ARTICLE OPEN
Synap ic ulne abili y o amyloid-βand au pa hologies
di e en ially dis up s emo ional and memo y neu al ci cui s
Ma ia Dolo es Capilla-López
1,2
, Angel Dep ada
1,2
, Yuniesky And ade-Tala e a
3
, I ene Ma ínez-Gallego
3
,
He ibe o Coa l-Cuaya
3
, Paula So illo
1,2
, José Rod íguez-Al a ez
1,2
, An onio Rod íguez-Mo eno
3
,
A naldo Pa a-Damas
1,2
✉and Ca los A. Sau a
1,2
✉
© The Au ho (s) 2025
Alzheime ’s disease (AD) is cha ac e ized by memo y loss and neu opsychia ic symp oms associa ed wi h ce eb al amyloid-β(Aβ)
and au pa hologies, bu whe he and how hese ac o s di e en ially dis up neu al ci cui s emains unclea . He e, we in es iga ed
he ulne abili y o memo y and emo ional ci cui s o Aβand au pa hologies in mice exp essing mu an human amyloid p ecu so
p o ein (APP), Tau o bo h APP/Tau in exci a o y neu ons. APP/Tau mice de elop age- and sex-dependen Aβand phospho yla ed
au pa hologies, he la e exace ba ed a ea ly s ages, in ulne able b ain egions. Ea ly memo y defici s we e associa ed wi h
hippocampal au pa hology in Tau and APP/Tau mice, whe eas anxie y and ea appea ed linked o in acellula Aβin he
basola e al amygdala (BLA) o APP and APP/Tau mice. T ansc ip ome hippocampal p ofiling e ealed gene changes a ec ing
myelina ion and RNA p ocessing in Tau mice, and inflamma ion and synap ic- ela ed pa hways in APP/Tau mice a 6 mon hs. A
9 mon hs, we de ec ed common and egion-specific changes in as ocy ic, mic oglia and 63 AD-associa ed genes in he
hippocampus and BLA o APP/Tau mice. Spa ial lea ning defici s we e associa ed wi h synap ic au accumula ion and synapse
dis up ion in he hippocampus o Tau and APP/Tau mice, whe eas emo ional dis u bances we e linked o Aβpa hology bu no
synap ic au in he BLA. In e es ingly, Aβand au exhibi ed syne gis ic de imen al e ec s in long- e m po en ia ion (LTP) in he
hippocampus bu hey coun e ac wi h each o he o mi iga e LTP impai men s in he amygdala. These findings indica e ha Aβ
and au pa hologies cause egion-specific e ec s and syne gize o induce synap ic dys unc ion and immune esponses,
con ibu ing o he di e ing ulne abili y o memo y and emo ional neu al ci cui s in AD.
Molecula Psychia y; h ps://doi.o g/10.1038/s41380-025-02901-9
INTRODUCTION
Alzheime ’s disease (AD), he majo cause o memo y loss in he
elde ly, is accompanied by ea ly neu opsychia ic symp oms ha
ac as ea ly isk ac o s o con e sion o demen ia [1,2]. Cogni i e
and emo ional dis u bances a e associa ed wi h pa hological
changes in he hippocampus and amygdala a ea ly AD s ages [3],
bu how hese b ain egions coope a e o cause emo ional- ela ed
memo y changes in AD is unclea . Age and sex a ec di e en ially
egional accumula ion o amyloid-β(Aβ) and au pa hologies, and
cogni i e and mood dis u bances (dep ession, anxie y, ea ,
apa hy) [4–7]. In AD mouse models, emales wi h enhanced
neu opa hology a e fi s a ec ed by anxie y and cogni i e defici s
[8–14]. Young 3xTg-AD emales show highe anxie y and educed
memo y e en ion compa ed o males [13], whe eas sexual
dimo phism in beha io , neu opa hology and inflamma o y
molecules a e also e iden in a auopa hy mouse model [14].
Howe e , how Aβand au in e ac o cause cogni i e and
neu opsychia ic symp oms emain unclea , in pa because
emo ional ac o s a e la gely unde ep esen ed in basic, pa holo-
gical and clinical s udies [15].
Recen e idence in human demons a es dissocia ion e ec s
bu also a c oss alk o pa hological au- and amyloid- ela ed
neu al ci cui dysconnec i i y in memo y impai men s o AD
pa ien s [16–21]. Spa io empo al pa hological changes o au and
Aβa e linked wi h cogni i e decline in aging, and hey cause
selec i e egional ulne abili y associa ed wi h di e en ial gene
p ofiles and cogni i e dys unc ion in AD [22,23]. Se e al s udies
indica e ha Aβand/o amyloid plaques inc ease au phospho -
yla ion, agg ega ion and seeding [24–28], and au inac i a ion
amelio a es Aβ-induced memo y defici s independen ly o amy-
loid pa hology [29,30]. Impo an ly, Aβand au induce synapse
dys unc ion and loss [31], a pa hological ea u e ha igh ly
co ela es wi h cogni i e decline [32–34]. Synap ic au induces
synapse dys unc ion, ins abili y and loss [35–38], and i domina es
o e Aβon neu al ci cui dis up ion [39,40]. Aβand au a ec
bo h p esynap ic and pos synap ic mechanisms impai ing exci a-
o y glu ama e gic ansmission [35,41,42], bu whe he synap ic
au is esponsible and/o syne gize wi h Aβ o induce synapse
pa hology and beha io al changes is unclea . Mo eo e ,
he mechanisms o neu opa hological c oss alk be ween Aβ
Recei ed: 19 Ap il 2024 Re ised: 22 Decembe 2024 Accep ed: 16 Janua y 2025
1
Ins i u de Neu ociències, Depa men de Bioquímica i Biologia Molecula , Facul a de Medicina, Uni e si a Au ònoma de Ba celona, Bella e a, Ba celona, Spain.
2
Cen o de
In es igación Biomédica en Red En e medades Neu odegene a i as (CIBERNED), Mad id, Spain.
3
Depa men o Physiology, Ana omy and Cell Biology, Uni e sidad Pablo de
Ola ide, Se illa, Spain. ✉email: a nald[email p o ec ed]; ca los.sau [email protected]
www.na u e.com/mp
Molecula Psychia y
1234567890();,:
and au leading o synap ic dys unc ion and ulne abili y o
emo ional and memo y neu al ci cui s a e s ill unknown.
To be e elucida e he cellula mechanisms and ac o s
con ibu ing o dis up ion o cogni i e- and emo ion- ela ed neu al
ci cui s in AD, we gene a ed no el double APP/Tau ansgenic mice
ha ecapi ula e ea ly synap ic, beha io al, and ansc ip ional
al e a ions linked o AD pa hophysiology. In e es ingly, whe eas
spa ial lea ning and memo y defici s we e associa ed wi h synap ic
au pa hology in he hippocampus, emo ional dis u bances we e
linked o Aβin he basola e al amygdala (BLA). T ansc ip ional
p ofiling e ealed hippocampal gene signa u es en iched in
inflamma o y and synap ic pa hways a ec ed by ea ly concomi an
Aβ/ au pa hology bu no by Aβo au alone. A la e s ages, specific
and coo dina ed ansc ip ional esponses, including as ocy ic,
mic oglia and AD isk genes iden ified by genome-wide associa ion
s udies (GWAS) occu in he hippocampus and amygdala o APP/
Tau mice. Ou s udy e eals ha Aβand au a ec di e en ially
emo ional and memo y neu al ci cui s by exe ing dis inc e ec s a
he ansc ip ional, unc ional and beha io al le els.
MATERIALS AND METHODS
Beha io al, biochemical, ansc ip omic, immunohis iochemical and elec-
ophysiological me hods a e ex ensi ely desc ibed in Supplemen a y
In o ma ion.
Mice
Con ol (WT), APP, Tau and APP/Tau ansgenic mice we e ob ained by
c ossing he e ozygous APP
Sw,Ind
(line J9; C57BL/6) and Tau P301S (line PS19,
JAX #008169; C57BL/6) mice [36,43], and housed unde s anda d condi ions
(n=4–6/cage; 22 ± 2 °C, 12 h ligh :da k cycle). Sex, age and geno ypes a e
indica ed in he figu e legends. Expe imen al p ocedu es we e app o ed by
he Animal and Human E hical Commi ee (CEEAH) o he Uni e si a
Au ònoma de Ba celona and local go e nmen (CEEAH/DMAH: 2895/10571,
4750/10839) ollowing Eu opean Union egula ions (2010/63/EU).
Beha io al es s
Gene al and anxie y-like beha io s we e s udied in he open field and da k/ligh
box (DLB) es s [44,45] (see Supplemen a y In o ma ion). Fo cued ea
condi ioning (CFC), 6 mon h-old mice we e exposed in con ex A o a
condi ioned sound s imulus (CS, 2800 Hz, 80 dB; 30 s) ollowed by an elec ic
oo shock (uncondi ioned s imulus, US,0.8mA;2s).F eezingbeha io was
au oma ically eco ded (Video F eeze So wa e, Med Associa es) immedia ely
a e he shock (2 min) and again 24 h la e in a no el chambe (con ex B) be o e
(p e-CS; 2 min) and du ing (CS; 3 min) one p esen a ion [45]. Mice a
9–10 mon hs we e condi ioned wi h wo CS-US pai ings (con ex A; US, 1 mA;
2 s). F eezing beha io was eco ded du ing he 2-min in e al be ween pai ings
and immedia ely a e he shock. A 24 h, 3-min p e-CS pe iod was ollowed by
16 CS p esen a ions (30 s each, 5-s in e -CS) in con ex B [46].
Immunohis ochemis y
Depa a finized co onal b ain sec ions (5 μm) we e an igen- e ie ed wi h
ci a e bu e (10 mM, pH 6.0) o au o wi h o mic acid (60%, 5 min) o
Aβo Aβ/ au, and incuba ed wi h an i-phospho yla ed (p) au (Se 202/
Th 205, AT8, 1:50; Se 202, CP13, 1:50) o an i-Aβ(6E10, 1:1000) an ibodies
be o e bio in-conjuga ed an i-mouse seconda y an ibodies (1:200), DAB
pe oxidase s aining (Vec o labo a o ies) and imaging (Nikon Eclipse 80i
mic oscope). Fo as ocy ic and mic oglial s ainings, depa a fined co onal
sec ions (5 μm) we e an igen- e ie ed (ci a e bu e ) and incuba ed wi h
an i-GFAP (Dako Z0334; 1:500) o an i-Iba1 (Wako 019-19741, 1:250)
an ibodies ollowed by AlexaFluo -488/594-conjuga ed goa IgGs (1:400)
and Hoechs (1:5000) be o e imaging (Con ocal Zeiss LSM 700 mic oscope).
Immunofluo escence s aining wi h APP/Aβ, an i-Aβ42, pTau and neu onal
ma ke s, Iba1 and GFAP quan ifica ion [47] and b ain a ophy and
expansion me hods a e desc ibed in Supplemen a y In o ma ion.
Bulk RNA ansc ip ional p ofiling
Bulk RNA-sequencing (RNA-seq) o hippocampus and BLA om 6 and 9
mon h-old emales was pe o med on an Illumina Nex Seq 6000/2000.
RNA-seq da a analyses, including alignmen o he e e ence genome and
di e en ial exp ession analysis we e pe o med using QuasR/Rhisa 2 and
DESeq2 packages, and Gene on ology (GO) and unc ional en ichmen
analyses using en ichR in Bioconduc o [48–50].
S a is ical analysis
S a is ical analysis was pe o med using pa ame ic one- o wo-way
Analysis o Va iance (ANOVA) o non-pa ame ic K uskal-Wallis es s
acco ding o D’Agos ino-Pea son omnibus no mali y es (P ism so wa e,
G aphPad 8.0.2). Fo mul iple compa isons, we used Sidak’s o Tukey’s pos
hocs o pa ame ic es s and Dunn’s pos hoc o non-pa ame ic es .
Pa ame ic unpai ed wo- ailed S uden ’s - es o non-pa ame ic Mann-
Whi ney es , acco ding o D’Agos ino-Pea son omnibus no mali y es ,
we e used when wo g oups we e compa ed. P alues less han 0.05 we e
conside ed significan . Adjus ed P alues (padj) less han 0.1 we e
conside ed significan in he ansc ip omic analysis. The sample size was
calcula ed using he “Powe and P ecision”so wa e o ensu e adequa e
powe while adhe ing o he 3Rs p inciple. Randomiza ion was pe o med
by an independen in es iga o who assigned animal codes and sample
o de , ensu ing unbiased alloca ion and blindness o he geno ypes du ing
da a acquisi ion. G ubbs’ es was used o iden i y ou lie s.
RESULTS
Aβand au accumula ion in exci a o y neu ons esul s in age-
dependen ce eb al AD pa hology
To in es iga e he specific con ibu ion o Aβand au on memo y and
emo ional dis u bances, we c ossed APP
Sw,Ind
and Tau P301S
ansgenic mice o gene a e double APP/Tau ansgenic mice ha
exp ess human APP and Tau ansla ed ansc ip s in hippocampal
exci a o y CaMKIIαbu no inhibi o y pa albumin neu ons (Supple-
men a y Fig. 1A). Immunofluo escence analysis e ealed high localiza-
ion o hAPP/Aβ(6E10) and o al/phospho yla ed (p) Tau in exci a o y
neu ons (CaMKIIα, Glu 1, L-glu ama e) bu no inhibi o y in e neu ons
(GAD-67, pa albumin, soma os a in) in APP/Tau hippocampus, and
some colocaliza ion o Aβwi h GAD-67 in he BLA (Supplemen a y
Fig. 1B). APP and APP/Tau mice a 6 mon hs show ele a ed human APP
(∼2 old, APP CTF an ibody;∼17 old, 6E10), APP α-CTFs and
in aneu onal Aβ42 (MOAB-2) bu unchanged APP β-CTFs (Supple-
men a y Fig. 2A, B). Tau was simila in Tau and APP/Tau mice (∼25 old)
bu phospho yla ed (p)Tau (Se 202/Th 205) le els we e significan ly
inc eased in APP/Tau mice (Supplemen a y Fig. 2A; P<0.01).
APP and APP/Tau mice show simila in acellula Aβs aining in CA1/
CA3 hippocampus wi h no s aining in he en o hinal co ex (EC) and
BLA a 6 mon hs (K uskal-Wallis es , CA1: P< 0.0001; CA3: P< 0.0001;
EC: P> 0.05; BLA: P> 0.05; Fig. 1A). A 9 mon hs, male and emale APP
and APP/Tau mice show simila numbe o Aβ-posi i e neu ons and
amyloid plaques in hippocampal and co ical egions, excep o an
inc ease o Aβ-posi i e neu ons in he EC o emales (Two-way ANOVA,
EC, geno ype e ec : F (3, 37) =12.78, P< 0.0001; sex e ec : F
(1, 37) =10.10, P< 0.01; in e ac ion: F (3, 37) =3.43, P< 0.05; Fig. 1B;
Supplemen a y Fig. 3). The numbe o pTau (Se 202/Th 205)-posi i e
neu ons was significan ly inc eased in APP/Tau mice a 6 mon hs
(K uskal-Wallis es , CA1: P<0.001;CA3:P< 0.001; one-way ANOVA, EC:
P< 0.001; BLA: P< 0.001; Fig. 1A).A 9mon hs,pTau(Se 202,CP13)was
p ominen ly inc eased in neu onal cell bodies and fibe s o hippocam-
pus and co ex o male and emale Tau and APP/Tau mice (∼20–40 old;
Fig. 1B, Supplemen a y Fig. 3), whe eas, compa ed o males, APP/Tau
emales show ele a ed pTau in CA3 (Two-way ANOVA, CA3, geno ype
e ec : F (3, 35) =43.57, P< 0.0001; sex e ec : F (1, 35) =7.87, P< 0.01)
(Fig. 1B). In e es ingly, APP/Tau mice de elop p ominen hippocampal
a ophy a 9 mon hs (P<0.01–0.0001; Supplemen a y Fig. 2C). These
esul s demons a e ha APP/Tau mice de elop age-dependen Aβand
au pa hologies wi h no majo sex di e ences.
Hippocampal-dependen lea ning/memo y defici s in young
Tau and APP/Tau mice
Analysis o gene al beha io in he open field e ealed no
significan di e ences among geno ypes a 6 mon hs in o al
M.D. Capilla-López e al.
2
Molecula Psychia y
a elled dis ance, ime inac i e and pe cen age o ime in he
cen e (C) and pe iphe y (P) a days 1 and 2 (Supplemen a y
Fig. 4A). DLB and cued ea condi ioning (CFC) es s e ealed
no significan changes in anxie y, neophobia, and ea memo y
among he ansgenic lines a his age (K uskal-Wallis es
and wo-way ANOVA, P> 0.05; Fig. 2A, B). In he Mo is
wa e maze (MWM), all g oups showed simila swimming
speeds (P> 0.05) and dec eased la encies du ing spa ial aining
(Two-way ANOVA, aining e ec : F (4, 390) =37.40, P< 0.0001),
al hough Tau and APP/Tau mice exhibi ed significan ly longe
la encies s a ing a day 2 (Geno ype e ec : F (3, 390) =25.95,
P< 0.0001; Fig. 2C, D). In he p obe ial, con ol mice displayed a
p e e ence o he a ge quad an (P< 0.05), whe eas APP, Tau
and APP/Tau mice showed educed a ge quad an occupancies
(Two-way ANOVA, in e ac ion e ec : F (3, 156) =5.86, P< 0.001)
(Fig. 2C, D). This esul sugges s ha spa ial lea ning defici s a e
associa ed wi h hippocampal au pa hology in Tau and APP/Tau
mice.
Fig. 1 Aβpo en ia es ea ly au pa hology in AD ulne able b ain egions in APP/Tau mice. Co onal b ain sec ions o con ol (WT), APP, Tau
and APP/Tau mice a 6 mon hs (A) and 9 mon hs (B) we e s ained wi h an i-human Aβ/APP (6E10; op) and pTau (Se 202, AT-8 (A) o CP13 (B);
bo om) an ibodies. Le images: Rep esen a i e low and high (inse s) magnified images o Aβ- and pTau-s ained neu ons in CA1 and CA3
hippocampus, en o hinal co ex (EC) and basola e al amygdala (BLA). Amyloid plaques in a 9 mon h-old APP/Tau mouse a e isualized in he
op igh inse o CA3 egion. Objec i e: 20×. Scale ba : 50 μm. Righ diag ams: Quan ifica ion o Aβ-posi i e cells in WT (whi e symbols), APP
( ed symbols), Tau (blue symbols), and APP/Tau (g ey symbols) mice. Da a ep esen mean numbe ± SEM o APP/Aβ-and pTau-posi i e
cells/μm
2
(×10
−5
)(n=3–4 slices/mouse). Numbe o mice (male/ emale), 6 mon hs: WT (2/6), APP (1/5), Tau (0/5) and APP/Tau (3/5); 9 mon hs:
WT (6/6), APP (5/6), Tau (5/5) and APP/Tau (4-5/6-7). S a is ical analysis was pe o med using one-way ANOVA o non-pa ame ic K uskal-Wallis
es s acco ding o he D’Agos ino-Pea son omnibus no mali y es , ollowed by Tukey’s o Dunn’s pos hoc es , espec i ely (A), and wo-way
ANOVA ollowed by Sidak’s (sex compa ison) and Tukey’s (geno ypes compa ison wi hin each sex) mul iple compa ison es s (B).
*
P< 0.05,
**
P< 0.01,
***
P< 0.001,
****
P< 0.0001 s WT o he indica ed g oup.
M.D. Capilla-López e al.
3
Molecula Psychia y
Age-dependen emo ional dis u bances in APP and APP/
Tau mice
Gene al beha io e alua ed in he open field e ealed no significan
di e ences in a elled dis ances in all male and emale g oups a
9–10 mon hs, excep o inc eased ac i i y on day 1 in APP/Tau
emales, indica ing simila ambula o y locomo o ac i i ies (Supple-
men a y Fig. 4B). In he DLB es , 9 mon h-old APP and APP/Tau mice
o bo h sexes, bu no Tau mice, showed educed en ies (One-way
ANOVA, male: F (3, 36) =6.88, P< 0.001; emales: F (3, 29) =5.16,
P< 0.01), and highe la encies (excep male APP/Tau mice; male: F
(3, 36) =2.98, P< 0.05; emales: F (329) =4.47, P< 0.05) in o he
ligh compa men (Fig. 2E, G), sugges ing inc eased anxie y in APP
and APP/Tau mice. In CFC, inna e eezing esponses be o e shock
(neophobia) and immedia ely a e shock a e simila in emale
g oups (Two-way ANOVA, geno ype e ec : P> 0.05), whe eas males
show a main significan e ec o geno ype (F (3, 108) =5.23,
P< 0.01) and ea men (F (2, 108) =15.70, P< 0.0001) (Fig. 2F, H). A
24 h, all g oups showed simila and significan CS-induced
eezing esponses indica ing ea memo y consolida ion.
Howe e , he e was a significan e ec o geno ype and one
Fig. 2 Di e en ial Aβand au e ec s on spa ial memo y and emo ional dis u bances in AD ansgenic mice. A–DBeha io o AD
ansgenic mice a 6 mon hs. ANumbe o en ies and la encies o ligh zone in he DLB es . BAssocia i e memo y in he ea condi ioning
es . F eezing esponses in he CFC du ing condi ioning (day 1, le ) and aining (day 2, igh ) o WT, APP, Tau and APP/Tau mice. CSpa ial
memo y aining in he MWM o 5 days. D% Time in a ge quad an s o he s on day 5 p obe ial in he MWM. Hea maps show dec eased
occupancies and ajec o ies o ansgenic mice in he a ge quad an (ma ked in whi e). Da a ep esen mean ± SEM. Fo A,Bnumbe o
mice (male/ emale): WT (4/6), APP (5/5), Tau (2/8) and APP/Tau (4-5/4). Fo C,Dnumbe o mice (male/ emale): WT (12/15), APP (11/6), Tau (7/8) and
APP/Tau (9/14). E–HBeha io o AD ansgenic mice a 9 mon hs. E,GAnxie y beha io o male (E) and emale (G) mice in he da k/ligh es . F,H
Associa i e memo ies o male (F) and emale (H) mice in he CFC es . F eezing esponses du ing condi ioning (day 1) and aining (CS 1:
condi ioned s imulus 1) o ex inc ion (CS 2–8) on day 2 in he CFC. The mean ± SEM o ime eezing (%) du ing CS (pe iod om 2 o 8) is
ep esen ed a he igh . In all cases, da a ep esen mean ± SEM. Numbe o mice (male/ emale) o E–H: WT (14/8, whi e ba s/symbols) APP (9/8,
ed ba s/symbols), Tau (10/9, blue ba s/symbols) and APP/Tau (7/8, g ey ba s/symbols). S a is ical analysis was de e mined by one-way o non-
pa ame ic K uskal-Wallis es s ollowed by Tukey’so Dunn’s pos hoc es (A,E,G) o wo-way ANOVA (B–D,F,H) ollowed by Tukey’s pos hoc
es o geno ype compa isons.
*
P< 0.05,
**
P< 0.01,
***
P< 0.001,
****
P< 0.0001 s he indica ed g oup o con ol mice (C,F,H).
M.D. Capilla-López e al.
4
Molecula Psychia y
du ing ex inc ion aining (males: geno ype e ec : F (3, 324) =9.57,
P< 0.0001; one e ec : F (8, 324) =9.51, P< 0.0001; emales:
geno ype e ec : F (3, 306) =12.56, P< 0.0001; one e ec :
F (8, 306) =11.48, P< 0.0001) (Fig. 2F, H). Pa icula ly, APP and
APP/Tau mice o bo h sexes show enhanced eezing esponses
compa ed o con ol and Tau mice (One-way ANOVA, male: F
(3, 24) =7.00, P< 0.01; emales: F (3, 24) =12.51, P< 0.0001)
(Fig. 2F, H). These esul s indica e ha ea - ela ed emo ional
symp oms a e associa ed wi h Aβaccumula ion in he amygdala o
APP and APP/Tau mice.
M.D. Capilla-López e al.
5
Molecula Psychia y
Ea ly Aβ/ au-induced ansc ip ional esponses in he
hippocampus a e ela ed o synapse dis up ion and
inflamma ion
To iden i y ansc ip ional esponses ela ed o biological pa h-
ways a ec ed by ea ly Aβand au pa hologies, we pe o med
genome-wide bulk RNAseq analysis ollowed by pa hway en ich-
men analysis in he hippocampus o li e ma e emale WT, APP,
Tau and APP/Tau mice a 6 mon hs. Gene exp ession analysis
e ealed di e en ially exp essed genes (DEGs) in he hippocam-
pus o APP (94: 39 up, 55 down), Tau (1551: 743 up, 808 down) and
APP/Tau (1129: 808 up, 321 down) mice (Fig. 3A). In APP mice, we
ound significan up egula ion o genes ela ed o me abolic
p ocesses (Idh1,Ug 2b34) and oxida i e s ess (P dx1,Gs m1), he
la e sha ed wi h APP/Tau mice bu no Tau mice. GO analysis
e ealed ha he op significan ly en iched pa hways o he
up egula ed genes we e ela ed o gliogenesis and axon
enshea hmen /myelina ion in Tau mice (Bmp2,Dlx2,Cs 7,I gax,
Opalin, Tenm4), and immune esponses in APP/Tau mice (Ai 1,C3,
Cd9,Clu, Cnp, Cxcl5, Sox10). By con as , down egula ed genes
we e ela ed o RNA p ocessing/splicing in Tau mice (Cdk9,Cel 3,
Cel 5, Me l3, S s 7, Ybx2), and synapse o ganiza ion, anspo e s
and ecep o s in APP/Tau mice (Bdn ,Cacnb3,Fg 13,Dlg4,Neg 1,
Np x ). The common up egula ed (178) and down egula ed (54)
genes sha ed be ween Tau and APP/Tau mice we e ela ed o
myelina ion/glial di e en ia ion and synapse o ganiza ion/ ans-
mission pa hways, espec i ely (no shown). Consis en ly, cellula
pa hways ela ed o immune/leukocy e esponses (up egula ion)
and synapse unc ion/ion anspo e s/neu o ansmi e ecep o s
(down egula ion) we e al e ed in APP/Tau hippocampus a 6 and
9 mon hs o age (Supplemen a y Fig. 5A). These esul s indica e
ha Aβand au coope a e o induce synap ic dis up ion and
immune esponses a he ansc ip ional le el.
We nex examined p e- and pos -synap ic p o eins and au in
pu ified hippocampal synap osomes and/o b ain sec ions o 6
mon h-old mice. Biochemical analysis e ealed no majo changes
in synap ic p o eins in hippocampal lysa es o ansgenic mice
despi e ∼45% o APP/Tau mice showed dec ease o hese p o eins
(Fig. 3B). In pu ified hippocampal synap osomes, we ound
dec ease o Home -1 in APP mice, and synap ophysin, Home -1
and PSD95 in Tau mice, whe eas all analyzed synap ic p o eins
and β-ac in we e dec eased in APP/Tau mice (Fig. 3B). Tissue
expansion e ealed high colocaliza ion o pTau (Se 202) wi h he
glu ama e gic pos synap ic p o ein Home 1 in he hippocampus
(53%) bu no in he BLA (15%) o APP/Tau mice (Fig. 3C). Changes
in p e- and pos -synap ic p o eins occu in pa allel wi h synap ic
au accumula ion (Fig. 3B), which sugges s ha synap ic
pa hological au con ibu es o hippocampal synapse pa hology
in APP/Tau mice.
Tau and Aβcoope a e o induce di e en ial b ain egional
e ec s in synap ic plas ici y
We nex in es iga ed he unc ional e ec s o neu onal amyloid
and au pa hologies on hippocampal and amygdala synap ic
ansmission and plas ici y. Inpu -ou pu cu es in Scha e
colla e al CA3/CA1 synapses we e simila ly dec eased in all
mu an ansgenic emale mice a 6 mon hs (Two-way ANOVA,
s imulus e ec : F (8, 180) =19.90, P< 0.0001; geno ype e ec : F
(3, 180) =34.34, P< 0.0001; in e ac ion e ec : F (24,180) =2.24,
P< 0.01; Fig. 4A). Sho - e m synap ic plas ici y (STP), and long-
e m po en ia ion (LTP) induc ion, measu ed as ea ly-LTP (60 min)
and la e-LTP (120 min) we e significan ly impai ed in APP/Tau
hippocampus compa ed wi h he es o g oups (One-way ANOVA,
STP: F (3, 24) =4.06, P< 0.05; E-LTP: F (3, 24) =3.41, P< 0.05; L-LTP:
F (3, 24) =7.00, P< 0.01; Fig. 4A–C). Pai ed-pulse acili a ion (PPF)
a io was educed in APP and APP/Tau mice (F (3, 21) =5.37,
P< 0.01; Fig. 4D). Whole-cell pa ch-clamp eco dings e ealed
impai ed NMDA esponses and NMDA/AMPA a io in APP/Tau
mice (F (3, 36) =3.44, P< 0.01; Fig. 4E). These esul s demons a e
ha au and Aβcoope a e o dis up hippocampal synap ic
plas ici y h ough pos synap ic mechanisms.
In halamic-LA synapses, basal synap ic ansmission was simila ly
dec eased in APP, Tau and APP/Tau mice a 6 mon hs (Two-way
ANOVA, s imulus e ec : F (12,481) =3.45, P< 0.0001; geno ype
e ec : F (3, 481) =10.11, P< 0.0001; Fig. 4F). Field eco dings
e ealed dec eased STP induc ion in APP, Tau and APP/Tau mice
(F (3, 34) =9.01, P< 0.001), and E-LTP and L-LTP defici s in APP and
Tau mice bu no in APP/Tau mice (E-LTP: F (3, 34) =7.61, P<0.001;
L-LTP: F (3, 34) =11.29, P< 0.0001; Fig. 4G, H). Basal PPF was
educed in APP mice and eco e ed in APP/Tau mice (F
(3, 36) =5.13, P< 0.01; Fig. 4I). NMDA/AMPA a ios we e no
significan ly a ec ed in he ansgenic g oups (F (3, 40) =0.41,
P> 0.05; Fig. 4J). Toge he , hese findings indica e ha , in con as
o he hippocampus, Aβand au coun e ac each o he o main ain
synap ic plas ici y in he amygdala o young mice.
Coo dina ed and di e gen ansc ip ional changes ela ed o
inflamma ion and AD-associa ed genes occu in he
hippocampus and BLA o APP/Tau mice
To iden i y gene signa u es and ne wo ks di e en ially a ec ed by
la e Aβ/ au pa hology in he hippocampus and BLA, we nex
assessed DEGs o emale WT and APP/Tau mice a 9 mon hs. We
iden ified 5985 hippocampal genes (3260 up, 2725 down) and
2168 BLA genes (1263 up, 905 down) de egula ed in APP/Tau
mice, wi h a high numbe o DEGs (1423: 24% hippocampus, 65%
BLA) coo dina ely de egula ed in bo h egions (Fig. 5A, B). The
mos significan en iched pa hways a ec ed in bo h egions we e
ela ed o neu ophil immune esponse (up egula ion: A hgap9,
Cd33, Cd68, I gam, Lamp1, N am1, S k10, Ticam2) and neu o ans-
mission/glu ama e ecep o (down egula ion: Camk2a,Dlgap1/4,
G ia2/3,G in1, Home 1, L k2, Mapk8ip2, Mink1, Shank1, Syn1)
(Fig. 5C). By con as , he mos significan ly di e gen en iched
pa hways o up egula ed genes in APP/Tau mice we e a y acid
me abolism (Echs1,Eci1,Echdc3,Hadha,Hadhb,Hsd17b4/10) in he
hippocampus, and Aβbinding, GTPase egula ion and oxido -
educ ase/NAD (Ad b2,Bdh2,Hspg2, I gb2,Nc 1/2/4, Rnls,Si 2,Si 5,
Tl 2) in he BLA (Fig. 5C; Table 1). In addi ion, genes highly
exp essed in eac i e as ocy es (G ap, ApoE) and disease-
associa ed mic oglia (DAM) (Ccl6,Cd33,Cs 1, Cs 7, Cx3c 1,I gax,
T em2,Ty obp) we e up egula ed in hippocampus and/o BLA o
APP/Tau mice (Supplemen a y Table 1). Acco dingly, GFAP and
Fig. 3 Di e en ial ansc ip ional p ofiles and synapse pa hology in he hippocampus o AD ansgenic mice. A Venn and gene on ology
diag ams illus a ing he di e en ially exp essed genes and en iched cellula pa hways al e ed in he hippocampus o 6 mon h-old emale
APP (n=8), Tau (n=8) and APP/Tau mice (n=10) compa ed wi h con ols (WT, n=11). BBiochemical analysis o p e- and pos -synap ic
p o eins and au (D1M9X an ibody) in hippocampal lysa es and pu ified synap osomes o WT, APP, Tau and/o APP/Tau mice a 6 mon hs
(n=8–11 mice/g oup). Da a ep esen mean le els ( old change) no malized o β- ubulin ± SEM. Numbe o mice (male/ emale): 11 WT (6/5,
whi e symbols), 8 APP (5/3, ed symbols), 8 Tau (5/3, blue symbols) and 9 APP/Tau (5/4, g ey symbols). CCon ocal mic oscope images o
clea ed expanded hippocampal and BLA sec ions (expansion ac o 3.63x) o 6 mon h-old WT and APP/Tau mice showing colocaliza ion
(yellow; quan ifica ion a he igh g aph) o pos synap ic Home 1 (g een) and pTau (Se 202; ed) in he CA1 hippocampus bu no in he BLA
o APP/Tau mice. Righ op inse s a e magnified images o he indica ed squa ed egions. Scale ba s: 10 μm and 36.30 μm (expanded).
S a is ics we e de e mined by one-way ANOVA o K uskal-Wallis es s ollowed by Dunne ’s o Dunn’s mul iple compa isons es s acco ding o
he no mali y es , espec i ely (B) o by wo- ailed S uden ’s - es (C).
*
P< 0.05,
**
P< 0.01,
***
P< 0.001 s non- ansgenic con ol g oup.
M.D. Capilla-López e al.
6
Molecula Psychia y
Fig. 4 B ain egional specific e ec s o au on Aβ-induced synap ic plas ici y impai men s. A, F Inpu -ou pu cu es showing EPSC
ampli udes (A) and EPSP slope (F) s he applied s imulus in ensi y in CA3/CA1 hippocampal (A) and halamus-la e al amygdala (LA) (F)
synapses o con ol (WT, whi e), APP ( ed), Tau (blue), and APP/Tau (g ey) mice. B, G T aces ( op) and ime cou se (bo om) o EPSPs be o e
and a e LTP induc ion in CA3-CA1 hippocampal (B) and halamus-LA synapses (G). C, H His og ams showing sho - e m synap ic plas ici y
(STP), ea ly long- e m po en ia ion (E-LTP), la e LTP (L-LTP) in CA3-CA1 hippocampal (C) and halamus-LA synapses (H). D, I Pai ed-pulse
acili a ion (PPF) does no change a e LTP induc ion in he hippocampus (D) o amygdala (I), and baseline PPFs o APP a e dec eased in he
hippocampus and amygdala compa ed o con ols. Howe e , PPF o APP/Tau mice is significan ly dec eased only in hippocampus. E, J NMDA/
AMPA a io was dec eased in hippocampal neu ons o APP/Tau mice compa ed o con ol, APP, and Tau mice (E). In amygdala (J), NMDA/AMPA
a io was no a ec ed in he ansgenic mice. Da a ep esen mean ± SEM om elec ophysiological eco dings in CA3-CA1 and halamic-LA
synapses o 6 mon h-old emale con ol (n=7–8), APP (n=5–6), Tau (n=7–8), and APP/Tau (n=6) mice. S a is ical analysis was de e mined
by one-way ANOVA o K uskal-Wallis es s acco ding o he D’Agos ino-Pea son omnibus no mali y es , ollowed by Tukey’s o Dunn’s pos
hoc es , espec i ely (D, E, I, J) o wo-way ANOVA (A, F) ollowed by Tukey’s pos -hoc es .
*
P< 0.05,
**
P< 0.01,
***
P< 0.001,
****
P< 0.0001 s he
indica ed g oup.
M.D. Capilla-López e al.
7
Molecula Psychia y
Iba1 s ainings we e significan ly ele a ed in he CA1/CA3
hippocampus and BLA o APP/Tau mice (One-way ANOVA,
F(3, 22) =3.51, P< 0.05; BLA: F(3, 22) =5.44, P< 0.01;
F(3, 20) =3.82, P< 0.05; BLA: F(3, 22) =4.10, P< 0.05) (Fig. 5E).
Impo an ly, 63 mouse o hologs o human AD-associa ed genes
p e iously iden ified in GWAS (e.g, APOE,BIN1,CD33,CLU,MS4A4A,
PICALM,PLCG2, PTK2B, SLC24A4, SORL1,TREM2,USP6NL) we e
de egula ed in he hippocampus and/o BLA o emale APP/Tau
mice (Fig. 5C; Table 1)[51,52]. In iguingly, AD-associa ed genes
ela ed o synapse unc ion and ion anspo (NCS1, NKAIN2,
M.D. Capilla-López e al.
8
Molecula Psychia y
PDE7B, SLC24A4, SLC4A8, TSPAN13) we e specifically down egu-
la ed in APP/Tau hippocampus a 9 mon hs. In addi ion, 23 o
hese genes we e dis inc ly al e ed in he hippocampus o APP,
Tau and APP/Tau mice a 6 mon hs (Supplemen a y Fig. 5B).
Rema kably, en mouse o hologs o hese human AD- ela ed
genes (AHNAK,ARHGAP20,BCL3,CELF1,CLU,INPP5D,NKAIN2,
PDE7B,STK32B, TREM2) we e sha ed and changed in he same
di ec ion in he hippocampus o APP/Tau mice a 6 and 9 mon hs
bu no in Tau mice (Fig. 5D; Supplemen a y Table 2). These esul s
sugges ha coo dina ed specific ansc ip ional esponses ela ed
o inflamma ion and synapse unc ion occu in esponse o Aβ/ au
pa hology in he hippocampus and BLA.
DISCUSSION
Recen compelling e idence indica es ha Aβand au pa hologies
exe syne gis ic e ec s on synap ic dys unc ion and memo y loss,
sugges ing ha he apeu ic app oaches a ge ing only one o hese
ac o s may no be su ficien o achie ing clinical benefi s [31,53].
Acco dingly, immuno he apy clinical ials a ge ing Aβo au showed
p omising bioma ke educ ions bu limi ed cogni i e benefi s
[54,55], likely because mul iple pa hological ac o s con ibu e o
he disease p ocess [56]. Disce ning he pa hological mechanisms o
Aβand au c oss alk is c i ical o de elop e ec i e AD he apies.
Impo an ly, unde s anding how Aβand au pa hologies a e
mechanis ically linked in specific b ain ci cui s may help o elucida e
hei ole in cogni i e and neu opsychia ic symp oms. He e, we show
ha APP/Tau mice eplica e he key pa hological ea u es o AD,
including ea ly in acellula Aβand po en ia ion o au pa hology,
synapse dys unc ion, inflamma ion, neu on loss and de egula ion o
disease- ela ed gene signa u es. Comp ehensi e analysis o single and
double ansgenic mice, indica e ha Aβenhances neu onal
pa hological au a ea ly disease s ages, coinciding wi h au-induced
dis up ion o synap ic unc ion, memo y, and al e ed ansc ip ional
p ofiles ela ed o inflamma ion and synap ic pa hways. Con e sely,
al e ed emo ional esponses we e linked o he p esence o Aβin he
amygdala a la e pa hological s ages.
Ou s udy e eals dis inc pa hological ulne abili y o memo y
and emo ional ci cui s, wi h he p esence o pa hological au in
hippocampal glu ama e gic neu ons linked o spa ial memo y
defici s, and Aβaccumula ion in he BLA associa ed wi h ea
emo ional dis u bances. Hippocampal au pa hology is a s ong
p edic o o memo y decline by disconnec ing neu al ne wo ks
[18,57], and syne gizes wi h Aβ o dis up hippocampal unc ion
and memo y pe o mance in olde adul s [16,17]. The egion-
specific suscep ibili y o pa hology may be exace ba ed by he
e ec o Aβon he p og ession o neu onal au pa hology, as
p e iously shown in double APP-V717I/Tau-P301L mice [58], which
can be media ed by cell- o non-cell-au onomous (e.g. mic oglia)
e ec s [59]. As epo ed in AD [32,33,41,60], synap ic au was
associa ed wi h educed glu ama e gic synap ic p o eins and
impai ed hippocampal-dependen lea ning in bo h Tau and APP/
Tau mice. This simila pheno ype occu s despi e significan
di e ences in ea ly hippocampal synap ic plas ici y, inflamma ion,
and dys egula ion o ansc ip ional p og ams, sugges ing ha
al e na i e cellula and molecula mechanisms con e ge o induce
hippocampal dys unc ion in Tau and APP/Tau mice. Based on ou
ansc ip omic analysis, hese al e na i e mechanisms may impli-
ca e molecula changes in RNA p ocessing/splicing and synap ic
genes/pa hways linked o demen ia and in ellec ual disabili y (e.g.
Bdn , Cacnb3,Dlg4,Fg 13…) in Tau and APP/Tau mice, espec i ely
[61–63].
The p esence o Aβpa hology in he EC and BLA coincided wi h
anxie y and emo ional dis u bances in 9–10 mon h-old APP and
APP/Tau mice, u he emphasizing he de imen al e ec s o Aβon
emo ional and mood beha io s in AD [45,64,65]. Amygdala
pa hology a ec s key cellula pa hways ela ed o inflamma ion,
neu o ansmission, Aβbinding, GTPase ac i i y, and oxido educ-
ase/NAD+ac i i y. Gi en he link among emo ional symp oms,
inflamma ion, and neu onal exci abili y, i is plausible ha
pha macological in e en ions aimed o enhance GABAe gic
neu o ansmission o o mi iga e inflamma ion and oxida i e s ess
could ep esen p omising he apeu ic s a egies o neu opsychia-
ic symp oms in demen ia [45,64]. No ably, au coun e ac ed he
Aβ-induced synap ic ansmission and plas ici y defici s in he
amygdala a ea ly s ages. The mechanisms unde lying au-
media ed p o ec ion o synap ic unc ion emain unknown;
howe e , i is in iguing ha his coincides wi h he absence o
synap ic au in his egion. As he pa hology ad ances, defici s in
his neu al ci cui y may a ise alongside dys egula ion o glu ama-
e gic synap ic genes, as obse ed in APP knock-in mice [66].
No ably, emale APP/Tau mice a 9 mon hs exhibi ed a egion-
specificinc easeo Aβin EC and pTau in CA3 compa ed o males.
This finding is pa icula ly significan conside ing ha women a e
mo e suscep ible o emo ional dis u bances and expe ience as e
cogni i e decline. The molecula ac o s d i ing sex-specific
di e ences in emo ional and memo y changes emain poo ly
unde s ood and wa an u he in es iga ion. A po en ial explana-
ion may be he dis inc cellula ulne abili y o women o
pa hological changes, including Aβand/o neu ofib illa y angles
[67–70].
The empo al and egional associa ions o Aβand au
pa hologies wi h specific beha io al changes do no exclude he
possibili y ha bo h ac o s may con ibu e o memo y and
emo ional decline [17]. Besides he s ong co ela ion be ween
ce eb al Aβdeposi ion and anxie y in elde ly non-demen ed
indi iduals [71,72], au and Aβ/ au pa hologies a e closely
associa ed wi h neu opsychia ic symp oms in demen ia [73,74].
Conside ing ha anxie y is e e sed by an i- au he apeu ic
ea men s [75–77], he appa en lack o au e ec s on emo ional
symp oms could be explained by dominan e ec s o Aβo e au,
di e ences in cell-specific ansgene exp ession and/o age-
dependen pa hological di e ences in APP/Tau mice. In his
con ex , gene p ofiling analysis iden ified c i ical genes ela ed o
inflamma ion and synap ic unc ion in wo ulne able b ain
egions in AD, which is pa icula ly ele an conside ing he s ong
co ela ion be ween inflamma o y and synap ic genes and he
p og ession o AD [78–80]. None heless, ou ansc ip omic
Fig. 5 Bulk RNAseq e eals di e en ial and common gene exp ession signa u es associa ed wi h AD in hippocampus and BLA o APP/
Tau mice. A Numbe and Venn diag am illus a ing he o al, up and down di e en ially exp essed ansc ip s in he hippocampus (HPC) and
BLA o 9 mon h-old emale APP/Tau mice (n=9) compa ed o con ols (WT, n=11 o HPC; n=10 o BLA). BVolcano plo s showing he old
change o genes di e en ially exp essed in he hippocampus and BLA o APP/Tau mice ( cu -o : P alue < 0.01, log2FC > 0.5). Some
significan de egula ed AD isk genes a e indica ed (see also Table 1). CFunc ional se en ichmen o di e en ially exp essed genes in APP/Tau
hippocampus and BLA. The plo s show he op significan ly en iched pa hways om he GWAS ca alog (2019) (le ) and GO Biological P ocess/
Molecula Func ion (2021) ( igh ) da abases. Human o hologs we e used in he analysis wi h GWAS ca alog da abase. DVenn diag am
illus a ing he AD-associa ed genes iden ified p e iously by GWAS sha ed in he hippocampus o 6 and 9 mon h-old emale APP/Tau and/o
Tau mice. ECon ocal mic oscope images (le ) and quan ifica ion ( igh g aphs) o pe cen age o GFAP (magen a; op images) and Iba1
(magen a; bo om images) a eas, and Hoechs (blue) s aining in CA1 hippocampus and BLA sec ions o emale 9 mon h-old WT (n=7), APP
(n=7), Tau (n=4−5) and APP/Tau (n=6−7) mice. Inse s a e magnified images o he indica ed egions. Scale ba s: 30 μm. S a is ical analysis
was de e mined by one-way ANOVA ollowed by Tukey’s pos -hoc es .
*
P< 0.05,
#
P=0.06 s he indica ed con ol g oup.
M.D. Capilla-López e al.
9
Molecula Psychia y
