Reversing EphB2 depletion rescues cognitive functions …basicmed.med.ncku.edu.tw/admin/up_img/0513-1.pdf · Reversing EphB2 depletion rescues cognitive functions in Alzheimer model

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


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



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

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



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



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



www.nature.com/nature

www.nature.com/reprints



mailto:[email protected]

!"#$%&

!"#$%&'()(*'(*+,(*(-./)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


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'&



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).



!"#$%&'()#*+, - - ... - -

/

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



!"#$ %&'()*(+(,-.#$ %/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



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.



!"#$%& ' '(#)!"#$%& * '

' '''

+,-)!"#$%& ' '(#)!"#$%& * '

' '''

.

/ 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



!

"

# $

%&'()

*+"+,-.

%&'()

*+"+,-.

/'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.



!

"

"#$%

"#%"

"#&%

'#""

'#$%

()*+,-./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



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



!

"

#$%&'$()

*+,-./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



!

"#$%&'()*+,-

.

./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.



!

!"#$%

&$'() *'+,-.

/.

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



!"#$%

&%'()*+,#(

-!.%&/#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.



! " #

$%&'

()*+,)-*+,)./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.



! !

!

"#$%&'()*

+,-(#,$)%./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



!"#$%&'()*&+,

-&++,'.&%'/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



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.



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.



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)



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



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



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



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)



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



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).



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



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



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+-



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



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



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



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).



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).



Documents

Reversing EphB2 depletion rescues cognitive functions …basicmed.med.ncku.edu.tw/admin/up_img/0513-1.pdf · Reversing EphB2 depletion rescues cognitive functions in Alzheimer model