Reversal of learning deficits in a Tsc2+/– mouse model oftuberous sclerosisDan Ehninger1, Sangyeul Han2, Carrie Shilyansky1, Yu Zhou1, Weidong Li1, David J Kwiatkowski3,Vijaya Ramesh2 & Alcino J Silva1

Tuberous sclerosis is a single-gene disorder caused by

heterozygous mutations in the TSC1 (9q34) or TSC2 (16p13.3)

gene1,2 and is frequently associated with mental retardation,

autism and epilepsy. Even individuals with tuberous sclerosis

and a normal intelligence quotient (approximately 50%)3–5 are

commonly affected with specific neuropsychological problems,

including long-term and working memory deficits6,7. Here we

report that mice with a heterozygous, inactivating mutation in

the Tsc2 gene (Tsc2+/– mice)8 show deficits in learning and

memory. Cognitive deficits in Tsc2+/– mice emerged in the

absence of neuropathology and seizures, demonstrating that

other disease mechanisms are involved5,9–11. We show that

hyperactive hippocampal mammalian target of rapamycin

(mTOR) signaling led to abnormal long-term potentiation in the

CA1 region of the hippocampus and consequently to deficits in

hippocampal-dependent learning. These deficits included

impairments in two spatial learning tasks and in contextual

discrimination. Notably, we show that a brief treatment with the

mTOR inhibitor rapamycin in adult mice rescues not only the

synaptic plasticity, but also the behavioral deficits in this

animal model of tuberous sclerosis. The results presented here

reveal a biological basis for some of the cognitive deficits

associated with tuberous sclerosis, and they show that

treatment with mTOR antagonists ameliorates cognitive

dysfunction in a mouse model of this disorder.

To address the mechanisms of tuberous sclerosis–related cognitivedeficits, we first determined whether Tsc2+/– mice had learning andmemory impairments. Because the tuberous sclerosis genes are highlyexpressed in adult brain, including hippocampus12, and hippocampaldysfunction could contribute to the tuberous sclerosis phenotype, westudied spatial learning in the Morris water maze. Initially, mice weretested in a hippocampus-dependent version of the Morris water maze,in which they use distal spatial cues to navigate and locate a hiddenescape platform (Fig. 1a). To assess whether the mice had learned theposition of the escape platform, we gave a probe trial (during whichthe platform was removed from the pool) at the end of training. Probe

trial data show that searches from wild-type (WT) mice were moreselective than those from Tsc2+/– mice, indicating spatial learningdeficits in these mutants (Fig. 1a). Studies with the hippocampus-independent, visible version of the water maze (SupplementaryTable 1 online) revealed no differences between Tsc2+/– mice andWT controls, suggesting that motivational, perceptual or motordifferences are unlikely to account for the observed spatial learningphenotype. Moreover, additional behavioral characterization revealednormal motor skills, anxiety, exploratory activity and social approachbehavior in Tsc2+/– mice (Supplementary Table 1). These results showthat spatial learning in the Morris water maze was poorer in Tsc2+/–

mice than in WT controls.Because humans affected by tuberous sclerosis may have working

memory deficits7, and because the hippocampus can contribute toworking memory13, we next tested Tsc2+/– mice on a hippocampus-dependent win-shift version of the eight-arm radial maze (Fig. 1b), acommonly used working memory task14. In this task, food-deprivedmice were first allowed to retrieve food pellets from four accessiblearms of an eight-arm maze (phase A). After a retention interval of2 min, mice were brought back to the maze (phase B) and given accessto all eight arms (only the four previously blocked arms were nowbaited). In this task, mice can commit two types of errors: within-phase errors are committed when mice enter an arm previously visitedin the same phase, and across-phase errors are committed when miceenter an arm in phase B that they had already visited in phase A.Although Tsc2+/– mice did not differ from wild-type controls in thenumber of within-phase errors (data not shown), they did makesignificantly more across-phase errors (P ¼ 0.0364; Fig. 1b). Thispattern of errors in the radial arm maze is consistent with deficits inhippocampal function13,14.

Discriminating between similar contexts is known to also dependon hippocampal function15. Mice were trained in context condition-ing with three 0.75-mA shocks and tested in the training context or anovel context (detailed in Methods). WT mice showed clear discrimi-nation between both contexts (Fig. 1c). The freezing responses ofTsc2+/– mice, however, lacked context specificity, confirming hippo-campal dysfunction in these mutants15.

©20

08 N

atu

re P

ub

lish

ing

Gro

up

htt

p:/

/ww

w.n

atu

re.c

om

/nat

ure

med

icin

e

Received 19 November 2007; accepted 27 May 2008; published online 22 June 2008; doi:10.1038/nm1788

1Departments of Neurobiology, Psychiatry & Biobehavioral Sciences, Psychology and the Brain Research Institute, University of California, Los Angeles, 695 Charles E.Young Drive South, Los Angeles, California 90095, USA. 2Center for Human Genetic Research, Massachusetts General Hospital, Harvard Medical School, Richard B.Simches Research Center, 185 Cambridge Street, Boston, Massachusetts 02114, USA. 3Genetics Laboratory, Division of Translational Medicine, Brigham andWomen’s Hospital, Harvard Medical School, 221 Longwood Avenue, Boston, Massachusetts 02115, USA. Correspondence should be addressed to A.J.S.(silvaa@mednet.ucla.edu).

NATURE MEDICINE ADVANCE ONLINE PUBLICATION 1

L E T T ERS

The behavioral data presented above demonstrate deficits in threedistinct hippocampal-dependent tasks. Therefore, next we testedwhether TSC protein levels and TSC-mTOR signaling were alteredin the hippocampus of Tsc2+/– mice. Although genetic analyses oftuberous sclerosis have revealed a large number of different mutationsin the tuberous sclerosis genes16, the majority results in prematuretruncation of TSC proteins. Nevertheless, there are also large deletionsand rearrangements that encompass the whole TSC1 or TSC2 gene. Inthe Tsc2+/– mouse model, the Tsc2 gene is disrupted by insertion of aselection cassette in its second coding exon8. We found that tuberin(Tsc2) abundance in whole hippocampal extracts of Tsc2+/– mice wasreduced to about 75% of WT levels, whereas hamartin (Tsc1)abundance was unchanged (Supplementary Fig. 1 online). Tuberinis a GTPase-activating protein that regulates the small G protein Rheb,a strong activator of mTOR17. Consequently, loss of hippocampaltuberin should result in disinhibition of mTOR signaling. We deter-mined the phosphorylation status of the S6 ribosomal protein atSer235 and Ser236 in whole hippocampal extracts, as phosphorylationat these sites is regulated by the mTOR pathway and correlates withtranslational rates. We found higher levels of phosphorylated S6 (p-S6,on Ser235 and Ser236) in hippocampi of Tsc2+/– mice than in WTcontrols (Supplementary Fig. 1), indicating that mTOR signaling wasdisinhibited in Tsc2+/– hippocampus, a result consistent with theinhibitory role of TSC proteins in mTOR signaling17.

Hippocampal activity–dependent synaptic plasticity such as long-term potentiation (LTP) has a major role in learning processes. Todetermine whether synaptic plasticity is altered in Tsc2+/– mice, weperformed standard extracellular field recordings in acute hippocam-pal slices from adult Tsc2+/– mice and WT controls. Basal synaptictransmission, paired-pulse facilitation and early-phase LTP (E-LTP) atthe Schaffer collateral–CA1 synapse were normal in Tsc2+/– mice(Supplementary Fig. 2 online), demonstrating that several aspectsof synaptic function were normal in these mice.

As mTOR signaling plays a part in the regulation of the late-phaseof LTP (L-LTP)18, we next studied this form of synaptic plasticity.Because we detected disinhibited mTOR signaling, we started by usingan E-LTP induction paradigm (100-Hz, 1-s tetanus) and recorded

responses for 4 h. As expected, this paradigm caused decaying LTP inWT slices but was sufficient to induce stable L-LTP in Tsc2+/– slices(Fig. 2). The lower threshold for L-LTP induction in the mutantscould cause a lower threshold for memory consolidation. This couldlead to inappropriate storage of unrelated or unprocessed informa-tion, thus resulting in learning impairments.

The studies presented above show that Tsc2+/– mice are a usefulmodel of some of the learning and memory deficits associated withtuberous sclerosis, and they suggest that enhanced mTOR signalingleads to abnormal L-LTP thresholds that account for these deficits. Totest this hypothesis, we next determined whether rapamycin couldrescue the learning and L-LTP abnormalities of Tsc2+/– mice. Wesuppressed mTOR signaling by administering rapamycin before learn-ing. A treatment regimen (described in Methods) was established thateffectively reduced hippocampal p-S6 (on Ser235 and Ser236) abun-dance (Supplementary Fig. 1) and did not measurably affect behaviorin WT mice (Supplementary Fig. 3 online). After rapamycin treat-ment, mice were trained and tested in context discrimination asdescribed above. Rapamycin did not affect context discrimination inWT controls; strikingly, however, this treatment reversed the contextdiscrimination deficits of Tsc2+/– mice (Fig. 3a).

Next, we tested whether rapamycin treatment could also amelioratethe learning deficits of Tsc2+/– mice in the Morris water maze(Fig. 3b). Before daily training, we treated WT and mutant micewith rapamycin or vehicle and assessed spatial learning with a probetrial after completion of training. Measurements of quadrant occu-pancy and target crossings indicated selective searching in WT mice onrapamycin and vehicle, but not in vehicle-treated Tsc2+/– mice(Fig. 3b). Notably, rapamycin-treated Tsc2+/– mice showed selectivesearching, as shown by their preference for the target quadrant andtarget position (Fig. 3b). Collectively, these results show that thelearning impairments described for Tsc2+/– mice are caused byenhanced mTOR signaling and that a brief pharmacological treatmentwith rapamycin can reverse these deficits in adult mice. Unlikerapamycin, the 3-hydroxy-3-methyl-glutaryl-CoA (HMG-CoA)reductase inhibitor lovastatin (which rescued learning deficits in amouse model of neurofibromatosis type 1; see Supplementary Fig. 4online) did not rescue the spatial learning deficits of Tsc2+/– mice(Supplementary Fig. 4), a result that attests to the specificity of therapamycin effects.

To explore the hypothesis that the change in L-LTP thresholdscontributes to the learning deficits of Tsc2+/– mice, we next testedwhether rapamycin also reverses the abnormal LTP in Tsc2+/– mice. As

©20

08 N

atu

re P

ub

lish

ing

Gro

up

htt

p:/

/ww

w.n

atu

re.c

om

/nat

ure

med

icin

e

****

0

10

20

30

40

50a

b c

Per

cent

age

time

spen

t in

quad

rant

s

WT

AROALT

0

1

2

3

4

Num

ber

of ta

rget

cro

ssin

gsWT

*

****

n.s.

Fre

ezin

g (%

tim

e)

WT

Test in novel contextTest in training contextBaseline

010203040506070

8090

100n.s.***

Num

ber

of a

cros

s-ph

ase

erro

rs

1 2 3 4 5 6 7 80.5

1

1.5

2

2.5

3WT

*

Session

Tsc2+/– Tsc2 +/–

Tsc2+/–

Tsc2 +/–

Figure 1 Tsc2 +/– mice show learning deficits in three hippocampus-

dependent tasks. (a) Quadrant occupancy and target crossings during the

probe trial given after completion of Morris water maze training (n ¼ 12

mice per genotype; two-way ANOVA for quadrant occupancy with genotype

as between-subjects factor and pool quadrant as within-subjects factor,

genotype � pool quadrant interaction: F (3,88) ¼ 4.763, P ¼ 0.004; two-

way ANOVA for target crossings with genotype as between-subjects factor

and pool quadrant as within-subjects factor, effect of genotype: F (1,88) ¼4.278, P ¼ 0.0415). Pool quadrants: target quadrant (T), adjacent right

(AR), adjacent left (AL), opposite quadrant (O). Dashed line marks chance

performance in the Morris water maze. (b) Number of across-phase errors in

eight-arm radial maze plotted against training session (n ¼ 19 mice per

genotype; one-way repeated-measures ANOVA with genotype as between-

subjects factor: F(1,36) ¼ 4.724, P ¼ 0.0364). (c) Context discrimination:

freezing scores before shock (baseline) and during the test in the trainingcontext (WT mice: n ¼ 11 mice; Tsc2+/– mice: n ¼ 9 mice) or the

novel context (WT mice: n ¼ 10 mice; Tsc2+/– mice: n ¼ 9 mice).

*P o 0.05, ***P o 0.001, n.s., not significant (P 4 0.05). Data

represent means ± s.e.m.

L E T TERS

2 ADVANCE ONLINE PUBLICATION NATURE MEDICINE

above, LTP was induced with a 100-Hz, 1-s tetanus and responses wererecorded for 4 h (slices were perfused with rapamycin or vehicle).Vehicle-treated Tsc2+/– slices showed stable L-LTP after induction withthe E-LTP stimulation paradigm, whereas LTP in vehicle-treated WTslices decayed throughout recording (Supplementary Fig. 5 online).As expected, rapamycin did not affect the E-LTP of WT slices; however,this treatment reversed the abnormal induction of L-LTP in Tsc2+/–

slices (Supplementary Fig. 5). These findings indicate that treatmentwith rapamycin rescues the abnormal L-LTP of Tsc2+/– mutants, just asit rescues their learning deficits, and that enhanced mTOR signalingcauses both phenotypes.

To test the efficacy of rapamycin in modulating other tuberoussclerosis brain phenotypes, we studied a homozygous conditionaldeletion of Tsc1 in neurons of the postnatal forebrain: floxed Tsc1mice (Tsc1cc)19 were crossed with mice expressing Cre recombinaseunder the control of the promoter for aCaMKII (aCaMKII-Cre)20.Not unexpectedly, the resulting mice had a much more profoundlyimpaired phenotype than heterozygous Tsc2+/– mice (Fig. 4,Supplementary Fig. 6 online, Supplementary Table 1 and data

not shown). The majority of Tsc1cc–aCaMKII-Cre mice died withinthe first postnatal weeks (data not shown). Rare surviving Tsc1cc–aCaMKII-Cre mice (B4% of what would be predicted byMendelian ratios) developed extreme macroencephaly, with brainweights being B2.5 times that of control brains (Supple-mentary Fig. 6). Brain enlargement was largely due to massiveneuronal hypertrophy and was accompanied by astrogliosis (Supple-mentary Fig. 6). Tsc1cc–aCaMKII-Cre mice showed a pathologicalhindlimb clasping reflex (Supplementary Fig. 6) and wereseverely hypoactive.

Rapamycin treatment substantially increased survival ofTsc1cc–aCaMKII-Cre mice (Fig. 4a; to B70% of what would beexpected by Mendelian ratios), decreased brain enlargement(Fig. 4b) and improved the neurological findings in Tsc1cc–aCaMKII-Cre mice (Fig. 4c,d). These data attest to the efficacyof rapamycin in modulating brain phenotypes associated withmutations in Tsc genes. Our findings suggest that rapamycinmay be able to ameliorate putative developmental deficits,potentially contributing to the profound cognitive abnormalitiespresent in a subset of individuals with tuberous sclerosis4.Of note, a null mutation of the Pten gene caused some of the samephenotypes described for the Tsc1cc–aCaMKII-Cre mice, andmTOR antagonists also rescued these phenotypes21. This resultis consistent with the role of PTEN as an upstream regulator of theTSC-mTOR pathway.

©20

08 N

atu

re P

ub

lish

ing

Gro

up

htt

p:/

/ww

w.n

atu

re.c

om

/nat

ure

med

icin

e

–10 0

1 × 100 Hz

40 12080 160 200

Time (min)

2400

25

50

75

100

125

150

175

200

fEP

SP

slo

pe (

% o

f bas

elin

e)Tsc2+/ –,control

Tsc2 +/ –,tetanized

WT,control

WT,tetanized

0

10

20

30

40

50

Per

cent

age

time

spen

t in

quad

rant

s

Vehicle 1 55

Rapamycin (mg/kg)

Vehicle 1

Rapamycin (mg/kg)

Vehicle 15 5

Rapamycin (mg/kg)

Vehicle 1

Rapamycin (mg/kg)

WT

0

0.5

1

1.5

2

2.5

3

3.5

Num

ber

of ta

rget

cro

ssin

gs

WT

*** *** n.s.** ** ** * *n.s. * *n.s.

AROALT

0

WT Tsc2 +/–

RapamycinVehicle RapamycinVehicle

n.s.

1020

3040506070

8090

100

Fre

ezin

g (%

tim

e)

Test in novel contextTest in training contextBaseline

*********

a b

Tsc2 +/– Tsc2+/–

Figure 3 Rapamycin reversed context discrimination and spatial learning deficits in Tsc2 +/– mice. (a) Context discrimination: freezing scores before shock

(baseline) and during the test in the training context (vehicle-treated WT mice, n ¼ 12; rapamycin-treated WT mice, n ¼ 13; vehicle-treated Tsc2+/– mice,

n ¼ 9; rapamycin-treated Tsc2+/– mice, n ¼ 9) or the novel context (vehicle-treated WT mice, n ¼ 12; rapamycin-treated WT mice, n ¼ 12; vehicle-treated

Tsc2+/– mice, n ¼ 10; rapamycin-treated Tsc2+/– mice, n ¼ 11). (b) Quadrant occupancy and target crossings during the probe trial that was given after

completion of Morris water maze training (vehicle-treated WT mice, n ¼ 17; WT mice treated with 1 mg/kg rapamycin, n ¼ 15; WT mice treated with5 mg/kg rapamycin, n ¼ 16; vehicle-treated Tsc2+/– mice, n ¼ 16; Tsc2+/– mice treated with 1 mg/kg rapamycin, n ¼ 15; Tsc2+/– mice treated with 5 mg/kg

rapamycin, n ¼ 14; three-way ANOVA for quadrant occupancy with genotype and treatment as between-subjects factors and pool quadrant as within-subjects

factor, genotype � treatment � quadrant interaction: F (6,348) ¼ 2.168, P ¼ 0.0456; three-way ANOVA for target crossings with genotype and treatment as

between-subjects factors and pool quadrant as within-subjects factor, genotype � treatment � quadrant interaction: F(6,348) ¼ 2.908, P ¼ 0.0088).

Dashed line marks chance performance in the Morris water maze. *P o 0.05, **P o 0.01, ***P o 0.001. Data represent means ± s.e.m.

Figure 2 An E-LTP stimulation paradigm elicited L-LTP in Tsc2+/– mice.

Initial fEPSP slopes recorded from hippocampal slices are shown before

(baseline) and following LTP induction (with a 1-s, 100-Hz tetanus) for the

tetanized pathway and for a separate, untetanized pathway (control). Data

are plotted in 2-min blocks for baseline and 8-min blocks for time after LTP

induction (WT slices, n ¼ 9 slices from 9 mice; Tsc2+/– slices, n ¼ 7 slices

from 7 mice; one-way repeated-measures ANOVA with genotype as between-

subjects factor, measure � genotype interaction: F (29,406) ¼ 2.436,

P o 0.0001; t-test, last 10 min of recording, P ¼ 0.0328). Sample traces

show responses (ten responses were averaged) during baseline and the last

10 min of recording, respectively. Scale: vertical bar, 1 mV; horizontal bar,

10 ms. Data represent means ± s.e.m.

L E T T ERS

NATURE MEDICINE ADVANCE ONLINE PUBLICATION 3

The neurocognitive phenotype associated with tuberous sclerosis iscomplex. The distribution of intelligence quotient (IQ) scores ofindividuals with tuberous sclerosis indicates the existence of twodistinct subpopulations3,4: a profoundly impaired group with severemental retardation (30%) and a group with normally distributed IQswith a slight reduction in mean IQ (70%). This group is affected bycertain neuropsychological impairments, such as specific learningdisabilities and memory problems5–7. Similarly, the Tsc2+/� mutationin mice also resulted in a specific behavioral phenotype: it disruptedhippocampal-dependent learning while leaving most other testedbehaviors unaffected (Supplementary Table 1). It will be useful todetermine whether individuals with tuberous sclerosis also showabnormal hippocampal function. Deficient context discrimination isa key component of hippocampal dysfunction in Tsc2+/– mice. It ispossible that deficits in contextual discrimination lead to overgener-alized fear and contribute to anxiety disorders, frequently encounteredin individuals with tuberous sclerosis22. As rapamycin rescued thesecognitive deficits in mice, it may also ameliorate cognitive impair-ments in individuals with tuberous sclerosis. Nevertheless, eventhough our studies showed that rapamycin rescued phenotypesother than spatial learning deficits (that is, Tsc1cc–aCaMKII-Crephenotypes), it is unclear whether rapamycin will reverse cognitivedeficits caused by cerebral malformations or infantile spasms23,24.

Consistent with a recent study11, our studies show that tuberoussclerosis–related cognitive deficits emerged in the absence of sponta-neous seizures and neuropathology, suggesting that other mechanismsare involved in causing cognitive dysfunction. Indeed, we found thatincreased mTOR signaling changed the induction of the late phase ofLTP, a finding that most likely accounts for the learning and memorydeficits in the Tsc2+/� mutants. Notably, our results are consistent withprevious findings that implicated TSC-mTOR signaling in synapticplasticity. First, the TSC-mTOR pathway regulates the local translationof specific proteins known to have crucial roles in LTP18,25. Second,Tsc1–/– hippocampal neurons were shown to have morphological and

functional synaptic abnormalities26. Third, changes in the induction ofthe late phase of LTP as well as learning deficits have also beenreported in mice lacking 4E-BP2, which also show increased mTOR-dependent translational initiation27. Altogether, these results stronglysuggest that the change in L-LTP that we found accounts for thelearning deficits of Tsc2+/� mice.

A key finding reported here is that a brief rapamycin treatment ofadult Tsc2+/� mice rescued not only their physiological abnormalities,but, more importantly, their learning and memory deficits. Our resultsalso suggest that there is an optimal range for translational activationduring learning and memory processes: decreases in translation (forexample, with high doses of rapamycin; Supplementary Fig. 3)28 aswell as hyperactivation of translation (for example, by the Tsc2+/�

mutation) disrupt learning and memory. Excessive mTOR-dependentneuronal protein synthesis is also involved in other neurogeneticsyndromes linked to mental retardation and autism29,30. Identifyingthe crucial mechanisms disrupted by changes in translation will notonly provide insights into the role of this process in learning andmemory, but also may inform treatment development.

A recent study reported hippocampal-dependent learning deficits inTsc1+/– mice11. Unlike Tsc2+/� mutants, however, Tsc1+/� mice alsoshow abnormal social behavior. This difference could be due either tothe specific gene disrupted or to the genetic background. Becausethere is a close functional interaction between the Tsc1 and Tsc2proteins, it is more likely that genetic background8,11 accounts for thedifferences in social behavior. Genetic background could also beresponsible for the high variability of the human tuberous sclerosisphenotype5. Another major model of tuberous sclerosis, the Eker rat,shows synaptic plasticity abnormalities, including LTP and LTDdeficits10. Of note, these rats show many features associated withtuberous sclerosis, but no behavioral deficits9. Instead, behavioralanalysis reveals enhancements in episodic-like memory9, a resultconsistent with the idea that modifier genes in the genetic backgroundinfluence the expressivity of tuberous sclerosis phenotypes5,8,11.

In summary, our results show that Tsc2+/– mice model some of theneurocognitive deficits associated with tuberous sclerosis. As predictedpreviously5, we found that treatment with rapamycin in adult micecan reverse these deficits. Contrary to prevailing hypotheses23,24, butin agreement with a recent study11 and earlier predictions5,9,10, ourdata show that tuberous sclerosis–related cognitive deficits are notnecessarily the consequence of neuroanatomical abnormalities (that is,tubers) or seizures. Thus, it is possible that abnormalities in mTORsignaling and the associated physiological phenotypes (that is, abnor-mal synaptic plasticity) have a key role in some of the neurocognitivedeficits associated with tuberous sclerosis. Our studies reveal thatdisinhibited mTOR signaling leads to a lower threshold for theinduction of L-LTP, which probably underlies the behavioral deficitsdescribed here (for example, abnormal potentiation leading to inap-propriate information storage). Altogether, these findings demonstrate

©20

08 N

atu

re P

ub

lish

ing

Gro

up

htt

p:/

/ww

w.n

atu

re.c

om

/nat

ure

med

icin

e

Bra

in w

eigh

t (g)

Rapamycin

Untreated

Rapamycin

Untreated

***

Controls Tsc1c+

αCaMKII-Cre

Tsc1cc

αCaMKII-Cre

Tsc1c+

αCaMKII-Cre

Tsc1cc

αCaMKII-CreUnt

reat

ed

Rapam

ycin

Per

cent

age

of Tsc1

cc–α

CaM

KII-

Cre

mic

e w

ith h

indl

imb

clas

ping

Tsc1

cc–

αCaM

KII-

Cre

mic

e(%

of e

xpec

ted)

0

20

40

60

80

100

0

250

500

750

1,000

1,250

1,500

1,750

2,000

2,250

2,500

Dis

tanc

e m

oved

(cm

)

Controls

***

0

20

40

60

80

100

Untre

ated

Rapam

ycin

a b

c d

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1Figure 4 Rapamycin treatment rescues lethality, reduces abnormal brain

enlargement and improves neurological findings in Tsc1cc–aCaMKII-Cre

mice. (a) Percentage of untreated and rapamycin-treated Tsc1cc–aCaMKII-

Cre mice that survived to the age of three months. (b) Brain weight at the

age of 3 months (untreated Tsc1cc–aCaMKII-Cre mice, n ¼ 3 mice;

rapamycin-treated Tsc1cc–aCaMKII-Cre mice, n ¼ 4 mice; t-test, P o0.001). (c) Percentage of untreated and rapamycin-treated Tsc1cc–aCaMKII-

Cre mice that showed a pathological hindlimb clasping reflex at the age of

3 months. (d) Ambulatory distance in the open field assessed at the age of

3 months (untreated Tsc1cc–aCaMKII-Cre mice, n ¼ 3 mice; rapamycin-

treated Tsc1cc–aCaMKII-Cre mice, n ¼ 7 mice; t-test, P o 0.001).

***P o 0.001. Data represent means ± s.e.m.

L E T TERS

4 ADVANCE ONLINE PUBLICATION NATURE MEDICINE

the role of TSC-mTOR signaling in the modulation of L-LTP,learning and memory; they also raise the possibility that mTORinhibitors could be used to treat cognitive deficits associated withtuberous sclerosis.

METHODSMice. Tsc2+/– mice were generated as previously described8 (genetic back-

ground: C57BL/6NCrl). We used a balanced ratio of 3–6-month-old males and

females. The generation of aCaMKII-Cre mice20 (genetic background: C57BL/

6NTac) and floxed Tsc1 (Tsc1c allele) mice19 (mixed genetic background

composed of C57BL/6J, 129/SvJae, BALB/cJ) were also previously described.

For the generation of experimental mice, we crossed floxed Tsc1 mice and

aCaMKII-Cre transgenic mice. We show data separately for the Tsc1cc–

aCaMKII-Cre and Tsc1c+–aCaMKII-Cre groups. When we found no significant

differences between the remaining control groups, we pooled the data and show

them as a single control group. All experiments were conducted blind to

genotype and treatment condition. The Chancellor’s Animal Research Com-

mittee at the University of California, Los Angeles approved the research

protocols used here.

Rapamycin. We freshly dissolved rapamycin (LC Laboratories) in vehicle

solution (specified below) before use. For the Tsc2 in vivo experiments, we

administered rapamycin intraperitoneally once daily at a dose of 5 mg/kg or

1 mg/kg (vehicle: 100% DMSO; DMSO dose: 2 ml/kg; volume of a single

injection did not exceed 50 ml). We gave injections (5 mg/kg) for 5 d before and

on the day of fear conditioning (3 h before conditioning). In the water maze

experiment, we injected mice daily 3 h before training (5 mg/kg or 1 mg/kg).

For hippocampal slice physiology, we used rapamycin at a concentration of

200 nM; vehicle was 0.1% DMSO in oxygenated artificial CSF (ACSF). We

continuously perfused slices with rapamycin (or vehicle) throughout the

recording and for at least 30 min before tetanization. For the Tsc1cc–aCaM-

KII-Cre study, we injected rapamycin subcutaneously at a dose of 0.2 mg/kg per

day (vehicle: 20% DMSO and 80% sterile physiological saline). We adminis-

tered a single daily injection starting at postnatal day 1.

Behavior. For the hidden version of the Morris water maze, we trained mice

with four training trials per day for 5 consecutive d. The escape platform was

hidden 1 cm under the water surface in a constant location of the pool. We

released mice into the pool from one of seven starting locations. Training trials

ended when the mouse was on the platform or 60 s had elapsed, whichever

came first. Mice remained for 15 s on the platform before they were removed

from the pool. We gave training trials in blocks of two spaced about 90 min

apart. We assessed spatial learning with a probe trial (during which the

platform was removed from the pool) given after completion of training. We

used ANOVAs and t-tests for analysis of data (quadrant occupancy, target

crossings). After initial ANOVA analysis, we determined specificity of searching

by comparing target quadrant measures to average measures of the other

quadrants (paired t-test).

For the win-shift version of the eight-arm radial maze, we deprived mice of

food until they reached 85% of their free-feeding body weight, shaped them to

consume 20 mg food pellets and habituated them to the maze. On phase A of

the task, we allowed mice to retrieve food pellets from four accessible arms of

the eight-arm maze (access to the remaining arms was blocked). After a 2-min

delay interval (in home cage), for phase B, we returned mice to the maze and

allowed them to freely explore all eight arms. Food rewards were only located in

the four previously blocked arms. We scored entries into arms that were visited

on phase A as across-phase errors. We scored re-entries into arms that were

visited earlier during the same phase as within-phase errors. Mice received one

daily session for 8 consecutive d. We used one-way repeated-measures ANOVA

with genotype as the between-subjects factor for statistical comparison of data

(number of across-phase and within-phase errors).

For context fear conditioning, we fear-conditioned mice to the training

context with three 0.75 mA shocks (2 s each). Training sessions were 5 min in

duration. We assessed the conditioned fear response (as measure by the

percentage of the time spent freezing during test, as scored by an experienced

observer) to the training context 24 h after training (5 min session). To test for

context specificity of the conditioned response, we tested a separate group of

mice in a novel context. This context was situated in another room with

different distal cues and with some shared features (horizontal floor, ethanol

scent, white background noise) with the training context, but was otherwise

distinct (plastic instead of metal grid floor, angled instead of rectangular walls,

red instead of white light, different distal spatial cues in room). We used t-tests

to compare percentage time spent freezing to training and novel context for the

different experimental conditions.

Hippocampal slice physiology. After killing mice with isoflurane, we extracted

brains and completed slice preparation in ice-cold ACSF that was saturated

with 95% O2 and 5% CO2 and contained (in mM): 120 NaCl, 20 NaHCO3,

3.5 KCl, 2.5 CaCl2, 1.3 MgSO4 1.25 NaH2PO4 and 10 D-glucose. We obtained

400-mm sagittal slices with a Leica VT1000S vibratome and allowed them to

recover at 23 1C for at least 70 min. We made recordings in a submerged

chamber continuously perfused with oxygenated ACSF (29 1C) at a rate of

3 ml/min. We made extracellular recordings of field excitatory postsynaptic

potentials (fEPSPs) with platinum-iridium electrodes in CA1 stratum radia-

tum. We stimulated Schaffer collaterals (pulse duration: 100 ms) with bipolar

platinum electrodes. For the input-output function, we stimulated slices with

incrementally increasing current (from 10 mA to 100 mA in 10-mA increments)

and recorded fEPSP responses. For paired-pulse facilitation, we applied two

pulses with different interstimulus intervals (10, 20, 50, 100, 200 or 400 ms).

We induced LTP with theta burst stimulation (four bursts) or 1-s, 100-Hz

stimulation. We determined initial fEPSP slopes and normalized them to the

average baseline fEPSP slopes. We used repeated-measures ANOVAs on all

responses after LTP induction and t-tests (on the average of the last 10 min of

recording) for statistical analysis.

Statistical Analyses. Statistical tests (ANOVAs or t-tests) and parameters for

data analysis were chosen a priori and are described in more detail in the

sections above. P o 0.05 was considered statistically significant.

Additional methods. Detailed methodology is described in Supplementary

Methods online.

Note: Supplementary information is available on the Nature Medicine website.

ACKNOWLEDGMENTSThe authors would like to thank B. Wiltgen, A. Matynia, Y.-S. Lee, R. Czajkowski,G. Ehninger and G. Kempermann for helpful comments on an earlier version ofthe manuscript and for valuable discussions, J.N. Crawley for helpful suggestionsregarding the social interaction paradigm, M. Meredyth-Steward for editing help,I. Roder for statistical advice and R. Chen and K. Cai for technical support. Thiswork was supported by the following grants: Deutsche ForschungsgemeinschaftEH223/2-1 to D.E., US National Institutes of Health R01-NS38480 to A.J.S.,US National Institutes of Health NS24279 and Autism Speaks to V.R.

AUTHOR CONTRIBUTIONSD.E. and A.J.S. conceptualized the research; D.E. performed behavioralexperiments and the Tsc1cc–aCaMKII-Cre study (Figs. 1, 3 and 4; SupplementaryFigs. 3, 4, 6; Supplementary Table 1); D.E., C.S. and Y.Z. contributed to slicephysiology experiments (Fig. 2; Supplementary Figs. 2 and 5); S.H., V.R., D.E.and W.L. contributed to western blot experiments (Supplementary Fig. 1);D.E. analyzed the data; D.J.K. provided Tsc1cc and Tsc2+/– founders for themouse colony; D.E. and A.J.S. wrote the manuscript.

Published online at http://www.nature.com/naturemedicine/

Reprints and permissions information is available online at http://npg.nature.com/

reprintsandpermissions/

1. European Chromosome 16 Tuberous Sclerosis Consortium. Identification and charac-terization of the tuberous sclerosis gene on chromosome 16. Cell 75, 1305–1315(1993).

2. van Slegtenhorst, M. et al. Identification of the tuberous sclerosis gene TSC1 onchromosome 9q34. Science 277, 805–808 (1997).

3. Joinson, C. et al. Learning disability and epilepsy in an epidemiological sampleof individuals with tuberous sclerosis complex. Psychol. Med. 33, 335–344(2003).

4. de Vries, P.J. & Prather, P.A. The tuberous sclerosis complex. N. Engl. J. Med. 356, 92,author reply 93–94 (2007).

5. de Vries, P.J. & Howe, C.J. The tuberous sclerosis complex proteins—a GRIPP oncognition and neurodevelopment. Trends Mol. Med. 13, 319–326 (2007).

©20

08 N

atu

re P

ub

lish

ing

Gro

up

htt

p:/

/ww

w.n

atu

re.c

om

/nat

ure

med

icin

eL E T T ERS

NATURE MEDICINE ADVANCE ONLINE PUBLICATION 5

6. Harrison, J.E., O’Callaghan, F.J., Hancock, E., Osborne, J.P. & Bolton, P.F. Cognitivedeficits in normally intelligent patients with tuberous sclerosis. Am. J. Med. Genet. 88,642–646 (1999).

7. Ridler, K. et al. Neuroanatomical correlates of memory deficits in tuberous sclerosiscomplex. Cereb. Cortex 17, 261–271 (2007).

8. Onda, H., Lueck, A., Marks, P.W., Warren, H.B. & Kwiatkowski, D.J. Tsc2+/� micedevelop tumors in multiple sites that express gelsolin and are influenced by geneticbackground. J. Clin. Invest. 104, 687–695 (1999).

9. Waltereit, R. et al. Enhanced episodic-like memory and kindling epilepsy in a rat modelof tuberous sclerosis. J. Neurochem. 96, 407–413 (2006).

10. von der Brelie, C., Waltereit, R., Zhang, L., Beck, H. & Kirschstein, T. Impairedsynaptic plasticity in a rat model of tuberous sclerosis. Eur. J. Neurosci. 23, 686–692(2006).

11. Goorden, S.M., van Woerden, G.M., van der Weerd, L., Cheadle, J.P. & Elgersma, Y.Cognitive deficits in Tsc1+/� mice in the absence of cerebral lesions and seizures. Ann.Neurol. 62, 648–655 (2007).

12. Murthy, V. et al. Developmental expression of the tuberous sclerosis proteins tuberinand hamartin. Acta Neuropathol. 101, 202–210 (2001).

13. Floresco, S.B., Seamans, J.K. & Phillips, A.G. Selective roles for hippocampal,prefrontal cortical and ventral striatal circuits in radial-arm maze tasks with or withouta delay. J. Neurosci. 17, 1880–1890 (1997).

14. Olton, D., Becker, J. & Handelmann, G. Hippocampus, space and memory. Behav.Brain Sci. 2, 313–365 (1979).

15. Frankland, P.W., Cestari, V., Filipkowski, R.K., McDonald, R.J. & Silva, A.J. The dorsalhippocampus is essential for context discrimination but not for contextual condition-ing. Behav. Neurosci. 112, 863–874 (1998).

16. Sancak, O. et al. Mutational analysis of the TSC1 and TSC2 genes in a diagnosticsetting: genotype-phenotype correlations and comparison of diagnostic DNAtechniques in tuberous sclerosis complex. Eur. J. Hum. Genet. 13, 731–741(2005).

17. Kwiatkowski, D.J. & Manning, B.D. Tuberous sclerosis: a GAP at the crossroads ofmultiple signaling pathways. Hum. Mol. Genet. 14, R251–R258 (2005).

18. Tang, S.J. et al. A rapamycin-sensitive signaling pathway contributes to long-termsynaptic plasticity in the hippocampus. Proc. Natl. Acad. Sci. USA 99, 467–472(2002).

19. Kwiatkowski, D.J. et al. A mouse model of TSC1 reveals sex-dependent lethality fromliver hemangiomas and up-regulation of p70S6 kinase activity in Tsc1 null cells. Hum.Mol. Genet. 11, 525–534 (2002).

20. Dragatsis, I. & Zeitlin, S. CaMKIa-Cre transgene expression and recombinationpatterns in the mouse brain. Genesis 26, 133–135 (2000).

21. Kwon, C.H., Zhu, X., Zhang, J. & Baker, S.J. mTor is required for hypertrophy of Pten-deficient neuronal soma in vivo. Proc. Natl. Acad. Sci. USA 100, 12923–12928(2003).

22. Lewis, J.C., Thomas, H.V., Murphy, K.C. & Sampson, J.R. Genotype and psychologicalphenotype in tuberous sclerosis. J. Med. Genet. 41, 203–207 (2004).

23. O’Callaghan, F.J. et al. The relation of infantile spasms, tubers and intelligence intuberous sclerosis complex. Arch. Dis. Child. 89, 530–533 (2004).

24. Raznahan, A. et al. Biological markers of intellectual disability in tuberous sclerosis.Psychol. Med. 37, 1293–1304 (2007).

25. Jaworski, J. & Sheng, M. The growing role of mTOR in neuronal development andplasticity. Mol. Neurobiol. 34, 205–219 (2006).

26. Tavazoie, S.F., Alvarez, V.A., Ridenour, D.A., Kwiatkowski, D.J. & Sabatini, B.L.Regulation of neuronal morphology and function by the tumor suppressors Tsc1 andTsc2. Nat. Neurosci. 8, 1727–1734 (2005).

27. Banko, J.L. et al. The translation repressor 4E–BP2 is critical for eIF4F complexformation, synaptic plasticity, and memory in the hippocampus. J. Neurosci. 25,9581–9590 (2005).

28. Dash, P.K., Orsi, S.A. & Moore, A.N. Spatial memory formation and memory-enhancingeffect of glucose involves activation of the tuberous sclerosis complex–Mammaliantarget of rapamycin pathway. J. Neurosci. 26, 8048–8056 (2006).

29. Vanderklish, P.W. & Edelman, G.M. Differential translation and fragile X syndrome.Genes Brain Behav. 4, 360–384 (2005).

30. Bear, M.F., Dolen, G., Osterweil, E. & Nagarajan, N. Fragile X: translation in action.Neuropsychopharmacology 33, 84–87 (2008).

©20

08 N

atu

re P

ub

lish

ing

Gro

up

htt

p:/

/ww

w.n

atu

re.c

om

/nat

ure

med

icin

eL E T TERS

6 ADVANCE ONLINE PUBLICATION NATURE MEDICINE

******

*

p-S

6(%

ofW

T/H

c/ve

hicl

egr

oup)

WT

Rapamycin Vehicle Rapamycin Vehicle

350

300

250

200

150

100

50

0

Home cage (Hc)

Context conditioning (Cc)

e

**

b

0

20

40

60

80

100

120

Tube

rin(%

ofW

Tgr

oup)

WT Tsc2+/ –

c

0

20

40

60

80

100

120H

amar

tin(%

ofW

Tgr

oup)

WT0

20

40

60

80

100

120

S6

(%of

WT

grou

p)WT

d

aTuberin

Hamartin

p-S6(Ser235/ 236)

S6

tubulin

Hc Cc Hc Cc Hc Cc Hc Cc

WT Tsc2+/ –

Vehicle

WT

Rapamycin

Tsc2+/ –

Tsc2+/ – Tsc2+/ –

Tsc2+/ –

∝

60

80

100

120

140

160

180

200

Faci

litat

ion

ratio

10Inter-stimulus interval (ms)

20 40050 100 200

Tsc2+/ -WTa

-200

0

200

400

600

800

1000

1200

fEP

SP

slop

e(m

V/

ms)

100908070605040302010Stimulation intensity (µA)

-0.050

0.05

0.10.15

0.20.25

0.3

0.350.4

0.45

Fibe

rvo

lley

(mV

)

100908070605040302010Stimulation intensity (µA)

WTb

-10 0 20 6040

Time (min)

↑TBS (4 bursts)

0

25

50

75

100

125

150

175

200

fEP

SP

slop

e(%

ofba

selin

e)

c

WT

-10 0 20 6040

Time (min)

↑1 × 100 Hz

0

25

50

75

100

125

150

175

200

fEP

SP

slop

e(%

ofba

selin

e)

d

WT

Tsc2+/ -

Tsc2+/ -Tsc2+/ -

01020

3040506070

8090

100

Free

zing

(%tim

e)

Vehicle Rapamycin(150 mg/kg)

**

b

01020

3040506070

8090

100Fr

eezi

ng(%

time)

Vehicle Anisomycin(150 mg/kg)

**

c

a

01020

3040506070

8090

100

Test

Baseline

Rapamycin(5 mg/kg)

Vehicle

Free

zing

(%tim

e)

0

10

20

30

40

50

Per

cent

age

time

spen

tin

quad

rant

s

LovastatinVehicle LovastatinVehicle

WT Tsc2+/ -

** ** n.s. n.s.

AR

O

AL

T

a

b

0

0.5

1

1.5

2

2.5

3

3.5

4

Num

ber

ofta

rget

cros

sing

s

LovastatinVehicle LovastatinVehicle

WT

* n.s.* n.s.

AR

O

AL

T

Tsc2+/ -

-10 0 40 12080 160

Time (min)

200 240

↑1 × 100 Hz

0

25

50

75

100

125

150

175

200

fEP

SP

slop

e(%

ofba

selin

e)

200 nM rapamycin or vehicle

Tsc2+/-,rapamycin,control

Tsc2+/-,rapamycin,tetanized

WT,rapamycin,control

Tsc2+/-,vehicle,control

Tsc2+/-,vehicle,tetanized

WT,vehicle,tetanized

WT,vehicle,control

WT,rapamycin,tetanized

Bra

inw

eigh

t(g

)

***

00.10.2

0.30.40.50.60.7

0.80.9

1

Tsc1c+ CaMKII-Cre∝Controls

b

d

0

5

10

15

20

25

Bod

yw

eigh

t(g

)

Controls Tsc1c+CaMKII-Cre∝

***

a c

Neo

corte

x(N

issl

)H

ippo

cam

pus

(Nis

sl)

Stria

tum

(Nis

sl)

Stria

tum

(GFA

P)

Neo

corte

x(G

FAP

)

Tsc1ccCaMKII-Cre∝

Tsc1cc CaMKII-Cre∝

Tsc1ccCaMKII-Cre∝

Tsc1ccCaMKII-Cre∝

Tsc1ccCaMKII-Cre∝

Control

Control

Control

3 weeks old

2 weeks old

3 months old

Supplementary Table 1

Motor skills, exploratory behavior, anxiety and social approach behavior were normal

in Tsc2+/– mice.

WT littermates Tsc2+/–

n Mean +/– SEM n Mean +/– SEM

(a) Accelerating rotarod

Latency to fall (s)

Trial 1

Trial 2

Trial 3

Trial 4

11

208.2 +/– 19.0

271.8 +/– 18.5

303.2 +/– 27.4

291.0 +/– 20.6

11

175.2 +/– 29.5

232.5 +/– 25.2

267.3 +/– 28.7

272.9 +/– 11.8

(b) Open field

Path length (cm)

Ratio path in outer zone/total path

17

1474.2 +/– 80.0

0.23 +/– 0.012

17

1449.2 +/– 71.2

0.24 +/– 0.017

(c) Elevated plus maze

% time spent in open arms

11

13.4 +/– 2.0

11

14.8 +/– 2.6

(d) Visible version of the

Morris water maze

Escape latency (s)

Block 1

Block 2

Block 3

Block 4

8

52.4 +/– 3.7

49.1 +/– 5.0

37.4 +/– 4.3

24.2 +/– 5.3

8

54.7 +/– 5.3

35.3 +/– 4.1

36.4 +/– 7.4

30.9 +/– 5.7

(e) Shock reactivity

Baseline activity (cm/s)

Activity burst (cm/s)

21

12.3 +/– 1.9

99.1 +/– 7.0

18

10.2 +/– 1.8

104.4 +/– 9.0

(f) Context conditioning

1 × 0.5 mA shock (% time freezing

on test)

1 × 0.75 mA shock (% time freezing

on test)

11

11

40.3 +/– 6.4

56.7 +/– 4.9

12

9

43.6 +/– 5.4

60.8 +/– 6.4

(g) Social interaction

Time exploring conspecific (s)

Time exploring empty cup (s)

12

63.6 +/– 11.7

29.3 +/– 3.0

14

52.8 +/– 7.6

26.0 +/– 4.2

(h) Brain weight

at age 6 weeks (mg)

12

420 +/– 5

18

424 +/– 4

(i) Body weight

at age 6 weeks (g)

8

18.4 +/– 0.6

11

18.6 +/– 4.9

(a) Latencies to fall from the accelerating rod did not significantly differ between genotypes (one-way

repeated-measures ANOVA with genotype as between-subjects factor: F(1,20) = 1.647; P = 0.2141),

indicating that gross motor skills were comparable in Tsc2+/– mice and WT controls.

(b) Exploratory activity was recorded in a 20 min open field session. Tsc2+/– mice and WT littermate

controls covered similar distances (t-test: P = 0.8198), indicating that activity levels were not different.

The ratio between the path length in the outer zone and the total path did not differ between genotypes

(t-test: P = 0.5585), suggesting that anxiety levels were similar in Tsc2+/– mice and WT controls.

(c) Anxiety-related behaviors were also studied in the elevated plus maze. Percent time spent in the open

arms was not significantly different in Tsc2+/– mice and WT controls (t-test: P = 0.6741).

(d) Escape latencies on a visible version of the Morris water maze were not different in Tsc2+/– mice and

WT controls (one-way repeated-measures ANOVA with genotype as between-subjects factor: F(1,14) =

0.116; P = 0.738)

(e) Velocity (cm/s) during the 2 s prior to the onset of a 0.75 mA, 2 s shock (baseline activity) and during

the shock (activity burst) was measured and not found to be different in Tsc2+/– mice and WT controls

(activity burst: t-test: P = 0.6445). This suggests that pain processing and responding was not altered in

Tsc2+/– mice.

(f) We trained Tsc2+/– mice and WT controls in context conditioning with different shock intensities (1 ×

0.5 mA or 1 × 0.75 mA). The results revealed no differences between mutants and controls (test in

training context: 1 × 0.5 mA: t-test: P = 0.6951; 1 × 0.75 mA: t-test: P = 0.6147), indicating that the

contextual discrimination deficits of Tsc2+/– mice were not due to abnormal conditioning responses (i.e.

enhanced conditioning responses).

(g) A large proportion of TSC patients is diagnosed with autism (20-60%), accounting for about 1-4% of

all cases of autism spectrum disorder5,6. Therefore, we compared social approach behavior in Tsc2+/– mice

and WT controls, using a social choice paradigm7. Tsc2+/– mice and WT controls spent both more time

exploring a conspecific than an inanimate object (paired t-tests of time spent exploring conspecific vs.

time spent exploring empty cup; WT: P = 0.0134; Tsc2+/–: P = 0.0025).

(h) Brain weight was not significantly different in Tsc2+/– mice and WT controls (t-test: P = 0.5993).

(i) Body weight of Tsc2+/– mice did not differ from WT controls (t-test: P = 0.717).

Supplemental methods

Behavior

Visible version of the Morris water maze: Mice were given 4 daily

training trials for 4 consecutive days in the visible version of the water

maze. The position of the visible platform and the animals’ starting

position varied across trials. Trials were completed when animals

reached the platform or 60 s had elapsed, whichever came first. One-

way repeated-measures ANOVA with genotype as between-subjects

factor was used for statistical analysis of the escape latencies.

Open field: Activity was tested in the open field assay. Mice were

placed for 20 min in a square open field made of acrylic, and activity

was recorded by an automated system (Med Associates). Total

distance moved (cm) was the primary measure for exploratory

activity. Zone analysis was performed to determine outer sector

occupancy of mice; high scores in this measure can indicate high

anxiety levels. For statistical analysis either t-test or ANOVA were

used, as described in the text.

Elevated plus maze: Mice were placed on an elevated plus maze for a

5 min session and behavior was videotaped for offline analysis.

Percent time spent on the open arms was determined and compared

across groups by t-test.

Accelerating rotarod: Motor coordination skills were tested with four 10

min trials on the accelerating rotarod (inter-trial interval: about 1 h).

The rod accelerated from 4 to 40 rpm within the initial 5 min and

stayed at 40 rpm for another 5 min. Latency to fall or adopt passive

cycling (for > 2 consecutive rounds) was recorded and statistical

analysis was performed with one-way repeated-measures ANOVA with

genotype as the between-subjects factor.

Social interaction: Social approach behavior was studied in Tsc2+/–

mice and WT controls using a 3-compartment apparatus, as previously

described7. Mice (all male) were habituated to the center of the

apparatus (5 min) and subsequently given access to the side

compartments (test phase: 10 min). One side compartment contained

a stimulus mouse (ovarectomized female; constrained by a

transparent, perforated acrylic cup), the control compartment an

empty cup. Time actively investigating (approaching, sniffing, rearing)

the mouse/empty cup was measured and compared by paired t-test.

Western Blots

To determine p-S6 levels (normalized to total S6), hippocampi were

extracted from naïve mice under resting condition (home cage) or

following context conditioning (15 min after a 5 min conditioning

session involving 3 × 0.75 mA shocks, as described above). Mice were

treated with rapamycin (5 mg/kg i.p. for 5 d prior to and including the

day of context conditioning) or vehicle (100% DMSO; DMSO dose: 2

ml/kg; volume of a single injection did not exceed 50 µl) prior to

behavioral procedures and hippocampal extraction, using the same

protocol as described above for the behavioral study. To determine

tuberin, hamartin and total S6 levels (normalized to α-tubulin; similar

results were obtained when total S6 served as loading control for

tuberin and hamartin, data not shown), hippocampi were extracted

from Tsc2+/– mice and WT controls under resting conditions. Whole

hippocampal extracts were homogenized in lysis buffer containing 150

mM NaCl, 50 mM Tris (pH7.5), 50 mM NaF, 0.5% NP-40, 2 mM EDTA,

1 mM sodium orthovanadate and 1 × complete protease inhibitor

cocktail (Roche). Protein concentrations were determined by DC

Protein Assay (Biorad). Equal amounts of lysates (100 µg) were

separated by 4-15% SDS-PAGE (Biorad), and transferred to

nitrocellulose membrane (Biorad). Membranes were incubated with

antibodies to p-S6 (S235/236), S6 (both Cell Signaling Technology),

hamartin (HF6)8, tuberin (TSDF)8 or α-tubulin (Sigma), followed by

horseradish peroxidase-conjugated secondary antibodies (Chemicon),

and then visualized using the ECL system (Amersham Biosciences).

Experimental conditions (antibody concentrations, protein loading,

exposure times etc.) were optimized to ensure detection in the linear

range and this has been confirmed by serial dilutions. For western blot

quantification of protein levels, samples were run on gels in a pseudo-

random order (experimental groups were counterbalanced across

gels). Signal intensities were quantified from these quantitative blots.

Protein levels were normalized to loading control for each sample

(total S6 or α-tubulin) and data were presented as percentage of the

WT/home cage/vehicle group or WT group, respectively. Two-way

ANOVA with genotype and behavioral group assignment as between-

subjects factor was used to test for genotype effects on resting or

post-conditioning p-S6 levels. To test whether rapamycin reduced p-

S6 levels, three-way ANOVA with genotype, drug treatment and

behavioral group assignment as between-subject factors was used for

statistical comparisons. T-tests were used to compare hippocampal

hamartin, tuberin and total S6 levels across genotypes. Representative

examples were chosen to generate representative blots (these blots

were not used for quantification), on which samples are shown sorted

by experimental condition.

Drugs

Lovastatin: HMG-CoA reductase inhibitor lovastatin was purchased

from Sigma, stored according to the manufacturer’s instructions and

freshly dissolved, as previously described4 . Lovastatin (10 mg/kg) or

vehicle was injected once daily s.c. for 3 d prior to and during Morris

water maze training (6 h prior to training). As in the other

experiments, animals were trained in the hidden version of the Morris

water maze for 5 d (4 trials/d) after which a probe trial was given.

Anisomycin: Anisomycin (Sigma) was stored according to the

manufacturer’s instructions and freshly dissolved in 0.9% saline (1N

HCl was used to adjust pH to 7.4). Anisomycin (150 mg/kg) or vehicle

were injected intraperitoneally 30 min prior to context conditioning (3

× 0.75 mA shocks), as previously described3. Context fear was tested

one day later. The experiment was performed using C57BL/6NCrl WT

mice.

Rapamycin (see also Methods): For the 5 mg/kg experiments,

rapamycin or vehicle (100% DMSO; 2 ml/kg; volume of a single

injection did not exceed 50 µl) was administered once daily i.p. for 5 d

prior to and including the day (3 h before conditioning) of context

conditioning. For the 150 mg/kg experiment, rapamycin or vehicle

(100% DMSO; 4 ml/kg; volume of a single injection did not exceed

100 µl) was i.p injected once 3 h prior to context conditioning. Context

conditioning involved 3 × 0.75 mA shocks and conditioned fear was

tested one day later (5 min session in the training context). These

experiments were done in C57BL/6NCrl WT mice.

Immunohistochemistry

Mice were killed with an overdose of sodium pentobarbital and

perfused transcardially with 4% paraformaldehyde in cold 0.1 M

phosphate buffer (pH 7.4). Brains were left in fixative overnight and

then transferred to 30% sucrose. Forty µm coronal slices were

prepared with a cryostat (Leica) and stored in a cryoprotectant

solution (25% ethylene glycol, 25% glycerin and 0.05 M phosphate

buffer). After washing slices in TBS and a blocking step in TBS-plus

(TBS pH 8.5 with 0.1% Triton X-100 and 3 % donkey serum), slices

were incubated in antibody to GFAP (Sigma; 1:2000) for 48 h at 4°C.

Next, after washing and blocking steps, slices were incubated in Alexa-

488-conjugated secondary antibody (Molecular Probes; 1:250; for 4 h

at room temperature, protected from light). Slices were washed and

coverslipped in polyvinyl alcohol with diazabicyclo-octane as anti-

fading agent.

Supplemental references 1. Tischmeyer, W., et al. Rapamycin-sensitive signalling in long-term consolidation

of auditory cortex-dependent memory. Eur J Neurosci 18, 942-950 (2003). 2. Dash, P.K., Orsi, S.A. & Moore, A.N. Spatial memory formation and memory-

enhancing effect of glucose involves activation of the tuberous sclerosis complex-Mammalian target of rapamycin pathway. J Neurosci 26, 8048-8056 (2006).

3. Abel, T., et al. Genetic demonstration of a role for PKA in the late phase of LTP and in hippocampus-based long-term memory. Cell 88, 615-626 (1997).

4. Li, W., et al. The HMG-CoA reductase inhibitor lovastatin reverses the learning and attention deficits in a mouse model of neurofibromatosis type 1. Curr Biol 15, 1961-1967 (2005).

5. Smalley, S.L. Autism and tuberous sclerosis. J Autism Dev Disord 28, 407-414 (1998).

6. Fombonne, E. Epidemiological surveys of autism and other pervasive developmental disorders: an update. J Autism Dev Disord 33, 365-382 (2003).

7. Moy, S.S., et al. Sociability and preference for social novelty in five inbred strains: an approach to assess autistic-like behavior in mice. Genes Brain Behav 3, 287-302 (2004).

8. Han, S., et al. Phosphorylation of tuberin as a novel mechanism for somatic inactivation of the tuberous sclerosis complex proteins in brain lesions. Cancer Res 64, 812-816 (2004).