Upload
vannga
View
220
Download
0
Embed Size (px)
Citation preview
ARTICLEdoi:10.1038/nature09635
Reversing EphB2 depletion rescuescognitive functions in Alzheimer modelMoustapha Cisse1,2, Brian Halabisky1,2, Julie Harris1,2, Nino Devidze1, Dena B. Dubal1,2, Binggui Sun1,2, Anna Orr1,2, Gregor Lotz1,2,Daniel H. Kim1, Patricia Hamto1, Kaitlyn Ho1, Gui-Qiu Yu1 & Lennart Mucke1,2
Amyloid-b oligomers may cause cognitive deficits in Alzheimer’s disease by impairing neuronal NMDA-type glutamatereceptors, whose function is regulated by the receptor tyrosine kinase EphB2. Here we show that amyloid-b oligomersbind to the fibronectin repeats domain of EphB2 and trigger EphB2 degradation in the proteasome. To determine thepathogenic importance of EphB2 depletions in Alzheimer’s disease and related models, we used lentiviral constructs toreduce or increase neuronal expression of EphB2 in memory centres of the mouse brain. In nontransgenic mice,knockdown of EphB2 mediated by short hairpin RNA reduced NMDA receptor currents and impaired long-termpotentiation in the dentate gyrus, which are important for memory formation. Increasing EphB2 expression in thedentate gyrus of human amyloid precursor protein transgenic mice reversed deficits in NMDA receptor-dependentlong-term potentiation and memory impairments. Thus, depletion of EphB2 is critical in amyloid-b-inducedneuronal dysfunction. Increasing EphB2 levels or function could be beneficial in Alzheimer’s disease.
Soluble amyloid-b oligomers may contribute to learning and memorydeficits in Alzheimer’s disease by inhibiting NMDA-receptor-dependent long-term potentiation (LTP)1–3, thought to underliememory formation4. In Alzheimer’s disease, hippocampal NMDA-receptor-subunit levels are reduced5, and protein levels and the phos-phorylation status of NMDA-receptor subunits NR1, NR2A and NR2Bcorrelate with cognitive performance6. Human amyloid precursorprotein (hAPP) transgenic mice with high brain levels of amyloid-boligomers have reduced hippocampal levels of tyrosine-phosphorylatedNMDA receptors and key components of NMDA-receptor-dependentsignalling pathways7,8. Alzheimer’s disease patients and hAPP micehave hippocampal depletions of the receptor tyrosine kinase EphB29,which regulates NMDA-receptor trafficking and function by interact-ing with NMDA receptors and Src-mediated tyrosine phosphoryla-tion10–13. EphB2 regulates NMDA-receptor-dependent Ca21 influxand downstream transcription factors involved in LTP formation12,such as Fos, which is depleted in the dentate gyrus of hAPP mice.Mice lacking EphB210,14 or Fos15 have impaired NMDA-receptor-dependent LTP and memory deficits. We hypothesized that EphB2depletion in Alzheimer’s disease-related models is caused by amy-loid-b oligomers and that reductions in EphB2 contribute to amy-loid-b-induced deficits in synaptic plasticity and cognitive functions(Supplementary Fig. 1). Here we confirm these hypotheses and showthat reversing EphB2 depletion in the dentate gyrus of hAPP micereverses LTP and memory impairments.
Amyloid-b oligomers bind to EphB2To determine if amyloid-b oligomers interact directly with EphB2, wemeasured binding of biotinylated synthetic amyloid-b1–42 oligomersto a purified recombinant EphB2–Fc chimaera. Biotinylated amyloid-b oligomers and EphB2–Fc were pulled down together by avidinagarose beads (Supplementary Fig. 2a, b) and co-immunoprecipitatedunder cell-free conditions (Supplementary Fig. 2c, d). EphB2 andamyloid-b oligomers also co-immunoprecipitated from homogenatesof primary neurons (Supplementary Fig. 2e–g). Thus, amyloid-b oli-gomers may interact directly with the extracellular region of EphB2.
This region comprises a ligand-binding (LB) domain, a cysteine-rich (CR) domain, and a fibronectin type III repeats (FN) domain(Fig. 1a). To determine which domain mediates the interaction withamyloid-b oligomers, we generated EphB2–GST deletion mutantslacking the LB domain (DLB-EphB2) or the FN domain (DFN-EphB2) (Fig. 1a). Amyloid-b oligomers bound to FL-EphB2 andDLB-EphB2 but not DFN-EphB2 (Fig. 1b, c), indicating that the FNdomain is critical for their interaction with EphB2.
Deleting the FN domain did not affect trafficking of EphB2 to thecell surface (Supplementary Fig. 3a). FL-EphB2 andDFN-EphB2 bothphosphorylated the NMDA-receptor subunit NR1 after stimulationwith the EphB2 ligand, Fc-ephrin-B2 (Supplementary Fig. 3b–d).Thus, deleting the FN domain did not eliminate the kinase functionof EphB2. Deleting the LB domain prevented Fc-ephrin-B2-inducedphosphorylation of NR1 (Supplementary Fig. 3b–d).
Mechanisms of amyloid-b-induced EphB2 depletionAt 3–4 but not 2 months of age, EphB2 messenger RNA and protein levelsin the hippocampus were lower in hAPP mice than in nontransgenic con-trols, and were lower in humans with Alzheimer’s disease than in non-demented controls (data not shown), consistent with previous findings9.
As reported by others16, we observed a doublet of putative EphB2carboxy-terminal fragments (CTFs) of 45–50 kDa in hippocampi ofhAPP mice and nontransgenic controls on western blots (not shown).Relative to nontransgenic controls, hAPP mice showed a comparabledecrease in CTFs and FL-EphB2 (not shown) and no difference in theratio of CTF11CTF2:FL-EphB2 (hAPP, 2.7 6 0.36; nontransgenic,2.3 6 0.59; P 5 0.55 by t test). Thus, pathologically raised levels ofamyloid-b do not affect EphB2 cleavage into CTFs.
Treating primary neuronal cultures from wild-type rats with naturallysecreted amyloid-b oligomers caused severe EphB2 depletions by 3 days(Fig. 1d–f ). Amyloid-b oligomers reduced EphB2 mRNA levels(Fig. 1g), but the reduction was subtle and unlikely to account for thesevere EphB2 protein depletion.
Amyloid-b-induced depletion of EphB2 was blocked by the protea-some inhibitor lactacystin (Fig. 1h, i). Bafilomycin, an inhibitor of
1Gladstone Institute of Neurological Disease, San Francisco, California 94158, USA. 2Department of Neurology, University of California, San Francisco, California 94158, USA.
6 J A N U A R Y 2 0 1 1 | V O L 4 6 9 | N A T U R E | 4 7
Macmillan Publishers Limited. All rights reserved©2011
endosomal acidification, had no effect (Supplementary Fig. 4b, c). Com-pared with amyloid-b treatment alone, treatment of cells with lactacystinalone or together with amyloid-b increased ubiquitinated EphB2(Supplementary Fig. 4a). These results indicate that amyloid-b depletesneuronal EphB2 mainly by enhancing its proteasomal degradation.
EphB2 depletion impairs synaptic plasticityTo determine if EphB2 depletion interferes with NMDA-receptor-dependent functions, we generated lentiviral vectors expressing greenfluorescent protein (GFP) and anti-EphB2 shRNA (Lenti-sh-EphB2–GFP) or scrambled control shRNA (Lenti-sh-SCR–GFP). In neuronalcultures, Lenti-sh-EphB2–GFP reduced EphB2 mRNA and proteinlevels (Fig. 2a, b) and surface levels of NR1 (Fig. 2c–e). In cultures co-infected with a mutant EphB2 construct whose mRNA is resistant tosh-EphB2 (Lenti-mut-EphB2–Flag) and Lenti-sh-EphB2–GFP, EphB2and surface NR1 were not reduced (Supplementary Fig. 5), excludingan off-target effect. Next we examined the effects of sh-EphB2 onexpression of the immediate-early gene c-fos, which depends onNMDA receptors and is regulated by EphB212. Anti-EphB2 shRNAprevented Fc-ephrin-B2-induced increases in Fos expression in
neurons expressing wild-type EphB2, but not in neurons expressingmutant EphB2 (Fig. 2f ). Thus, depleting EphB2 reduces surface NR1expression and impairs NMDA-receptor-dependent gene expression.
To explore whether EphB2 depletion accounts for LTP deficits inhAPP mice8, we reduced EphB2 in the dentate gyrus of nontransgenicmice. Although granule cells are not very susceptible to degenerationin Alzheimer’s disease, perforant path to granule cell synapses areaffected early and severely17,18.
Two anti-EphB2 shRNAs reduced EphB2 mRNA and protein levelsin neuronal culture (Supplementary Fig. 6). Mice injected with lentiviralvectors expressing sh-EphB2-308–GFP (Fig. 3a, b) or sh-EphB2-306–GFP (Supplementary Fig. 7a, b) had lower EphB2 mRNA levels in the
FL-EphB2–GST
ΔLB-EphB2–GST
ΔFN-EphB2–GST
b
Sig
nal in
tensity (A
U)
00 0.5
Amyloid-β (μg ml–1)
1.0
5,000
10,000
15,000
ΔLB-EphB2–GSTΔFN-EphB2–GST
FL-EphB2–GST
Rela
tive E
phB
2 level
Amyloid-β:0
2,000
4,000
10,000
– – ++
Lactacystin: – + +–
*
*
**
**
Rela
tive E
phB
2m
RN
A level
0
2.0
Veh
g
2 64
Amyloid-β (days)
* *** ***
IP: anti-GSTWB: anti-amyloid-β
IP: anti-EphB2WB: anti-EphB2
Input
WB: anti-amyloid-β
Amyloid-β(μg):
IP: anti-GST WB: anti-GST
Lactacystin: – +
Amyloid-β: – –
– +
++
Veh 1Biotin: + +
2 3+– +
Amyloid-β (days)
Rela
tive t
ota
lE
phB
2 level
0
2,500
5,000
7,500
0.5
1.0
1.5
10,000
Veh
f
1 32
Amyloid-β (days)
*
**
****
*****
Rela
tive s
urf
ace
Ep
hB
2 level
0
5,000
10,000
15,000
Veh
e
c d
a
h i
1 32
Amyloid-β (days)
**
***
EphB2
DimerTrimer
Tubulin
EphB2
Surface
EphB2
Total
EphB2
DimerTrimer
pcDNA3FL
-EphB
2–
GST ΔLB
-EphB
2–
GST ΔFN
-EphB
2–
GST
– – 0.51.0 1.0 – 0.5 1.0 – 0.5 1.0
LB CR FN
TM
KD SAM PDZ
*
Figure 1 | Amyloid-b oligomers bind to the fibronectin repeats domain ofEphB2 and cause degradation of EphB2 in the proteasome. a, Domainstructure of full-length (FL) EphB2 and deletion constructs. Ligand-binding (LB)domain, cysteine-rich (CR) region, fibronectin type III repeats (FN) domain,transmembrane (TM) region, tyrosine kinase (KD) domain, sterile alpha motif(SAM) domain, and PSD95, DLG and ZO1 (PDZ) domain. b, Binding of amyloid-bdimers and trimers to different EphB2 constructs. See Supplementary Table 2 forexperimental details pertaining to data shown in figures. AU, arbitrary units.c, Representative western blot (WB). IP, immunoprecipitation. d–f, Amyloid-b-induced depletion of EphB2. Primary rat neurons were treated with amyloid-b orvehicle (Veh), and surface and total levels of EphB2 were determined by westernblots. Representative western blots are shown in d. e, f, Quantification of surface(e) and total (f) levels of EphB2. g, EphB2 mRNA levels in primary neurons treatedwith amyloid-b or vehicle. h, i, Lactacystin blocks amyloid-b-induced depletion ofEphB2 in primary neurons. Representative western blot (h) and quantification ofsignals (i). For all experiments, n 5 3–6 wells per condition from threeindependent experiments. *P , 0.05, **P , 0.001, ***P , 0.0001 versus emptybars or as indicated by brackets (Tukey test). Values are means 6 s.e.m.
a
c
b sh-SCR sh-EphB2
Rela
tive F
os level
0
2,500
5,000
7,500
10,000
***
***
*****
******
Rela
tive E
phB
2m
RN
A level
sh-EphB2:0
50
100
150
– + –sh-SCR:
sh-EphB2:
sh-SCR:
Surface NR1
Total NR1
– – +
– + +
– – –
–
+
*
***
d
f
Rela
tive s
urf
ace
NR
1 level
sh-EphB2:0
5,000
10,000
15,000
0
5,000
10,000
15,000
– – +
sh-SCR: – + –
sh-EphB2: –Fc-ephrin-B2: –
––
–+
++
––
–+
EphB2 mut-EphB2
++
***
e
Rela
tive t
ota
lN
R1 level
sh-EphB2: – – +
sh-SCR: – + –
**
GFP
EphB2
Merged
Figure 2 | Knockdown of EphB2 reduces surface NR1 levels and Fc-ephrin-B2-dependent Fos expression. a, b, EphB2 expression is reduced in primaryneurons infected with Lenti-sh-EphB2–GFP as determined by RT–qPCR (a) orEphB2 immunostaining (b). Scale bar, 20mm. c–e, Reduction of EphB2 levelsby Lenti-sh-EphB2–GFP and effect on surface NR1 levels. f, shRNA againstwild-type but not mutated EphB2 reduces Fc-ephrin-B2-dependent Fosexpression. Primary rat neurons were co-infected or not with Lenti-sh-EphB2–GFP (sh-EphB2) in combination with either Lenti-EphB2 encoding wild-typeEphB2 or Lenti-mut-EphB2 (mut-EphB2) encoding a mutated EphB2 mRNAthat is not recognized by sh-EphB2. Four days later, cells were stimulated withclustered multimeric recombinant Fc-ephrin-B2 ligand to activate EphB2.n 5 3–6 wells per condition from three independent experiments. *P , 0.05,**P , 0.001, ***P , 0.0001 versus empty bar or as indicated by brackets(Tukey’s test). Values are means 6 s.e.m.
RESEARCH ARTICLE
4 8 | N A T U R E | V O L 4 6 9 | 6 J A N U A R Y 2 0 1 1
Macmillan Publishers Limited. All rights reserved©2011
dentate gyrus than controls. Transduction efficiencies (SupplementaryFig. 8) were 50–74% (mean 6 s.e.m., 62.4 6 6.2; n 5 7 mice), consistentwith other reports19,20.
Field (Fig. 3c) and whole-cell patch-clamp recordings (Fig. 3e)from dentate gyrus granule cells in acute hippocampal slices fromLenti-sh-EphB2–GFP-injected nontransgenic mice revealed prom-inent LTP deficits similar to those in untreated hAPP J20 (Fig. 3d,f ) and other lines of hAPP mice21,22. Lenti-sh-SCR–GFP-injectednontransgenic mice had robust LTP in the dentate gyrus (Fig. 3c,e). Whole-cell recordings from individual GFP-negative granule cellsin Lenti-sh-ephB2–GFP-injected mice revealed no LTP deficits, com-pared with GFP-negative granule cells in untreated nontransgenicmice and GFP-positive granule cells in Lenti-sh-SCR–GFP-injectedmice (P . 0.1 by repeated-measures ANOVA, n 5 6 neurons from 3mice per group; data not shown).
EphB2 depletion reduces synaptic strengthLTP at the medial perforant path to granule cell synapses depends onNMDA-receptor activity23. We determined whether impaired synaptic
plasticity in sh-EphB2-treated nontransgenic and untreated hAPPmice is related to a selective impairment of these glutamate receptors.NMDA-receptor-mediated, but not a-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA)-receptor-mediated synaptic trans-mission strength at this synapse was affected in sh-EphB2-treatednontransgenic mice (Fig. 3g) and untreated hAPP mice (Fig. 3h), asdetermined by field recordings and analysis of input-output (I/O)curves. These alterations markedly reduced ratios of NMDA-receptor-to AMPA-receptor-mediated synaptic strength in both groups(Fig. 3j). Similar results were obtained by whole-cell recordings fromindividual granule cells (Fig. 3i, k). To exclude a contribution of altera-tions in AMPA-receptor currents to the altered ratios, we recordedpharmacologically isolated, AMPA-receptor-mediated miniatureexcitatory synaptic currents (mEPSCs). The four groups of mice hadcomparable mEPSC peak amplitudes (Supplementary Fig. 9). Thus,like amyloid-b, EphB2 depletion probably reduces LTP by impairingNMDA-receptor function.
EphB2 rescues synaptic functions in hAPP miceTo determine if increasing EphB2 expression in the dentate gyrus ofhAPP mice reverses LTP deficits, we used a lentivirus expressingEphB2–Flag (Lenti-EphB2–Flag). Lenti-EphB2–Flag-treated hAPPand nontransgenic mice had comparable EphB2–Flag expressionlevels in the dentate gyrus (Fig. 4a and Supplementary Fig. 10).Lenti-empty-treated nontransgenic mice and Lenti-EphB2–Flag-treated hAPP mice had comparable dentate gyrus levels of total(endogenous and exogenous) EphB2 (Fig. 4b), indicating thatEphB2 levels in hAPP mice were normalized. EphB2 levels were lowerin Lenti-empty-injected hAPP mice and higher in Lenti-EphB2–Flag-injected nontransgenic mice (Fig. 4b). Increasing dentate gyrusEphB2 levels reversed LTP deficits in two independent cohorts ofhAPP mice but did not alter LTP in nontransgenic mice (Fig. 4c).
Lenti-EphB2–Flag-treated mice showed a trend towards loweramyloid-b levels in the dentate gyrus (Supplementary Fig. 11), butthis trend did not reach statistical significance. At analysis, hAPP micewere 4–5-months old and had not yet formed plaques, excludingEphB2 effects on plaque formation. To determine if LTP rescue wasdue to improved NMDA-receptor function, we again measuredAMPA-receptor- and NMDA-receptor-mediated synaptic strength.Increasing EphB2 levels in the dentate gyrus of hAPP mice reverseddeficits in NMDA-receptor-mediated synaptic strength without
0
1.0
–15 0 15 30 45 60
1.2
1.4
1.6
1.8
Time (minutes)
No
rmaliz
ed
fE
PS
P s
lop
eE
vo
ked
EP
SC
peak a
mp
litud
e(n
orm
aliz
ed
)
Evo
ked
EP
SC
peak a
mp
litud
e(n
orm
aliz
ed
)
2.0
2.2
0
1.0
–15 0 15 30 45 60
1.2
1.4
1.6
1.8
Time (minutes)
No
rmaliz
ed
fE
PS
P s
lop
e
2.0
2.20.2 mV
5 ms
50pA
10 ms
50pA
10 ms
NTG: sh-EphB2NTG: sh-SCR
**
AMPAR NMDAR AMPAR NMDAR
NTG
: sh-
SCR
NTG
: sh-
EphB2
*** ***
NTG
: Unt
hAPP: U
nt
NTG
: sh-
SCR
NTG
: sh-
EphB2
NTG
: Unt
hAPP: U
nt0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
NM
DA
R:A
MP
AR
ba
c
g
d
e f
Rela
tive E
phB
2m
RN
A level
0
0.2
0.4
0.6
0.8
1.0
*
hAPP: UntNTG: Unt
NTG: sh-EphB2NTG: sh-SCR
hAPP: UntNTG: Unt
NTG
:
sh-S
CR N
TG:
sh-E
phB2
h
i j k
0.2 mV1 ms
0.2 mV
10 ms
00.6
0.2
0.4
0.6
0.8
1.0
1.2
1.4
0.4
Fibre volley (mV)
0.20 0.60.4
Fibre volley (mV)
0.20 0.60.4
Fibre volley (mV)
0.20 0.60.4
Fibre volley (mV)
0.20
NTG: sh-EphB2NTG: sh-SCR NTG: Unt
hAPP: Unt0
25
50
75
100
AM
PA
R f
EP
SP
slo
pe (V
ms
–1)
NM
DA
R f
EP
SP
slo
pe (m
V m
s–1)
0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
0
25
50
75
100
AM
PA
R f
EP
SP
slo
pe (V
ms
–1)
NM
DA
R f
EP
SP
slo
pe (m
V m
s–1)
0
0.5
1.0
1.5 *** ***
NM
DA
R:A
MP
AR
–15 0 15 30 45 600
0.5
1.0
1.5
2.0
2.5
3.0
Time (minutes)
–15 0 15 30 45 600
0.5
1.0
1.5
2.0
2.5
3.0
Time (minutes)
******
25 pA
20 ms
NTG: sh-SCR NTG: sh-EphB2
NTG: Unt hAPP: Unt
–80 mV+50 mV
–80 mV+50 mV
Figure 3 | Knockdown of EphB2 reduces LTP in dentate gyrus granule cellsof nontransgenic mice. a, Anti-GFP immunostaining of dentate gyrusshowing infected neurons in Lenti-sh-EphB2–GFP injected mice. Right panelshows higher magnification image of boxed region on left. Scale bars: 100mm(left), 25mm (right). b, EphB2 mRNA levels in the entire dentate gyrus(reflecting levels in infected and uninfected cells) (n 5 5–7 mice per condition).*P , 0.001 versus sh-SCR (t test). NTG, nontransgenic mice. c–f, LTP at themedial perforant path to granule cell synapse measured by field recordings(c, d) or by whole-cell patch clamp from individual GFP-positive cells (e, f ) inthe dentate gyrus. LTP was impaired in nontransgenic mice treated with Lenti-sh-EphB2–GFP (sh-EphB2) compared to nontransgenic mice treated withLenti-sh-SCR–GFP (sh-SCR) (c, e). Similar LTP impairments were observed inuntreated (Unt) hAPP mice (d, f ) (NTG: sh-EphB2 versus hAPP: Unt).*P , 0.05, ***P , 0.001 (repeated-measures ANOVA and Bonferroni post-hoc test on the last 10 min of data). n 5 8–9 slices from 3–4 mice per treatment(c) or genotype (d). g, h, Comparison of AMPA-receptor (AMPAR)-mediated(left) and NMDA-receptor (NMDAR)-mediated (right) input-output (I/O)relationships in the medial perforant path to granule cell synapses ofnontransgenic mice treated with sh-EphB2 versus sh-SCR (g) and of untreatednontransgenic (NTG: Unt) versus hAPP (hAPP: Unt) mice (h). i, Exampletraces of evoked glutamate receptor currents from individual granule cellsvoltage clamped at –80 or 50 mV to measure AMPA-receptor- and NMDA-receptor-mediated currents, respectively. j, k, Summary plot of the ratios ofNMDA-receptor I/O relationships to AMPA-receptor I/O relationshipsmeasured by field recordings (i) or by individual granule cells (k). ***P , 0.001(two-way ANOVA and Bonferroni post-hoc test). n 5 8–9 slices from 3–4 miceper group. Values are means 6 s.e.m.
ARTICLE RESEARCH
6 J A N U A R Y 2 0 1 1 | V O L 4 6 9 | N A T U R E | 4 9
Macmillan Publishers Limited. All rights reserved©2011
changing AMPA-receptor-mediated synaptic strength (Fig. 4d, e),normalizing the balance between them (Fig. 4f ). OverexpressingEphB2 did not alter NMDA-receptor- or AMPA-receptor-mediatedsynaptic strength in nontransgenic mice (Fig. 4d–f ).
Increasing EphB2 expression in granule cells did not reverseimpairments in paired pulse modification at perforant path to granulecell synapses (Supplementary Fig. 12a) or in synaptic strength atSchaffer collateral to CA1 pyramidal cell synapses (SupplementaryFig. 12b, c).
EphB2 ameliorates cognitive deficits in hAPP miceTo determine if increasing EphB2 levels in the dentate gyrus also reverseslearning and memory deficits in hAPP mice24–27, we injected Lenti-EphB2–Flag or Lenti-empty bilaterally into the dentate gyrus of hAPPand nontransgenic mice and analysed them behaviourally 2 months later.
Spatial learning and memory in the Morris water maze is stronglyaffected by dentate gyrus impairments28. In the spatial, hidden-platformcomponent, Lenti-EphB2–Flag-treated but not Lenti-empty-treatedhAPP mice performed at control levels (Fig. 5a, b). OverexpressingEphB2 did not alter learning in nontransgenic mice (Fig. 5a, b). Allgroups of mice learned similarly well in the cued-platform component(data not shown).
In a probe trial, Lenti-empty-treated but not Lenti-EphB2–Flag-treated hAPP mice took longer to reach the original platform locationthan Lenti-empty-treated nontransgenic controls (Fig. 5c). Lenti-EphB2–Flag-treated nontransgenic mice performed slightly worsethan Lenti-empty-treated nontransgenic mice (Fig. 5c) (P 5 1.0 byone-way ANOVA and Bonferroni post-hoc test).
In the novel object recognition test, Lenti-EphB2-treated but notLenti-empty-treated hAPP mice spent more time exploring the novelobject (Fig. 5d). In the novel place recognition test, Lenti-EphB2-treated but not Lenti-empty-treated hAPP mice spent more timeexploring the object whose location had changed (Fig. 5e). Thus,increasing EphB2 expression in the dentate gyrus of hAPP mice ame-liorates deficits in both spatial and nonspatial learning and memory.
Finally, we assessed passive avoidance learning, which depends atleast partly on hippocampal functions29,30. During training, escape
latencies were similar across groups (Fig. 5f ). However, 24 h later,Lenti-empty-treated hAPP mice were severely impaired, whereas allother groups performed well (Fig. 5f ). Increasing dentate gyrusEphB2 levels in hAPP mice did not reverse behavioural deficits thatwere probably caused by impairments of other brain regions, includ-ing hyperactivity in the open field and disinhibition in the elevatedplus maze (Supplementary Fig. 13).
DiscussionOur study shows that EphB2 depletion contributes to amyloid-b-induced neuronal deficits and cognitive dysfunction. Reducing neuronalEphB2 levels caused functional deficits similar to those caused by amy-loid-b, including deficits in NMDA-receptor-dependent synapticstrength and gene expression and impaired LTP and memory.Increasing neuronal EphB2 levels in hAPP mice reversed these deficits,indicating that EphB2 impairment is necessary and sufficient to elicitthem and that increasing EphB2 activity counteracts amyloid-b-inducedneuronal dysfunction. Consistent with a previous report9, EphB2 deple-tion in memory-related brain regions was detected not only in hAPPmice, but also in humans with Alzheimer’s disease, underlining thepotential clinical relevance of our findings. Our data further indicatethat the depletion of EphB2 by amyloid-b oligomers involves directbinding of amyloid-b oligomers to the FN repeats domain of EphB2and EphB2 degradation in the proteasome. Reduction of EphB2 mRNAmay have an additional role.
Our results and those of others indicate that neuronal EphB2 deple-tion causes deficits in learning and memory by impairing NMDA-receptor functions (Supplementary Fig. 1). EphB2 modulates NMDAreceptors by tyrosine phosphorylation and recruits active NMDAreceptors to excitatory synapses10–12. EphB2-deficient mice haveLTP deficits10,14 and fewer NR1 subunits in the postsynaptic density10.Our results are consistent with these findings, although LTP deficitsafter shRNA knockdown of EphB2 in adult nontransgenic mice weremore severe than those in EphB2-deficient mice. Other members ofthe large Eph family might partially compensate for EphB2 ablationduring early development. The more severe deficits after acute EphB2knockdown in adults probably reflect the lack of such compensation.
a
b
d e f
c
00.6
0.2
0.4
0.6
0.8
1.0
1.2
1.4
NTG hAPP
***
NTG
Empty EphB2
hAPP
Tubulin
0.4Fibre volley (mV)
0.20 0.60.4Fibre volley (mV)
0.20 0.60.4Fibre volley (mV)
0.20 0.60.4Fibre volley (mV)
0.20
EphB2
NT
G:
em
pty
hA
PP
: E
phB
2
hA
PP
: em
pty
NT
G:
Ep
hB
2
hA
PP
: em
pty
NT
G:
em
pty
hA
PP
: em
pty
NT
G:
Ep
hB
2
hA
PP
: E
phB
2
0
1.0
–15 0 15 30 45 60
1.2
1.4
1.6
1.8
Time (minutes)N
orm
aliz
ed
fE
PS
P s
lop
e NTG: EphB2
NTG: empty
hAPP: EphB2
hAPP: empty2.0
NTG: emptyhAPP: empty
0.1 mV
2 ms
**
***
0.2 mV
1 ms
0.2 mV
10 ms
AMPAR
0
25
50
75
100
NMDAR
AM
PA
R f
EP
SP
slo
pe (V
ms
–1)
NM
DA
R f
EP
SP
slo
pe (m
V m
s–1)
0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
NTG: EphB2hAPP: EphB2
0
25
50
75
100
AM
PA
R f
EP
SP
slo
pe (V
ms
–1)
NM
DA
R f
EP
SP
slo
pe (m
V m
s–1)
AMPAR NMDAR
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
NM
DA
R:A
MP
AR
WB: anti-Flag
EphB2
IP: anti-EphB2 WB: anti-EphB2
Figure 4 | Increasing EphB2 expression rescuessynaptic plasticity in hAPP mice. a, b, Levels ofEphB2–Flag (a) and total EphB2 (b) in the dentategyrus of nontransgenic and hAPP mice injectedwith Lenti-empty or Lenti-EphB2–Flag (n 5 9–12mice per genotype and treatment).c, Normalization of LTP (measured as in Fig. 3c, d)in hAPP mice treated with EphB2–Flag.**P , 0.01 (repeated-measures ANOVA andBonferroni post-hoc test on the last 10 min of data).The following ratios represent the numbers of slicesper number of mice from which the recordingswere obtained: nontransgenic:empty, 8/4;hAPP:empty, 6/3; nontransgenic:EphB2, 13/6;hAPP:EphB2, 20/8. d, e, Comparison of AMPA-receptor-mediated (left) and NMDA-receptor-mediated (right) I/O relationships in the medialperforant path to granule cell synapses ofnontransgenic and hAPP mice treated with Lenti-empty (d) or Lenti-EphB2–Flag (e). Recordingconditions were as in Fig. 3g, h. f, Summary plot ofthe ratios of NMDA-receptor I/O relationships toAMPA-receptor I/O relationships. ***P , 0.001(two-way ANOVA and Bonferroni post-hoc test).Number of slices per number of mice were:nontransgenic:empty, 8/4; hAPP:empty, 6/3;nontransgenic:EphB2, 6/3; hAPP:EphB2, 8/4.Values are means 6 s.e.m.
RESEARCH ARTICLE
5 0 | N A T U R E | V O L 4 6 9 | 6 J A N U A R Y 2 0 1 1
Macmillan Publishers Limited. All rights reserved©2011
Modulation of other LTP-related proteins also results in different out-comes, depending on when it is initiated31.
Amyloid-b may impair LTP by inducing internalization of NMDAreceptors32,33. We found that depletion of EphB2 contributes to theamyloid-b-induced decrease in NMDA receptors and that increasingEphB2 expression markedly improves LTP and memory even in thepresence of high amyloid-b levels. Increased EphB2 levels probablyincreasesurfaceNMDA-receptorexpression.Indeed, increasingneuronalEphB2 expression reversed amyloid-b-induced deficits in NMDA-receptor-mediated synaptic strength.
Opposition of amyloid-b-induced surface depletion of NMDA recep-tors is the most parsimonious interpretation of the EphB2-mediated
rescue effects (Supplementary Fig. 1). However, amyloid-b may impairLTP and memory through alternative processes, and increased expres-sion of EphB2 may counteract amyloid-b effects also through down-stream signalling mechanisms.
Manipulating individual functional hubs of neurons can profoundlyaffect a larger network34,35. Even manipulating an individual neuroncan affect the global brain state36. Thus, improving the function of asubset of neurons might allow an impaired brain region to bettersupport specific behaviours. The current study supports this hypo-thesis: increasing EphB2 expression in a subset of granule cellsimproved dentate gyrus LTP and learning and memory in hAPP mice.It remains to be determined whether EphB2 depletions contribute toamyloid-b-dependent impairments in other brain regions and whetherincreasing neuronal EphB2 levels in these regions is tolerated as well asit was in the dentate gyrus. If so, pharmacological treatments might beused to increase EphB2 expression or activity. Our results indicateadditional entry points for interventions (Supplementary Fig. 1). Forexample, it may be possible to identify small molecules that block thebinding of amyloid-b oligomers to EphB2’s FN repeats domain, pre-vent proteasomal degradation of EphB2, or improve its interactionswith NMDA receptors.
METHODS SUMMARYGeneral. Unless indicated otherwise, all data reported in this paper were obtainedin blind-coded experiments, in which the investigators who obtained the datawere unaware of the specific genotype and treatment of mice, brain slices and cellcultures. For number of mice, slices and cell cultures analysed in each experiment,refer to Supplementary Table 1. For experimental details related to each figurelegend, refer to Supplementary Table 2.Experimental models. Heterozygous transgenic and nontransgenic mice werefrom hAPP line J207,8,37,38. Primary neuronal cultures from wild-type rats weretreated with medium conditioned by CHO cells that do or do not produce humanamyloid-b oligomers (Supplementary Figs 14 and 15 and refs 39, 40.).Experimental manipulations. Lentiviral constructs directing neuronal expres-sion of no transgene products, EphB2–Flag, or GFP in combination with anti-EphB2 shRNAs or scrambled control shRNA were injected stereotactically intothe dentate gyrus of mice20,41. Neuronal cultures were infected with some of theseconstructs and stimulated with Fc-ephrin-B2 or Fc control12,42.Outcome measures. The interaction between biotinylated or naturally secretedamyloid-b oligomers and EphB2 was assessed under cell-free conditions and inneuronal cultures of primary neurons or HEK cells by pull-down with avidinagarose beads43 or immunoprecipitation and western blot44. EphB2 and NR1levels in brain tissues or neuronal cultures were determined by immunoprecipi-tation and western blot or western blot alone44. Corresponding transcripts weremeasured by quantitative polymerase chain reaction with reverse transcription(RT–qPCR). Fos expression in neuronal cultures was determined by westernblot44. Field recordings8 or whole-cell patch-clamp recordings45 from acute hip-pocampal slices were used to determine synaptic strength (fEPSP I/O relation-ships; mediated by either AMPA receptors or NMDA receptors), synapticplasticity (LTP), and NMDA-receptor:AMPA-receptor ratios of EPSCs at themedial perforant path to dentate gyrus granule cell synapses. Learning and memorywere assessed in the Morris water maze, novel object recognition test, novel placerecognition test, and passive avoidance test46–49. Amyloid-b levels in the dentategyrus of hAPP-J20 mice were determined by ELISA50.
Received 25 November 2009; accepted 8 November 2010.
Published online 28 November 2010.
1. Walsh, D. M. & Selkoe, D. J. Deciphering the molecular basis of memory failure inAlzheimer’s disease. Neuron 44, 181–193 (2004).
2. Shankar, G. M. et al. Amyloid-b protein dimers isolated directly from Alzheimer’sbrains impair synaptic plasticity and memory. Nature Med. 14, 837–842 (2008).
3. Kamenetz, F. et al. APP processing and synaptic function. Neuron 37, 925–937(2003).
4. Malenka, R. C. & Bear, M. F. LTP and LTD: an embarrassment of riches. Neuron 44,5–21 (2004).
5. Ikonomovic, M. D. et al. Distribution of glutamate receptor subunit NMDAR1 in thehippocampus ofnormal elderly andpatientswith Alzheimer’sdisease. Exp.Neurol.160, 194–204 (1999).
6. Sze, C., Bi, H., Kleinschmidt-DeMasters, B. K., Filley, C. M. & Martin, L. J. N-Methyl-D-aspartate receptor subunit proteins and their phosphorylation status are alteredselectively in Alzheimer’s disease. J. Neurol. Sci. 182, 151–159 (2001).
a
01 1 2 3 4 5
Trial Session
6 7 8 9 10
10
20 NTG: empty
NTG: EphB2hAPP: empty
hAPP: EphB2
30
40
50
60 hAPP: empty
hAPP: EphB2
NTG: empty
NTG: EphB2
Late
ncy (s)
b
T
ime (%
)
Late
ncy t
o t
arg
et
pla
tfo
rm lo
catio
n (s)
0
20
40
60
80 Familiar object
NTG
Empty
*****
**
hAPP NTG hAPPNTG
Empty
hAPP NTG
EphB2
hAPP
EphB2
Novel object
Tim
e (%
)
0
20
40
60
80 Familiar place
NTG
Empty
**** **
hAPP NTG hAPP
EphB2
Novel place
Late
ncy t
o e
nte
r (s
)
0
50
100
150
200
250 Training
NTG
Empty
** **
*
**
hAPP NTG hAPP
EphB2
Test
0
10
20
30
40 ***c d
e f
Figure 5 | Increasing EphB2 expression in the dentate gyrus ameliorateslearning and memory deficits in hAPP mice. a, Learning curves duringspatial training in the Morris water maze. The latency for each mouse to reachthe hidden platform was recorded. Trial 1 represents performance on the firsttrial, and subsequent sessions represent the average of two training trials. Lenti-empty treated hAPP mice had longer latencies and travelled farther (notshown) to find the hidden platform than all other groups (P , 0.0001, repeated-measures ANOVA). b, Representative paths from the last session of hidden-platform training. c, Time it took mice to reach the target platform locationduring a probe trial (platform removed) 24 h after the last hidden-platformtraining. *P , 0.05, **P , 0.01 versus first bar or as indicated by bracket (one-way ANOVA followed by Bonferroni post-hoc test). d, Object recognitionmemory as reflected by the percentage of time mice spent exploring a familiarversus a novel object during a 10-min test session. **P , 0.01, ***P , 0.001versus familiar object (paired t test). e, Spatial location memory as reflected bythe percentage of time mice spent exploring familiar objects whose locationswere or were not altered. **P , 0.01 versus familiar place (t test). f, Passiveavoidance memory. *P , 0.05, **P , 0.01, versus training or as indicated bybracket (one-way nonparametric Kruskal–Wallis test followed by Dunn’s posttest). n 5 9–12 mice per genotype and treatment. Values are means 6 s.e.m.
ARTICLE RESEARCH
6 J A N U A R Y 2 0 1 1 | V O L 4 6 9 | N A T U R E | 5 1
Macmillan Publishers Limited. All rights reserved©2011
7. Palop, J. J. et al. Vulnerability of dentate granule cells to disruption of Arcexpression in human amyloid precursor protein transgenic mice. J. Neurosci. 25,9686–9693 (2005).
8. Palop, J. J. et al. Aberrant excitatory neuronal activity and compensatoryremodeling of inhibitory hippocampal circuits in mouse models of Alzheimer’sdisease. Neuron 55, 697–711 (2007).
9. Simon, A. M. et al. Early changes in hippocampal Eph receptors precede the onsetof memory decline in mouse models of Alzheimer’s disease. J. Alzheimers Dis. 17,773–786 (2009).
10. Henderson, J. T. et al. The receptor tyrosine kinase EphB2 regulates NMDA-dependent synaptic function. Neuron 32, 1041–1056 (2001).
11. Dalva, M. B. et al. EphB receptors interact with NMDA receptors and regulateexcitatory synapse formation. Cell 103, 945–956 (2000).
12. Takasu, M. A., Dalva, M. B., Zigmond, R. E. & Greenberg, M. E. Modulation of NMDAreceptor-dependent calcium influx and gene expression through EphB receptors.Science 295, 491–495 (2002).
13. Chen, Y., Fu, A. K. & Ip, N. Y. Bidirectional signaling of ErbB and Eph receptors atsynapses. Neuron Glia Biol. 4, 211–221 (2008).
14. Grunwald, I. C. et al. Kinase-independent requirement of EphB2 receptors inhippocampal synaptic plasticity. Neuron 32, 1027–1040 (2001).
15. Fleischmann, A. et al. Impaired long-term memory and NR2A-type NMDAreceptor-dependent synaptic plasticity in mice lacking c-Fos in the CNS.J. Neurosci. 23, 9116–9122 (2003).
16. Litterst, C. et al. Ligand binding and calcium influx induce distinct ectodomain/c-secretase-processing pathways of EphB2 receptor. J. Biol. Chem. 282,16155–16163 (2007).
17. Wakabayashi, K., Honer, W. G. & Masliah, E. Synapse alterations in thehippocampal-entorhinal formation in Alzheimer’s disease with and without Lewybody disease. Brain Res. 667, 24–32 (1994).
18. Scheff, S. W. & Price, D. A. Alzheimer’s disease-related alterations in synapticdensity: neocortex and hippocampus. J. Alzheimers Dis. 9, 101–115 (2006).
19. Mueller-Steiner, S. et al. Anti-amyloidogenic and neuroprotective functions ofcathepsin B: implications for Alzheimer’s disease. Neuron 51, 703–714 (2006).
20. Sun, B. et al. Imbalance between GABAergic and glutamatergic transmissionsimpairs adult neurogenesis in an animal model of Alzheimer’s disease. Cell StemCell 5, 624–633 (2009).
21. Shemer, I. et al. Non-fibrillar b-amyloid abates spike-timing-dependent synapticpotentiation at excitatory synapses in layer 2/3 of the neocortex by targetingpostsynaptic AMPA receptors. Eur. J. Neurosci. 23, 2035–2047 (2006).
22. Ashe, K. H. & Zahs, K. R. Probing the biology of Alzheimer’s disease in mice. Neuron66, 631–645 (2010).
23. Colino, A. & Malenka, R. C. Mechanisms underlying induction of long-termpotentiation in rat medial and lateral perforant paths in vitro. J. Neurophysiol. 69,1150–1159 (1993).
24. Harris, J. A. et al. Many neuronal and behavioral impairments in transgenic mousemodels of Alzheimer’s disease are independent of caspase cleavage of theamyloid precursor protein. J. Neurosci. 30, 372–381 (2010).
25. Sanchez-Mejia, R. O. et al. Phospholipase A2 reduction ameliorates cognitvedeficits in mouse model of Alzheimer’s disease. Nature Neurosci. 11, 1311–1318(2008).
26. Meilandt, W. J. et al. Enkephalin elevations contribute to neuronal and behavioralimpairments in a transgenic mouse model of Alzheimer’s disease. J. Neurosci. 28,5007–5017 (2008).
27. Roberson, E. D. et al. Reducing endogenous tau ameliorates amyloid b-induceddeficits in an Alzheimer’s disease mouse model. Science 316, 750–754 (2007).
28. Nguyen, P. V., Abel, T., Kandel, E. R. & Bourtchouladze, R. Strain-dependentdifferences in LTP and hippocampus-dependent memory in inbred mice. Learn.Mem. 7, 170–179 (2000).
29. Nakajima, R. et al. Comprehensive behavioral phenotyping of calpastatin-knockout mice. Mol. Brain 1, 7 (2008).
30. Potter, M. C. et al. Reduction of endogenous kynurenic acid formation enhancesextracellular glutamate, hippocampal plasticity, and cognitive behavior.Neuropsychopharmacology 35, 1734–1742 (2010).
31. Terashima, A. et al. An essential role for PICK1 in NMDA receptor-dependentbidirectional synaptic plasticity. Neuron 57, 872–882 (2008).
32. Snyder, E. M. et al. Regulation of NMDA receptor trafficking by amyloid-b. NatureNeurosci. 8, 1051–1058 (2005).
33. Kurup, P. et al. Ab-mediated NMDA receptor endocytosis in Alzheimer’s diseaseinvolves ubiquitination of the tyrosine phosphatase STEP61. J. Neurosci. 30,5948–5957 (2010).
34. Bonifazi, P. et al. GABAergic hub neurons orchestrate synchrony in developinghippocampal networks. Science 326, 1419–1424 (2009).
35. Han, J. H. et al. Selective erasure of a fear memory. Science 323, 1492–1496(2009).
36. Li, C. Y., Poo,M.M.&Dan,Y.Burst spikingofa singlecorticalneuronmodifiesglobalbrain state. Science 324, 643–646 (2009).
37. Rockenstein, E. M. et al. Levels and alternative splicing of amyloid b proteinprecursor (APP) transcripts in brains of transgenic mice and humans withAlzheimer’s disease. J. Biol. Chem. 270, 28257–28267 (1995).
38. Mucke, L. et al. High-level neuronal expression of Ab1–42 in wild-type humanamyloid protein precursor transgenic mice: synaptotoxicity without plaqueformation. J. Neurosci. 20, 4050–4058 (2000).
39. Koo, E. H. & Squazzo, S. L. Evidence that production and release of amyloidb-protein involves the endocytic pathway. J. Biol. Chem. 269, 17386–17389(1994).
40. Walsh,D.M.et al.Naturally secretedoligomersofamyloidbproteinpotently inhibithippocampal long-term potentiation in vivo. Nature 416, 535–539 (2002).
41. Franklin, K. B. J. & Paxinos, G. The Mouse Brain in Stereotaxic Coordinates.(Academic, 1997).
42. Xia, Z., Dudek, H., Miranti, C. K. & Greenberg, M. E. Calcium influx via the NMDAreceptor induces immediate early gene transcription by a MAP kinase/ERK-dependent mechanism. J. Neurosci. 16, 5425–5436 (1996).
43. Lauren, J., Gimbel, D. A., Nygaard, H. B., Gilbert, J. W. & Strittmatter, S. M. Cellularprion protein mediates impairment of synaptic plasticity by amyloid-b oligomers.Nature 457, 1128–1132 (2009).
44. Alfa Cisse, M. et al. M1 and M3 muscarinic receptors control physiologicalprocessing of cellular prion by modulating Alzheimer’s disease AM17phosphorylation and activity. J. Neurosci. 27, 4083–4092 (2007).
45. Wu, J., Rush, A., Rowan, M. J. & Anwyl, R. NMDA receptor- and metabotropicglutamate receptor-dependent synaptic plasticity induced by high frequencystimulation in the rat dentategyrus in vitro. J. Physiol. (Lond.) 533,745–755 (2001).
46. Raber, J. et al. Hypothalamic-pituitary-adrenal function in Apoe2/2 mice: possiblerole in behavioral and metabolic alterations. J. Neurosci. 20, 2064–2071 (2000).
47. Raber, J., LeFevour, A., Buttini, M. & Mucke, L. Androgens protect againstApolipoprotein E4-induced cognitive deficits. J. Neurosci. 22, 5204–5209 (2002).
48. Dere, E., Huston, J. P. & De Souza Silva, M. A. Episodic-like memory in mice:simultaneous assessment of object, place and temporal order memory. Brain Res.Protoc. 16, 10–19 (2005).
49. Benice, T., Rizk, A., Kohama, S., Pfankuch, T. & Raber, J. Sex-differences in age-related cognitive decline in C57BL/6J mice associated with increased brainmicrotubule-associated protein 2 and synaptophysin immunoreactivity.Neuroscience 137, 413–423 (2006).
50. Johnson-Wood, K.et al. Amyloid precursorproteinprocessingand Ab42 depositionin a transgenic mouse model of Alzheimer disease. Proc. Natl Acad. Sci. USA 94,1550–1555 (1997).
Supplementary Information is linked to the online version of the paper atwww.nature.com/nature.
Acknowledgements We thank I. Ethell for the plasmid encoding the Flag-taggedEphB2 receptor; D. J. Selkoe and D. Walsh for CHO-7PA2 cells; S. Finkbeiner for theplasmid encoding the NMDAreceptor subunitNR1; J. Palop for comments;H. Solanoy,M. Thwin and X. Wang for technical support; G. Howard and S. Ordway for editorialreview; J. Carroll for preparation of graphics; and M. Dela Cruz for administrativeassistance. The study was supported by NIH grants AG011385, AG022074 andNS041787 to L.M., a fellowship from the McBean Family Foundation to M.C., and theNational Center for Research Resources Grant RR18928-01 to the GladstoneInstitutes.
Author Contributions M.C. and L.M. conceptualized the study. M.C., B.H., J.H. and N.D.performed experiments, and all authors participated in designing experiments and inanalysing and interpreting data. M.C., B.H. and L.M. wrote the manuscript. L.M.supervised the project.
Author Information Reprints and permissions information is available atwww.nature.com/reprints. The authors declare competing financial interests: detailsaccompany the full-text HTML version of the paper at www.nature.com/nature.Readers are welcome to comment on the online version of this article atwww.nature.com/nature. Correspondence and requests for materials should beaddressed to L.M. ([email protected]).
RESEARCH ARTICLE
5 2 | N A T U R E | V O L 4 6 9 | 6 J A N U A R Y 2 0 1 1
Macmillan Publishers Limited. All rights reserved©2011
!"#$%&
!"#$%&'()(*'(*+,(*(-./)0(1123*
()402*&56
!"#$%
Ca2+Ca2+
789
$ -:;2<3=(01
.)456
789
R
Ca2+ Ca2+
!"#$%&'()(*'(*+,(*(-./)0(1123*
.)456
'"()*(+%&,-#)%($)*(.
!"#$%&'()(*'(*+,(*(-./)0(1123*
Ca2+ Ca2+ Ca2+
789
Ca2+Ca2+ Ca2+
>?-903+(@13=(#(<0@'@+23*
P
A
B6
/&01)+$)#2.,3+.)%.)C@ >
()402*&56
Supporting Figures
Figure S1. Model of how Aβ oligomers may affect NMDAR function and synaptic
plasticity through alterations of EphB2. a, Under physiological conditions, interactions
between EphB2 and its natural ligand, ephrin-B2, regulate NMDAR function through
tyrosine phosphorylation of NR1 and/or NR2B subunits, promoting calcium-dependent
signaling, NMDAR-dependent gene expression, and formation of LTP. EphB2 may also
be involved in the recruitment of NMDARs into synapses from perisynaptic or
intracellular pools51. b, Aβ oligomers decrease surface levels of NMDARs through the
induction of LTD-like mechanisms52-54. They also bind to EphB2 via its fibronectin III
repeats domain and may thereby directly prevent EphB2 from interacting with NMDARs.
Binding of Aβ to EphB2 triggers the internalization of EphB2 and its degradation in the
proteasome. In light of decreased EphB2 mRNA levels in brain tissues of hAPP mice
and humans with AD, Aβ probably also decreases EphB2 gene expression or mRNA
stability. Depletion of EphB2 by these mechanisms further reduces the activation and/or
SUPPLEMENTARY INFORMATIONdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 1
retention of NMDARs in the synapse, fueling a vicious cycle and contributing to the
inhibition of NMDAR-dependent gene expression and LTP formation. c, Potential
treatments for AD. The regulation of NMDAR functions by EphB2 may be improved even
in the presence of high Aβ levels by (1) blocking the interaction of Aβ oligomers with
EphB2, (2) increasing the neuronal expression of EphB2 (this study), or (3) enhancing
the interaction between EphB2 and NMDARs.
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 2
Figure S2. Interaction of Aβ oligomers with EphB2. a, b, Different amounts of
biotinylated synthetic Aβ oligomers and recombinant Fc-EphB2 were incubated in
binding buffer and rotated at 4°C for 4 h. Avidin-agarose beads (~40 µl of 75% slurry)
!"# $"!!"%!"&'()*(+,- ./(0)1234567089
:
;
2 < 0
=
>?//,=)@+A-B(=(+
C'A5 ,DEF'&
>?//,=)@+A-B(=(+
C'A5 --
!"# $"!!"%!"&G;,DEF'&56708
H+E?*
$%5IJ9
$!5IJ9
$%5IJ9
H>A5 ,DEF'&C'A5 ,-
$!5IJ9$%5IJ9$!5IJ9
$%!5IJ9
K3(123J(123L)+)123
K3(123J(123L)+)123
K3(123J(123
G;,DEF'&
H+E?*
M* H>A5,DEF'&
C'A5,-
H+E?*
M* '()*(+,-
H+E?*
M* H>A5,-
C'A5,DEF'&
H0N4
K3(123
G;,DEF'&
DEF'&
H0N
H0NJ(123$%5IJ9
$%!5IJ9
$%!5IJ9
$!!5IJ9
$!5IJ9L)+)123
H+E?*
M* '()*(+,-
OP+*F
M* Q2F
M* #>-&
H>A5 ,DEF'&C'A5 ,-
H>A5 ,-C'A5 ,DEF'&
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 3
were added to each sample and rotated at 4°C for 2 h. Beads were washed and
resuspended in loading buffer. Bound Aβ monomers, dimers, and trimers (a) and EphB2
(b) were detected by WB with antibodies against Aβ (6E10) or EphB2, respectively. c, d,
Equal amounts (1 µg) of biotinylated synthetic Aβ oligomers and recombinant Fc-EphB2
were mixed in a binding buffer and incubated overnight at 4°C. Samples were
immunoprecipitated (IP) with anti-EphB2 (c) or a mixture of anti-Aβ antibodies (6E10 and
4G8) (d). Immunoprecipitated proteins were analyzed by WB with anti-Aβ (6E10) (c) or
anti-EphB2 (d). Immunoprecipitations without primary antibodies served as a negative
control (Ct). e-g, Primary cortical and hippocampal rat neurons (7 DIV) were pretreated
with biotinylated synthetic Aβ oligomers (e, f) or with 7PA2-conditioned medium (7PA2)
or untransfected CHO cell-conditioned medium (Veh) (g) for 1 h at 37°C. Cells were
harvested and lysed in a lysis buffer. Samples were spun at 13,000 rpm to remove cell
debris, and supernatants were immunoprecipitated with anti-EphB2 antibody (e, g) or
mixed anti-Aβ antibodies (4G8 and 6E10) (f). Immunoprecipitations without primary
antibodies served as a negative control (Ct). Immunoprecipitated proteins were analyzed
by WB with anti-Aβ (6E10) (e, g) or anti-EphB2 (f).
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 4
!"#$%&'()#*+, - - ... - -
/
01'2/"$3%&*+
456/73%&*+
%"89:;!<#3%&*+#=04
<*#3%&*+#=04
!9#3%&*+#=04
!<#3%&*+#=04
<*#3%&*+#=04
!9#3%&*+#=04
!"#$%&'()#*+, - -.. .%"89:;
!<#3%&*+#=04
- -.
<*#3%&*+#=04%"89:;
456/7>9?@
456/7>9?@
%A#9?@
!"#$%&'()#*+, - -..%"89:;
B177#C5D),06'$%6/E(C()*$/CFG*,> #=04
G*,> #=04
H
"
C !9#3%&*+#=04
Figure S3. Levels of full-length EphB2 and EphB2 deletion mutants in HEK cells.
a-d, We generated and coupled with GST two deletion mutants of mouse EphB2 lacking
either the ligand binding domain (ΔLB-EphB2-GST) or the fibronectin type III repeats
domain (ΔFN-EphB2-GST) and compared them with GST-coupled full-length EphB2
(FL-EphB2-GST). HEK cells were transfected with ΔLB-EphB2-GST, ΔFN-EphB2-GST,
FL-EphB2-GST, or pcDNA3 without insert, either individually (a) or in combination with a
pcDNA3 construct encoding NR1 (b-d). Two days later, some cultures were biotinylated
to compare surface and total EphB2 (a), while others were or were not pretreated with
Fc-ephrin-B2 (500 ng/ml, 1 h) to induce EphB2-dependent phosphorylation of NR1 (b-d).
a, Representative WB illustrating that comparable levels of surface and total EphB2
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 5
!"#$ %&'()*(+(,-.#$ %/01.2
!"#$ %/01.2-.#$ %/01.2
.34(56789(,# : ;< # : ;
;;:
=39+398>+(,# : :< # : ;
; ;;:
3
'
9
?*'*5(,
/01.2
&'()*(+(,@/01.2
< #
.34(56789(,#
: : ;;: ; ;:
AAAA
B
2BBB
CBBB
DBBB
EBFBBB
$$$$GH53+(IH
$$$$$$$$$/01.2$=HIH5
:
A
were expressed by cells transfected with the different EphB2 constructs. b,
Representative WB demonstrating that deletion of EphB2’s ligand binding domain
precluded Fc-ephrin-B2-induced tyrosine phosphorylation of NR1, whereas deletion of
the fibronectin type III repeats did not. c, d, Representative WBs documenting that cells
expressing different EphB2 constructs had comparable levels of total NR1 and that
these levels were unaffected by stimulation with Fc-ephrin-B2.
Figure S4. Aβ induces the degradation of EphB2 in the proteasome. a-c, Primary
cortical and hippocampal rat neurons (5 DIV) were pretreated with synthetic Aβ
oligomers or vehicle for 36 h, followed by treatment with lactacystin (10 µM) to block the
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 6
proteasome (a), bafilomycin (1.0 µM) to block endosomal acidification (b, c), or vehicle
to the medium and incubation for another 12 h. Cells were then lysed and 100 µg of
protein extracts immunoprecipitated with anti-ubiquitin (a) or anti-EphB2 (b, c) and
immunoblotted with anti-EphB2 (a-c). a, Representative WB demonstrating that
lactacystin protected against Aβ-induced EphB2 depletion. Note that ubiquitinated
EphB2 levels were increased when neurons were treated with lactacystin alone or
lactacystin in combination with Aβ relative to treatment of neurons with Aβ alone. * in (a)
represents EphB2 cleavage products or nonspecific bands. b, c, Representative WB (b)
and densitometric quantitation of WB signals (c) demonstrating that Aβ reduced
neuronal EphB2 levels in the presence of bafilomycin. n = 3 wells per condition from
three independent experiments. **P < 0.01, versus empty bars (Tukey test). Values are
means ± s.e.m.
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 7
!"#$%& ' '(#)!"#$%& * '
' '''
+,-)!"#$%& ' '(#)!"#$%& * '
' '''
.
/ 0
1
2
3
4
56-.789:
;,<3.0289:
56-.789:
;,<3.0289:
(#)!"#$%&!"#$%&
* '' '
====>
?>>>
:>@>>>
:?@>>>
927.-AB2
;,<3.02C89:CD2B27
(#)!"#$%&+,-)!"#$%&
* '' '
>
%>>>
E>>>
F>>>
927.-AB2
;,<3.02C89:CD2B27
(#)!"#$%&+,-)!"#$%&
* '' '
>
%>>>
E>>>
F>>>
927.-AB2
56-.7C89:CD2B27
(#)!"#$%&!"#$%&
* '' '
>
%>>>
E>>>
F>>>
G>>>
927.-AB2
56-.7C89:CD2B27
!"#$%&(#)!"#$%&
+,-)!"#$%&(#)!"#$%&
!"#$% +,-)!"#$%
'* ''* *
'* ''* *
Figure S5. shRNA against wild-type, but not mutated, EphB2 reduces surface
levels of NR1. Primary cortical and hippocampal rat neurons (3 DIV) were co-infected or
not with Lenti-sh-EphB2/GFP (sh-EphB2) in combination with either Lenti-EphB2
encoding wild-type EphB2 or Lenti-mut-EphB2/GFP (mut-EphB2) encoding a mutated
EphB2 mRNA that is not recognized by the anti-EphB2 shRNA. Four days later, surface
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 8
!
"
# $
%&'()
*+"+,-.
%&'()
*+"+,-.
/'0123 /'3450678
/'0123 /'3450679
):;,:<=.>-0%&'()
):;,:<=.>-0%&'()
7
7?@
A?7
A?@
/'0123
/'03450678
/'0123
/'03450679
BBB
7
7?@
A?7
A ?@
BB
8 98)9)
8 98)9)
CD,E ,FG
CD,E ,FG
3=,!>-H=:%&'():
::I345:<=H=,
3=,!>-H=:%&'():
::I345:<=H=,
levels of NR1 were determined by biotinylation and subsequent WB analysis, and total
levels of NR1 by WB analysis alone. a, Representative WBs showing that sh-EphB2
reduces neuronal levels of wild-type, but not mutant, EphB2. b, c, Representative WBs
showing that sh-EphB2 reduces surface NR1 levels in neurons expressing wild-type
EphB2 but not in neurons expressing mutant EphB2. d-g, Densitometric quantitation of
WB signals, demonstrating that sh-EphB2 reduces surface (d), but not total (e), NR1
levels in the presence of wild-type EphB2, but neither surface (f) nor total (g) NR1 levels
in the presence of mutant EphB2. Tubulin served as a loading control. n = 3 wells per
condition from three independent experiments. ****P < 0.0001 vs empty bar (t test).
Values are means ± s.e.m.
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 9
!
"
"#$%
"#%"
"#&%
'#""
'#$%
()*+,-./01
()*+,-.23-4$
55516789:;6523-4$
<1(=5>6;67
8
?
Figure S6. shRNA-mediated knockdown of EphB2 in cell culture. a-d, Primary
cortical and hippocampal rat neurons (5 DIV) were infected with Lenti-EphB2-Flag in
combination with lentiviral constructs encoding anti-EphB2 shRNA (shRNA-306 or
shRNA-308, both shRNAs were GFP tagged) or shRNA of a random sequence (sh-
SCR). Four days later, cells were harvested, protein extracts (25 µg) analyzed by WB
with anti-Flag (a, b), and mRNAs extracted for qRT-PCR analysis (c, d). Tubulin and
GAPDH served as loading controls for WBs and qRT-PCRs, respectively. a, b,
Representative WBs showing efficient knockdown of EphB2-Flag by shRNA306 (a) and
shRNA308 (b). c, d, qRT-PCR quantitation of EphB2-Flag mRNA reduction by
shRNA306 (c) and shRNA308 (d). n = 6 wells per condition from three independent
experiments. **P < 0.001, ***P < 0.0001 vs empty bar (t test). Values are means ± s.e.m.
Figure S7. Knockdown of EphB2 prevents LTP in DG GCs of NTG mice. NTG mice
received bilateral injections of Lenti-sh-EphB2/GFP (sh-EphB2) or Lenti-sh-SCR/GFP
(sh-SCR) into the DG at 4–5 months of age. Three weeks later, the infected brain
regions were analyzed by acute slice electrophysiology, qRT-PCR, or immunostaining
and fluorescence microscopy. Untreated (Unt) age-matched NTG and hAPP mice were
analyzed in parallel. a, Anti-GFP immunostaining of DG showing infected neurons in
Lenti-sh-EphB2/GFP treated mice. Right panel shows higher magnification image of
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 10
10.154J (10.153Q.Micro)
10.154J (10.153O.Micro)
10.154J (10.153R.Micro)
10.154J (10.153T.Micro)
10.154J (10.153S.Micro)
10.154J (10.153P.Micro)
!
"
#$%&$'()*+,
boxed region above. Scale bars: 100 µm (left), 25 µm (right). b, Quantitation of EphB2
mRNA by qRT-PCR demonstrating the reduction of EphB2 levels in the DG (reflecting
levels in infected and uninfected cells) (n = 6–12 slices, seven mice per condition). *P <
0.001 vs sh-SCR (t test).
Figure S8. Proportion of GCs expressing GFP after injection of Lenti-SCR/GFP
into the DG of NTG mice. a, b, NTG mice (n = 7) received bilateral injections of Lenti-
SCR/GFP into the dentate gyrus (DG) at 3 months of age. Three weeks later, coronal
sections of DG were immunostained for calbindin (CB, red), which is expressed at high
levels in mature granule cells (GCs). Both CB and GFP (green) were analyzed by
fluorescence microscopy. a, Representative photomicrographs. Scale bar: 100 µm. b,
Higher magnification images of boxed regions above. Scale bar: 25 µm. To determine
the proportion of GCs expressing GFP, every 10th hippocampal section (30 µm thick,
300 µm apart) was collected from each hemibrain between –2.54 and –3.16 mm from
bregma. Four sections that were selected to be matched for rostrocaudal level were
analyzed per mouse. CB-positive and GFP-expressing GCs were counted separately
within the upper and lower blades of the DG in each section. The percentage of viral
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 11
!
"
#$%&'$()
*+,-./01&22-./0
*+,-)13456*+,-)137%189
*+,-)13456
*+,-./0 1&22-./0
*+,-)137%189
0/:;5$<=>0"?:
(:5
@@
@A9
@AB
@AC
@AD
'A@
# '@ '#(7245$2<"E$&(%?>0:F<$G%&H
9@ 9# I@
transduction was calculated by multiplying the ratio of double-labeled to CB-positive cells
by 100. Immature GCs do not express CB and, thus, were not included in this
calculation. They accounted for less than 5% of GCs per section.
Figure S9. Normal post-synaptic AMPAR-mediated currents in DG GCs of Lenti-
sh-EphB2/GFP treated NTG mice and untreated hAPP mice. Four-month-old NTG
and hAPP mice did or did not receive bilateral injections of Lenti-sh-EphB2/GFP (sh-
EphB2) or Lenti-sh-SCR/GFP (sh-SCR) into the dentate gyrus (DG). Three weeks later,
acute slices of DG from treated and untreated (Unt) mice were analyzed by whole-cell
patch-clamp recordings. a, Typical mEPSCs recordings from granule cells (GCs) voltage
clamped at –60 mV in the presence of 1 µM TTX in the indicated groups of mice. b,
Cumulative histogram of the peak amplitude of 500 individual mEPSCs from five GCs
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 12
!
"#$%&'()*+,-
.
./0
1/.
1/0
2 2 33*455- 2 3 32
"#$%&'#6)%7-"#$%&'()*+,-
*455-
"#$%&'#6)%7-3 3 222 2 332 3 32
3 3 22
8
.
,...
9...
:...
;...
<<=#>!%&?#<()*+,'@>!A
6=B4<"#?#>
=#>!%&?#
<()*+,'@>!A<"#?#>
(100 consecutive mEPSCs per GC) in five slices from two to three mice per group (P >
0.4 by Kolmogorov-Smirnoff test).
Figure S10. Detection of EphB2-Flag in the DG of hAPP mice. a,b, Two-month-old
hAPP and NTG mice received bilateral injections of Lenti-EphB2-Flag or Lenti-empty
into the dentate gyrus (DG). Two months later, levels of exogenous EphB2 mRNA (a)
and protein (b) in DG were determined by qRT-PCR with probes designed to Flag
sequence or WB with anti-Flag, respectively. GAPDH and tubulin served as loading
controls for qRT-PCR and WB, respectively. n = 5 mice per treatment and genotype. No
significant differences in DG EphB2-Flag mRNA or protein levels were identified
between Lenti-EphB2-Flag treated hAPP and NTG mice (t test). Values are means ±
s.e.m.
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 13
!
!"#$%
&$'() *'+,-.
/.
0..
0/.
-..1
2 034
!"#$%
&$'() *'+,-.
-/
/.
5/
0..6
2 037-
89(1:
&$'() *'+,-.
.;0
.;-
.;<
.;7
.;/
.;=>
/2 037-
+2?? +2?? +2??
2 034
!"#$%&'() *+,,$%&'()
!"# *+,,
--- ---
!"#%&'() .'*/0
*+,,1
120
123
124
125
621
620
623
624
625
,%7897:;(<,:(*<,:=7%><,?@A%<B:(=9
C
1124
12612012D12312E12412F12512G
123H=C%7<I9@@%)<+&'@=(?>%<J&IK
1201
LM*:88%7<N9@@:(%7:@<8.,L,
L@9'%<JI$AK
120<&I0<&A
!"#$.'*/0 *+,,$.'*/0M
1124
12612012D12312E12412F12512G
123H=C%7<I9@@%)<+&'@=(?>%<J&IK
1201
LM*:88%7<N9@@:(%7:@<8.,L,
L@9'%<JI$AK
:
--- ---
Figure S11. Aβ levels in the DG of hAPP mice treated with Lenti-EphB2-Flag. a-c,
Two-month-old hAPP mice received bilateral injections of Lenti-EphB2-Flag (EphB2) or
Lenti-Empty (Empty) in the dentate gyrus (DG). Two months later, levels of Aβ1–x (a)
and Aβ1–42 (b) and Aβ1–42/Aβ1–x ratios (c) in both DG were determined by ELISA55. n
= 5 mice per treatment. Student's t test did not reveal significant differences between
groups. Values are means ± s.e.m.
Figure S12. Increasing expression of EphB2 in the DG of hAPP mice does not
improve their deficits in short-term synaptic plasticity and synaptic strength.
Besides LTP deficits in the dentate gyrus (DG), hAPP mice also have deficits in paired
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 14
!"#$%
&%'()*+,#(
-!.%&/#0
1233 -!.4/156
12337
877
9777
9877
6777
6877
:777,
%&;!*
&<2*$%/
=*>
-!.%&/#0
1233 -!.4/156
12337
97
67
:7
?7
87 @@@@@@@@A
pulse ratios (PPR) (indicative of short-term synaptic plasticity) in the DG and in input-
output (I/O) relationship (indicative of synaptic transmission strength) along the Schaffer
collateral to CA1 pyramidal cell synapse56,57. To determine whether these abnormalities
are reversed by increasing the expression of EphB2 in the DG, 2-month-old hAPP and
NTG mice were given bilateral injections of Lenti-EphB2-Flag (EphB2) or Lenti-empty
(Empty) into the DG (n = 3–4 mice per genotype and treatment). Two months later,
acute hippocampal slices were prepared from these mice and analyzed by field
recordings. a, PPR at the medial perforant pathway were reduced to a similar extent in
hAPP mice treated with Lenti-EphB2-Flag or Lenti-empty, as compared to NTG controls.
***P < 0.001 vs similarly treated NTG (two-way ANOVA and Tukey test). b, c, The I/O
relationship along the Schaffer collateral-CA1 synapse was reduced to a similar extent in
hAPP mice treated with Lenti-empty (b) or Lenti-EphB2-Flag (c), as compared to NTG
controls. ***P < 0.001 (repeated measures ANOVA and Bonferroni test). The number of
slices/mice analyzed were NTG/empty: 8/4, hAPP/empty: 6/3, NTG/EphB2-Flag: 6/3,
hAPP/EphB2-Flag: 8/4).
Figure S13. Increasing EphB2 expression in the DG of hAPP mice does not
improve their behavioral abnormalities in the open field and elevated plus maze.
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 15
! " #
$%&'
()*+,)-*+,)./0/+,)
1234567
89:
;98
;9:
498
0%<+=
8
0%<+=
8
48
>8
?8
567 1234567 1234
3 ;@>4 3 ;@A
Four- to five-month-old NTG and hAPP mice underwent behavioral studies 2 months
after receiving bilateral injections of Lenti-EphB2-Flag (EphB2) or Lenti-empty (empty)
into the DG. a, The open field test reveals similar levels of hyperactivity in hAPP mice
treated with Lenti-EphB2-Flag or Lenti-empty, as compared with NTG controls. b, The
elevated plus maze reveals a similar level of disinhibition in hAPP mice treated with
Lenti-EphB2-Flag or Lenti-empty, as compared with NTG controls. No significant
difference in time spent in the open arms was identified between NTG/empty and
NTG/EphB2 mice. n = 9 mice per genotype and treatment. Two-way ANOVA revealed
an effect of genotype but not of treatment and no genotype x treatment interaction. *P <
0.01, **P < 0.001, ***P < 0.0001. Values are means ± s.e.m.
Figure S14. Detection of Aβ oligomers in 7PA2 cell-conditioned medium. a,
Medium conditioned for 24 h were collected from 7PA2 and untransfected CHO control
cell cultures and analyzed by immunoprecipitation with an antibody against Aβ amino
acid residues 17–24 and western blotting with the same antibody. b-c, Levels of soluble
Aβ1-42 (b) and Aβ1-x (c) in 7PA2- and control CHO-conditioned media determined by
ELISA55.
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 16
! !
!
"#$%&'()*
+,-(#,$)%./01+23
%$
,
)4516
7
87
977
987
:;<
-+==
:;<
>=+?
:;<+ 9@A
>=+?
>=+?
:;< >=+?
:;< >=+?
:;< >=+?
9B 9> 9C 9D ?7 9B 9>99 9? 9E 9F 98"#$%&'()/GH1,.#"#$%&'()/GH1,.#
"#$%&'()/GH1,.#"#$%&'()/GH1,.#
99 9? 9E 9F 98
E9 ? F 8E9 ? F 8
9 E 8 > D 99 9E 98 9> 9D ?9?E?8?>?D E9EEE8E>
IJK
IJK
IJK
IJK
LM)&N&N.&'%
///////+
LM)&N&N.&'%
///////+
LM)&N&N.&'%
///////+
LM)&N&N.&'%
///////+
9C//9D//?7
B > C D 97 B > C D 97
Figure S15. Purification of Aβ oligomers by size-exclusion chromatography. a,
7PA2- and control CHO cell-conditioned media were applied separately to a
Superdex 75 column, and elution was monitored at an absorbance of 254 nm. The
elution volume of each fraction was 1 ml. Peaks represent Aβ assemblies of different
sizes. b, Selected fractions were separated on SDS-PAGE and immunoblotted for
Aβ. M, monomer; D, dimer; T, trimer. Asterisk indicates nonspecific band. Synthetic
Aβ oligomerized as described58 was used as a positive control. Fractions 11–15 from
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 17
!"#$%&'()*&+,
-&++,'.&%'/0*1"2"0*
3*1&.&*1&*2'45.&%"6&*2,
7$6)*'8%)"*'9)6.+&,'.&%'/0*1"2"0*
:0$,&'8%)"*'9)6.+&,'.&%';%0$.'
!"#$%&'()*&+,
<02)+'=$6>&%'0?':"@&
:"@&'.&%'45.&%"6&*2
:"@&'.&%';&*02A.&')*1'<%&)26&*2
:"@&'.&%';&*02A.&
:"@&'.&%'<%&)26&*2
9+"@&,'.&%'<%&)26&*2')*1B0%';&*02A.&
!"#$%&'()*&+,
<02)+'=$6>&%'0?':"@&
/0C0%2,'0?':"@&
:"@&'.&%'/0C0%2
:"@&'.&%';&*02A.&')*1'<%&)26&*2
DE'>F?G'CHE'@F?9IE'>G'@9JE'>F#
K
K
KE'@G'1G'#9DHE )F@
IJ
DI
K
I
LFM
JE')F?9DKE')G'>
DNH
D
IH
MFDH
DE'#HE')9OE')F1
O
K
IE'@G'?
HD
KFL
OFHN
H
ON
DJ
KE'>9PE'>9L9DNE')G'>9DD
IFDN
9ME')G'>
DN
KFI
K
J
IFO
!"#$%&'"()*+
,-&$)*#./%+("#-#0+
!&%12"#*
7PA2 cell-conditioned medium, which contained Aβ oligomers, and corresponding
fractions from CHO control cell-conditioned medium were collected, lyophilized,
pooled (within each cell source), and resuspended in artificial cerebrospinal fluid
(ACF). Fractions 6–10 from 7PA2 cell-conditioned medium contained the secreted
ectodomain of APP (sAPP) and were discarded. c, ELISA measurements of Aβ1-x
levels in fractions 11–15 from 7PA2 cell- and CHO cell-conditioned media.
Table 1
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 18
Cissé et al, Supplementary Table 2
FIGURE 1
FIGURE 2
Panels Methods
bThe binding of dimers and trimers to EphB2 constructsexpressed in HEK cells was quantitated by immunoprecipitationwith anti-GST antibodies and densitometric analysis of anti-Aβ (6E10) WB signals.
d,f
Primary rat neurons were treated with vehicle (veh, medium conditioned by untransfected control CHO cells, 3 days) or Aβ (equivalent of 60 ng/mlor 12.5 nM in 7PA2 cell-conditioned medium, for indicated times). Surfacelevels of EphB2 were determined by biotinylation and subsequent WB analysis, and total levels by WB analysis alone.
g Primary rat neurons were treated with vehicle for 6 days or with Aβ for the indicated times. EphB2 mRNA levels were determined by qRT-PCR.
h,i
Primary rat neurons were pretreated for 36h with synthetic Aβ oligomers orvehicle, followed by addition of lactacystin (10μM) or veh to the culturemedium and incubation for another 12h. Cells were then lysed and 100μgextracts immunoprecipitated with anti-EphB2 and immunoblotted withanti-EphB2
Primary rat neurons were co-infected or not with Lenti-sh-EphB2/GFP (sh-EphB2) in combination with either Lenti-EphB2 encoding wild-type EphB2 or Lenti-mut-EphB2 (mut-EphB2) encoding a mutated EphB2 mRNA that is not recognized by sh-EphB2. Four days later, cells were stimulated with clustered multimeric recombinant Fc-ephrin-B2 ligand to activate EphB2. WB signals were quantitated by densitometry.
f
FIGURE 3
b
NTG mice received bilateral injections of Lenti-sh-EphB2/GFP (sh-EphB2) or Lenti-sh-SCR/GFP (sh-SCR) into the dentate gyrus (DG) at 4–5 months of age. Three weeks later, the infected brain regions were analyzed by acute slice electrophysiology, qRT-PCR, or immuno-staining and fluorescence microscopy. Untreated (Unt) age-matched NTG and hAPP mice were analyzed in parallel. EphB2 mRNA levels were determined by qRT-PCR.
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 19
Cissé et al, Supplementary Table 2
FIGURE 3
FIGURE 4
Panels Methods
c-f
Two-month-old NTG and hAPP mice received bilateral injections of Lenti-empty or Lenti-EphB2-Flag into the dentate gyrus (DG). Two months after the injection, DG were microdissected for determination of levels of EphB2-Flag and total EphB2 by WB analysis with anti-Flag andanti-EphB2 antibodies, respectively.
a,b
FIGURE 5
a
Four- to 5-month-old NTG and hAPP mice were analyzed behaviorally 2 months after they received bilateral injections of Lenti-empty or Lenti-EphB2-Flag in the DG. Trial 1 represents performance on the first trial, and subsequent sessions represent the average of two training trials. Lenti-empty treated hAPP mice had longer latencies and traveled farther (not shown) to find the hidden platform than all other groups.
LTP was induced by theta burst stimulation (TBS) and measured by field recordings or by whole-cell patch clamp from individual GFP-positive cells in the DG. Three consecutive responses were averaged for each slice and these data were then averaged for all slices in a group to generate each point on the graph. Top traces depict the average of ten synaptic responses from a single neuron before and after TBS LTP.
g,hTraces at the top show example fEPSPs for AMPAR-mediated responses or NMDAR-mediated responses. Fiber volley strengthswere placed into 0.1 mV bins; fEPSP slopes were then averaged from each bin to generate the points on the graphs below.
fPassive avoidance memory was assessed 24 h after training mice in a light/dark chamber as reflected by the time it took them to re-enter the dark chamber during a 5-min test session.
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 20
FULL METHODS
hAPP transgenic mice. Heterozygous transgenic and NTG mice were from line J20,
which expresses an alternatively spliced hAPP minigene encoding hAPP695, hAPP751
and hAPP770 with the Swedish and Indiana familial AD mutations directed by the PDGF
β-chain promoter57,59-63.
Preparation of Aβ oligomers. Naturally secreted Aβ oligomers. Stably hAPP-
transfected CHO-7PA2 cells, which produce Aβ oligomers, were cultured as
described64,65. Briefly, untransfected CHO cells and CHO-7PA2 cells were grown to 80%
confluency in 150-mm dishes, washed with PBS, and incubated for ~24 h in serum-free
Neurobasal A medium. The medium was collected and spun at 1000 rpm for 10 min to
eliminate cell debris. Supernatants were concentrated 10-fold with a Centriprep YM-3
(Millipore), collected as 1-ml aliquots in 1.5-ml Eppendorf tubes and stored at –80°C.
After size-exclusion chromatography (Supplementary Fig. 17) to remove secreted APP,
1-ml aliquots of conditioned medium were lyophilized and reconstituted in artificial
cerebrospinal fluid.
Synthetic Aβ oligomers: Synthetic biotinylated Aβ1-42 peptides (rPeptide) were
lyophilized in hydroxyfluroisopropanol (HFIP), reconstituted in dimethyl sulfoxide at 2.2
mM, diluted in Neurobasal A medium, pH 7.4 (Invitrogen) to 1 µg/ml, incubated at 4°C
for 48 h, and stored at –80°C until use58. For treatment of cells, stock solutions of Aβ
peptides were diluted in fresh Neurobasal A/N2 medium to final concentrations of 1
µg/ml (equivalent in total Aβ content to a 0.22 µM solution of monomeric Aβ).
Primary neuronal culture and pharmacology. Cortex and hippocampus of wild-type
rat pups (P0) were digested with papain. Cells were plated in polylysine-coated wells
and maintained in serum-free Neurobasal medium supplemented with B27 (Invitrogen)
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 21
and antibiotics. Half the medium was changed after 5 days in culture. Cells were used
after 5-11 days in culture. More than 95% of the cells were neurons, as determined by
staining with an antibody against the neuron-specific marker MAP2 (data not shown).
Neuronal cultures were treated with Aβ oligomer fractions from 7PA2-conditioned
medium, control fractions from untransfected CHO cells, synthetic Aβ or vehicle,
clustered recombinant Fc-ephrin-B2- (R&D Systems), or control Fc (Jackson
ImmunoResearch Labs) as described in the text. For detection of Fos, cells were
pretreated wtih tetrodotoxin (TTX, 1 µM, 48 h) and 2,3-dihydroxy-6-nitro-7-sulfamoyl-
benzo[f]quinoxaline-2,3-dione (NBQX, 40 µM, 48 h) to reduce endogenous synaptic
activity66. Fc-ephrin-B2 and control Fc were preclustered with anti-human Fc antibody at
50 ng/ml in Neurobasal medium at room temperature for 1 h and applied at final
concentrations of 500 ng/ml. Treatment with anti-Fc antibodies served as an additional
control. Inhibitors were used at the following concentration in the indicated vehicle:
lactacystin (10 µM in water), bafilomycin (1.0 µM in DMSO). After treatment, cells were
harvested in lysis buffer A (10 mM Tris/HCl, pH 7.5, 150 mM NaCl, 0.5% Triton X-100,
0.5% deoxycholate, 5 mM EDTA), spun at 13,000 rpm for 5 min, and frozen at –80°C for
subsequent determination of protein concentration and western blot analyses.
Biotinylation assay. Rat primary neurons were surface biotinylated as described67.
Briefly, primary neurons were cultured for 7 DIV, placed on ice, and rinsed three times in
ice-cold PBS. Neurons were then incubated in ice-cold PBS containing 2 mg/ml sulfo-
NHS-LC-biotin (Pierce) for 30 min at 4°C, rinsed twice in PBS, and lysed in 250 µl of
PBS (for each well of a 6-well plate) containing complete protease inhibitor cocktail
(Roche), 0.1% sodium dodecyl sulfate (SDS), and 1% Triton X-100. Samples were then
briefly sonicated. Ten percent of the cell lysate was saved to determine total protein
concentration by Bradford assay. To isolate biotinylated proteins, the other 90% of the
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 22
cell lysate (approx. 250 µg of protein per sample) was incubated overnight with 50 µl of
Avidin-agarose beads (Pierce) in PBS containing 1% NP-40 to avoid nonspecific
binding. Isolated proteins were rinsed three times in PBS and boiled in 50 µl of sample
buffer. Western blots were then carried out, and data were quantified by comparing the
ratio of biotinylated to total protein for a given culture and normalizing to control
untreated cultures.
Pull-down assay. Cell-free condition. Different amounts of synthetic biotinylated Aβ1-42
oligomers (rPeptide) and recombinant mouse Fc-EphB2 chimera (R&D Systems) were
mixed in 400 µl of binding buffer (50 mM Tris, pH 7.5, 200 mM NaCl, 0.1% NP-40) and
rotated overnight at 4°C. Avidin-agarose beads (40 µl of 75% slurry; Pierce) were added,
and the tubes were rotated at 4°C for 2 h and spun at 13,000 rpm for 30 s. The
supernatant was discarded. Beads were washed twice with 500 µl of PBS and
resuspended in 30 µl of 2X loading buffer. Samples were boiled at 90°C and loaded onto
a NuPAGE 4–12% Bis-Tris gel for western blot analysis.
Cell culture condition. HEK cell line: Cells grown on 12-well plates were transiently
transfected with full-length EphB2 or EphB2 lacking either its LB domain or its FN
repeats domain. Empty pcDNA3 served as a negative control. Seventy-two hours after
transfection, cells were treated or not with different amounts of synthetic Aβ for 2h. After
incubation, cells were washed with PBS to remove unbound Aβ and then lysed with
buffer A supplemented with a protease inhibitor mixture (Sigma). Bound Aβ was
analyzed by IP using an antibody against GST, followed by SDS-PAGE and
immunoblotting with an antibody directed against Aβ. Primary neurons: Rat primary
neurons grown on 12-well plates for 7 DIV were treated with different amounts of
synthetic Aβ oligomers for 2 h. Cells were washed with PBS to remove unbound Aβ and
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 23
lysed with buffer A supplemented with a protease inhibitor mixture (Sigma). Samples
were then analysed by immunoprecipitation and SDS-PAGE followed by immublotting.
Immunohistochemistry. Immunofluorescence staining. Rat primary neurons were
grown on coverslips for 7 DIV. Cells were rinsed with ice-cold PBS, fixed with 4%
paraformaldehyde in PBS for 30 min, then rinsed in 0.1% PBS-Triton X-100 for 10 min.
Coverslips were incubated in blocking solution (10% normal donkey serum in 0.01%
PBS-Triton X-100) for 30 min at room temperature and overnight at 4°C with anti-rabbit
EphB2 antibody (H-80, 1:200, Santa Cruz Biotechnology) diluted in blocking solution.
After rinses with 0.01% PBS-Triton X-100, cells were incubated with appropriate Alexa-
conjugated secondary antibodies (1:300, Invitrogen) diluted in 10% normal donkey
serum in PBS for 1 h at room temperature. Coverslips were rinsed extensively with PBS
and mounted with Vectashield mounting medium (Vector Laboratories). For analysis,
digitized images were obtained with a DEI-470 digital camera (Optronics) on a BX-60
microscope (Olympus). DAB staining. Tissue preparation and immunohistochemistry
were performed as described62. Primary antibodies used included the following: rabbit
anti-calbindin (1:15,000; Swant), rabbit anti-Fos (1:10,000; Ab-5, Oncogene).
Generation of EphB2 deletion and point mutants. Deletion mutant: Cloning of full-
length EphB2 and deletion mutants lacking the LB domain or the FN repeats domain
was performed using polymerase chain reaction. Each construct was designed with a
carboxy terminal GST-tag by cloning synthetic genes into NdeI digested pET41a(+)
derivative lacking its multiple cloning sites. The resulting GST-tagged genes were then
inserted between the XbaI and xhoI sites of a pcDNA3 vector and their expression was
tested in HEK cells. Point mutant: To generate EphB2 point mutant, Wild-type EphB2
cDNA (encoding for Flag-tagged mouse EphB2) was used as the template to introduce
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 24
point mutations using the QuickChange site-directed mutagenesis kit (Stratagene)
according to the manufacturer’s protocol. The resulting EphB2 gene bears the following
mutations: T6836C, C6839A and G6842A. These mutations were introduced into the
EphB2 sequence targeted by sh-EphB2-308, without altering its amino acid sequence.
The EphB2 mutant was used to produce active lentiviral particles by cotransfecting the
transfer vector with two helper plasmids, delta8.9 (packaging vector) and VSV-G
(envelope vector), into HEK293T cells.
Lentivirus production and stereotaxic injection. Lentiviral vectors were based on
FUGW68. EphB2 expression was reduced with two different shRNAs targeting mouse
EphB2 placed under the U6 promoter. The target sequences were 5′-
ACGAGAACATGAACACTAT-3′ (sh-EphB2-306), 5′-TGAACAGTATCCAGGTGAT-3′
(sh-EphB2-308). The U6-shRNA expression cassette (pSilencer 2.0, Ambion) was
inserted between the PacI and NheI sites of a modified FUGW lentiviral backbone,
placing the shRNA cassette upstream of an ubiquitin C promoter directing expression of
enhanced GFP. A similar construct expressing a scrambled shRNA was used as a
control. To increase expression of EphB2, a sequence encoding EphB2-Flag was
inserted between the NotI sites of the FUW backbone. Because EphB2 cDNA is ~3 kB
and large inserts can lead to packaging problems and low viral titers, we did not include
GFP in this construct but used the short Flag tag instead. Active lentiviral particles were
generated by cotransfecting the transfer vector with two helper plasmids, delta8.9
(packaging vector) and VSV-G (envelope vector), into HEK293T cells. The viral particles
were purified from the culture medium by ultracentrifugation. An empty virus was used
as control. Viral titers were determined by p24 ELISA55.
Two- to 4-month-old NTG and hAPP-J20 mice were anesthetized by intraperitoneal
injection with Avertin (tribromoethanol, 250 mg/kg) or a mixture of ketamine (75 mg/kg)
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 25
and medetomidine (1 mg/kg). Mice were placed in a stereotaxic frame, and lentiviral
vectors were stereotactically injected bilaterally into the DG (2–3 ml/site; 1
site/hemisphere) at the following coordinates69: a/p, –2.1, m/l ±1.8, d/v, –2.0. After
surgery, anesthesia was reversed with atipamezole (1 mg/kg). Behavioral assays were
carried out 4–8 weeks after lentiviral injections. Hemibrains from replicate groups of mice
injected with lentiviral vectors as described above were used after a similar interval to
prepare acute hippocampal slices for electrophysiological measurements; the opposite
hemibrains were snap-frozen at –80°C and homogenized in lysis buffer for biochemical
analyses.
While the small size of shRNAs allowed us to incorporate GFP into shRNA-encoding
lentiviral constructs, the large size of the EphB2 cDNA made this strategy impossible for
EphB2-encoding lentiviral constructs. Consequently, we were able to use GFP to
document typical transduction efficiencies and expression patterns only for the former
(Supplementary Fig. 8) but not the latter. Based on the results obtained with Lenti-sh-
SCR/GFP, we estimate that on average ~60% of GCs are transduced. Similar
transduction efficiencies and expression patterns were observed when lentiviral
constructs were used to express other factors in DG GCs70,71, making it likely that the
transduction efficiency and expression pattern of Lenti-EphB2-Flag were not much
different.
Protein extraction from tissues. Total tissue lysates from mouse or human brain were
obtained by homogenizing entire mouse hemibrains or microdissected hippocampus in
ice-cold lysis buffer (10 mM Tris/HCl, pH 7.5, 150 mM NaCl, 0.5% Triton X-100, 0.5%
deoxycholate, 5 mM EDTA) supplemented with a protease inhibitor mixture (Sigma).
Samples were centrifuged at 1000 x g for 10 min at 4°C. The supernatant was placed on
ice and the pellets were re-homogenized in 500 µl of lysis buffer and centrifuged at 1000
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 26
x g for 10 min at 4°C. The supernatant was combined with the first supernatant collected
and centrifuged at 100,000 x g for 1 h at 4°C. Supernatant from this last centrifugation
was then collected and used to determine the protein concentration of the samples and
for western blot analyses.
ELISA analysis of Aβ levels. Whole hemibrains were microdissected, and the DG was
isolated. DG tissues homogenized in 5 M guanidine buffer were analyzed by ELISA for
levels of human Aβ1-x and Aβ1–42 as described55.
Immunoblotting. For detection of Fos and NR1, 25 µg of protein was loaded into each
well of a 4–12% gradient SDS-PAGE gel. Gels were transferred to nitrocellulose
membranes and immunoblotted with rabbit anti-Fos (1:500, Santa Cruz Biotechnology)
or mouse anti-NR1 (1:1000, Millipore). For detection of EphB2 in mouse samples, 250
µg of proteins were immunoprecipitated with an anti-mouse EphB2 antibody (2 µg, R&D
Systems) and analyzed by WB. For detection of EphB2 in human samples, 100 µg of
proteins were directly immunoblotted with a rabbit polyclonal antibody against amino
acids 255–334 in the N-terminal extracellular domain of human EphB2 (H-80, 1:200,
Santa Cruz Biotechnology) in blocking buffer (Tris-buffered saline/0.1% Tween/5% milk,
pH 7.6) overnight. For detection of ubiquitinated EphB2, 100 µg of proteins were
immunoprecipitated with anti-ubiquitin (P4D1, 2 µg, Santa Cruz Biotechnology) and
analyzed by WB with anti-mouse EphB2 (R&D Systems). Tubulin signals were obtained
by loading 15 µg of protein per well from corresponding samples and immunoblotting
with an anti-tubulin antibody. Goat anti-rabbit or anti-mouse antibodies (1:5000,
Chemicon; room temperature, 2 h) were used as secondary antibodies. Protein bands
were visualized with an ECL system (Pierce) and quantified densitometrically with Image
J software (National Institutes of Health).
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 27
qRT-PCR. For quantitative fluorogenic RT-PCR, total RNA was isolated from frozen
brain tissues with RNeasy Mini kits with an on column RNase-free DNase I treatment
(Qiagen). Total RNA was reverse transcribed with random hexamers and
oligo(dT) primers. Diluted reactions were analyzed with SYBR green PCR reagents
and an ABI Prism 7700 sequence detector (Applied Biosystems). Human EphB2 mRNA
levels were normalized to 18S RNA, whose levels did not differ between AD cases and
nondemented controls (not shown). Endogenous mouse EphB2 and exogenous EphB2-
Flag mRNA levels were normalized to GAPDH. cDNA levels of EphB2, Flag-EphB2, 18S
and GAPDH were determined relative to standard curves from pooled samples. The
slope of standard curves, control reactions without RT, and dissociation curves of
products indicated adequate PCR quality. The following primers were used: mouse
EphB2 forward, 5 ′-GTGTGGAGCTATGGCATCGT-3′; reverse, 5′- TGGGCG
GAGGTAGTCTGTAG-3′. Human EphB2 forward, 5'-TGCAATGTCTTTGAGTCAA GCC-
3'; reverse, 5'-ATGCGG TGGGCGCC-3'. Human 18S forward, 5’-
ATCAACTTTCGATGGTAGTCG-3’; reverse, 5’- TCCTTGGATGTGGTAGCCG-3’. Flag
forward, 5'-ATTCTGCTGGCTGCTGCT-3’; reverse, 5’-CGTTGCTGTCGTAGAGTCC-3’.
Electrophysiology in acute slice preparations. NTG and/or hAPP (J20 line) mice (2–5
months old) were anesthetized with Avertin (tribromoethanol, 250 mg/kg) and
decapitated 4–8 weeks after the injection with lentivirus. For NTG mice injected with
Lenti-sh-EphB2/GFP or Lenti-sh-Scramble/GFP, half of the brain was used to measure
levels of EphB2 mRNA by qRT-PCR, and the other half from the same mice was used
for electrophysiology recordings. For hAPP mice and NTG controls injected with Lenti-
EphB2-Flag or lenti-Empty, half of the brain was used for biochemical measurements
(WB, immunohistochemistry) and the other half from the same mice was used for
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 28
electrophysiology measurements. Brains were quickly removed and placed in ice-cold
solution containing (in mM) 2.5 KCl, 1.25 NaPO4, 10 MgSO4, 0.5 CaCl2, 26 NaHCO3, 11
glucose, and 234 sucrose (pH, ~7.4; 305 mOsmol). Coronal 350-µm slices were cut with
a vibratome and collected in the above solution. Slices were then incubated for 30 min in
standard artificial cerebrospinal fluid (30°C) containing (in mM) 2.5 KCl, 126 NaCl, 10
glucose, 1.25 NaH2PO4, 1 MgSO4, 2 CaCl2, and 26 NaHCO3 (290 mOsmol; gassed with
95% O2–5% CO2, pH ~7.4). Subsequently, slices were maintained at room temperature
for 30 min before recording. Individual slices were transferred to a submerged recording
chamber, where they were maintained at 30°C and perfused with artificial cerebrospinal
fluid at a rate of 2 ml/min. No recordings were made on slices > 5 h after dissection.
For whole-cell patch-clamp recordings, EGFP-expressing GCs were identified under
epifluorescence, and voltage-clamp recordings were obtained under infrared differential
interference contrast video microscopy. The intracellular patch pipette solution contained
(in mM) 120 Cs-gluconate, 10 HEPES, 0.1 EGTA, 15 CsCl2, 4 MgCl2, 4 Mg-ATP, and
0.3 Na2-GTP, pH 7.25, adjusted with 1 M CsOH (285–290 mOsm; patch electrode
resistance: 3–6 MΩ). EPSCs were evoked with a theta-glass pipette filled with 1 M NaCl
and 25 mM HEPES, pH 7.3, placed in the medial perforant path in the dorsal blade of
the DG. A stable 15-min baseline of EPSCs evoked at ~30% of maximum peak
amplitude was established before LTP was induced by theta burst stimulation (TBS; 10
theta bursts were applied at 15 s intervals, each theta burst consisted of 10 bursts at 200
ms intervals, and each burst consisted of four 100-Hz pulses). Miniature EPSCs were
isolated by focally applying TTX (1 µM) to the DG through a local perfusion system
(AutoMate Scientific). Miniature EPSCs were analyzed by event detection software
(wDetecta; Dr. John Huguenard, Stanford University). Amplitude measurements were
determined from isolated miniature EPSCs uncontaminated by other EPSCs, and 100
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 29
miniature EPSCs from each granule cell were pooled for each experimental group to
generate cumulative histograms.
Field excitatory postsynaptic potentials (fEPSPs) were recorded with glass
electrodes (~3 ΜΩ tip resistance) filled with 1 M NaCl and 25 mM HEPES, pH 7.3, and
were evoked every 20 s with a parallel bipolar tungsten electrode (FHC). The stimulating
electrode was placed in the same location (halfway between the end of the GC layer and
the vertex of the two blades of the DG, ~75 µm from the GC layer) of the medial
perforant path in the dorsal blade of the DG for all slices. The recording electrode was
also placed in the medial perforant path but ~150 µm closer to CA3 than the recording
electrode and also ~75 µm from the GC layer. fEPSPs were recorded in the presence of
50 µM picrotoxin (Tocris). Measures of synaptic strength and plasticity assessed in each
slice consisted of input-output (I-O) relationships, paired pulse ratios, and LTP; these
measures were recorded in the order listed. Synaptic transmission strength was
assessed by generating I-O curves for fEPSPs; input was the peak amplitude of the fiber
volley and the output was the initial slope of the fEPSP. For each slice, we measured the
fiber volley amplitude and initial slope of the fEPSP responses to a range of stimulation
from 25 to 800 µA, and a response curve was generated for both values. Following the I-
O curve, stimulus strength was then adjusted to be ~30% of the maximal fEPSP. Paired
pulse ratios were determined by evoking two fEPSPs 50 ms apart and dividing the initial
slope of the second fEPSP by the initial slope of the first fEPSP (fEPSP2/fEPSP1). After
measurement of paired-pulse ratios, a 15-min stable baseline was established, and LTP
was induced by theta burst stimulation. Measurements of AMPAR- and NMDAR-
mediated synaptic strength were performed on naïve slices (i.e., no LTP protocol was
performed before or after I/Os). First, measurements of AMPAR-mediated synaptic
strength were recorded in normal ACSF where the overwhelming majority of the initial
fEPSP slope is mediated by AMPARs in a range of stimulation from 25 to 800 µA. Mg2+-
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 30
free ACSF containing 20 µM NBQX was then washed in to relieve blockade of NMDARs
and block AMPARs, respectively. fEPSPs were continued to be evoked every 20 s until
a stable 10-min baseline was reached, indicating all AMPARs were blocked and there
was no further removal of Mg2+ blockade of NMDARs. It typically took ~ 15 min to reach
the beginning of the stable baseline. NMDAR-mediated fEPSPs were then evoked using
the exact same set of stimulus strengths used for the AMPAR I/O curve.
Patch and recording electrodes (3–6 MΩ) were pulled from borosilicate glass
capillary tubing (World Precision Instruments) on a horizontal Flaming-Brown
microelectrode puller (model P-97, Sutter Instruments). Whole-cell voltage-clamp data
were low-pass filtered at 6 kHz (–3 dB, eight-pole Bessel), digitally sampled at 20 kHz
with a Multiclamp 700A amplifier (Molecular Devices), and acquired with a Digidata-
1322A digitizer and pClamp 9.2 software (Molecular Devices). Field recordings were
filtered at 2 kHz (−3 dB, eight-pole Bessel) and digitally sampled data were analyzed
offline with pClamp9 software and OriginPro 8.0 (OriginLab).
Behavioral tests. Morris water maze: The maze consisted of a pool (122-cm diameter)
filled with water (21 ± 1 oC) made opaque with nontoxic white tempera paint powder; the
pool was located in a room surrounded by distinct extra-maze cues. Before hidden
platform training, mice were given four pre-training trials in which they had to swim in a
rectangular channel (15 cm x 122 cm) and mount a platform hidden 1.5 cm below the
surface in the middle of the channel. Mice that did not mount the platform were gently
guided to it and were allowed to sit on it for 10 sec before being removed by the
experimenter. The maximum time allowed per trial in this task was 90 sec. The day after
pre-training, mice were trained in the circular water maze. For hidden platform training,
the platform (14 x 14 cm) was submerged 1.5 cm below the surface. The platform
location remained the same throughout hidden-platform training, but the drop location
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 31
varied semi-randomly between trials. Mice received two training sessions with a 3-h
intersession interval for 5 consecutive days. Each session consisted of two trials with a
10-min intertrial interval. The maximum time allowed per trial in this task was 60 sec. If a
mouse did not find the platform, it was guided to it and allowed to sit on it for 10 sec. For
probe trials, the platform was removed and mice were allowed to swim for 60 sec before
they were removed. The drop location for probe trials was 180o from where the platform
was located during hidden-platform training. After the probe trial, mice were allowed to
rest for 1 day before visible platform training was performed. In the latter task, the
platform location was marked with a visible cue (15 cm tall black-and-white striped pole)
placed on top of the platform. Mice received two training sessions per day with a 3- to 4-
h intersession interval. Each session consisted of two training trials with a 10-min
intertrial interval. The maximum time allowed per trial in this task was 60 sec. For each
session, the platform was moved to a new location, and the drop location varied semi-
randomly between trials.
Novel object recognition: Mice were transferred to the testing room and acclimated
for at least 1 h before testing. The testing was performed in a white round plastic
chamber 35 cm in diameter under a red light. On day 1, mice were habituated to the
testing arena for 30 min. On day 2, each mouse was presented with two identical objects
in the same chamber and allowed to explore freely for 10 min. Three hours after this
training session, mice were placed back into the same arena for the test session, during
which they were presented with an exact replica of one of the objects used during
training and with a novel, unfamiliar object of different shape and texture. Object
locations were kept constant during training and test sessions for any given mouse, but
objects were changed semi-randomly between mice. Arenas and objects were cleaned
with 70% ethanol between each mouse. Behavior was recorded with a video tracking
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 32
system (Noldus). Frequency of object interactions and time spent exploring each object
were recorded for subsequent data analysis.
Novel place recognition: Mice were transferred to the testing room and acclimated for
at least 1 h before testing. The testing was performed in a white plastic chamber (40 x
20 x 20 cm) under red light. On the first day, mice were habituated to the testing arena
for 30 min. On the second day, each mouse was presented with two identical objects
and allowed to explore freely for 10 min. Three hours after training, mice were presented
with the same two objects, only this time one of the objects had been moved to a new
location. Arenas and objects were cleaned with 70% ethanol between each mouse.
Behavior was recorded with a video tracking system (Noldus). Frequency of object
interactions and time spent exploring each object were recorded for subsequent data
analysis.
Passive avoidance: The apparatus consisted of a two-compartment dark/light shuttle
box separated by a guillotine door (Gemini, Avoidance System, San Diego Instruments).
The dark compartment had a stainless-steel shock grid floor. During the acquisition trial,
each mouse was placed in the lit chamber. After a 15-s habituation period, the door
separating the light and dark chambers was opened, and the time before mice entered
the dark chamber was recorded. Immediately after mice entered the dark chamber, the
door was closed and an electric foot shock (0.5 mA, 2 s) was delivered by the floor grids.
Ten seconds later, the mouse was removed from the dark chamber and returned to its
home cage. After 24 h, the re-entrance latency was measured as in the acquisition trial,
except that no foot shock was delivered. The latency to enter the dark chamber was
recorded up to a maximum of 300 s.
Open field: Spontaneous locomotor activity in an open field was measured in an
automated Flex-Field/Open Field Photobeam Activity System (San Diego Instruments,
San Diego, CA). Before testing, mice were transferred to the testing room and
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 33
acclimated for at least 1-hour. Mice were tested in a clear plastic chamber (41 x 41 x 30
cm) for 15 min, with two 16 x 16 photobeam arrays detecting horizontal and vertical
movements. The apparatus was cleaned with 70% alcohol between testing of each
mouse. Total movements (ambulations) in the outer periphery and center of the open
field were recorded for further data analysis.
Elevated plus maze: The elevated plus maze consisted of two open (without walls)
and two enclosed (with walls) arms elevated 63 cm above the ground (Hamilton-Kinder,
Poway, CA). Mice were allowed to habituate in the testing room under dim light for 1 h
before testing. During testing, mice were placed at the junction between the open and
closed arms of the plus maze and allowed to explore for 5 min. The maze was cleaned
with 70% alcohol between testing of each mouse. Total distance traveled and time spent
in both the open and closed arms were calculated for data analysis.
Statistical analyses. Statistical analyses were performed with GraphPad Prism or
SPSS v13.0 (SPSS). Data distribution was assessed by Kolmogorov-Smirnoff non-
parametric test of equality. Differences between two means were assessed by paired or
unpaired t test. Differences among multiple means were assessed, as indicated, by one-
way, two-way or repeated-measures ANOVA, followed by Bonferroni’s, Dunn’s, Kruskal-
Wallis’s or Tukey’s post-hoc test. Error bars represent s.e.m. Null hypotheses were
rejected at the 0.05 level.
REFERENCES
51 Henderson, J. T. et al. The receptor tyrosine kinase EphB2 regulates NMDA-dependent synaptic function. Neuron 32, 1041–1056 (2001).
52 Kessels, H. W. & Malinow, R. Synaptic AMPA receptor plasticity and behavior. Neuron 61, 340–350 (2009).
53 Hsieh, H. et al. AMPAR removal underlies Aβ-induced synaptic depression and dendritic spine loss. Neuron 52, 831–843 (2006).
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 34
54 Li, S. et al. Soluble oligomers of amyloid Beta protein facilitate hippocampal long-term depression by disrupting neuronal glutamate uptake. Neuron 62, 788–801 (2009).
55 Johnson-Wood, K. et al. Amyloid precursor protein processing and Aβ42 deposition in a transgenic mouse model of Alzheimer disease. Proc. Natl. Acad. Sci. USA 94, 1550–1555 (1997).
56 Fitzjohn, S. M. et al. Age-related impairment of synaptic transmission but normal long-term potentiation in transgenic mice that overexpress the human APP695SWE mutant form of amyloid precursor protein. J. Neurosci. 21, 4691–4698 (2001).
57 Palop, J. J. et al. Aberrant excitatory neuronal activity and compensatory remodeling of inhibitory hippocampal circuits in mouse models of Alzheimer’s disease. Neuron 55, 697–711 (2007).
58 Klein, W. L. Aβ toxicity in Alzheimer's disease: Globular oligomers (ADDLs) as new vaccine and drug targets. Neurochem. Int. 41, 345–352 (2002).
59 Cheng, I. et al. Accelerating amyloid-β fibrillization reduces oligomer levels and functional deficits in Alzheimer disease mouse models. J. Biol. Chem. 282, 23818–23828 (2007).
60 Rockenstein, E. M. et al. Levels and alternative splicing of amyloid b protein precursor (APP) transcripts in brains of transgenic mice and humans with Alzheimer’s disease. J. Biol. Chem. 270, 28257–28267 (1995).
61 Mucke, L. et al. High-level neuronal expression of Aβ1-42 in wild-type human amyloid protein precursor transgenic mice: Synaptotoxicity without plaque formation. J. Neurosci. 20, 4050–4058 (2000).
62 Palop, J. J. et al. Neuronal depletion of calcium-dependent proteins in the dentate gyrus is tightly linked to Alzheimer’s disease-related cognitive deficits. Proc. Natl. Acad. Sci. USA 100, 9572–9577 (2003).
63 Cheng, I. et al. Aggressive amyloidosis in mice expressing human amyloid peptides with the Arctic mutation. Nat. Med. 10, 1190–1192 (2004).
64 Koo, E. H. & Squazzo, S. L. Evidence that production and release of amyloid b-protein involves the endocytic pathway. J. Biol. Chem. 269, 17386–17389 (1994).
65 Walsh, D. M. et al. Naturally secreted oligomers of amyloid β protein potently inhibit hippocampal long-term potentiation in vivo. Nature 416, 535–539 (2002).
66 Xia, Z., Dudek, H., Miranti, C. K. & Greenberg, M. E. Calcium influx via the NMDA receptor induces immediate early gene transcription by a MAP kinase/ERK-dependent mechanism. J. Neurosci. 16, 5425–5436 (1996).
67 Snyder, E. M. et al. Regulation of NMDA receptor trafficking by amyloid-β. Nat. Neurosci. 8, 1051–1058 (2005).
68 Lois, C., Hong, E. J., Pease, S., Brown, E. J. & Baltimore, D. Germline transmission and tissue-specific expression of transgenes delivered by lentiviral vectors. Science 295, 868–872 (2002).
69 Franklin, K. B. J. & Paxinos, G. The mouse brain in stereotaxic coordinates. (Academic Press, Inc., 1997).
70 Mueller-Steiner, S. et al. Anti-amyloidogenic and neuroprotective functions of cathepsin B: Implications for Alzheimer's disease. Neuron 51, 703–714 (2006).
71 Sun, B. et al. Imbalance between GABAergic and glutamatergic transmissions impairs adult neurogenesis in an animal model of Alzheimer’s disease. Cell Stem Cell 5, 624–633 (2009).
SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature09635
WWW.NATURE.COM/NATURE | 35