compe i i e lab
condi ions. He e we demons a e ha ansgenic AOX
exp ession in D osophila unde he α- ubulin p omo e
does cause a clea de imen o male ep oduc i e suc-
cess, using a s ingen spe m-compe i ion assay. This
was accompanied by a dec eased and diso ganized
p oduc ion o ma u e spe ma ozoids. Al hough some
spe m cells appea o comple e he p ocess o mo -
phological di e en ia ion, ollowing indi idualiza ion
hey appea o emain lodged in he p oximal egion
o he es is, in he ou e shea h o which AOX is
highly exp essed.
AOX, when exp essed di ec ly unde he α- ubulin
p omo e o ia he ubGS d i e , is p esen a high
abundance in he pigmen cells o he es is shea h and
SV, and is also ound a e y low le els in some soma ic
cys cells. Howe e , i is below he le el o de ec ion in
he ge mline. The α- ubulin p omo e has been epo ed
o d i e ge mline exp ession o ansgenes up o he la e
spe ma ocy e phase [29]. Clea ly, his does no apply o
he cons uc we used [7]. Su p isingly we also obse ed
no abno mali ies du ing spe ma ogenesis in he mi o-
chond ial ne wo k (Addi ional ile 1: Fig. S11), he
eo ganiza ion o which p o ides he s uc u al pla o m
o spe m axoneme o ma ion [37]. On he o he hand,
he ele a ed numbe and al e ed dis ibu ion o WBs in
he es es o AOX-exp essing males may e lec in-
c eased ecycling o dys unc ional ge m cells, whe e di -
e en ia ion has ailed. We posi ha he speci ic
dis u bance in spe ma ogenesis p oduced by AOX ex-
p ession is mos likely due o de anged signaling om
soma ic cells, a he han low-le el AOX exp ession in
he ge mline i sel .
The p ecu so cells o he es is and SV shea hs ha e
wo dis inc de elopmen al o igins: he pigmen cells
o igina e in he gonad imaginal disc, along wi h he es-
es, whe eas he muscle cells and he SVs o igina e in
he geni al disc [38, 39]. When he wo o gans s a us-
ing, mig a ion o pigmen and muscle cells p oceeds in
opposi e di ec ions wi h he gonad and he SVs each ac-
qui ing an ou e laye o pigmen cells and an inne
Fig. 4 Indi idualized spe m cells emain lodged in he es es o AOX-exp essing males. Fluo escence imaging o es es o 10-day old males o geno-
ype X/Y; UAS-AOX
F6
/2; ubGS/DJ-GFP, unde uninduced (uppe panels) and mi ep is one-induced (lowe panels) condi ions, was pe o med using
phalloidin-TRITC o de ec ac in and he DJ-GFP luo escence o de ec he ails o elonga ed spe ma ids and ma u e spe ma ozoids. No e ha in he
p oximal egion o he es is ( luo escen con ocal images, igh , and ed boxes on he whi e-ligh mic oscopy images, le ) o he uninduced con ol
sample (AOX-nonexp esso ) only he bundles o no -ye -indi idualized elonga ed spe ma ids a e obse ed, whe eas in he same egion in AOX-
exp essing males hese cells appea indi idualized, angled and/o diso ganized, accoun ing o he accumula ed whi ish ma e ial shown in Fig. 2
Saa i e al. BMC De elopmen al Biology (2017) 17:9 Page 7 o 11
laye o muscle [30]. The ailu e o p oduce pigmen
cells du ing de elopmen is known o lead o abe an
muscle-cell mig a ion, causing abno mal es is mo ph-
ology and male s e ili y, wi hou a ec ing he ge mline
o soma ic cell- ypes ha suppo spe ma ogenesis [30].
This in e dependen mig a ion implies he exis ence o a
signaling ne wo k be ween he pigmen and he muscle
cells. Ou indings sugges ha communica ion be ween
hese cell- ypes migh con inue a e es is de elopmen is
comple e. We sugges ha AOX exp ession in he pigmen
cells migh in some way dis up hei abili y o deli e sig-
nals equi ed o he comple ion o spe ma ogenesis, and/
o o signal muscle con ac ion needed o guide indi idual-
ized ma u e spe ma ozoids om he p oximal end o he
es is in o he SV, leading o he accumula ion o cys s in
la e s ages o ma u a ion h oughou he es is. This is sup-
po ed by he ac ha a low pe cen age o es es o 3X ub-
AOX males (~5% o o al es es analyzed, Addi ional ile 1:
Fig. S15) p esen ed mal o ma ions e y simila o hose e-
po ed o males unable o de elop pigmen cells [30]. Al-
hough he majo i y o es es a e mo phologically no mal in
mos 3X ub-AOX males a eclosion, hey emain hick o e
he i s 10 days o adul li e (Fig. 2B), indica ing a unc-
ional de ec in spe m p oduc ion and deli e y ha would
be su icien o accoun o he dec eased ep oduc i e suc-
cess unde compe i i e condi ions (Fig. 1). To p o e ha
AOX exp ession in he shea h pigmen cells was he cause
o he obse ed spe m and ep oduc i e de ec s would e-
qui e a d i e ha enables i s exp ession speci ically in
hese cells. Since no such d i e is cu en ly a ailable, we
mus ely ins ead on he s ongly sugges i e co ela ions we
ha e documen ed he e.
The e ec o AOX on cell-cell communica ion could
esul om one o se e al me abolic dis u bances con-
e ed by he enzyme. I cons i u i ely ac i a ed by local
me abolic condi ions in speci ic cells, AOX could pa -
icipa e signi ican ly in espi a o y elec on low, impai -
ing ATP p oduc ion o o he downs eam p ocesses
dependen on espi a o y ene gy, such as he bu e ing
o calcium ( e iewed in [40]). A calcium signaling pa h-
way has been implica ed in he up egula ion o Wn -2
[41], which egula es di e en ia ion in many cellula
con ex s, including he es is pigmen cells du ing D os-
ophila de elopmen , p omo ing muscle cell mig a ion
and a achmen [30]. D osophila males mu an o Wn -
2 a e s e ile due o se e e es is mal o ma ions [30],
whe eas males exp essing AOX a e e ile and appea
mo phologically no mal a eclosion, wi h a unc ional
de ec only e ealed by a s ingen compe i ion assay
and by he appa en e en ion o spe m wi hin he es is
du ing he i s 10 days o adul li e.
An in e e ence wi h cell signaling may also in ol e
ROS. AOX has been epo ed o dec ease mi ochond ial
ROS p oduc ion e en unde condi ions o no mal
oxida i e phospho yla ion ac i i y [6, 42]. In ei he case,
he e y sub le pheno ype, wi h all o he physiological
and de elopmen al unc ions appa en ly una ec ed, p e-
sen s a challenge o unde s anding he mechanism.
Clea ly, much emains o be lea ned abou he physio-
logical e ec s o AOX exp ession. Ou indings indica e
he impo ance o s udying hese e ec s and hei mech-
anisms in de ail, be o e implemen ing AOX in gene he -
apy. Some o i s p ope ies may need o be modi ied,
be o e he enzyme can be sa ely deployed.
Conclusions
AOX o e s a po en ial he apeu ic s a egy o bu e ing
pa hological s esses in mi ochond ia. Whils i is, in-
deed, able o compensa e o a numbe o pa hological
insul s, i has ema kably li le, i any e ec on no mal
physiology o model o ganisms, including D osophila.
T ansgenic lies ubiqui ously exp essing AOX we e ea l-
ie ound o exhibi no mal de elopmen , ep oduc ion
and li espan.
In he p esen s udy we ook his analysis o a deepe
le el, by applying a mo e s ingen es o ep oduc i e
capaci y, he well-es ablished spe m-compe i ion assay.
Despi e being e ile, AOX-exp essing males showed a
speci ic de imen in his es , which co ela es wi h o al
ansgene exp ession dose. Mo phologically, we ound
his o be co ela ed wi h spa ially de anged spe ma o-
genesis and ailu e o elease ma u e spe m in o he
s o age o gan (seminal esicle) in no mal quan i ies, al-
hough he de ec is in e ed o a ise in he soma ic pig-
men cells o he es is, no he ge mline. This is he i s
demons a ion o a de elopmen al de ec caused by
AOX exp ession in a me azoan, and needs also o be
aken in o accoun when conside ing he apeu ic uses o
AOX in humans.
Me hods
Fly s ocks and main enance
S anda d lines w
1118
, CyO and TM3,Sb balance s, UAS-
S inge GFP,daGAL4, ubGS, and DJ-GFP we e ob ained
om s ock cen es. The lines UAS-AOX
F6
[6], ub-
AOX
35
, ub-AOX
112
, ub-AOX
7
,2X ub-AOX,3X ub-
AOX [7], UAS-AOXw
8.1
,UAS-AOXmu ,UAS-emp y
2nd
,
and UAS-emp y
3 d
[27] we e desc ibed p e iously. All ly
lines we e backc ossed in o he w
1118
backg ound o
six- o- en gene a ions, and we e main ained in s anda d
die a 25 °C [6]. RU486 (mi ep is one, Sigma, USA) was
added o he die a he indica ed concen a ions o in-
duc ion o ansgene exp ession using he ubGS line.
Spe m-compe i ion assays
Spe m-compe i ion assays we e conduc ed using wo ap-
p oaches. In he de ensi e pa adigm, app oxima ely 50
i gin w
1118
emales (whi e eyes), aged be ween 3 and
Saa i e al. BMC De elopmen al Biology (2017) 17:9 Page 8 o 11
7 days, we e ini ially allowed o ma e indi idually wi h
5–8 day-old ( i gin) ansgenic males ( ed eyes, w+) o
3 days. The males we e disca ded and he emales ans-
e ed o new ials and allowed o ma e indi idually wi h
5–8 day-old w
1118
males (whi e eyes), o a u he 3 days,
a e which he second males we e also disca ded. The
emales we e again ans e ed o new ials, allowed o
lay eggs o 3 days, hen ans e ed inally o u he
egg laying o e 5 days. In he o ensi e pa adigm, ap-
p oxima ely 50 i gin w
1118
emales (whi e eyes) we e
i s c ossed wi h w
1118
males (whi e eyes), and hen wi h
ansgenic males ( ed eyes, w+), ollowing he same ma -
ing and egg laying scheme o he de ensi e pa adigm.
Rep oduc i e success was measu ed by coun ing he
numbe o ed- and whi e-eyed p ogeny, which ep e-
sen , espec i ely, he p ogeny o ansgenic and w
1118
males. The small numbe o ials de i ed om any sin-
gle emale ha con ained p ogeny exclusi ely o one eye
colou we e conside ed o ep esen only a single ma ing
and we e excluded om he analysis. Da a we e plo ed
as mean numbe o p ogeny pe emale o each eye
colou , in he ou successi e ma ing/egg laying ial se s,
± s anda d de ia ion.
Dissec ion and imaging o male ep oduc i e o gans
Rep oduc i e o gans we e dissec ed by anes he izing
males o he indica ed age and geno ype wi h CO
2
p io o
ans e ing o a dissec ion boa d con aining phospha e-
bu e ed saline (PBS). Using hin dissec ion o ceps, he
in e nal o gans we e emo ed om he abdomen by dis-
loca ing he ex e nal geni alia, ollowed by manual isola-
ion o he ep oduc i e o gans, which we e imaged
immedia ely using he Nikon SMZ 745 T sys em unde
whi e ligh . The images we e analyzed and es is and SV
hickness was measu ed using NIS Elemen s D4.20 so -
wa e (Nikon Ins umen s So wa e, Ne he lands).
Fo immuno luo escence imaging, dissec ed o gans
we e ixed in 4% pa a o maldehyde o 20 min a oom
empe a u e, and hen ans e ed o PBS p io o
pe meabiliza ion. The o gans we e insed wice wi h
PBS con aining 0.1% T i on X-100 and 0.1% BSA, hen
pe meabilized and blocked wi h PBS con aining 1% T i-
on X-100 and 1% BSA o 30 min a oom empe a u e,
ollowed by o e nigh incuba ion a 4 °C wi h he p i-
ma y an ibody dilu ed in PBS con aining 0.3% T i on X-
100 and 0.5% BSA. The samples we e hen insed wice
and washed h ee imes (30 min each) wi h PBS con ain-
ing 0.1% T i on X-100 and 0.1% BSA a oom
empe a u e, ollowed by o e nigh incuba ion a 4 °C
wi h he seconda y an ibody dilu ed in PBS con aining
0.3% T i on X-100 and 0.5% BSA. Washes we e pe -
o med as desc ibed abo e, and samples we e insed
wi h PBS and hen Milli-Q wa e , p io o ans e ing o
op ically clea 35 mm glass bo om dishes (Ma Tek,
USA) con aining P oLong Gold An i ade Moun ing
medium wi h DAPI (Molecula P obes, Li e Technolo-
gies). P ima y an ibodies used we e abbi polyclonal
an i-AOX (1:10,000, [6]), mouse monoclonal an i-
ATP5A (1:1000, Abcam, UK), abbi polyclonal an i-
ac i a ed caspase-3 (1:200, Cell Signaling Technology,
USA), abbi polyclonal an i-Sex Pep ide (1:300, gi om
D . Shanjun Chen) and abbi polyclonal an i-emp y spi-
acles (1:400, An ibody Ve i y, USA). Fluo escen sec-
onda y an ibodies we e Alexa 488-conjuga ed goa an i-
abbi IgG (1:1000) and Alexa 568-conjuga ed ga an i-
mouse IgG (1:1000) (Fishe Scien i ic, USA). Ac in was
s ained using phalloidin-TRITC (1 μg/ml, Sigma, USA).
Samples we e imaged using an Ando Spinning Disc o
Zeiss LSM 780 lase -scanning con ocal mic oscope, e-
spec i ely wi h Ando iQ 3.0 o ZEN 2011 SP3 (black edi-
ion) so wa e. Images we e analyzed wi h ImageJ so wa e
(Na ional Ins i u es o Heal h, USA). No signals we e ob-
se ed when he p ima y an ibodies we e omi ed.
Immunoblo ing
P o ein ex ac s om whole adul males (20–30) and
om dissec ed es es and SVs o 50–100 males we e p e-
pa ed by gen le g inding o issues in 50–200 μl PBS
con aining 1.5% T i on X-100 and Comple e, Mini,
EDTA- ee P o ease Inhibi o Cock ail (Roche,
Swi ze land). The suspension was cen i uged a 16,000
g
max
o 10 min a 4 °C, and supe na an p o ein concen-
a ion was measu ed by he B ad o d me hod. C ude
mi ochond ial p epa a ions we e ob ained by homogen-
izing 50–100 adul males in ice-cold isola ion bu e
(250 mM suc ose, 5 mM T is, 2 mM EGTA, pH 7.4)
ollowed by cen i uga ion a 200 g
max
o 3 min a 4 °C.
The supe na an was ans e ed o a new ube and cen-
i uged a 9000 g
max
o 10 min a 4 °C. The pelle was
esuspended in 50 μl o isola ion bu e , and p o ein con-
cen a ions we e de e mined using he B ad o d me hod.
100 μg o o al p o ein ex ac s and 40 μgo mi ochon-
d ial p epa a ions we e mixed wi h 5× Laemmli bu e
(10% SDS, 50% glyce ol, 25% 2-me cap oe hanol, 0.02%
b omophenol blue and 0.3125 M T is-HCl, pH 6.8), dena-
u ed a 95 °C o 5 min and esol ed by SDS-PAGE on
4–20% o Any kD C i e ion P ecas Gels (Bio-Rad) a
120 V o app oxima ely 1.5 h. P o eins we e ans e ed
o ni ocellulose memb anes using he iBlo Gel T ans e
sys em (In i ogen, USA) o 7 min a oom empe a u e.
Memb anes we e blocked in TBST (0.15 M NaCl, 50 mM
T is-HCl, 0.05% Tween 20, pH 7.6) con aining 5% d ied
non a milk o 2 h a oom empe a u e o o e nigh a
4 °C. P ima y an ibodies ( abbi polyclonal an i-AOX
[1:10,000, [6]], mouse monoclonal an i-ATP5A [1:5000,
Abcam, UK], mouse monoclonal an i-PDH E1α[1:5000,
Abcam, UK], and mouse monoclonal an i-GAPDH
[1:5000, Abcam, UK]) we e incuba ed o a leas 1 h a
Saa i e al. BMC De elopmen al Biology (2017) 17:9 Page 9 o 11