ORIGINAL PAPER
Mo phological Di e en ia ion Towa ds Neu onal Pheno ype
o SH-SY5Y Neu oblas oma Cells by Es adiol, Re inoic Acid
and Choles e ol
Heidi Teppola
1
•Je a-Riina Sa kanen
2
•Tuula O. Jalonen
3
•Ma ja-Leena Linne
1
Recei ed: 19 July 2015 / Re ised: 16 Oc obe 2015 / Accep ed: 17 Oc obe 2015 / Published online: 30 Oc obe 2015
ÓThe Au ho (s) 2015. This a icle is published wi h open access a Sp inge link.com
Abs ac Human SH-SY5Y neu oblas oma cells main ain
hei po en ial o di e en ia ion and eg ession in cul u e
condi ions. The induc ion o di e en ia ion could se e as a
s a egy o inhibi cell p oli e a ion and umo g ow h. P e-
ious s udies ha e shown ha di e en ia ion o SH-SY5Y
cells can be induced by all- ans- e inoic-acid (RA) and
choles e ol (CHOL). Howe e , signaling pa hways ha lead
o e minal di e en ia ion o SH-SY5Y cells a e s ill la gely
unknown. The goal o his s udy was o examine in he RA
and CHOL ea ed SH-SY5Y cells he addi i e impac s o
es adiol (E
2
) and b ain-de i ed neu o ophic ac o (BDNF)
on cell mo phology, cell popula ion g ow h, synap ic esicle
ecycling and p esence o neu o ilamen s. The abo e ea-
u es indica e a highe le el o neu onal di e en ia ion. Ou
da a show ha ea men o 10 days in i o (DIV) wi h RA
alone o when combined wi h E
2
(RE) o CHOL (RC), bu
no when combined wi h BDNF (RB), signi ican ly
(p 0.01) inhibi ed he cell popula ion g ow h. Synap ic
esicle ecycling, induced by high-K
?
depola iza ion, was
signi ican ly inc eased in all ea men s whe e RA was
included (RE, RC, RB, RCB), and when all agen s we e
added oge he (RCBE). Speci ically, ou esul s show o he
i s ime ha E
2
ea men can alone inc ease synap ic
esicle ecycling in SH-SY5Y cells. This wo k con ibu es o
he unde s anding o he ways o imp o e supp ession o
neu oblas oma cells’ popula ion g ow h by inducing ma u-
a ion and di e en ia ion.
Keywo ds B ain-de i ed neu o ophic ac o 
Choles e ol Di e en ia ion Es adiol Re inoic acid 
SH-SY5Y
Abb e ia ions
AChE Ace ylcholines e ase
A c Ac i i y- egula ed cy oskele on associa ed
p o ein
BDNF B ain-de i ed neu o ophic ac o
CHOL Choles e ol
CO
2
Ca bon dioxide
CTNF Co ec ed o al neu o ilamen luo escence
DIV Days in i o
E
2
Es adiol
ER Es ogen ecep o
LTP Long- e m po en ia ion
NF-68 Neu o ilamen 68 kD
PBS Phospha e-bu e ed saline
RA All- ans e inoic acid
RB All- ans e inoic acid wi h BDNF
RC All- ans e inoic acid wi h CHOL
RCB All- ans e inoic acid wi h CHOL and BDNF
RCBE All- ans e inoic acid wi h CHOL, BDNF and E
2
RE All- ans e inoic acid wi h E
2
RT Room empe a u e
T kB T opomyosin- ela ed kinase B
Tuula O. Jalonen and Ma ja-Leena Linne a e Co-senio au ho s.
&Heidi Teppola
[email p o ec ed]
1
Depa men o Signal P ocessing, Tampe e Uni e si y o
Technology, P.O. Box 553, 33101 Tampe e, Finland
2
Depa men o Cell Biology, School o Medicine, Uni e si y
o Tampe e, Tampe e, Finland
3
Depa men o Physiology and Neu oscience, S . Geo ge’s
Uni e si y School o Medicine, S . Geo ge’s,
G enada, Wes Indies
123
Neu ochem Res (2016) 41:731–747
DOI 10.1007/s11064-015-1743-6
In oduc ion
Neu oblas oma is he mos common ex a-c anial solid
malignan umo o sympa he ic ne ous sys em in in an s
and young child en [1]. Rega dless o i s s age, un il oday,
he e is no cu e o ea men , which could o e good
p ognosis o pa ien s [2,3]. Human SH-SY5Y cell line,
used in his s udy, is a subclone o SK-N-SH cell line
which was isola ed om a bone ma ow o a 4 yea -old
emale pa ien [4]. SH-SY5Y cells main ain hei po en ial
o eg ession, which esul s in agg essi e p oli e a ion o
hese cells [5]. No el he apeu ic ea men s inducing di -
e en ia ion in o neu onal cell ypes could help o imp o e
he p ognosis o child en su e ing om neu oblas oma [6].
The induc ion o di e en ia ion could se e as a s a egy o
inhibi cell popula ion g ow h and e en ually s op he
umo g ow h, as well as induce heal hy ma u e neu ons in
pa ien s.
P e ious s udies ha e p esen ed ha di e en ia ion o
SH-SY5Y cells can be induced by dibu y yl cyclic AMP
[5,7], 12-o- e adecanoyl-pho bol-13-ace a e [8–13], all-
ans- e inoic-acid (RA) [9,14–16], b ain-de i ed neu-
o ophic ac o (BDNF) [17–19], anada e [20], ne e
g ow h ac o [21,22], choles e ol (CHOL) [23], i amin
D3, and neu egulin be a1 [24]. The signaling pa hways ha
lead o e minal di e en ia ion o SH-SY5Y cells, how-
e e , a e s ill la gely unknown.
The e inoic acid (RA) is a po en cell di e en ia ing
ac o , which h ough i s nuclea ecep o s a ec s a as
ange o p omo e si es in he neu onal and glial cells in
e e y s ep o emb yonic and pos na al li e [25]. RA-in-
duced di e en ia ion has been shown o inhibi cell p o-
li e a ion [9], change cellula sodium conduc ance [15],
enhance he ou g ow h o neu i es [16], inc ease he
ace ylcholines e ase (AChE) ac i i y [26,27], and enhance
he synap ic esicle ecycling [23]. Howe e , clinical ials
ha e demons a ed ha ea men wi h RA alone, o in
combina ion wi h in e e on alpha, is no enough agains
ecu en neu oblas oma in child en [28,29]. The e o e,
new al e na i e esou ces o mo e e ec i e neu onal di -
e en ia ion a e needed.
Choles e ol (CHOL) is a necessa y componen in cell
memb anes and impo an o synap ic s uc u e and
unc ion [30]. I is syn hesized by neu ons hemsel es o
hei su i al and g ow h. The de elopmen o ac i e
synapses equi es addi ional amoun o choles e ol ha is
shown o be sec e ed by glial cells (speci ically as ocy es)
in he cen al ne ous sys em [30–32] and by Schwann
cells in he pe iphe al ne ous sys em [33]. The glia-
de i ed choles e ol has also been shown o be c ucial o
di e en ia ion o dend i es, synap ogenesis, inc ease in
synap ic p o ein exp ession (synapsin-1) and neu onal
ac i i y, and o ansmi e elease [30,32,34,35]. In pu e
human SH-SY5Y cell cul u es, he glia-de i ed choles e ol
is non-exis ing, and addi ion o choles e ol is needed in
o de o achie e condi ions esembling no mal neu onal
en i onmen wi h su ounding glial cells [23].
The b ain-de i ed neu o ophic ac o (BDNF) has been
shown o suppo he su i al o neu ons and s imula e he
g ow h and di e en ia ion o new neu ons and synapses
[36]. BDNF is a ligand o opomyosin- ela ed kinase B
(T kB) ecep o , exp ession o which is lacking in naı
¨ e
neu oblas oma cells. Howe e , T kB ecep o exp ession
and esponsi eness o BDNF is induced by RA [37]. The
ac i a ion o T kB by BDNF has been sugges ed o
enhance neu oblas oma cell su i al and esis ance o
chemo he apy [38]. BDNF has also been shown o expose
only a modes bene i o RA-induced a es in a do man
s a e [6]. Howe e , he sequen ial ea men o SH-SY5Y
cells wi h RA and BDNF has been epo ed o induce
di e en ia ed, neu o ophic ac o -dependen neu on-like
cells [18] and sus ained ea men has been epo ed o
enhance neu onal di e en ia ion o neu al p ogeni o cells
[39]. Mo eo e , RA-BDNF ea men induces signi ican
inc ease in he exp ession o synap ic genes, b ain miRNA,
miRNA biogenesis machine y, and AChE ac i i y, in
compa ison o sole RA ea men [19]. These s udies
s imula e he in e es o u he examine he po en ial
he apeu ic compe ence o BDNF o RA-induced SH-
SY5Y cell di e en ia ion and o ea men o neu oblas-
oma. Dis up ion o BDNF and i s downs eam signaling
pa hways ha e been obse ed in many neu odegene a i e
diseases such as Alzheime ’s, Pa kinson’s and Hun ing-
on’s diseases [40–42], unde lining he impo ance o
BDNF. Howe e , he esul s o he ole o added exoge-
nous BDNF o di e en ia ion o RA ea ed SH-SY5Y
cells a e s ill con o e sial.
Es adiol (17-be a-es adiol, E
2
), a o m o es ogen
ho mone, has bo h acu e and long- e m e ec s on a a ie y
o neu onal cell ypes. Inc ease in he numbe o dend i ic
spines and numbe o exci a o y synapses, which a e he
slow long- e m e ec s o es adiol, we e i s de ec ed [43–
45]. The acu e e ec s, which al e he in insic and synap ic
physiology o neu ons wi hin minu es ( e e sible depola -
iza ion and inc eased inpu esis ance wi h a la ency o
1 min in 19.8 % o CA1 neu ons es ed) we e de ec ed
la e [46,47]. Se e al s udies ha e shown ha es ogen
enhances synap ogenesis and synap ic plas ici y [48–55],
which p ope ies may be c ucial o example in enhancing
memo y consolida ion [43,56]. Addi ionally, es ogen has
been demons a ed o induce synap ic connec i i y [52],
enhance NMDA ecep o exp ession and ac i i y [57–66],
and long- e m po en ia ion (LTP) [58,59,64,67].
Se e al ea lie s udies ha e add essed possible signaling
732 Neu ochem Res (2016) 41:731–747
123
mechanisms associa ed wi h es ogen-induced cellula
unc ions. Es ogen ac i a es hese unc ions h ough he
ac i a ion o es ogen ecep o s (ERs), ERaand ERb,
which se e as ansc ip ion ac o s modi ying he ac i i y
o a ge genes [68,69]. Es ogen has been shown o
inc ease he key synap ic p o eins, e.g. PSD-95, ia ei he
ac i a ion o ERa[70], ERb[51], o bo h [71]. Es ogen is
hough o use bo h nuclea ERs and plasma memb ane
ERs which a e usually e e ed o as classical genomic and
non-genomic pa hways [69]. In classical genomic ac ion,
ERs a e hough o ansloca e in o nucleus in ligand-de-
penden manne and ac ing as a ansc ip ion ac o o
a ge genes a e p olonged es ogen exposu e [72]. In
non-genomic ac ion, es ogen has been shown o ac i a e
he memb ane ERs, which apidly s imula e he memb ane-
associa ed signaling molecules such as PI-3K and MAPK,
esul ing in quick inc ease in p o ein exp ession [73,74].
SH-SY5Y cells ha e been shown o exp ess bo h ERaand
ERb[75]. Es ogen has also been shown o s imula e he
ac i i y- egula ed cy oskele on associa ed p o ein (A c)
exp ession ia he MAPK and PI-3K dependen pa hway in
SH-SY5Y cells [48]. A c is known o be induced by neu-
onal ac i i y and playing a key ole in ac i i y-dependen
synap ic plas ici y [76]. I s knockdown has been shown o
lead o impai men o long- e m memo y [77,78]. How-
e e , he speci ic ole o E
2
o inducing di e en ia ion in
human SH-SY5Y neu oblas oma cells is s ill unknown.
The main goal o he cu en s udy was o ind a unc-
ional combina ion o subs ances o e ec i e induc ion o
di e en ia ion o he SH-SY5Y cells. Based on ou ea lie
s udies, we used RA and CHOL as p ima y di e en ia ion
ea men s [23]. We u he in es iga ed he abili y o E
2
and BDNF o suppo , and possible enhance, he RA and
CHOL induced neu onal di e en ia ion. We quan i ied he
indi idual and addi i e impac s o BDNF and E
2
on he RA
and CHOL-induced neu i e ou g ow h, p esence o neu o-
ilamen 68, synap ic esicle ecycling and a es in he
popula ion g ow h a e o SH-SY5Y neu oblas oma cells
in i o.
Me hods
Main enance and Di e en ia ion o Cell Cul u es
The human SH-SY5Y neu oblas oma cell line (CRL-2266;
Ame ican Type Cul u e Collec ion, Manassas, VA, USA)
was cul u ed as p e iously desc ibed [23]. B ie ly, he cells
we e pla ed a passage 29–30 wi h densi y o 5000 cells/
cm
2
on 48-well cul u e dishes. Cells we e cul u ed and
main ained in 5 % CO
2
humidi ied incuba o a 37 °Cin
1:1 nu ien mix u e F-12K Kaighn’s modi ica ion, and
minimum essen ial medium supplemen ed wi h 10 % e al
bo ine se um, 2 mM/L L-glu amine, 1 % an ibio ic–an-
imyco ic mix u e and 1 % non-essen ial amino acids (all
eagen s om GIBCO, In i ogen, Ca lsbad, CA, USA,
unless o he wise s a ed). Cell di e en ia ion was induced
wi h 10 lM/L all- ans e inoid-acid (RA), 1 nM/L
17-be a-es adiol (E
2
), 50 ng/mL b ain-de i ed neu-
o ophic ac o (BDNF), 10 lg/mL choles e ol (3b-hy-
d oxy-5-choles ene, CHOL), o wi h combina ions such as
(i) 5 lM/L RA wi h 5 lg/mL CHOL (RC), (ii) 5 lM/L RA
wi h 50 ng/mL BDNF (RB), (iii) 5 lM/L RA wi h 1 nM/L
E
2
(RE), (i ) 5 lM/L RA wi h 5 lg/mL CHOL, and 50 ng/
mL BDNF (RCB), and ( ) 5 lM/L RA wi h 5 lg/mL
CHOL, 50 ng/mL BDNF, and 1 nM/L E
2
(RCBE) o 10
DIV (all di e en ia ion eagen s om Sigma-Ald ich, S
Louis, MO, USA, unless o he wise s a ed). S ock solu ions
o di e en ia ion subs ances we e dilu ed in 96 % e hanol;
he inal e hanol concen a ion ne e exceeded 0.1 % in
cell cul u e. Con ol cells we e ea ed wi h 0.1 % e ha-
nol. All used subs ance concen a ions we e ca e ully
e alua ed acco ding o al eady published li e a u e. Sui-
able, leas oxic concen a ions, also used by o he labo-
a o ies, we e used o enable compa ison o ou esul s wi h
o he s. All di e en ia ion subs ances (excep BDNF when
used in combina ions) we e applied wi h medium exchange
a 1, 3 and 7 DIV. BDNF was applied a 4 and 7 DIV when
used oge he wi h RA (RB), RA and CHOL (RCB) o RA,
CHOL and E
2
(RCBE). The cell g ow h, condi ion, and
mo phology we e obse ed wi h cul u e mic oscope
(Olympus CK40) and images we e aken a 10 DIV DP10
mic oscope digi al came a sys em (Olympus, Tokyo,
Japan).
Neu o ilamen S aining
Fo de ec ing he le el o di e en ia ion in he neu oblas-
oma cell cul u es, he cells we e s ained a 10 DIV wi h
neu onal ma ke NF-68 o neu o ilamen ligh polypep ide
(68 kDa, Sigma-Ald ich). Cells we e i s ixed o 20 min
wi h 4 % pa a o maldehyde (Sigma-Ald ich) in phospha e
bu e ed saline solu ion (PBS), washed h ee imes wi h
PBS and pe meabilized in 0.5 % T i on X-100 (J.T. Bake ,
Phillipsbu g, NJ, USA) o 15 min. A e washing wi h
PBS, he non-speci ic an ibody binding si es we e blocked
wi h 10 % bo ine se um albumin (GIBCO) in PBS o
30 min o educe he backg ound. Cells we e hen incu-
ba ed wi h he p ima y an ibody mouse monoclonal an i-
NF-68 1:200 o 1 h a oom empe a u e (RT; ?22 °C),
insed h ee imes wi h PBS, and hen incuba ed wi h a
seconda y an ibody FITC-conjuga ed goa an i-mouse IgG
1:100 (Sigma-Ald ich) o 30 min a RT. Fluo escence was
isualized wi h Nikon Eclipse TS100 mic oscope equipped
wi h Nikon DS Came a Con ol Uni DS L-1 and images
we e o ganized wi h Visio 2010 (Mic oso , WA, USA).
Neu ochem Res (2016) 41:731–747 733
123
The in ensi y o o al neu o ilamen luo escence (NF-68)
and he in ensi y o o al backg ound luo escence we e
measu ed om each luo escence image wi h ImageJ
so wa e (Na ional Ins i u e o Men al Heal h, Be hesda,
Ma yland, USA) [79]. Co ec ed o al neu o ilamen luo-
escence (CTNF) was calcula ed om he ga he ed da a in
Excel 2010 (Mic oso , WA, USA) wi h he me hod used
p e iously [80,81], as ollows: The luo escence o he
neu o ilamen s o in e es was selec ed using he selec ion
ool. A ea o in e es , in eg a ed densi y, and mean g ay
alue we e calcula ed om selec ed a eas wi h ImageJ
so wa e. A egion nex o he selec ed neu o ilamen was
selec ed as a backg ound alue. The CTNF was calcula ed
by using he ollowing equa ion CTNF =in eg a ed den-
si y -(a ea o selec ed neu o ilamen s 9mean luo es-
cence o backg ound eadings).
Quan i ica ion o Cell Popula ion G ow h
The subs ance-induced changes in he g ow h a e we e
quan i ied by coun ing he nuclei o 10 DIV cul u ed SH-
SY5Y cells in each ea men g oup. Cell nuclei we e
s ained wi h 10 lg/mL Hoechs 33258 (Sigma-Ald ich) o
5 min. Cul u es we e washed i e imes in PBS and
moun ed on co e slips. Fluo escence esul s we e isual-
ized wi h Nikon DS Came a Con ol Uni DS L-1. Images
o each ea men g oup we e analyzed wi h CellC analysis
so wa e [82], which co ec s he image backg ound o
au o- luo escence by i ing a wo-dimensional quad a ic
polynomial o he image and sub ac s he i ed polynomial
su ace om he o iginal image. A e his he algo i hm
sepa a es he nuclei pixels om backg ound pixels by
global h esholding and p oduces a bina ized image wi h
whi e nuclei on a black backg ound. I u he mo e sepa-
a es clus e ed nuclei om each o he by ma ke -con olled
wa e shed segmen a ion, which is based on nuclei in en-
si y. E en ually he so wa e emo es a i ac s, such as
s aining esidues by disca ding objec s smalle han 1/10 o
he mean size o all objec s. Images we e o ganized wi h
Mic oso Visio 2010. The ob ained nuclei coun s and
s a is ics (see sec ion ‘‘S a is ical Analysis’’) we e analyzed
and plo ed in MATLAB ( e sion 2013b, The Ma hwo ks
Inc., MA, USA).
Quan i ica ion o Neu i e Leng h
The SH-SY5Y cells we e cul u ed in CTRL, CHOL, E
2
,
BDNF, RA, RE, RB, RC and RCBE condi ions a 10 DIV.
Neu i es we e aced om phase con as images o each
ea men g oup wi h Neu onG ow h plugin [83] o he
ImageJ so wa e (Na ional Ins i u e o Men al Heal h,
Be hesda, Ma yland, USA) [79]. The Neu onG ow h
au oma ically coun s he leng h o aced neu i es in pixels
and expo s he da a. The ga he ed da a and s a is ics we e
analyzed and plo ed in MATLAB ( e sion 2013b, The
Ma hwo ks Inc., MA, USA).
Synap ic Vesicle Recycling
The le el o synap ic esicle ecycling was e i ied by
measu ing he numbe o luo escence punc a in 10 DIV
cul u ed SH-SY5Y cells. Cells we e ea ed ei he wi h
0.1 % e hanol (CTRL), RA, CHOL, BDNF, E
2
, o wi h
hei combina ions. Cul u es we e s ained wi h AM1-43
s y yl dye (Bio ium, Haywa d, CA, USA) o de ec ing
synap ic exo/endocy osis in cells. AM1-43 is a ixable
ne e e minal p obe. I is no able o pass h ough he
memb anes, bu ins ead, when cells a e depola ized wi h
high po assium (K
?
)-Ty ode solu ion, AM1-43 s y yl dye
a aches inside hose esicles, which a e ongoing exocy-
osis. S aining was modi ied om me hod p e iously
desc ibed [23,84,85]. In he cu en expe imen s, he cells
we e incuba ed o 1 min wi h 4 lmol/L AM1-43,
acco ding o manu ac u e ’s p o ocol, wi h he depola izing
Ty ode solu ion including 80 mmol/L K
?
(80 mmol/L
KCl, 29 mmol/L NaCl, 2 mmol/L MgCl
2
, 30 mmol/L
glucose, 25 mmol/L HEPES, 2 mmol/L CaCl
2
). Cells we e
u he washed se e al imes wi h SCAS quenche solu ion
(Bio ium, Haywa d, CA, USA) a RT o educe backg ound
luo escence. Cells we e ixed o 20 min wi h 4 %
pa a o maldehyde (GIBCO), pe meabilized in 0.01 %
T i on X-100 (J.T. Bake ) o 12 min and washed h ee
imes o 1 min in cold PBS. All eagen s we e om Sigma
Ald ich unless o he wise s a ed. The luo escence was
isualized wi h Nikon Eclipse TS100 mic oscope equipped
wi h Nikon DS Came a Con ol Uni DS L-1 and images
we e o ganized wi h Mic oso Visio 2010. Fluo escence
images o each ea men g oup we e analyzed wi h ImageJ
analysis so wa e [79] using he ollowing p ocedu al s eps
speci ically designed o his s udy: backg ound o he
image was sub ac ed by se ing a olling ball adius o 50
pixels, a e image was sha pened, and hen he maxima o
luo escence punc a we e ound wi h noise ole ance o 20
and wi h he poin selec ion s yle. This p ocedu e was
e alua ed by isual inspec ion and i was ound o be he
bes o inding he co ec numbe o AM1-43 punc a om
luo escence images. The ob ained coun s o luo escence
punc a pe image we e di ided by he ob ained median
nuclei numbe (see sec ion ‘‘Quan i ica ion o Cell Popu-
la ion G ow h’’) in pa icula cul u e in 10 DIV ea men s.
These ob ained coun s o luo escence punc a pe median
nuclei numbe and s a is ics (see sec ion ‘‘S a is ical
Analysis’’) we e analyzed and plo ed in MATLAB ( e -
sion 2013b).
734 Neu ochem Res (2016) 41:731–747
123
Summa y o he Le el o Di e en ia ion
Resul s a e summa ized in Table 1, which shows he le el
o di e en ia ion induced by di e en ea men s. The le el
o di e en ia ion was assessed a leas om h ee samples
om wo sepa a e expe imen s by analyzing he ollowing
ea u es; neu i e leng h, p esence o neu o ilamen s, inhi-
bi ion in cell popula ion g ow h a e and synap ic esicle
ecycling. Neu i e leng hs we e de ec ed bo h isually and,
by using au oma ed me hods o suppo he isual de ec-
ion. O he ea u es we e de ined acco ding o he me hods
explained abo e (see sec ions ‘‘Neu o ilamen S aining,
Quan i ica ion o Cell Popula ion G ow h and Synap ic
Vesicle Recycling’’).
S a is ical Analysis
S a is ical analysis was pe o med using One-way ANOVA
in MATLAB ( e sion 2013b). Di e ences we e conside ed
o be signi ican when p 0.01 o p 0.05, di e en
signi icances a e indica ed wi h ** o * in he igu es,
espec i ely.
Resul s
Mo phology o Di e en ia ed SH-SY5Y Cells
Phase con as images o SH-SY5Y cul u es a 10 DIV
we e i s isually analyzed o mo phological assessmen .
In he isual analysis, con ol cells showed no pa icula
neu i e ou g ow h (Fig. 1a), whe eas CHOL- ea ed cells
had a numbe o sho neu i es (g een a ows in Fig. 1b; see
also Table 1). Cells ea ed wi h E
2
had e y sho neu i es
(Fig. 1c), which we e bo h ewe and sho e han he
CHOL-induced neu i es. The mo phology o BDNF ea ed
SH-SY5Y was ela i ely pola and cells g ew mo e sp ead
in he cul u e dish (Fig. 1d). This di e ed om con ol
cells, which g ew in clus e s (Fig. 1a). No signi ican ou -
g ow h o neu i es was obse ed a e BDNF ea men .
T ea men wi h RA alone (Fig. 1e) as well as wi h com-
bina ions such as RE, RB, RC, RCB o RCBE, induced
b anching o longe neu i es and de ec able ne wo k o -
ma ion (Fig. 1 –j, espec i ely).
The da a show o he mo phological di e ences be ween
he ea men s as well. The RE ea ed SH-SY5Y cells had
Table 1 Summa y o di e en ia ion
T ea men

Neu i e leng h To al neu o ilamen luo escence Inhibi ion o cell popula ion g ow h Synap ic esicle ecycling
CTRL -- - -
CHOL ??** -- -
E
2
?*-- ?*
BDNF -- - -
RA ???** ???** ???** ??**
RE ???** ??*???** ??**
RC ???** ???** ???** ???*
RB ???** ???** ?? ??**
RCB ???**
àà
???*
RCBE ???** ???**
à
???*
The c i e ia o ca ego izing he neu i e leng h we e as ollows: -neu i es simila o con ol, ?sho neu i es wi hou b anching and signi i-
can ly longe han in con ol [p 0.05 (*)], ?? in e media e neu i es wi hou b anching and signi ican ly longe han in con ol [p 0.01 (**)],
??? long neu i es wi h b anching and signi ican ly longe han in con ol [p 0.01 (**)] and in CHOL o E
2
ea men condi ions [p 0.01
(**)]. The c i e ia o ca ego izing he o al neu o ilamen luo escence we e as ollows: -no neu o ilamen luo escence, ?? signi ican ly
[p 0.05 (*)] inc eased neu o ilamen luo escence compa ed o con ol, ??? signi ican ly [p 0.01 (**)] inc eased neu o ilamen luo es-
cence compa ed o con ol. The c i e ia o ca ego izing he inhibi ion o cell popula ion g ow h we e as ollows: - he numbe o cells has no
changed a e he ea men , ?? he numbe o cells dec eased (no signi ican ly), ??? he numbe o cells signi ican ly [p 0.01 (**)]
dec eased. The c i e ia o ca ego izing he amoun o synap ic esicle ecycling we e as ollows: -no signi ican change in he coun s o punc a
in compa ison o con ol, ?signi ican [p 0.05 (*)] inc ease in he coun s o luo escence punc a in compa ison o con ol, ?? signi ican
[p 0.01 (**)] inc ease in he coun s o luo escence punc a in compa ison o con ol, ??? signi ican [p 0.05 (*)] inc ease in he coun s o
luo escence punc a in compa ison o RE and RB

Con ol (CTRL, 0.1 % e hanol), choles e ol (CHOL; 10 lg/ml), 17-be a-es adiol (E
2
; 1 nM/L), b ain de i ed neu o ophic ac o (BDNF;
50 ng/mL), all- ans e inoic acid (RA; 10 lg/mL), RA wi h E
2
(RE; RA 5 lg/mL, E
2
1 nM/L), RA wi h CHOL (RC; RA 5 lg/mL, CHOL 5 lg/
mL), RA wi h BDNF (RB; RA 5 lg/mL, BDNF 50 ng/mL), RA wi h CHOL and BDNF (RCB; RA 5 lg/mL, CHOL 5 lg/mL, BDNF 50 ng/
mL), RA wi h CHOL, BDNF and E
2
(RCBE; RA 5 lg/mL, CHOL 5 lg/mL, BDNF 50 ng/mL, E
2
1 nM/L)
à
Conclusi e da a no a ailable
Neu ochem Res (2016) 41:731–747 735
123

Fig. 1 Mo phology and
ne wo k o ma ion o SH-SY5Y
neu oblas oma cells a 10 DIV.
aSH-SY5Y cells we e g own
o 10 DIV in con ol condi ions
(CTRL) and wi h bcholes e ol
(CHOL), ces adiol (E
2
),
db ain-de i ed neu o ophic
ac o (BDNF), eall- ans
e inoic-acid (RA), o wi h hei
combina ions –jRE, RB, RC,
RCB, RCBE, espec i ely. Da a
show ha he CHOL ea men
induced sho neu i es (g een
a ows) wi h many a icosi ies
( ed a ows)(b). E
2
induced
ew e y sho neu i es in
compa ison o CHOL induced
neu i es (c). BDNF ea men
did no induce signi ican
g ow h o neu i es o ne wo k
o ma ion (d). RA ea men
gene a ed hin b anching
neu i es and p omo ed ne wo k
o ma ion (blue a ows,e).
Ne wo ks o cells wi h cell- o
cell con ac s (blue a ows) we e
always de ec ed when ea ed
wi h RA oge he wi h E
2
(RE), gBDNF (RB), hCHOL
(RC), iCHOL and BDNF
(RCB), and jCHOL, BDNF,
and E
2
(RCBE). Fla subs a e-
adhe en (S- ype) cells we e
de ec ed especially when cells
we e ea ed wi h E
2
o BDNF
(o ange a ows)(c,d) bu also
when cells we e ea ed wi h RE
(da a no shown) o RB (o ange
a ows)(g). The RE ea ed
cells ( ) had hinne neu i es
han hose ea ed wi h RCBE
(j) (Colo igu e online)
736 Neu ochem Res (2016) 41:731–747
123
ne wo ks o oundish cells wi h hin neu i es wi hou hea y
b anching (Fig. 1 ). A small numbe o subs a e-adhe en
(S- ype) la cells [86] we e obse ed in cul u es ea ed
wi h RE (da a no shown), RB (o ange a ows in Fig. 1g)
and RA (da a no shown). The RB ea men induced ne -
wo ks ha consis ed o ex ended con ac ing neu i es, as
well as o cells in di ec con ac wi h each o he wi hou
neu i es. The neu i es o he RB ea ed cells we e hicke
in compa ison o RE induced neu i es. The RC ea men
induced cells wi h long, b anching and connec ing neu i es
(blue a ows in Fig. 1h). O he choles e ol ea ed cul u es,
such as RCB and RCBE, con ained neu ons wi h long
b anching neu i es and ne wo k o ma ion wi hou S- ype
cells. Mo e a icosi ies ( ed a ows in Fig. 1b) and small
cell clus e s (da a no shown) we e obse ed in CHOL
ea ed cul u es (CHOL, RC, RCB, RCBE) in compa ison
o he con ol, RE and RB ea ed g oups, in which cells
we e mo e uni o mly dis ibu ed (da a no shown).
Inhibi ion o Cell Popula ion G ow h
The abili y o a subs ance o inhibi he popula ion g ow h
o human SH-SY5Y cells is one o he indica o s o
inc eased le el o di e en ia ion. The e o e, we coun ed
he numbe s o he Hoechs 33258 s ained nuclei a 10 DIV
in RA, CHOL, BDNF, E
2
, RE, RB, and RC ea ed cell
cul u es and compa ed he esul s o he numbe o nuclei
in con ol condi ions. The da a demons a ed ha CHOL,
E
2
o BDNF ea men s on hei own did no inhibi he cell
popula ion g ow h (Fig. 2), which, howe e , was seen
when ea ed wi h RA, as well as wi h RA oge he wi h
CHOL (RC) [p 0.01(**)] when compa ed o con ols.
Mo eo e , signi ican (p 0.01 (**) inhibi ion was also
de ec ed wi h RA oge he wi h E
2
(RE) ea men , when
compa ed o con ols (Fig. 2). In e es ingly, when cells
we e ea ed wi h RA and BDNF, no inhibi ion o g ow h
was de ec ed (Fig. 2).
Neu i e Leng hs
The neu i es o SH-SY5Y cells we e aced om phase
con as images aken om each expe imen a 10 DIV
wi h Neu onG ow h (see sec ion ‘‘Me hods’’), which p o-
ides suppo i e in o ma ion o he leng hs in addi ion o
he isual inspec ion o he cell mo phology. In addi ion o
RA [p 0.01 (**)], also wi h CHOL alone [p 0.01 (**)]
and E
2
alone [p 0.05 (*)] ea men s, induced a signi i-
can inc ease in he leng h o neu i es in compa ison o
con ol cells a 10 DIV (Fig. 3). Fu he mo e, he signi i-
can inc ease in he neu i e leng h was seen in all
0
50
100
150
200
250
300
350
400
CTRL CHOL ES BDNF RA RE RB RC
TREATMENTS
CELL NUCLEI COUNTS
** ** **
Fig. 2 Inhibi ion o cell popula ion g ow h. Cells we e cul u ed in
con ol condi ions [CTRL (n =7)], and wi h CHOL (n =6), E
2
(n =6), BDNF (n =6), RA (n =6), RE (n =6), RB (n =6), and
RC (n =8) (n is he numbe o analyzed mic oscopy images) and
he cell nuclei we e coun ed a 10 DIV. In he boxplo ep esen a ion
he ob ained median nuclei coun s, 25 h and 75 h pe cen iles,
ex eme da a poin s, and ou lie s o he da a a e shown wi h ed
line,blue edges,black whiske s, and ed as e isks, espec i ely.
Each di e en ia ion agen is shown on he x-axis and he cell nuclei
coun s on he y-axis. The s a is ically signi ican di e ences
(p 0.01) a e shown wi h as e isks (**). Signi ican ly lowe cell
numbe s we e de ec ed when cells we e ea ed wi h RA, RE, o RC
in compa ison o CTRL da a. Sligh inc ease in cell numbe s (nuclei
coun s) we e obse ed when cells we e ea ed wi h RB, in
compa ison o he cells ea ed solely wi h RA. The nuclei coun s
o RB ea ed cul u es we e no signi ican ly lowe in compa ison o
CTRL (Colo igu e online)
Neu ochem Res (2016) 41:731–747 737
123
combina ion ea men s such as RE, RB, RC, RCBE
[p 0.01 (**)] ela i e o con ol condi ions (Fig. 3). Wi h
he BDNF ea men alone no inc ease in he neu i e leng h
was de ec ed.
P esence o Neu o ilamen s in SH-SY5Y
Neu oblas oma Cells
The le el o di e en ia ion o he SH-SY5Y neu oblas oma
cells a 10 DIV was u he e i ied by imaging he NF-68
neu o ilamen s. The neu o ilamen luo escence was
de ined by isual inspec ion and by measu ing he in ensi y
o CTNF, when cells we e ea ed solely wi h RA, BDNF,
CHOL, o E
2
(Fig. 4b–e, espec i ely) o in combina ion
wi h RA and E
2
(RE), RA and BDNF (RB), RA and CHOL
(RC), and RA, CHOL, BDNF, and E
2
(RCBE) (Fig. 4g–j,
espec i ely). Bo h isual and au oma ed image analyses
showed ha he p esence o he NF-68 neu o ilamen s was
clea ly induced by RA (Fig. 4b). No majo inc ease in NF-
68 luo escence le els was obse ed isually o au oma i-
cally, when cells we e ea ed wi h BDNF, CHOL o E
2
(Fig. 4c–e). Howe e , he in ensi y o NF-68 luo escence
was signi ican ly inc eased in all o he combina ion
ea men s ela i e o con ol cells, as shown in Fig. 4g–j,
and in Fig. 5 o RA [p 0.01 (**)], RE [p 0.05(*)],
RC [p 0.01(**)], RB [p 0.01(**)] and RCBE
[p 0.01(**)].
Di e en ia ion-Induced Synap ic Vesicle Recycling
Ou g oup has ea lie shown ha RA and RC ea ed
human SH-SY5Y cells show in ense Synap ophysin I
(SypI) luo escence in cell soma a, along he neu i es and a
he si es o he cell- o-cell con ac s. Fu he mo e, we ha e
shown co-localiza ion o SypI and AM1-43 a he end o
he neu i es a he cell- o-cell con ac s o he RA and RC
di e en ia ed and high K
?
depola ized human SH-SY5Y
cells [23]. I has been also shown elsewhe e ha he SH-
SY5Y cells a e capable o depola iza ion wi h high K
?
s imula ion [87]. The e o e, high K
?
s imula ion was used
o s udying he s imula ion- ela ed synap ic esicle ecy-
cling also in his s udy. The SH-SY5Y neu oblas oma cells
we e incuba ed wi h E
2
, CHOL, BDNF, o RA o wi h hei
combina ions (RE, RB, RC, RCB and RCBE) and s ained
a 10 DIV wi h AM1-43, a luo escen s y yl dye (a ne e
e minal p obe) wi h he p esence o depola izing high K
?
-
Ty ode solu ion. The numbe o luo escen punc a,
e lec ing he ecycling synap ic esicles, was coun ed a e
depola iza ion (see Fig. 6and sec ion ‘‘Me hods’’). T ea -
men wi h CHOL o BDNF alone does no inc ease he
numbe o luo escence punc a in compa ison o CTRL.
Ou da a show o he i s ime, ha he ea men wi h E
2
alone [p 0.05(*)], o RA oge he wi h E
2
(RE), BDNF
(RB), BDNF and CHOL (RBC) o BDNF, CHOL and E
2
(RBCE), [p 0.01(**)] signi ican ly inc eases he numbe
CTRL CHOL E2 BDNF RA RE RB RC RCBE
200
400
600
800
1000
1200
TREATMENTS
NEURITE LENGTH (PIXELS)
**
** **
**
**
** *
Fig. 3 Neu i e leng hs in SH-SY5Y cells. Cells we e cul u ed in
CTRL condi ions, and wi h CHOL, E
2
, BDNF, RA, o wi h hei
combina ions; RE, RB, RC, RCBE. In he boxplo ep esen a ion he
ob ained median o neu i e leng hs, he 25 h and 75 h pe cen iles, he
ex eme da a poin s, and he ou lie s o he da a a e shown wi h ed
line,blue edges,black whiske s and ed as e isks, espec i ely. Each
ea men is shown on x-axis and he neu i e leng hs in pixels on
y-axis. The s a is ically signi ican di e ences (p 0.01) and
(p 0.05) a e shown wi h as e isks (**) and (*), espec i ely. The
neu i e leng hs we e signi ican ly longe when cells a e ea ed wi h
RA, CHOL, RE, RB, RC, and RCBE (p 0.01). In e es ingly, a 10
DIV, E
2
induces only sho neu i es, bu he inc ease in he neu i e
leng h is s ill signi ican in compa ison o con ols (p 0.05).
Mo eo e , RA induces signi ican ly longe neu i es when compa ed
o CHOL o E
2
induced neu i e leng hs (p 0.01) (Colo
igu e online)
738 Neu ochem Res (2016) 41:731–747
123
Fig. 4 The p esence o NF-68
neu o ilamen s in SH-SY5Y
neu oblas oma cells.
Neu o ilamen s we e de ec ed a
10 DIV by immunos aining o
NF-68 (g een) and cell nuclei
wi h Hoechs 33258 (blue). a–
eCombined double s aining o
neu o ilamen s and nuclei in he
SH-SY5Y cells in con ol
condi ions (CTRL), and when
ea ed wi h RA, BDNF, CHOL
and E
2,
show ha RA induces
NF-68, seen also in he long
b anching neu i es. a, No NF-
68 is seen in CTRL. g–jRE,
RB, RC and RCBE ea men s
show p esence o
neu o ilamen s simila o hose
wi h RA alone (Colo
igu e online)
Neu ochem Res (2016) 41:731–747 739
123
73. I ano a T, Mendez P, Ga cia-Segu a LM e al (2002) Rapid
s imula ion o he PI3-kinase/ak signalling pa hway in de el-
oping midb ain neu ones by oes ogen. J Neu oendoc inol
14:73–79
74. Mannella P, B in on RD (2006) Es ogen ecep o p o ein
in e ac ion wi h phospha idylinosi ol 3-kinase leads o ac i a ion
o phospho yla ed ak and ex acellula signal- egula ed kinase
1/2 in he same popula ion o co ical neu ons: a uni ied
mechanism o es ogen ac ion. J Neu osci 26:9439–9447
75. Bang OY, Hong HS, Kim DH e al (2004) Neu op o ec i e
e ec o genis ein agains be a amyloid-induced neu o oxici y.
Neu obiol Dis 16:21–28
76. S ewa d O, Wo ley PF (2001) A cellula mechanism o a -
ge ing newly syn hesized mRNAs o synap ic si es on dend i es.
P oc Na l Acad Sci U S A 98:7062–7068
77. Guzowski JF, Ly o d GL, S e enson GD e al (2000) Inhibi ion
o ac i i y-dependen a c p o ein exp ession in he a hip-
pocampus impai s he main enance o long- e m po en ia ion
and he consolida ion o long- e m memo y. J Neu osci
20:3993–4001
78. Guzowski JF, Se low B, Wagne EK e al (2001) Expe ience-
dependen gene exp ession in he a hippocampus a e spa ial
lea ning: a compa ison o he immedia e-ea ly genes a c, c- os,
and zi 268. J Neu osci 21:5089–5098
79. Schneide CA, Rasband WS, Elicei i KW (2012) NIH image o
ImageJ: 25 yea s o image analysis. Na Me hods 9:671–675
80. Ga e O, Pines J (2010) P og essi e ac i a ion o CyclinB1-
Cdk1 coo dina es en y o mi osis. De Cell 18:533–543
81. Bu gess A, Vigne on S, B ioudes E e al (2010) Loss o human
g ea wall esul s in G2 a es and mul iple mi o ic de ec s due o
de egula ion o he cyclin B-Cdc2/PP2A balance. P oc Na l
Acad Sci U S A 107:12564–12569
82. Selinummi J, Seppala J, Yli-Ha ja O e al (2005) So wa e o
quan i ica ion o labeled bac e ia om digi al mic oscope ima-
ges by au oma ed image analysis. Bio echniques 39:859–863
83. Fan i Z, De-Miguel FF, Ma inez-Pe ez ME (2008) A me hod
o semiau oma ic acing and mo phological measu ing o
neu i e ou g ow h om DIC sequences. Con P oc IEEE Eng
Med Biol Soc 2008:1196–1199
84. C o BG, Fo in GD, Co e a AT e al (2005) No mal biogenesis
and cycling o emp y synap ic esicles in dopamine neu ons o
esicula monoamine anspo e 2 knockou mice. Mol Biol
Cell 16:306–315
85. Pyle JL, Ka alali ET, Choi S e al (1999) Visualiza ion o
synap ic ac i i y in hippocampal slices wi h FM1-43 enabled by
luo escence quenching. Neu on 24:803–808
86. Biagio i T, D’Amico M, Ma zi I e al (2006) Cell enewing in
neu oblas oma: elec ophysiological and immunocy ochemical
cha ac e iza ion o s em cells and de i a i es. S em Cells
24:443–453
87. Daniele F, Di Cai ano ES, Mo e i S e al (2015) TIRFM and
pH-sensi i e GFP-p obes o e alua e neu o ansmi e esicle
dynamics in SH-SY5Y neu oblas oma cells: Cell imaging and
da a analysis. J Vis Exp (95). doi:10.3791/52267
88. Simon T, Be hold F, Bo kha d A e al (2011) T ea men and
ou comes o pa ien s wi h elapsed, high- isk neu oblas oma:
esul s o ge man ials. Pedia Blood Cance 56:578–583
89. Be hold F, Boos J, Bu dach S e al (2005) Myeloabla i e
mega he apy wi h au ologous s em-cell escue e sus o al
main enance chemo he apy as consolida ion ea men in
pa ien s wi h high- isk neu oblas oma: a andomised con olled
ial. Lance Oncol 6:649–658
90. Ma hay KK, Villablanca JG, Seege RC e al (1999) T ea men
o high- isk neu oblas oma wi h in ensi e chemo he apy,
adio he apy, au ologous bone ma ow ansplan a ion, and
13-cis- e inoic acid. Child en’s cance g oup. N Engl J Med
341:1165–1173
91. Pea son AD, Pinke on CR, Lewis IJ e al (2008) High-dose
apid and s anda d induc ion chemo he apy o pa ien s aged
o e 1 yea wi h s age 4 neu oblas oma: a andomised ial.
Lance Oncol 9:247–256
92. Nakamu a K, Ma in KC, Jackson JK e al (2006) B ain-de i ed
neu o ophic ac o ac i a ion o T kB induces ascula
endo helial g ow h ac o exp ession ia hypoxia-inducible
ac o -1alpha in neu oblas oma cells. Cance Res 66:4249–4255
93. Acheson A, Cono e JC, Fandl JP e al (1995) A BDNF au o-
c ine loop in adul senso y neu ons p e en s cell dea h. Na u e
374:450–453
94. Jaboin J, Kim CJ, Kaplan DR e al (2002) B ain-de i ed neu-
o ophic ac o ac i a ion o T kB p o ec s neu oblas oma cells
om chemo he apy-induced apop osis ia phospha idylinosi ol
3’-kinase pa hway. Cance Res 62:6756–6763
95. Ma sumo o K, Wada RK, Yamashi o JM e al (1995) Exp ession
o b ain-de i ed neu o ophic ac o and p145T kB a ec s su -
i al, di e en ia ion, and in asi eness o human neu oblas oma
cells. Cance Res 55:1798–1806
96. Nakagawa a A, Aza CG, Sca a da NJ e al (1994) Exp ession
and unc ion o TRK-B and BDNF in human neu oblas omas.
Mol Cell Biol 14:759–767
97. Nishida Y, Ada i N, Ozawa R e al (2008) Iden i ica ion and
classi ica ion o genes egula ed by phospha idylinosi ol 3-ki-
nase- and TRKB-media ed signalling pa hways du ing neu onal
di e en ia ion in wo sub ypes o he human neu oblas oma cell
line SH-SY5Y. BMC Res. No es 1:95-0500-1-95
98. Takahashi K, Piao S, Yama ani H e al (2011) Es ogen induces
neu i e ou g ow h ia ho amily GTPases in neu oblas oma
cells. Mol Cell Neu osci 48:217–224
99. Muley PD, McNeill EM, Ma zinke MA e al (2008) The a RA-
esponsi e gene neu on na iga o 2 unc ions in neu i e ou -
g ow h and axonal elonga ion. De Neu obiol 68:1441–1453
100. F ye RH, Kaplan DR, K ome LF (1997) T unca ed kB
ecep o s on nonneu onal cells inhibi BDNF-induced neu i e
ou g ow h in i o. Exp Neu ol 148:616–627
101. Messi E, Flo ian MC, Caccia C e al (2008) Re inoic acid
educes human neu oblas oma cell mig a ion and in asi eness:
e ec s on DCX, LIS1, neu o ilamen s-68 and imen in exp es-
sion. BMC Cance 8:30-2407-8-30
102. Rebhan M, Vacun G, Bay eu he K e al (1994) Al e ed gan-
glioside exp ession by SH-SY5Y cells upon e inoic acid-in-
duced neu onal di e en ia ion. Neu oRepo 5:941–944
103. Hu R, Cai WQ, Wu XG e al (2007) As ocy e-de i ed es ogen
enhances synapse o ma ion and synap ic ansmission be ween
cul u ed neona al a co ical neu ons. Neu oscience 144:1229–1240
104. Sodhi RK, Singh N (2014) Re inoids as po en ial a ge s o
Alzheime ’s disease. Pha macol Biochem Beha 120:117–123
105. P iege FW (2003) Role o choles e ol in synapse o ma ion
and unc ion. Biochim Biophys Ac a 1610:271–280
106. Deu sch JW, Kelly RB (1981) Lipids o synap ic esicles: ele ance
o he mechanism o memb ane usion. Biochemis y 20:378–385
107. Line i A, F a angeli A, Ta e na E e al (2010) Choles e ol
educ ion impai s exocy osis o synap ic esicles. J Cell Sci
123:595–605
108. Smi h AJ, Sugi a S, Cha l on MP (2010) Choles e ol-dependen
kinase ac i i y egula es ansmi e elease om ce ebella
synapses. J Neu osci 30:6116–6121
109. Renne M, Choque D, T ille A (2009) Con ol o he pos sy-
nap ic memb ane iscosi y. J Neu osci 29:2926–2937
110. Ao o J, Nam CI, Poon MM e al (2008) Synap ic signaling by
all- ans e inoic acid in homeos a ic synap ic plas ici y. Neu on
60:308–320
746 Neu ochem Res (2016) 41:731–747
123

111. Da is GW (2006) Homeos a ic con ol o neu al ac i i y: om
phenomenology o molecula design. Annu Re Neu osci
29:307–323
112. Rich MM, Wenne P (2007) Sensing and exp essing homeos a ic
synap ic plas ici y. T ends Neu osci 30:119–125
113. K ama EA, Babayan AH, Gall CM e al (2013) Es ogen p o-
mo es lea ning- ela ed plas ici y by modi ying he synap ic
cy oskele on. Neu oscience 239:3–16
114. Fes e L, Zhou L, Bu ow A e al (2009) Choles e ol-p omo ed
synap ogenesis equi es he con e sion o choles e ol o es a-
diol in he hippocampus. Hippocampus 19:692–705
115. Nilsen J, Mo G, Na olin F (1998) Raloxi ene induces neu i e
ou g ow h in es ogen ecep o posi i e PC12 cells. Menopause
5:211–216
116. Chen G, Li HM, Chen YR e al (2007) Dec eased es adiol
elease om as ocy es con ibu es o he neu odegene a ion in a
mouse model o niemann-pick disease ype C. Glia 55:1509–
1518
Neu ochem Res (2016) 41:731–747 747
123