LUP
Lund Uni e si y Publica ions
Ins i u ional Reposi o y o Lund Uni e si y
This is an au ho p oduced e sion o a pape
published in Neu obiology o Disease. This pape has
been pee - e iewed bu does no include he inal
publishe p oo -co ec ions o jou nal pagina ion.
Ci a ion o he published pape :
Miguel Bu guillos Ga cia, N Hajji, E Englund,
A Pe sson, Angela Cenci Nilsson, A Machado,
J Cano, B Joseph, J L Vene o
"Apop osis-inducing ac o media es dopamine gic
cell dea h in esponse o lps-induced in lamma o y
s imulus E idence in Pa kinson's disease pa ien s."
Neu obiology o Disease 2010 Sep 17
h p://dx.doi.o g/10.1016/j.nbd.2010.09.005
Access o he published e sion may equi e jou nal
subsc ip ion.
Published wi h pe mission om: Else ie
1
Regula a icle
APOPTOSIS-INDUCING FACTOR MEDIATES DOPAMINERGIC
CELL DEATH IN RESPONSE TO LPS-INDUCED INFLAMMATORY
STIMULUS. EVIDENCE IN PARKINSON’S DISEASE PATIENTS
M.A. Bu guillos1,2,5, N. Hajji2,3, E. Englund4 , A. Pe sson4, A. M. Cenci5, A. Machado1, J.
Cano1, B. Joseph2* and J.L. Vene o1*
1Depa amen o de Bioquímica y Biología Molecula . Facul ad de Fa macia. Uni e sidad de
Se illa, Spain
2 Depa men o Oncology-Pa hology, Cance Cen um Ka olinska (CCK), R8:03,
Ka olinska Ins i u e , SE-171 76 S ockholm, Sweden
3 P esen add ess: Depa men o Expe imen al Medicine and Toxicology, Di ision o
In es iga i e Science, Impe ial College London, Uni ed Kingdom.
4 Depa men o Pa hology, Di ision o Neu opa hology, Lund Uni e si y Hospi al, 221 85
Lund, Sweden.
5 P esen add ess: Basal Ganglia Pa hophysiology Uni , Depa men o Expe imen al
Medical Science. BMC F11, 22184 Lund, Sweden.
* BJ and JLV sha e co-senio ship o his pape .
Co esponding au ho :
D . Jose L. Vene o
Depa amen o de Bioquímica y Biología Molecula . Facul ad de Fa macia, Uni e sidad de
Se illa, C/ P o . Ga cía González, 2, 41012 Se illa, Spain
Phone: 0034 954 55 38 03
Fax: 0034 954 55 6752
E-mail: [email p o ec ed]
2
Abs ac
We show ha in anig al lipopolysaccha ide (LPS) injec ion, which p o okes speci ic
degene a ion o DA neu ons, induced caspase-3 ac i a ion in he a en al mesencephalon,
which was mos ly associa ed wi h glial cells. In con as , nig al DA neu ons exhibi ed AIF
nuclea ansloca ion in esponse o LPS. A signi ican dec ease o he Bcl-2/Bax a io in
nig al issue a e LPS injec ion was obse ed. We nex de eloped an in i o co-cul u e
sys em wi h he mic oglial BV2 and he DA neu onal MN9D mu ine cell lines. The
silencing o caspase-3 o AIF by small in e e ing RNAs exclusi ely in he DA MN9D cells
demons a ed he key ole o AIF in he LPS-induced dea h o DA cells. In i o chemical
inhibi ion o caspases, Poly(ADP- ibose)polyme ase-1, an ups eam egula o o AIF
elease and calpain, p o ed he cen al ole o he AIF-dependen pa hway in LPS-induced
nig al DA cell dea h. We also obse ed nuclea ansloca ion o AIF in he en al
mesencephalon o Pa kinson’s disease subjec s.
Keywo ds: apop osis-inducing ac o , caspase 3, apop osis, Pa kinson’s disease, subs an ia
nig a, dopamine.
3
In oduc ion
Pa kinson’s disease (PD) a ec s app oxima ely 1-3% o he popula ion, and is
cha ac e ised by a slow and p og essi e degene a ion o dopamine gic (DA) neu ons in he
subs an ia nig a (SN) (Obeso e al., 2000).
Recen e idence sugges s ha in lamma ion may play a cen al ole in he cell loss seen
in PD (Hi sch and Huno , 2009; Tansey e al., 2007). Epidemiological s udies ha e
demons a ed con incingly ha he incidence o idiopa hic PD is lowe in ch onic use s o an i-
in lamma o y d ugs (Chen e al., 2003, 2005; Esposi o e al., 2007).
We ha e p e iously demons a ed ha he in anig al injec ion o lipopolysaccha ide
(LPS) selec i ely induces he dea h o DA neu ons (Cas año e al., 1998). Gi en he po en ial
ole o in lamma ion in he pa hogenesis o PD, he elucida ion o he cellula /molecula
mechanisms leading o DA cell dea h in he LPS model is indispensable.
S udies in expe imen al models and PD pa ien s a e sugges i e o mul iple dea h
pa hways including in insic and ex insic apop osis and au ophagy (Le y e al., 2009; I che
and Pa k, 2009). Human pos mo em s udies sugges ha DA neu ons die by apop osis in PD
(Ha mann and Hi sch, 2001; Ta on e al., 2003). Fas ecep o (Mogi e al., 1996) and
au ophagic acuoles (Anglade e al., 1997) we e epo ed o be ele a ed wi hin he nig os ia al
sys em o PD pa ien s and in se e al expe imen al models o pa kinsonism (Fo nai e al.,
2004). Neu al cell dea h associa ed wi h se e al PD genes, including α-synuclein, pa kin,
PINK1, DJ-1, and LRRK2, in ol es caspase ac i a ion (Le y e al., 2009). Simila ly, he DA
neu o oxins MPTP, o enone, and 6OHDA p oduce caspase ac i a ion and mi ochond ial
elease o cy och ome c (Hol z e al., 2006; She e e al., 2002; Tu mel e al., 2001). Howe e ,
he abili y o caspase inhibi o s o con e p o ec ion has been inconsis en (Bilsland e al.,
2002; Chu e al., 2005; Ha mann e al., 2001; Liou e al.,2005; Lo ha ius e al., 1999; Yang e
4
al., 2004). This obse a ion sugges s ha al e na i e dea h pa hways, independen o caspase
ac i a ion, can be ac i a ed in esponse o inju y in nig os ia al DA neu ons.
Apop osis-inducing ac o (AIF) is an apop o ogenic 67-kDa p o ein ha esides in he
mi ochond ial in e memb ane space. A e an apop o ic insul , AIF ansloca es om he
mi ochond ia o he nucleus whe e he p o ein induces caspase-independen p og ammed cell
dea h (Susin e al., 1999). AIF elocaliza ion p ecedes cell dea h as indica ed by apop osis-
ela ed ch oma in condensa ion and la ge-scale DNA agmen a ion (Hong e al., 2004).
Inc easing e idence suppo s a p ominen ole o AIF in neu odegene a ion. The
nuclea ansloca ion o AIF appea s o play a key ole in cell dea h egula ion in auma ic
b ain inju y (Zhang e al., 2002), oxida i e s ess (Fon ia e al., 2002), ce eb al hypoxia-
ischemia (Zhu e al., 2003), epilepsy (Cheung e al., 2005) and Alzheime ’s disease (Reix e
al., 2007). Consequen ly, we ha e unde aken a de ailed s udy wi h he aim o cha ac e izing
he molecula mechanisms leading o he dea h o nig al DA neu ons in he LPS animal model
o PD along wi h examina ion o au opsy PD pa ien s. Ou in i o and in i o da a suppo s a
majo ole o AIF in inducing he dea h o DA neu ons.
Ma e ials and Me hods
Animals and lesions – Fo y- ou male Wis a a s (200-250g), housed in ou labo a o y,
we e used o hese s udies. The a s we e anes he ized wi h 400 mg/kg chlo al hyd a e and
posi ioned in a s e eo axic appa a us (Kop Ins umen s, Tuyunga, CA, USA) o con o m o
he b ain a las o Paxinos and Wa son (1986) and o p oceed wi h he in anig al injec ion o
2 µl o LPS (1µg/µl). The needle was lowe ed h ough a d ill hole 5.5 mm pos e io , 1.5
mm la e al and 8.3 mm en al o he b egma o he SN. Monas al blue (Sigma) dissol ed
in 0.9 % saline se ed as an ine ace o he injec ion si e. Fo expe imen s in ol ing in
i o inhibi o s, he animals we e di ided in i e expe imen al g oups (4 animals pe g oup

5
and pos lesion ime): (a) One g oup ecei ed 160 mg/kg o benzamide, a selec i e PARP-1
inhibi o (Sigma), dissol ed in saline 60 minu es be o e in anig al LPS injec ions and
u he injec ed wi h benzamide e e y 12 hou s in ape i oneally; (b) a second g oup was
injec ed in anig ally wi h 1.5 nmol Z-Val-Ala-Asp luo ome hyl ke one (Z-VAD- mk), a
pan caspase inhibi o oge he wi h LPS, dissol ed in saline con aining 0.5% DMSO; (c) a
hi d g oup was injec ed in anig ally wi h LPS oge he wi h 1.5 nmol Z-Asp(OMe)-
Glu(OMe)-Val-Asp(OMe)-FMK (DEVD- mk), a speci ic inhibi o o caspase-3/7 ac i i y,
dissol ed in saline con aining 0.5% DMSO; (d) a ou h g oup was injec ed in anig ally
wi h LPS ollowed by in a enous bolus injec ion o MDL-28170 (30 mg/kg; 0.5 ml/kg,
dissol ed in PEG 300 (9):e hanol (1)), (e) a i h g oup was injec ed wi h LPS in anig ally.
Since he di e en inhibi o s we e adminis e ed unde a ying condi ions, LPS-injec ed
animals we e subdi ided in o h ee di e en subg oups. In a i s subg oup (n=4), LPS was
dissol ed in saline, adminis e ed in anig ally ollowed by i.p. saline injec ions e e y 12 h.
This g oup was compa ed wi h benzamide-injec ed animals. In a second subg oup (n=4),
LPS was dissol ed in saline con aining 0.5% DMSO and adminis e ed in anig ally. This
g oup was compa ed wi h hose ecei ing caspase inhibi o s; i.e. Z-VAD- mk- and DEVD-
mk-injec ed animals. In a hi d subg oup (n=4), LPS was dissol ed in saline, adminis e ed
in anig ally ollowed by a single bolus injec ion o 0.5 ml/kg PEG 300 (9):e hanol (1). This
g oup was compa ed wi h MDL-28170-injec ed animals. The in eg i y o he nig al DA
sys em was always e alua ed 4 days a e he in anig al LPS injec ion. In addi ion,
benzamide ea men was also e alua ed 24h a e he LPS injec ion o disca d e ec s on he
mic oglial popula ion. The animals we e pe used wi h 4% o pa a o maldehyde unde deep
anes hesia and hei b ains we e emo ed and p epa ed o immunohis ological s udies. All
he expe imen s we e pe o med in acco dance wi h he guidelines o he Eu opean Union
6
Council (86/609/EU), ollowing he Spanish egula ions (BOE 67/8509-12,1988) o he use
o labo a o y animals and app o ed by he Scien i ic Commi ee o he Uni e si y o Se ille.
Cell cul u e and ea men s - Mu ine mic oglial BV2 cell line and mu ine DA MN9D cell line
we e used in his s udy (Joseph e al., 2003). BV2 cells and MN9D cells we e cul u ed and
main ained in DMEM media and DMEM/F12 media espec i ely, bo h supplemen ed wi h
10% FCS, 2 mM L-glu amine, penicillin (100 U/ml) and s ep omycin (100 mg/ml).
Expe imen s we e pe o med in educed 5% FCS media. Fo he co-cul u e expe imen s,
MN9D cells (80 × 104 cells/ml) we e pla ed i s . The ollowing day, MN9D cells we e
ans ec ed wi h lipo ec amine 2000 (In i ogen) ollowing manu ac u e ’s ins uc ions. A e 3
hou s o ans ec ion, cells we e s ained wi h he CellT acke ™ G een CMFDA o 30 min,
washed h ee imes wi h PBS and esh media added. Then BV2 (30 × 104 cells/ml) cells we e
seeded on hem and 1 µM e inoic acid was added o he co-cul u e o induce cell
di e en ia ion. Twen y ou hou s la e , LPS was added (1µg/ml) o ano he 24 hou s. All he
knockdown expe imen s we e pe o med on MN9D cells. Fo he ime-cou se s udy o caspase
3 silencing in MN9D cells, expe imen s we e s opped a 48 and 72 hou s. The siRNAs used
we e he ollowing: siRNA agains caspase-3 ON-TARGETplus SMARTpool (L-043042-00),
ON-TARGETplus SMARTpool Con ol siRNA Non- a ge ing siRNA (siRNA NT) #1 (D-
001810-01-05) and an oligonucleo ide speci ic o mouse AIF (sense an i-mouse AIF 5'-
AUGCAGAACUCCAAGCACGTT-3'.
Human b ain
Human b ain issues om pa ien s wi h PD o 5, 9 and 15 yea s’ du a ion, and age-ma ched
con ol cases (ca diac a es ic ims, no b ain disease) we e used in his s udy. The egion
in es iga ed was he an e io mesencephalon co e ing he subs an ia nig a. They we e
7
mic oscopically e iewed o e i ica ion o pa hology and in ac mo phology, espec i ely.
P io o he in es iga ion, he en i e collec ion o b ain sec ions, 15-20 pe case including he
mesencephalic sec ion, we e subjec ed o a neu opa hological whole b ain analysis o
clinical diagnos ic pu pose, acco ding o ou ine p ocedu es a he depa men o
Neu opa hology. The p ojec p ocedu es in ol ing human b ain issue we e app o ed by he
Regional E hical Re iew Boa d in Lund, Sweden.
Wes e n Blo Analysis – Sus an ia nig a and cell ex ac s we e p ocessed o immunoblo ing.
SDS-polyac ylamide gel elec opho esis was pe o med as desc ibed p e iously (Joseph e al.,
2002). Goa polyclonal an ibody di ec ed agains AIF (San a C uz Bio ech.), abbi polyclonal
p ocaspase-3 (Cell Signalling), abbi polyclonal clea ed PARP (Asp214), abbi polyclonal
Bcl-2 (San a C uz Bio ech.) and abbi polyclonal Bax (San a C uz Bio ech.) we e employed.
Mouse monoclonal β-ac in an ibody (Sigma) was used o e i y equal loading o he gel.
Seconda y ho se adish pe oxidase-conjuga ed an i- abbi , an i-goa and an i-mouse an ibodies
we e om Vec o labs.
Immunocy ochemis y - MN9D cells we e s ained wi h CellT acke ™ G een CMFDA
(In i ogen) ollowing manu ac u e ’s ecommenda ions o 30 min be o e co-cul u e wi h
BV2 cells. Cells we e g own on glass slides and ixed wi h 4% pa a o maldehyde du ing 15
minu es and hen washed once wi h PBS. Cell nuclei we e coun e s ained wi h Hoechs 33342
(1μg/ml) o DAPI 1μg/ml and apop o ic cell dea h was quan i ied as agmen ed and/o condensed
nuclei o s ained neu ons, compa ed o he o al numbe o neu ons and wi h con ol le els.
Immunohis ological s udies - Animals we e pe used h ough he hea unde deep
anaes hesia (chlo al hyd a e) wi h 100 ml o PBS con aining 10 U/ml hepa in ollowed by
8
150–200 ml o 4% pa a o maldehyde in phospha e bu e , pH 7.4. B ains we e emo ed and
hen imme sed in suc ose in PBS, pH 7.4, i s in 10% suc ose o 24 h, 20% suc ose o 24 h
and hen 30% suc ose un il sunk (2–5 days). Tissues we e hen ozen in isopen ane a −15°C
and 30 μm sec ions we e cu on a c yos a and moun ed in gela ine-coa ed slices. P ima y
an ibody used was a mouse-de i ed an i-TH (Chemicon, 1:1000), abbi -de i ed an i-AIF (Cell
Signaling, 1:50), abbi -de i ed an i-clea ed caspase-3 (Cell Signaling, 1:200), mouse-de i ed
an i-OX-6 (Se o ec, 1:200), mouse-de i ed an i-GFAP (Chemicon, 1:300) and abbi -de i ed
an i Iba1 (Wako, 1:500). All incuba ions and washes we e in PBS, pH 7.4, unless o he wise
no ed. All wo k was done a oom empe a u e. Sec ions we e hen incuba ed o 2 h wi h
bio inyla ed ho se an i-mouse o goa an i- abbi IgG (Vec o , 1:200) ollowed by a second 1-h
incuba ion wi h Ex A idin®-Pe oxidase solu ion (Sigma, 1:100) and o immuno luo escence
by a seconda y FITC-conjuga ed an i- abbi and Texas Red an i-mouse seconda y an ibody
(Vec o , 1:200). Fo colocaliza ion expe imen s, samples we e incuba ed i s ei he wi h OX6,
GFAP o TH an ibodies o e nigh , and hen u he incuba ed wi h clea ed caspase-3 o AIF
an ibodies. The di e en seconda y an ibodies we e hen added oge he .
Con ocal s udies - Fo ime lapse eco ding, he co-cul u e was g own in a glass chambe .
MN9D was ans ec ed wi h 1µg o pcDNA3.1 ec o exp essing enhanced g een luo escen
p o ein (EGFP) wi h AIF a i s C- e minus. Cell nuclei we e coun e s ained wi h 0.1 µg/ml
Hoeschs 33342 (In i ogen) and a Red- FLICATM ki o caspase-3 ac i a ion
(Immunochemis y echnologies) was also used. Mi ochond ia wi h an in ac ΔΨm we e
labelled wi h he po en ial-dependen dye TMRE (25 nM) (In i ogen). Chambe s we e placed
in he POC-Chambe / CTI Con olle / Hea ing inse P sys em o li e cell imaging. Time–
lapse was acqui ed a 15 minu es in e als o 8 h s a ing 6h a e LPS ea men . A minimum
o 100 cells we e analyzed (n=3).
15
(Fig 7). 24h a e LPS ea men he e was a obus dec ease in DA neu ons su i al (Fig 7D,
siRNA non- a ge ing) in ou co-cul u e sys em. Caspase-3 knockdown showed no p o ec ion in
undi e en ia ed DA cell su i al agains LPS ea men (Fig. 7D, siRNA caspase-3). We nex
pe o med a ime-cou se analysis o caspase 3 silencing in MN9D cells o demons a e ha
knock down o AIF in ou expe imen al condi ions was long-las ing. Ou analysis
demons a ed ha caspase 3 was e icien ly silenced a 48 and 72h upon ans ec ion (Fig. 7A).
In addi ion, caspase3 knockdown did no a ec MN9D neu onal cells di e en ia ion (Suppl.
Fig. 1).
AIF knockdown inc eased DA cell dea h in undi e en ia ed mixed cul u es in he
absence o LPS ea men (Fig. 7, siRNA AIF). This e ec suppo s he no ion ha AIF is
phylogene ically an old mi ochond ial NADH oxidase whose local edox unc ion is essen ial
o op imal oxida i e phospho yla ion and o an e icien an i-oxidan de ense (Modj ahedi e
al., 2006). In con as , AIF knockdown in di e en ia ed MN9D cells show no e ec in he
absence o he LPS ea men . Howe e and ema kably, di e en ia ed and undi e en ia ed
MN9D cells lacking AIF in co-cul u e wi h BV2 cells we e esis an o he LPS-induced cell
dea h (Figs. 7 and 8).
T ansloca ion o AIF in PD subjec s
Double immunolabeling o TH and AIF and subsequen con ocal imaging analysis we e
pe o med in he en al mesencephalon in pa a in sec ions om h ee PD pa ien s and h ee
age- and gende -ma ched heal hy con ol cases. AIF immunolabeling was widely dis ibu ed in
he en al mesencephalon o non-PD subjec s, wi h a p e e en ial dis ibu ion o e he nig al
DA neu ons (Fig. 9). Dis ibu ion o AIF wi hin en al mesencephalic neu ons was s ic ly
cy osolic in con ol subjec s. We ailed o de ec a single DA neu on exhibi ing nuclea AIF

16
immunolabeling in non-PD subjec s. In con as , scan y bu consis en nuclea localisa ion o
AIF wi hin degene a ing nig al DA neu ons could be obse ed in PD subjec s (Fig. 9).
Discussion
Ou da a demons a e ha AIF is a key egula o o in i o DA neu onal cell dea h in
esponse o LPS-induced in lamma o y challenge. Analysis o he en al mesencephalon o
PD pa ien s e ealed he occu ence o AIF ansloca ion in degene a ing DA neu ons. Taken
oge he , we p o ide compelling e idence ha AIF plays a signi ican ole in he complex
cascade o e en s leading o selec i e degene a ion o nig al DA neu ons.
The LPS-induced in lamma o y esponse p o oked a complex cascade o e en s
leading o bo h caspase-dependen and caspase-independen mechanisms in he SN. Ac i e
caspase-3 was p edominan ly associa ed wi h eac i e mic oglia soon a e he challenge. The
signi icance o caspase-3 ac i a ion in eac i e mic oglia seems o be associa ed wi h he ea ly
dea h o esiden mic oglia and in il a ed neu ophils in esponse o in anig al LPS injec ion
(Ji e al., 2007). On he o he hand, degene a ing nig al DA neu ons exhibi ed AIF nuclea
ansloca ion bu no caspase 3 ac i a ion in esponse o LPS ea men . AIF elease is
associa ed o inc easing mi ochond ial ou e memb ane pe meabiliza ion, which is egula ed by
he Bcl-2 p o eins (Chipuk and G een, 2008). Ou ime-cou se analysis o p o ein exp ession o
Bcl-2 and Bax in nig al homogena es a e LPS challenge demons a ed ha he a io Bcl-2 o
Bax was highly diminished wi hin he i s 4 days pos lesion, when dea h o DA neu ons akes
place. Since he balance o an i-apop o ic o p o-apop o ic Bcl-2 p o eins (i.e. Bcl-2 o Bax)
dic a es cellula a e by inc easing mi ochond ial ou e memb ane pe meabiliza ion and he
p o-apop o ic Bcl-2 membe Bax has been mechanis ically linked o AIF elease (C egan e al.,
2002), we p o ide compelling e idence ha he en al mesencephalon exhibi s biochemical
ea u es o ease AIF elease unde condi ions o in lamma ion.
17
Using a siRNA-media ed gene knockdown app oach, we con i med, using mic oglia
cells (BV2) and DA neu onal cells (MN9D), in a co-cul u e se up, ha AIF is speci ically
equi ed o DA neu ons o unde go cell dea h when he co-cul u e is exposed o LPS. In he
same in i o sys em, he selec i e silencing o caspase-3 in MN9D cells was unable o p e en
he LPS-induced dea h o DA neu ons. Since we could no exclude he possibili y ha caspase
3 silencing was sho -li ed due o compensa ing mechanisms in esponse o he LPS-induced
caspase 3 ac i a ion, we nex pe o med a ime-cou se analysis o caspase 3 silencing in
MN9D cells. Ou analysis demons a ed he e ec i eness and los -las ing silencing o caspase
3 o a leas 72h upon ans ec ion, hus excluding he possibili y ha caspase 3 would
eappea du ing he expe imen al ime window. By doing ha , we conclude ha caspase 3 is
no in ol ed in he LPS-induced dea h o dopamine gic MN9D cells in co-cul u e wi h
mic oglial BV2 cells.
Undi e en ia ed MN9D cells exp ess y osine hyd oxylase, elease and anspo DA,
and exp ess ol age-ac i a ed sodium channels (Choi e al., 1991). Despi e hese dopamine gic
cha ac e is ics, in an undi e en ia ed o m, hese cells ha e a nonneu onal mo phology wi h
a ound cell body gene ally lacking any p ocess ex ension. Di e en ia ion o MN9D cells has
been shown o al e cell mo phology, g ow h, gene exp ession and o induce cell cycle a es in
a way ha mo e closely esemble ma u e SN pa s compac a neu ons. (Cas o e al., 2001;
He manson e al., 2003). Since he use ulness o MN9D cells in s udies o PD may well depend
on how closely hey esemble ma u e dopamine gic nig al neu ons, we decided o es he
e ec o AIF silencing on he LPS-induced dea h o di e en ia ed MN9D cells in co-cul u e
wi h mic oglial BV2 cells. Unde hese condi ions, AIF silencing p e en ed he dea h o
dopamine gic MN9D cells in esponse o LPS, hus con i ming ou iew ha AIF plays a
c i ical ole in he in eg i y o dopamine gic cells in esponse o LPS-induced in lamma o y
challenge.
18
In a nex s ep and in o de o demons a e conclusi ely ha nig al DA neu ons die by a
caspase-independen AIF-dependen mechanism, we pe o med in i o chemical inhibi ion
s udies using DEVD- mk, an inhibi o o caspase-3-like ac i i ies, Z-VAD- mk, a pan caspase
inhibi o , benzamide, an inhibi o o PARP-1 ac i i y and MDL-28170, a calpain inhibi o .
The usage o he b oad ange caspase inhibi o Z-VAD- mk p o ec ed nig al DA neu ons
agains LPS-induced neu odegene a ion. Howe e , he inhibi ion o caspase-3/7-like ac i i ies
ailed o p o ec nig al DA neu ons unde iden ical expe imen al condi ions. Inhibi ion o ei he
PARP-1 o calpain ac i i ies by benzamide and MDL-28170 con e ed a nea -comple e
p o ec ion o he nig al DA sys em. Conside ing ha Z-VAD- mk is known o inhibi calpain
ac i i y besides caspase ac i i ies (Wood and Newcomb, 1999), hese in i o indings highly
sugges ha AIF is a key cell dea h media o in DA neu ons in esponse o a p o-in lamma o y
s imulus.
AIF knockdown pe se inc eased he dea h o MN9D DA cells in co-cul u e wi h BV2
mic oglial cells in undi e en ia ed bu no in di e en ia ed MN9D cells in he absence o LPS
ea men . Howe e , LPS was ine ec i e in inducing DA cell dea h in ou co-cul u e se up in
bo h o ms, hus con i ming ou in i o da a. Consequen ly, AIF showed a dual ole in ou in
i o cul u e model: an i-apop o ic in he absence o LPS and p o-apop o ic in esponse o LPS
ea men . Suppo ing his iew, AIF has been shown o possess a dual unc ion in he con ol
o s ess-induced cell dea h (Modj ahedi e al., 2006). AIF is a phylogene ically old
mi ochond ial NADH oxidase whose local edox unc ion is essen ial o op imal oxida i e
phospho yla ion and o an e icien an i-oxidan de ence. I should be no ed he ex emely
high sensi i i y o DA cells o espi a o y chain complex I de iciency. In ac , MPTP and
o enone, wo neu o oxins widely used o de elop animal models o PD, a e selec i e complex
I inhibi o s (Ayala e al., 2007). In ag eemen wi h he abo e, he silencing o AIF induces a
educ ion in complex I ac i i y in a ious cell lines along wi h inc eased ROS p oduc ion
19
(Apos olo a e al., 2006), which in u n could inc ease he suscep ibili y o cells o apop o ic
dea h igge . These e ec s gi e ise o a plausible explana ion o he inc eased dea h a e o
undi e en ia ed DA cells ans ec ed wi h a siRNA a ge ing AIF, obse ed in he co-cul u es.
As al eady men ioned, di e en ia ion o MN9D cells has been shown o al e cell mo phology,
g ow h, gene exp ession and o induce cell cycle a es , which may unde lie in he obse ed
di e ences be ween di e en ia ed and undi e en ia ed MN9D cells in esponse o AIF
silencing in absence o LPS. An al e na i e explana ion may be ela ed o he an ioxidan
capaci y o e inoic acid, which was used o inducing di e en ia ion o MN9D cells.
I is no ewo hy ha only a ew s udies ha e shown apop o ic-like ea u es in he
pa kinsonian SN (Anglade e al., 1997; Mochizuki e al., 1996). O he s udies, in con as , ha e
ques ioned he p esence o apop o ic cells in he pa kinsonian SN (D agunow e al., 1995;
Kosel e al., 1997; Bana i e al., 1998; Wullne e al., 1999; Jellinge , 2000). To ou
knowledge, ou s udy is he i s o show AIF ansloca ion in he en al mesencephalon o PD
pa ien s. I nig al DA cell dea h does occu by an apop o ic mechanism in PD, de ec ion o
hese neu ons becomes a di icul ask, gi en he ela i ely slow a e o SN DA cell loss and
he apid clea ance o apop o ic cells (Le y e al., 2009). I has been es ima ed ha , a a gi en
ime poin (1 day), no mo e han ~ 0.1–0.2% dying cells should be p esen in he SN o a PD
pa ien (Wullne e al., 1999). E en wi h hese limi a ions, we ound low bu signi ican nig al
DA neu ons exhibi ing AIF ansloca ion o he nucleus, hus suppo ing an impo an ole o
AIF in he dea h o nig al DA neu ons in PD.
The i s s ong e idence suppo ing caspase-independen cell dea h in nig al DA
neu ons came om a s udy by C ocke e al. (2003). They demons a ed ha calpain inhibi ion
p o ec s nig al DA neu ons agains MPTP-induced cell dea h in mice. The second majo
e idence came om s udies de eloped in mice lacking he PARP-1 gene, which we e highly
20
esis an o MPTP oxici y in e ms o nig al DA neu on su i al (Mandi e al., 1999). Bo h
mechanisms ha e been u he associa ed wi h AIF elease; i.e. PARP-1 and calpain ac i a ion
(Yu e al., 2006). In ac , he e is a sequen ial ac i a ion o PARP-1, calpain and AIF elease,
hus sugges ing a ela ionship be ween bo h mechanisms (Mouba ak e al., 2007). PARP-1 is a
DNA damage su eillance molecule and a media o o cell dea h a e in lamma o y p ocesses
since i media es AIF elease om mi ochond ia o cy osol (Yu e al., 2002). Howe e , i
should be s essed ha PARP-1 ac i a ion is clea ly linked o in lamma ion. Di e en ac o s
ha e been shown o con ibu e o he LPS-induced dea h o nig al DA neu ons including
in e leukin-1 (IL-1 ), umo nec osis ac o (TNF- ) and ac i a ion o iNOS and NADPH
oxydase ac i i ies (A imo o and Bing, 2003, Gayle e al., 2002; McCoy e al., 2006; Qin e al.,
2004). Suppo ing his iew, i) addi ion o neu alizing an ibodies o IL-1 o TNF- o
p ima y mesencephalic cul u es p o ec ed dopamine gic neu ons o LPS (Gayle e al., 2002);
ii) selec i e in i o inhibi ion o TNF- wi h XENP345 was pa ially e ec i e in p o ec ing
nig al DA neu ons agains in anig al LPS injec ion (McCoy e al., 2006); iii) The same was
ue wi h in i o adminis a ion o L-N(G)-ni oa ginine, a selec i e inhibi o o NOS (A imo o
and Bing, 2003). In keeping wi h his iew, we ha e demons a ed a co ela ion be ween LPS-
induced o ma ion o pe oxyni i es, a highly eac i e ee adical de i ed om NO and O2−
(To eilles e al., 1999) wi h nig al DA cell dea h (Tomas-Cama diel, 2004). Since one o he
bes -accep ed ways o ac i a ing PARP-1 is h ough pe oxyni i e-induced DNA damage
(Szabo and Dawson, 1998), we nex es ed he possibili y ha PARP-1 ac i a ion and
subsequen mi ochond ial AIF elease could be in ol ed in he dea h o nig al DA neu ons. In
i o PARP-1 inhibi ion using benzamide p o ec ed nig al DA neu ons agains LPS-induced
neu o oxici y comple ely, hus suppo ing he conclusion ha he AIF-media ed caspase-
independen pa hway plays a key ole in he dea h o adul DA neu ons in esponse o
in lamma ion. I has been ecen ly epo ed ha PARP-1 ac i a ion induces mi ochond ial

21
Ca2+ dys egula ion, which in u n media es calpain ac i a ion and subsequen AIF unca ion
and ansloca ion o he nucleus (Vosle e al., 2009). Ou da a a gues in a ou o his
ela ionship unde neu o oxic condi ions associa ed o in lamma ion.
In conclusion, we p o ide s ong e idence suppo ing a dea h ole o AIF in
expe imen al and eal PD. Elucida ing he molecula mechanisms in ol ed in he dea h o DA
neu ons in SN is a p io i y in o de o ind u u e pha macological in e en ions aimed a
slowing down he na u al cou se o he disease.
Acknowledgemen s: MN9D cells we e ob ained om D . A. Helle (Uni e si y o
Chicago, Chicago, IL, USA). This wo k was suppo ed by g an s om he Spanish
Minis e io de Ciencia y Tecnología SAF 2006-04119 and SAF 2009-13778 (JLV), he
Swedish Resea ch Council and he Swedish Cance Socie y (BJ). M.A.B. was a FPI ellow
o Spanish Minis e io de Educación, Cul u a y Depo e du ing he cou se o his s udy. Ou
g a i ude o P o . Alan Boobis o c i ically eading his manusc ip . Also we would like o
hank he ines imable help o D Modes o Ca ballo, D Albe Ga cía, D . J.L. Ribas and A.
Fe nández (Cen o de In es igación, Tecnología e Inno ación –CITIUS-) o hei ad ices
and echnical suppo .
22
Re e ences
Anglade, P., Vyas, S., Ja oy-Agid, F., He e o, M.T., Michel, P.P., Ma quez, J., Moua -
P igen , A., Rube g, M., Hi sch, E.C., Agid, Y., 1997. Apop osis and au ophagy in nig al
neu ons o pa ien s wi h Pa kinson's disease. His ol. His opa hol. 12, 25-31.
Apos olo a, N., Ce e a, A.M., Vic o , V.M., Cadenas, S., Sanjuan-Pla, A., Al a ez-
Ba ien os, A., Esplugues, J.V., McC ea h, K.J., 2006. Loss o apop osis-inducing ac o
leads o an inc ease in eac i e oxygen species, and an impai men o espi a ion ha can be
e e sed by an ioxidan s. Cell Dea h Di e . 13, 354-357.
A imo o, T., Bing, G., 2003. Up- egula ion o inducible ni ic oxide syn hase in he
subs an ia nig a by lipopolysaccha ide causes mic oglial ac i a ion and neu odegene a ion.
Neu obiol. Dis. 12, 35–45.
Ayala, A., Vene o, J.L., Cano, J., Machado, A., 2007. Mi ochond ial oxins and
neu odegene a i e diseases. F on . Biosci. 12, 986-1007.
Bana i, R.B., Daniel, S.E., Blun , S.B., 1998. Glial pa hology bu absence o apop o ic nig al
neu ons in long-s anding Pa kinson's disease. Mo . Diso d. 13, 221-227.
Bilsland, J., Roy, S., Xan houdakis, S., Nicholson, D.W., Han, Y., G imm, E., He i, F.,
Ha pe , S.J., 2002. Caspase inhibi o s a enua e 1-me hyl-4-phenylpy idinium oxici y in
p ima y cul u es o mesencephalic dopamine gic neu ons. J. Neu osci. 22, 2637-2649.
Cas ano, A., He e a, A.J., Cano, J., Machado, A., 1998. Lipopolysaccha ide in anig al
injec ion induces in lamma o y eac ion and damage in nig os ia al dopamine gic sys em. J.
Neu ochem. 70, 1584-1592.
Cas o, D.S., He manson, E, Joseph, B., Wallen, A., Aa nisalo, P., Helle , A., Pe lmann, T.,
2001. Induc ion o cell cycle a es and mo phological di e en ia ion by Nu 1 and e inoids
in dopamine MN9D cells. J. Biol. Chem. 276: 43277–43284.
23
Chen, H., Jacobs, E., Schwa zschild, M.A., McCullough, M.L., Calle, E.E., Thun, M.J.,
Asche io, A., 2005. Nons e oidal an iin lamma o y d ug use and he isk o Pa kinson's
disease. Ann. Neu ol. 58, 963-967.
Cheung, E.C., Melanson-D apeau, L., C egan, S.P., Vande lui , J.L., Fe guson, K.L.,
McIn osh, W.C., Pa k, D.S., Benne , S.A., Slack, R.S., 2005. Apop osis-inducing ac o is a
key ac o in neu onal cell dea h p opaga ed by BAX-dependen and BAX-independen
mechanisms. J. Neu osci. 25, 1324-1334.
Chipuk, J.E., G een, D.R., 2008. How do BCL-2 p o eins induce mi ochond ial ou e
memb ane pe meabiliza ion?. T ends Cell Biol. 18, 157-164.
Choi, H.K., Won, L.A., Kon u , P.J., Hammond, D.N., Fox, A.P., Waine , B.H., Ho mann,
P.C., Helle , A., 1991. Immo aliza ion o emb yonic mesencephalic dopamine gic neu ons
by soma ic cell usion. B ain. Res. 552, 67–76.
Chu, C.T., Zhu, J.H., Cao, G., Signo e, A., Wang, S., Chen, J., 2005. Apop osis inducing
ac o media es caspase-independen 1-me hyl-4-phenylpy idinium oxici y in dopamine gic
cells. J. Neu ochem. 94, 1685-1695.
C egan, S.P., Fo in, A., MacLau in, J.G., Callaghan, S.M., Cecconi, F., Yu, S.W., Dawson,
T.M., Dawson, V.L., Pa k, D.S., K oeme , G., Slack, R.S., 2002. Apop osis-inducing ac o
is in ol ed in he egula ion o caspase-independen neu onal cell dea h. J. Cell Biol. 158,
507-517.
C ocke , S.J., Smi h, P.D., Jackson-Lewis, V., Lamba, W.R., Hayley, S.P., G imm, E.,
Callaghan, S.M., Slack, R.S., Melloni, E., P zedbo ski, S., Robe son, G.S., Anisman, H.,
Me ali, Z., Pa k, D.S., 2003. Inhibi ion o calpains p e en s neu onal and beha io al de ici s
in an MPTP mouse model o Pa kinson's disease. J. Neu osci. 23, 4081-4091.
D agunow, M., Faull, R.L., Lawlo , P., Beilha z, E.J., Single on, K., Walke , E.B., Mee, E.,
1995. In si u e idence o DNA agmen a ion in Hun ing on's disease s ia um and
Alzheime 's disease empo al lobes. Neu o epo . 6, 1053-1057.
24
Esposi o, E., Di Ma eo, V., Benigno, A., Pie ucci, M., C escimanno, G., Di Gio anni, G.,
2007. Non-s e oidal an i-in lamma o y d ugs in Pa kinson's disease. Exp. Neu ol. 205, 295-
312.
Fon ia, E., Da e, E., Benelli, M., Sunol, C., Cecca elli, S., 2002. T ansloca ion o apop osis-
inducing ac o in ce ebella g anule cells exposed o neu o oxic agen s inducing oxida i e
s ess. Eu . J. Neu osci. 16, 2013-2016.
Fo nai, F., Lenzi, P., Gesi, M., Soldani, P., Fe ucci, M., Lazze i, G., Capobianco, L.,
Ba aglia, G., De Blasi, A., Nicole i, F., Papa elli, A., 2004. Me hamphe amine p oduces
neu onal inclusions in he nig os ia al sys em and in PC12 cells. J. Neu ochem. 88, 114-
123.
Gayle, D.A., Ling, Z.D., Tong, C.W., Lande s, T., Lip on, J.W., Ca ey, P.M., 2002.
Lipopolysaccha ide (LPS)-induced dopamine cell loss in cul u e: ole o umo nec osis
ac o -α, in e leukin-1β, and ni ic oxide. De . B ain Res. 133, 27-35.
Ha mann, A., Hi sch, E.C., 2001. Pa kinson's disease. The apop osis hypo hesis e isi ed.
Ad . Neu ol. 86, 143-153.
Ha mann, A., T oadec, J.D., Huno , S., Kikly, K., Faucheux, B.A., Moua -P igen , A.,
Rube g, M., Agid, Y., Hi sch, E.C., 2001. Caspase-8 is an e ec o in apop o ic dea h o
dopamine gic neu ons in Pa kinson's disease, bu pa hway inhibi ion esul s in neu onal
nec osis. J. Neu osci. 21, 2247-2255.
He manson, E., Joseph, B., Cas o, D., Lindq is , E., Aa nisalo, P., Wallen, A., Benoi , G.,
Henge e , B., Olson, L., Pe lmann, T., 2003. Nu 1 egula es dopamine syn hesis and
s o age in MN9D dopamine cells. Exp. Cell. Res. 288, 324–334.
He e a, A.J., Cas ano, A., Vene o, J.L., Cano, J., Machado, A., 2000. The single in anig al
injec ion o LPS as a new model o s udying he selec i e e ec s o in lamma o y eac ions
on dopamine gic sys em. Neu obiol. Dis. 7, 429-447.
Hi sch, E.C., Huno , S., 2009. Neu oin lamma ion in Pa kinson's disease: a a ge o
neu op o ec ion?. Lance Neu ol. 8, 382-397.
31
se e al mic oglial cells a 45 min (A) and memb ane blebbing a 60 min (A). No e ha ed
TMRE s aining in g een-labelled neu onal cells u ned in o o ange (see a ow in panel A). In
panel B, MN9D DA cells we e ans ec ed wi h ec o encoding GFP- agged AIF and
seeded oge he wi h BV2 cells. Cell nuclei we e s ained wi h 0.1 µg/ml Hoechs 33342 and
an aliquo o Red- FLICATM ki o caspase-3 ac i a ion was added o he media. Upon LPS
exposu e, whe eas he elease o AIF om mi ochond ia was obse ed in he dying neu ons,
no caspase-3 like ac i i ies could be de ec ed (see empo al cou se in panel B). No e how
wo DA cells exp essing AIF (a and b) a e g adually loosing g een labelling, an indica ion o
mi ochond ial AIF elease o s a exhibi ing ea u es o nuclea condensa ion and cell dea h.
Cell labelled as (c) in panel B is o mic oglial o igin which possesses ac i e caspase 3/7
ac i i y as labelled in ed. In panel C, MN9D DA neu onal cells we e co ans ec ed wi h an
exp ession ec o encoding EGFP agged AIF (g een) and he mi ochond ial ma ke DsRed
Mi o ( ed). MN9D cells we e hen co-cul u ed wi h mic oglia BV2 cells and he co-cul u e
ea ed wi h LPS. Rep esen a i e con ocal images o GFP-AIF, mi ochond ial DsRed
s aining and nuclea Hoechs 33342 s aining in un ea ed (C, le panel) and LPS ea ed co-
cul u e (C, igh panel) a e shown and analyzed he localiza ion o he h ee channels using
he con ocal so wa e. No ice how in he le panel, g een and ed channel ( o AIF and
mi hocond ia espec i ely) appea oge he , while in he igh panel he g een channel
mos ly coincides wi h he blue channel (nucleus) and ba ely wi h he ed channel
(mi ochond ia).
Figu e 7: Knock-down o AIF bu no o caspase 3 in undi e en ia ed MN9D DA cells
p o ec s hem in a co-cul u e se up wi h mic oglia BV2 cells in esponse o LPS. Panel A
shows a ime-cou se expe imen s o knock-down o caspase 3 a 48 and 72 hou in
dopamine gic MN9D cells. Immunoblo s o caspase-3 and AIF demons a e he e icien
supp ession o hei exp ession upon knockdown wi h siRNA (A and B). Rep esen a i e

32
pic u es o he MN9D-BV2 cells co-cul u e, whe e MN9D cells we e s ained wi h he
CellT acke ™ G een CMFDA showing condensed nuclei a e LPS ea men (C). Panel D
shows he quan i ica ion o he condensed nuclei in CellT acke ™ G een exp essing DA
neu ons in un ea ed and LPS ea ed co-cul u es upon silencing in MN9D cells o caspase-3
o AIF. Non- a ge ing siRNA (siRNA NT) was used as con ol and β-ac in used as a
p o ein loading con ol. S a is ical signi icance: *: p<0.01.
Figu e 8: Knock-down o AIF educed apop o ic cell dea h o di e en ia ed MN9D cells in
co-cul u e wi h BV2 cells in esponse o LPS. MN9D apop o ic cell dea h was assessed by
immunos aining (A). siRNAs we e di ec ed agains AIF o agains Non-Ta ge ing gene
(siRNA NT) in MN9D cells. Cells we e s ained wi h he CellT acke ™ G een CMFDA
be o e co-cul u e wi h BV2 cells, di e en ia ed wi h 1 µM e inoic acid and exposed o LPS
o no . A e ixing he cells, AIF was s ained ed wi h Alexa Fluo 594 and nuclei we e
s ained wi h 1 µg/ml DAPI. Whi e a ows indica e MN9D knock-ou AIF cells, “G” le e
indica es mic oglia cells (BV2) and “A” le e indica es apop o ic MN9D cells (A). Black
ba s ep esen siRNA agains AIF and g ay ba s ep esen siRNA NT. Bo h ba s indica e he
pe cen age o MN9D apop o ic cell dea h. Ba s ep esen he mean ± s anda d o h ee
indi idual expe imen s. *p < 0.01 compa ed o siRNA NT.
Figu e 9: Nuclea AIF ansloca ion in subs an ia nig a dopamine gic neu ons om PD
subjec s. High magni ica ion pho og aphs o TH (g een) and AIF ( ed) in he en al
mesencephalon in in pa a in sec ions om age- and gende -ma ched heal hy con ol cases
(A-C) and wo PD pa ien s (D-I). Me ge pic u es o e e y case is depic ed ( igh column).
Highe magni ica ion pho og aphs a e p o ided o he PD cases o ease isualiza ion o AIF
ansloca ion. Scale ba : (A-C): 80 µm; (D-I): 40 µms.
33
Supplemen a y Figu e 1: Di e en ia ion o MN9D cells is main ained a 48h and 72h a e
caspase3 knockdown. siRNAs we e di ec ed agains caspase3 (siRNA C3) o no (non-
a ge ing siRNA; siRNA NT) o 24h. Then, MN9D cells we e di e en ia ed wi h 1 µM
e inoic acid o 24 h. Pho og aphs we e aken a 48h and 72h a e siRNA ans ec ion.
C
35 kDa
19 kDa
P ocaspase 3
Clea ed caspase 3
24h 48h 96h72hSham STS
42 kDa
89 kDa
Ac in
Clea ed PARP
1.0
0.8
0.6
0.4
0.2
0
1.2
* **
**
*
Sham 24h 48h 72h 96h
Fold o con ol
Bcl-2
Bax
Ra io Bcl-2/Bax
**
**
*
Ac in 42 kDa
Bax
Bcl-2
23 kDa
26 kDa
24h 48h 96h72hSham
A
B
Figu e 1. Bu guillos e al
A B C
D E F
G H I
Figu e 2. Bu guillos e al
Clea ed caspase 3
Clea ed caspase 3
Clea ed caspase 3
TH
GFAP
OX-6
Me ged
Me ged
Me ged
Hoechs AIF Hoechs + AIF TH TH + AIF
Con ol
LPS
LPS
A B C D E
F G H I J
K L M N O
Figu e 3. Bu guillos e al
Clea ed caspase 3 Hoechs + Clea ed caspase 3 TH + Clea ed caspase 3
P Q R S T
U V W X Y
Con ol
LPS

F
B
C D
E
A G H
I J
K L
Sham
LPS
LPS+DEVD mk
Sham
LPS
LPS+DEVD mk
LPS+zVAD mk LPS+zVAD mk
LPS+Benzamide LPS+Benzamide
LPS+MDL28170 LPS+MDL28170
Figu e 4. Bu guillos e al
Figu e 5. Bu guillos e al
A B
LPS LPS+Benzamide
Reac i e mic oglia densi y
(cells/mm2)
LPS
Benzamide
+ +
+-
Fig. 1 Bu guillos e al.
C
35 kDa
19 kDa
P ocaspase 3
Clea ed caspase 3
24h 48h 96h72hSham STS
42 kDa
89 kDa
Ac in
Clea ed PARP
1.0
0.8
0.6
0.4
0.2
0
1.2
* *
*
*
*
*
Sham 24h 48h 72h 96h
F l o c
od on ol
Bcl-2
Bax
Ra io Bcl-2/Bax
**
*
*
*
Ac in 42 kDa
Bax
Bcl-2
23 kDa
26 kDa
24h 48h 96h72hSham
A
B
12000
0
10000
8000
4000
6000
2000
LPS
Benzamide
DEVD- mk
num o be
p nTH- osi i e eu ons
M
zVAD- mk
MDL 28170
-+ + + + +
-
-
-
-
- - - -
+
-+
- - -
-- - +-
-- - +
-
* *
Fig. 5 Bu guillos e al.
Con ol
A
LPS
C
LPS + DEVD mk
E
LPS + zVAD mk
G
B
Con ol
LPS
D
LPS + DEVD mk
F
LPS + zVAD mk
H
LPS + Benzamide
I
LPS + Benzamide
J
LPS + MDL28170
K
LPS + MDL28170
L
siRNA C3
72h
siRNA C3
Di e en ia ion
siRNA NT
Undi e en ia ion
48h
siRNA NT
Suppl. Figu e 1. Bu guillos e al