Hypoxic damage to the prefrontal cortex (PFC) has been implicated in the frontal lobe dysfunction found in various neuropsychiatric
disorders. The underlying subcortical mechanisms, however, have not been well explored. In this study, we induced a PFC-specific
hypoxia-like damage by cobalt-wire implantation to demonstrate that the role of the mediodorsal thalamus (MD) is critical for the
burst spikes, known to depend on T-type Ca2?channels, were increased in MD neurons. In vivo knockout or knockdown of the T-type
locomotor hyperactivity in this model. These results suggest a two-step model of prefrontal dysfunction in which the response to a
the onset of neurological and behavioral abnormalities. This study provides valuable insights into preventing the development of
Abnormal oscillations or excitability has been a key neurological
sign for frontal lobe dysfunction in neurological and psychiatric
disorders (McAllister and Price, 1987; Barry et al., 2003; Gucuy-
Beleza et al., 2009); for example, the 1–7 Hz oscillations seen in
attention deficit hyperactivity disorder (ADHD) (Barry et al.,
2003), and the frontal lobe-specific paroxysms in frontal lobe
hood (Hernandez et al., 2001; Gonzalez-Heydrich et al., 2007)
and schizophrenia-like symptoms in adults (Adachi et al., 2000;
Helmstaedter, 2001; Patrikelis et al., 2009).
Considering the critical role of thalamic neurons in controlling
1994; Steriade and Contreras, 1995; Steriade, 2006; Ernst et al.,
tal lobe dysfunction could be attributed to the thalamus. This hy-
Price, 1987; Stuss and Gow, 1992; Paradiso et al., 1999; Kellinghaus
oxia (Ishige et al., 1987; Choi, 1990), has been one of the most
tions (Damasio and Anderson, 1985; Owen et al., 1990, 1993; Shal-
lice and Burgess, 1991; Stuss and Gow, 1992; Pigula et al., 1993;
Ho ¨ckel and Vaupel, 2001). Based on the fact that cobalt (Co) ions
trigger signaling pathways which are activated by hypoxia (Piret et
implanting a wire made of Co in the prefrontal cortex (PFC) and
demonstrate how the role of the thalamus is involved in the devel-
Animal care and handling were performed according to the directives of
the Animal Care and Use Committee of Korea Advanced Institute of
Science and Technology. The CaV3.1?/? and wild-type littermate con-
Mid-Career Researcher Program of the National Research Foundation of Korea funded by the Korea government
TheJournalofNeuroscience,March16,2011 • 31(11):4063–4073 • 4063
trol mice were generated by mating heterozygotes from C57BL/6J back-
a 12 h light/dark cycle.
Mouse model of FLE generated using a wire form of cobalt
C57BL/6J male and female mice (10–20 weeks old) were prepared for
Co-wire implantation. Under avertin (0.2%) anesthesia (tribromoetha-
Alfa Aesar) with the radial orientation (vertical to the cortical surface)
ventral, 1.3 mm) using a stereotaxic device (David Kopf Instruments).
An epidural electrode for EEG recordings was implanted in the right
frontal cortex, where the cobalt was implanted, and another electrode
was implanted in the left frontal and temporal cortexes (right frontal:
anteroposterior, 2.8 mm; lateral, 0.8 mm; left frontal: anteroposterior,
2.8 mm; lateral, 0.8 mm; right temporal: anteroposterior, ?2.4 mm;
lateral, ?2.4 mm; left temporal: anteroposterior; ?2.4 mm, lateral, 2.4
skull. After 3 d of recovery, the EEG signal and video recording were
ber (a square-floor rectangular box). EEG signals were amplified (model
7H polygraph, Grass Technologies) and digitized at a sampling rate of
500 Hz (DIGIDATA 1320A, Molecular Devices). pClamp9.2 software
(Molecular Devices) was used for data acquisition. To induce secondary
jected intraperitoneally (550 mg/kg), 7 d after Co-wire implantation
(7.0 ? 0.8 d on average) (Walton and Treiman, 1988).
Magnetic resonance imaging
For magnetic resonance imaging (MRI), mice with Co-wire implanta-
tion (5 d after Co-wire implantation) were anesthetized and set into an
MR compatible cradle. During MRI, the animals were anesthetized by
breathing 2% isoflurane into oxygen-enriched air with a facemask. The
The MRI acquisitions were performed on a 4.7 T horizontal MRI (Bio-
spec, Bruker) with a shielded gradient coil 65 mm in diameter. The
experiments were performed using a 38 mm internal diameter birdcage
coil (Bruker). The measurements were performed using the parameters
given below. T1-weighted images were obtained using a fast spin-echo
sequence with the following parameters: pulse repetition time (TR) ?
(FOV) ? 25 mm; slice thickness (ST) ? 1 mm; pixel resolution ? 100
?100 ?m; and acquisition time for one set ? 6.7 min.
T2-weighted images were obtained using a fast spin-echo sequence
and acquisition time for one set ? 11.5 min.
Histology and immunoblotting
buffered 4% formalin for 24 h and embedded in paraffin for sectioning
on a microtome into 5 ?m slices, which were stained with hematoxylin
and eosin after deparaffination. The brain slices were visualized under a
bright-field microscope (Olympus). For immunoblotting, brains were
isolated on ice and cut into 1 mm slices using a brain matrix (RBMA-
tungsten-implanted group) with morphologically normal neurons in the surrounding area (image 2 in tungsten-implanted group). Left, Ipsilateral PFC (100?); middle, magnified image 1 of
4064 • J.Neurosci.,March16,2011 • 31(11):4063–4073Kimetal.•ThalamicFeedbackMediatesPrefrontalDysfunction
atlas, were separately removed and sonicated
in homogenization buffer containing protease
phosphatase inhibitor (Phostop, Roche). Pro-
teins (30–50 ?g) were loaded onto 10% poly-
acrylamide Tris-HCl gel and transferred to a
nitrocellulose membrane (Protran, Whatman).
The following antibodies used: ?-actin (sc-
47778, Santa Cruz Biotechnology), VEGFa (sc-
507, Santa Cruz Biotechnology). Anti-mouse
(Abcam) and anti-rabbit secondary antibodies
Open-field test. Each mouse was gently placed at
the center of the open-field test arena (a square-
40 ? 50 cm) in a dark room. The distance of
spontaneous movement during a 1.5 h time pe-
riod was monitored at 5 min intervals by digital
6:00 P.M. and 10:00 P.M. EthoVision (Noldus)
Fear-conditioning test. Fear conditioning
was conducted in conditioning chambers
(acrylic test station, dim light, metal grid floor;
San Diego Instruments) and a testing chamber
(a white plastic, circular cylinder). A video
camera was mounted on top of the chambers.
The fear-conditioning procedure was con-
ducted over 2 d. On day 1, each mouse was
acclimated to the conditioning chamber (4
min, 40 s) and then given three pairs of a con-
coterminated with an unconditional stimulus
(foot shock, 2 s, 0.7 mA). The trial interval was
60 s. On the day of testing, the freezing re-
sponses to the conditional stimulus were mea-
sured in the testing chamber with test tones
(180 s). To test contextual conditioning, the
mice were placed in the conditioning chamber
and were allowed to explore for 5 min. Behav-
ioral freezing was manually counted by an ob-
server blinded to the previous treatments.
C57BL/6J male and female mice (10–20 weeks
ethanol, 20 mg/ml) and placed in a stereotaxic
was made in the skull with a dental drill. Elec-
trolytic lesions were made in the mediodorsal
thalamus (MD) (anteroposterior, ?1.46; lat-
eral, ?0.25; ventral, 3.3 mm) with monopolar
electrodes (stainless steel insulated with Ep-
oxylite, Am Systems). Electrolytic lesions were
made by passing a DC current (10 mA; A365,
World Precision Instruments) for 15 s. A piece
of Co-wire (Alfa Aesar) was then implanted in
planted as described for the cobalt model of
PFC damage. After 4 d of recovery, the EEG
dose of avertin and fixed with 10% formalin.
Brain slices (40 ?M) were cut on a cryostat and
stained with cresyl violet.
Kimetal.•ThalamicFeedbackMediatesPrefrontalDysfunctionJ.Neurosci.,March16,2011 • 31(11):4063–4073 • 4065
Bicuculline methobromide test. C57BL/6J male
and female mice (10–20 weeks old) were pre-
pared for injection of bicuculline methobro-
mide (BMB). Under avertin (0.2%) anesthesia
(tribromoethanol, 20 mg/ml, i.p.), electrolytic
lesions were made in the contralateral and ip-
silateral sides of the MD thalamus. After 4 d of
recovery, the EEG signal was acquired for 2 h,
during which BMB (30 mg/kg) was injected
Antiepileptic drug test. C57BL/6J male and
female mice (10–20 weeks old) were pre-
pared for injection of ethosuximide (Coulter
et al., 1989). Basal EEG activity was recorded
for 8 ? 1.1 d after Co-wire implantation.
for 2 h. Zonisamide (60 mg/kg)and phenyt-
oin (100 mg/kg) were injected to test the ef-
fect on frontal seizure spikes.
C57BL/6J male and female mice (10–20 weeks
ethanol, 20 mg/ml) and placed in a stereotaxic
device (David Kopf instruments). A piece of
mm; ventral, 1.3 mm). After 4.2 ? 0.2 d, and
9.2 ? 0.2 d after cobalt implantation, mice
were anesthetized by intraperitoneal injection
of urethane (under light anesthesia; 1.35 g/kg)
and placed in a stereotaxic device (Thomas
RECORDING). The temperature was main-
tained at 37°C using a homothermic blanket
system (Harvard Apparatus). A single incision
above the region where the MD nuclei are located. Quartz-coated te-
trodes (0.5–2 M?; Thomas RECORDING) were positioned in the MD
mm). Signals were amplified ?95-fold and bandpass filtered at 300–
5000 Hz (for the measurement of multiunit activity) or at 0.50–50 Hz
[for the measurement of local field potential (LFP)] by an AC amplifier
(PGMA, Thomas RECORDING) digitized at the 10 kHz sampling rate
(DT3010, Neuralynx). A burst was defined by the following criteria in
interspike interval within a burst) must be ?6 ms; maximum interval to
end burst, 10 ms; minimum interval between bursts, 200 ms; minimum
duration of burst, 4 ms; and minimum number of spikes within a burst,
?100 ms, high-frequency spikes of 200–400 Hz with more than two
spikes, a shortening of the first interspike interval (1–4 ms), and a pro-
viously (Guido et al., 1992; Kim et al., 2003).
In vivo transduction of lentivirus
A lentiviral vector expressing short hairpin RNA (shRNA) to target the
CaV3.1 T-type calcium channel was constructed. A synthetic double-
oligonucleotide (5?–CGGAATTCCGGGAAGATCGTAGATA GCAAA-
was inserted into the shLentisyn3.4G lentiviral vector. The shLentisyn3.4G
National Center for Biotechnology Information, other than CaV3.1. A
CGGGTAAGTGAA CTGACAAGAAttcaagaga TTCTTGTCAGTTCACT-
TACT TTTTGATATCTAGACA–3?) was also inserted into the
shLentisyn3.4G vector and used as a control. This sequence did not have
anol, 20 mg/ml) and perfused transcardially with heparin solution (10
units/ml) followed by 4% formaldehyde dissolved in PBS. For CaV3.1
?m) were incubated with CaV3.1 primary antibody (Alomone Labs),
Seven to 10 d after infection with lentiviruses harboring shC or sh3.1,
ascorbate acid, and 3.0 mM pyruvate (bubbled with 95% O2/5% CO2).
The brain was then rapidly removed, and slices (350 ?m thick) were
made with a vibratome (Leica Microsystems). After incubation for at
least 45 min in 32°C, slices were kept at room temperature in a holding
chamber until they were transferred to a submersion-type recording
chamber held at room temperature. The solution used for slice incuba-
the MD neurons, cells were maintained at ?60 mV and the recording
4066 • J.Neurosci.,March16,2011 • 31(11):4063–4073 Kimetal.•ThalamicFeedbackMediatesPrefrontalDysfunction
solution was substituted with Ca2?current recording solution com-
posed of 115 mM NaCl, 3.0 mM KCl, 10 mM sucrose, 10 mM glucose, 26
mM NaHCO3, 2 mM MgCl2, 2.5 mM CaCl2, 0.5 mM 4-aminopyridine, 5
impermeant), 0.5 mM CaCl2, 1.0 mM MgCl2, 4 mM Mg-ATP, 0.3 mM
TEA-Cl, pH 7.3 (titrated with CsOH). Standard voltage protocols for
inducing T-currents were applied (Perez-Reyes, 2003). Only cells with
low membrane capacitance (40–100 pF) and high membrane resistance
(?400 M?) were included in the analysis.
Coherence analysis. The signals ?600 s were analyzed to measure the
analyzed using the higher-order spectral analysis toolbox and Matlab
R2008a (Chua et al., 2007).
right frontal cortex (FR) and MD thalamus. The GC analysis uses the
prediction error in autoregressive (AR) modeling of the signal. It can be
determined whether signal y influences signal x by comparing the pre-
diction error of x obtained from AR modeling with the prediction error
of x obtained from joint AR modeling of x and y. That is, when y influ-
ences x, the prediction error is significantly decreased for the joint AR
modeling. This is quantified by the following relative prediction im-
provement, called the Granger–Sargent statistic:
of the univariate AR model of x, and ?xy
multivariate mean error of predicting x ob-
tained from the joint AR model of x and y.
signal y influences signal x. The AR model or-
der was determined from Bayesian informa-
tion criteria considering the tradeoff between
racy of prediction (Pereda et al., 2005). For the
GC analysis, it is essential to determine the ep-
och in which the signal can be assumed to be
stationary (Granger, 1969). We could verify
this and obtained reliable AR models using 5 s
segmentation of our data. Therefore, the sig-
nals within 600 s were segmented into 5 s ep-
ochs (Sitnikova et al., 2008).
Event detection. For detection of spikes,
Clampfit 9.2 (Molecular Devices) was used.
number and property of each spike.
of knockdown after transfection of shRNA, a
Positive Pixel Count Algorithm (Aperio) was
used. All data were analyzed using the t test,
ANOVA, and the SigmaStat 3.1 software pack-
2is the mean-squared prediction error
2becomes large, and it can be assumed that
To induce frontal lobe-specific damage,
we implanted a Co-wire into the right
PFC of adult mice. After 5 d of implanta-
tion, a bloody scar reflecting angiogenesis
was formed where Co-wire had been im-
(MRI) showed that Co-wire implantation led to more profound
n ? 4). Such damage was limited to the right frontal cortex,
while no damage signals were found in the left frontal cortex
www.jneurosci.org as supplemental material). The damaged
area detected in MR images was further characterized by he-
matoxylin/eosin staining, and we found that Co-wire implan-
tation resulted in clear signs of hypoxic damage as measured
by the increase of ghost cells (Ito et al., 2006) and inflamma-
tory angiogenesis (Sharp and Bernaudin, 2004) visualized by
blood vessels filled with red blood cells (RBCs) (Fig. 1b,c;
supplemental Fig. S2a, available at www.jneurosci.org as sup-
plemental material) (nonlesion control, n ? 3; tungsten, n ?
3; cobalt, n ? 4; p ? 0.05).
To measure the Co-wire-induced damage at molecular level,
we examined vascular endothelial growth factor (VEGF), a key
angiogenic modulator in response to hypoxia (Sharp and Ber-
naudin, 2004) and found that VEGF proteins were significantly
increased in the PFC 5 d after Co-wire implantation when com-
pared with tungsten-wire-implanted groups (Fig. 1d; supple-
mental Fig. S2b, available at www.jneurosci.org as supplemental
material) (nonlesion control, n ? 2; tungsten, n ? 4, cobalt, n ?
3). These results suggest that Co-wire led to a hypoxia-like dam-
age in the PFC.
injection. c, *Co: comparison of the effects of contralateral and ipsilateral MD lesions on spike generation, 6 d after Co-wire
Lesions in the ipsilateral MD thalamus abolish frontal lobe-specific spikes in the cobalt-implanted group. a,
Kimetal.•ThalamicFeedbackMediatesPrefrontalDysfunctionJ.Neurosci.,March16,2011 • 31(11):4063–4073 • 4067
Next, we wondered whether the Co-wire-
induced PFC damage caused behavioral
abnormalities known to be reported in
patients with PFC dysfunctions (Benson,
2009). An open-field test revealed a pro-
gressive development of hyperactivity
(Fig. 2a,c) (control, n ? 6; cobalt ?6 d,
n ? 5; cobalt ?14 d, n ? 5; cobalt ?28 d,
n ? 9; two-way ANOVA, F(3,192) ?
30.018, p ? 0.001; two-tailed t test, p ?
0.05) when compared with control mice.
In addition, they showed a decreased ex-
ploration in center areas of an open-field
due to their stereotypic circling behaviors
along the wall of test box (Fig. 2b,e) (two-
tailed t test, p ? 0.05). The abnormal
hyperactivity and stereotypic behaviors
have been associated with attention prob-
lems in patients with PFC dysfunctions
(McGuire and Sylvester, 1987; Gainetdi-
nov et al., 2001; Kates et al., 2005; Gilby,
2008). Consistently, they also showed re-
duced learning in a fear-conditioning ex-
periment (supplemental Fig. S3, available
at www.jneurosci.org as supplemental
material) (two-tailed t test, context, p ?
0.01; cue, p ? 0.05), which requires care-
ful attention to cue or contextual stimuli
(Armony and Dolan, 2002; Han et al.,
2003). These results suggest that the im-
behavioral abnormalities similar to those
found in ADHD (Davids et al., 2003) and
FLE childhood (Hernandez et al., 2001;
Gonzalez-Heydrich et al., 2007).
Since frontal lobe dysfunctions are associated with abnormal
frontal lobe rhythms (Elbert et al., 1992; Barry et al., 2003) we
tried to examine cortical electroencephalogram (EEG) and
LFPs of MD thalamus (supplemental Fig. S4a, available at
www.jneurosci.org as supplemental material). Approximately
5 to 6 d after Co-wire implantation in the right PFC (5.8 ?
0.7 d on average), the right frontal lobe showed single spike
activities (Fig. 3a). However, other wires, made with metals
such as tungsten, copper, and aluminum, did not induce sei-
zure spikes when implanted in the PFC during the whole re-
cording period (supplemental Figs. S4c, Fig. S5a, available at
www.jneurosci.org as supplemental material) (cobalt, n ? 10;
after (9.0 ? 1.2 d on average), the frontal seizure spikes were
propagated to the left frontal lobe and to the MD thalamus
(Fig. 3a). The frontal lobe-specific spikes were continuously
examined during the whole recording period (Fig. 4a), al-
though the peak frequency of single-spike activities became
faster as time passed by (supplemental Fig. S5b, available at
30 d postimplantation, mice intermittently showed a second-
ary generalization of single spikes (onset, 11.2 ? 1.4 d on
average) or ictal discharges with whole-body convulsions (on-
set, 18.1 ? 2.5 d on average) (Fig. 3a).
By administration of HT (Walton and Treiman, 1988) at a
subconvulsive dose (550 mg/kg), the corticothalamic and corti-
mittent secondary generalizations was facilitated within an hour
with a similar pattern (Figs. 3b, 4b) (n ? 12). At that dose of HT,
mice without Co-wire implantation did not show any seizure
spikes (supplemental Fig. S6, available at www.jneurosci.org as
lobe seizures in the Co-wire model seems to be activity- and
neural pathway-dependent but not due to the time-dependent
diffusion of Co ions from the PFC to the MD or other brain
Since frontal lobe-specific seizure with secondary general-
ization is a well known symptom found in patients with FLE
induced seizures could be inhibited by anti-FLE drugs including
Hz) during development of frontal lobe spikes. contr, Contralateral; ipsi, ipsilateral. *p ? 0.05, **p ? 0.01. Data represent
4068 • J.Neurosci.,March16,2011 • 31(11):4063–4073 Kimetal.•ThalamicFeedbackMediatesPrefrontalDysfunction
zonisamide and phenytoin. Systemic administration of FLE
drugs robustly abolished the Co-wire-induced seizures (Fig.
4c,d). In addition, we found that ethosuximide, a drug for ab-
al., 1989), also suppressed the Co-wire-induced seizure spikes
(8 ? 1.1 d after Co-wire implantation) (Fig. 4d) (vehicle group,
n ? 4; zonisamide group, n ? 3; phenytoin group, n ? 3; etho-
suximide group, n ? 5; p ? 0.05) (Frampton and Scott, 2005).
These results suggest that the PFC damage by Co-wire leads to
temporally associated with the progressive increase of abnormal
hyperactivities in this model (Fig. 2). Consistently, the abnormal
The propagation of seizure spikes to the MD and the suppres-
sion of seizures by ethosuximide suggested a possibility that
the thalamus is involved in the development of FLE. To ad-
dress this issue, we examined the Co-wire-induced FLE in
mice with a right lesion in the MD (Fig. 5a) and found that
they were resistant to the generation of frontal spikes (Fig.
5b,c) (6 d after Co-wire implantation; contralateral, n ? 5;
ipsilateral, n ? 5; two-tailed t test, p ? 0.01) and secondary
generalizations when compared with the left MD lesion. The
right MD lesion, however, showed no effects on cortical sei-
induces cortical seizures through thalamus-independent mecha-
nisms (Steriade and Contreras, 1998) (Fig. 5b,c) (two-tailed t
test, p ? 0.05). These results suggest that the development of
Co-wire-induced FLE demands the role of thalamic neurons.
The hemisphere-specific interactions between the MD and
PFC neurons through reciprocal connections (Kuroda et al.,
1993) seem to be critical.
To probe the PFC–MD interactions, we
tween the two nuclei during the develop-
ment of frontal spikes. The bicoherence
between the two regions reached a peak
around theta frequencies before the onset
of seizure spikes (4.2 ? 0.3 d on average)
two-tailed t test, p ? 0.05). As frontal
spikes were initiated and their interevent
interval (IEI) was increased, the theta co-
herence between the two regions was de-
creased (IEI ?2, 5.8 ? 0.2 d on average;
IEI ?1.5, 7.6 ? 0.5 d on average) (Fig.
of the thalamocortical feedback in re-
and after the onset of frontal spikes,
Granger causality analysis was performed
between MD and PFC. Consistent with
coherence analysis, the MD–PFC interac-
tion was increased before the onset of
frontal spikes (4.2 ? 0.3 d after Co-wire
implantation) (Fig. 6c; supplemental Ta-
and the PFC at theta frequencies preceded the onset of frontal
Considering a previous study that thalamic burst firings are in-
volved in the generation of thalamocortical interactions (Jean-
monod et al., 1996, 2001), we examined the firing mode of MD
neurons in our model by in vivo extracellular recordings of the
MD under partial urethane anesthesia (1.35 mg/kg). Before the
onset of frontal lobe spikes (4.2 ? 0.2 d after Co-wire implanta-
tion), low-threshold burst spikes known to depend on the acti-
al., 1999) significantly increased in the MD without changes in
tonic spike activity (Fig. 7). In the period when the faster frontal
spikes fully developed (9.0 ? 0.2 d after Co-wire implantation),
[control, n ? 14; before, n ? 18; spike (IEI ?1.5), n ? 12; one-
way ANOVA: burst, p ? 0.05; tonic, p ? 0.05]. These results
suggest that burst firings of MD neurons support the PFC–MD
of frontal spikes (Fig. 6).
Considering that CaV3.1 is the major T-type Ca2?channel sub-
unit that supports burst firings in the thalamocortical relay neu-
rons (Crunelli et al., 1989), we introduced lentiviruses harboring
CaV3.1-specific shRNA into the MD; this infection significantly
reduced CaV3.1 proteins and T-type Ca2?currents in the MD
neurons (Fig. 8a,b; supplemental Table S2, available at www.
jneurosci.org as supplemental material) [lentivirus harboring
shRNA-Control (shC), n ? 5; lentivirus harboring shRNA
Kimetal.•ThalamicFeedbackMediatesPrefrontalDysfunction J.Neurosci.,March16,2011 • 31(11):4063–4073 • 4069
Cav3.1 (sh3.1), n ? 5; two-tailed ? test,
p ? 0.05]. Knockdown of CaV3.1 in the
MD also significantly ameliorated neu-
rological and behavioral abnormalities
found in this model. The PFC power at
theta range and MD–PFC cross talk was
significantly decreased before spiking
(4.6 ? 0.2 d) (Fig. 8c,d; supplemental Fig.
S7, available at www.jneurosci.org as sup-
plemental material) (shC, n ? 5; sh3.1,
n ? 3; two-tailed t test, p ? 0.01), com-
pared with the control group. The num-
two-tailed t test, p ? 0.05). MD-specific
knockdown of the CaV3.1 gene also ame-
liorated the locomotor hyperactivity seen
n ? 5; two-tailed t test, p ? 0.01).
In addition, the therapeutic effect of
the MD-specific knockdown of the
CaV3.1 gene on FLE-like symptoms was
also observed after knockout of the gene
(CaV3.1?/?), which is known to abolish
burst spikes in the thalamus (Kim et al.,
2001). CaV3.1?/?mice showed reduced
(supplemental Fig. S8a, available at www.
jneurosci.org as supplemental material)
(CaV3.1?/?, n ? 6; CaV3.1?/?, n ? 5;
two-tailed t test, p ? 0.01), a lack of fron-
tal lobe-specific spikes after the onset of
spiking, and an absence of locomotor hy-
peractivity (6 d after Co-wire implanta-
www.jneurosci.org as supplemental ma-
4; two-way ANOVA, p ? 0.001).
Since the first report on Phineas Gage,
who suffered a severe frontal lobe injury
in 1848 (Damasio et al., 1994; Macmillan, 2002), frontal lobe
dysfunction characterized by a spectrum of cognitive impair-
ments such as hyperactivity, inattention, impulsiveness, and, in
severe cases, personality changes (Mateer and Williams, 1991;
(Levin, 1984; Benson, 1991; Gedye, 1991; Shue and Douglas,
in the PFC in mice leads to neurological and behavioral abnor-
malities that have been reported in patients with FLE (Helms-
taedter, 2001; Kellinghaus and Lu ¨ders, 2004).
FLE does not usually cause severe convulsions, despite frontal-
lobe-specific spikes. Instead, patients with FLE show cognitive dys-
functions that have been reported in other neuropsychiatric
initiated in the frontal cortex and then leads to complex behavioral
the lack of robust experimental models that allow researchers to
induce frontal-lobe-specific spikes. In this study, we established a
Co-wire model of PFC damage, which leads to neurological and
behavioral symptoms that have been reported in patients with FLE
(Figs. 2, 3), and found that the activation of thalamic T-type Ca2?
PFC damage by Co-wire implantation mimics hypoxic condi-
tions including increased neuronal death, ghost cells (Fig. 1b; sup-
plemental Fig. S2a, available at www.jneurosci.org as supplemental
material), and enhanced VEGF expression (Fig. 1d). In addition to
VEGF, hypoxia induces many other molecules including Hif-1,
has been well known that hypoxic stress on brain slices increases a
Heinemann, 1992; Jensen et al., 1998). Those molecular and physi-
In this model, our results suggest that the activation of tha-
lamic T-type Ca2?channels within the MD is critical for the
showed decreased hyperactivity, compared with shC group (6 d after cobalt wire was implanted). *p ? 0.05. Data represent
4070 • J.Neurosci.,March16,2011 • 31(11):4063–4073 Kimetal.•ThalamicFeedbackMediatesPrefrontalDysfunction
onset of FLE (Figs. 5–8). In contrast to our results, other studies
show that the involvement of thalamic T-type Ca2?channels in
cortically induced seizures is not required, since the induction of
cortical seizures by local injection of BMB is not associated with
thalamic functions (Neckelmann et al., 1998; Steriade and Con-
treras, 1998; Destexhe et al., 1999). This difference can be ex-
plained, in part, by the capacity of BMB to directly increase the
synchronous seizure spikes in cortical neurons (Gutnick et al.,
1982). PFC damage caused by Co-wire, however, may not be
enough to induce cortical synchrony, but may also require tha-
lamic bursting activity modulated by T-type Ca2?channels.
How can PFC damage cause activation of thalamic T-type
Ca2?channels? From an anatomical perspective, PFC neurons
can stimulate two types of thalamic neurons [thalamocortical
through reciprocal connections (Groenewegen, 1988; Hugue-
neurotransmitter released into the MD by nRT neurons may be
involved in de-inactivation mechanisms that allow T-type Ca2?
channels to be easily activated in response to excitatory inputs
(Crunelli and Leresche, 1991; Cox et al., 1997). Consistent with
this idea, both the corticothalamic and thalamocortical drives
increased together (Fig. 6c) and at the same time as the low-
threshold burst spikes of MD neurons were enhanced (Fig. 7),
which reflects the activation of T-type Ca2?channels (Hugue-
nard, 1996; Kim et al., 2001; Crunelli and Leresche, 2002; Lee et
al., 2004; Ernst et al., 2009). Since the thalamic burst spikes are
known to be efficient in stimulating postsynaptic neurons (Lis-
man, 1997), the activation of T-type Ca2?channels in the MD
PFC cross talk and leading to frontal lobe-specific seizures at the
EEG level (Fig. 6).
In addition, thalamic burst spiking is thought to be a sensory
gating mechanism, which can terminate the relay of sensory in-
formation to the cortex (McCormick and Bal, 1994). In this re-
gard, it is plausible that the selective increases in burst spikes in
tion in the MD–PFC pathways, leading to frontal lobe dysfunc-
(Fig. 8) or drugs that weaken the thalamocortical interactions
may be a novel option for preventing further cognitive dysfunc-
this, ethosuximide, a T-type Ca2?channel blocker used to treat
absence seizures, efficiently abolished the onset of FLE in our
model (Fig. 4d) (ethosuximide, n ? 5; p ? 0.01).
Finally, our results suggest a two-step model of PFC dysfunc-
mal thalamocortical feedback driven by thalamic T-type Ca2?
channels, which, in turn, leads to the onset of neurological and
behavioral abnormalities. This model neatly explains why the
progressive development of cortical dysfunction in patients with
a lesion in the PFC takes time (Grattan and Eslinger, 1992; An-
derson et al., 2000; McKinlay et al., 2002; Englander et al., 2003)
and how the thalamus contributes to attention span when pa-
tients are awake, which is a controversial issue (Crick, 1984; Ste-
riade et al., 1993; Sherman, 2001).
(2000) Inter-ictal and post-ictal psychoses in frontal lobe epilepsy: a ret-
rospective comparison with psychoses in temporal lobe epilepsy. Seizure
Anderson SW, Damasio H, Tranel D, Damasio AR (2000) Long-term se-
Armony JL, Dolan RJ (2002) Modulation of spatial attention by fear-
conditioned stimuli: an event-related fMRI study. Neuropsychologia
Barry RJ, Clarke AR, Johnstone SJ (2003) A review of electrophysiology in
attention-deficit/hyperactivity disorder: I. Qualitative and quantitative
electroencephalography. Clin Neurophysiol 114:171–183.
BelezaP,BilginO,NoachtarS (2009) Interictalrhythmicalmidlinetheta
differentiates frontal from temporal lobe epilepsies. Epilepsia
BensonDF (1991) Theroleoffrontaldysfunctioninattentiondeficithyper-
activity disorder. J Child Neurol 6:S9–12.
Choi DW (1990) Cerebral hypoxia: some new approaches and unanswered
questions. J Neurosci 10:2493–2501.
Chua CK, Chandran V, Acharya R, Lim CM (2007) Higher order spectral
CoulterDA,HuguenardJR,PrinceDA (1989) Characterizationofethosux-
imide reduction of low-threshold calcium current in thalamic neurons.
Ann Neurol 25:582–593.
Cox CL, Huguenard JR, Prince DA (1997) Nucleus reticularis neurons me-
diate diverse inhibitory effects in thalamus. Proc Natl Acad Sci U S A
Crick F (1984) Function of the thalamic reticular complex: the searchlight
hypothesis. Proc Natl Acad Sci U S A 81:4586–4590.
CrunelliV,LerescheN (1991) AroleforGABABreceptorsinexcitationand
inhibition of thalamocortical cells. Trends Neurosci 14:16–21.
CrunelliV,LerescheN (2002) Childhoodabsenceepilepsy:genes,channels,
neurons and networks. Nat Rev Neurosci 3:371–382.
Crunelli V, Lightowler S, Pollard CE (1989) A T-type Ca2?current under-
lies low-threshold Ca2?potentials in cells of the cat and rat lateral genic-
ulate nucleus. J Physiol 413:543–561.
Daffner KR, Mesulam MM, Holcomb PJ, Calvo V, Acar D, Chabrerie A,
Kikinis R, Jolesz FA, Rentz DM, Scinto LF (2000) Disruption of atten-
tion to novel events after frontal lobe injury in humans. J Neurol Neuro-
surg Psychiatry 68:18–24.
Damasio A, Anderson S (1985) The frontal lobes. Clin Neuropsychol
DamasioH,GrabowskiT,FrankR,GalaburdaAM,DamasioAR (1994) The
return of Phineas Gage: clues about the brain from the skull of a famous
patient. Science 264:1102–1105.
Davids E, Zhang K, Tarazi FI, Baldessarini RJ (2003) Animal models of
attention-deficit hyperactivity disorder. Brain Res Rev 42:1–21.
Destexhe A, Contreras D, Steriade M (1999) Cortically-induced coherence
of a thalamic-generated oscillation. Neuroscience 92:427–443.
Elbert T, Lutzenberger W, Rockstroh B, Berg P, Cohen R (1992) Physical
aspects of the EEG in schizophrenics. Biol Psychiatry 32:595–606.
Bergman W (2003) Analyzing risk factors for late posttraumatic sei-
zures: a prospective, multicenter investigation. Arch Phys Med Rehabil
Ernst WL, Zhang Y, Yoo JW, Ernst SJ, Noebels JL (2009) Genetic enhance-
voltage-activated calcium current induces pure absence epilepsy.
J Neurosci 29:1615–1625.
FramptonJE,ScottLJ (2005) Zonisamide:areviewofitsuseinthemanage-
ment of partial seizures in epilepsy. CNS Drugs 19:347–367.
Gainetdinov RR, Mohn AR, Caron MG (2001) Genetic animal models: fo-
cus on schizophrenia. Trends Neurosci 24:527–533.
Gedye A (1991) Frontal lobe seizures in autism. Med Hypotheses
Gilby KL (2008) A new rat model for vulnerability to epilepsy and autism
spectrum disorders. Epilepsia 49 [Suppl 8]:108–110.
Gonzalez-Heydrich J, Dodds A, Whitney J, MacMillan C, Waber D, Faraone
SV, Boyer K, Mrakotsky C, DeMaso D, Bourgeois B, Biederman J (2007)
Psychiatric disorders and behavioral characteristics of pediatric patients
with both epilepsy and attention-deficit hyperactivity disorder. Epilepsy
Granger CWJ (1969) Investigating causal relations by econometric models
and cross-spectral methods. Econometrica 37:424–438.
Kimetal.•ThalamicFeedbackMediatesPrefrontalDysfunctionJ.Neurosci.,March16,2011 • 31(11):4063–4073 • 4071
Grattan LM, Eslinger PJ (1992) Long-term psychological consequences of
childhood frontal lobe lesion in patient DT. Brain Cogn 20:185–195.
Groenewegen HJ (1988) Organization of the afferent connections of the
mediodorsal thalamic nucleus in the rat, related to the mediodorsal-
prefrontal topography. Neuroscience 24:379–431.
Gucuyener K, Erdemoglu AK, Senol S, Serdaroglu A, Soysal S, Kockar AI
(2003) Use of methylphenidate for attention-deficit hyperactivity disor-
der in patients with epilepsy or electroencephalographic abnormalities.
J Child Neurol 18:109–112.
Guido W, Lu SM, Sherman SM (1992) Relative contributions of burst and
tonic responses to the receptive field properties of lateral geniculate neu-
rons in the cat. J Neurophysiol 68:2199–2211.
Gutnick MJ, Connors BW, Prince DA (1982) Mechanisms of neocortical
epileptogenesis in vitro. J Neurophysiol 48:1321–1335.
Han CJ, O’Tuathaigh CM, van Trigt L, Quinn JJ, Fanselow MS, Mongeau R,
Koch C, Anderson DJ (2003) Trace but not delay fear conditioning re-
quires attention and the anterior cingulate cortex. Proc Natl Acad Sci
U S A 100:13087–13092.
Helmstaedter C (2001) Behavioral aspects of frontal lobe epilepsy. Epilepsy
bergM (2007) Thefrequency,complicationsandaetiologyofADHDin
new onset paediatric epilepsy. Brain 130:3135–3148.
Hernandez MT, Sauerwein HC, de Guise E, Lortie A, Jambaque ´ I, Dulac O,
Lassonde M (2001) Neuropsychology of frontal lobe epilepsy in chil-
dren. Adv Behav Biol 50:103–111.
Ho ¨ckel M, Vaupel P (2001) Tumor hypoxia: definitions and current clini-
cal, biologic, and molecular aspects. J Natl Cancer Inst 93:266–276.
Huguenard JR (1996) Low-threshold calcium currents in central nervous
system neurons. Annu Rev Physiol 58:329–348.
Huguenard JR, Prince DA (1994) Intrathalamic rhythmicity studied in
J Neurosci 14:5485–5502.
Ishige N, Pitts LH, Hashimoto T, Nishimura MC, Bartkowski HM (1987)
Effect of hypoxia on traumatic brain injury in rats. Part 1: changes in
Ito U, Kuroiwa T, Nagasao J, Kawakami E, Oyanagi K (2006) Temporal
of the cerebral cortex: ultrastructure of neuronal remodeling. Stroke
Jeanmonod D, Magnin M, Morel A (1996) Low-threshold calcium spike
bursts in the human thalamus. Common physiopathology for sensory,
motor and limbic positive symptoms. Brain 119:363–375.
R, Ribary U, Kronberg E, Schulman J, Zonenshayn M (2001) Thalamo-
JensenFE,WangC,StafstromCE,LiuZ,GearyC,StevensMC (1998) Acute
and chronic increases in excitability in rat hippocampal slices after peri-
natal hypoxia in vivo. J Neurophysiol 79:73–81.
JobstBC,WilliamsonPD (2005) Frontallobeseizures.PsychiatrClinNorth
Kates WR, Lanham DC, Singer HS (2005) Frontal white matter reductions
in healthy males with complex stereotypies. Pediatr Neurol 32:109–112.
Kellinghaus C, Lu ¨ders HO (2004) Frontal lobe epilepsy. Epileptic Disord
Kim D, Song I, Keum S, Lee T, Jeong MJ, Kim SS, McEnery MW, Shin HS
(2001) Lack of the burst firing of thalamocortical relay neurons and re-
sistance to absence seizures in mice lacking ?1G T-type Ca2?channels.
Kim D, Park D, Choi S, Lee S, Sun M, Kim C, Shin HS (2003) Thalamic
control of visceral nociception mediated by T-type Ca2?channels. Sci-
Kuroda M, Murakami K, Oda S, Shinkai M, Kishi K (1993) Direct synaptic
sal thalamic nucleus (MD) and corticothalamic neurons the MD in the
prefrontal cortex. Brain Res 612:339–344.
Lee J, Kim D, Shin HS (2004) Lack of delta waves and sleep disturbances
during nonrapid eye movement sleep in mice lacking ?1G-subunit of
T-type calcium channels. Proc Natl Acad Sci U S A 101:18195–18199.
Levin S (1984) Frontal lobe dysfunctions in schizophrenia–II. Impairments
of psychological and brain functions. J Psychiatr Res 18:57–72.
Lisman JE (1997) Bursts as a unit of neural information: making unreliable
synapses reliable. Trends Neurosci 20:38–43.
Llina ´sRR,RibaryU,JeanmonodD,KronbergE,MitraPP (1999) Thalamo-
cortical dysrhythmia: a neurological and neuropsychiatric syndrome
characterized by magnetoencephalography. Proc Natl Acad Sci U S A
Luhmann HJ, Heinemann U (1992) Hypoxia-induced functional altera-
tions in adult rat neocortex. J Neurophysiol 67:798–811.
Luria A, Pribram K, Homskaya E (1964) An experimental analysis of the
behavioral disturbance produced by a left frontal arachnoidal endotheli-
oma (meningioma). Neuropsychologia 2:257–280.
Macmillan M (2002) An odd kind of fame: stories of Phineas Gage: MIT.
MateerC,WilliamsD (1991) Effectsoffrontallobeinjuryinchildhood.Dev
Max JE, Manes FF, Robertson BA, Mathews K, Fox PT, Lancaster J (2005)
Prefrontal and executive attention network lesions and the development
of attention-deficit/hyperactivity symptomatology. J Am Acad Child
Adolesc Psychiatry 44:443–450.
McAllister TW, Price TR (1987) Aspects of the behavior of psychiatric in-
patients with frontal lobe damage: some implications for diagnosis and
treatment. Compr Psychiatry 28:14–21.
McCormick DA, Bal T (1994) Sensory gating mechanisms of the thalamus.
Curr Opin Neurobiol 4:550–556.
McGuire TL, Sylvester CE (1987) Neuropsychiatry evaluation and treat-
ment of children with head injury. J Learn Disabil 20:590–595.
McKinlay A, Dalrymple-Alford JC, Horwood LJ, Fergusson DM (2002)
Long term psychosocial outcomes after mild head injury in early child-
hood. J Neurol Neurosurg Psychiatry 73:281–288.
Millichap JG (2008) Etiologic classification of attention-deficit/hyperactiv-
ity disorder. Pediatrics 121:e358–e365.
Neckelmann D, Amzica F, Steriade M (1998) Spike-wave complexes and
fast components of cortically generated seizures. III. Synchronizing
mechanisms. J Neurophysiol 80:1480–1494.
Niedermeyer E (1998) Frontal lobe functions and dysfunctions. Clin Elec-
Owen AM, Downes JJ, Sahakian BJ, Polkey CE, Robbins TW (1990) Plan-
ning and spatial working memory following frontal lobe lesions in man.
Owen AM, Roberts AC, Hodges JR, Summers BA, Polkey CE, Robbins TW
(1993) Contrasting mechanisms of impaired attentional set-shifting in
patients with frontal lobe damage or Parkinson’s disease. Brain
Paradiso S, Chemerinski E, Yazici KM, Tartaro A, Robinson RG (1999)
Frontal lobe syndrome reassessed: comparison of patients with lateral or
Patrikelis P, Angelakis E, Gatzonis S (2009) Neurocognitive and behavioral
functioning in frontal lobe epilepsy: a review. Epilepsy Behav 14:19–26.
Pereda E, Quiroga RQ, Bhattacharya J (2005) Nonlinear multivariate anal-
ysis of neurophysiological signals. Prog Neurobiol 77:1–37.
Perez-Reyes E (2003) Molecular physiology of low-voltage-activated t-type
calcium channels. Physiol Rev 83:117–161.
Perez-Reyes E, Cribbs LL, Daud A, Lacerda AE, Barclay J, Williamson MP,
FoxM,ReesM,LeeJH (1998) Molecularcharacterizationofaneuronal
low-voltage-activated T-type calcium channel. Nature 391:896–900.
Pigula FA, Wald SL, Shackford SR, Vane DW (1993) The effect of hypoten-
sion and hypoxia on children with severe head injuries. J Pediatr Surg
Piret JP, Mottet D, Raes M, Michiels C (2002) CoCl2, a chemical inducer of
hypoxia-inducible factor-1, and hypoxia reduce apoptotic cell death in
hepatoma cell line HepG2. Ann NY Acad Sci 973:443–447.
RollsET,HornakJ,WadeD,McGrathJ (1994) Emotion-relatedlearningin
patients with social and emotional changes associated with frontal lobe
damage. J Neurol Neurosurg Psychiatry 57:1518–1524.
Andermann F, Desbiens R, Keene D, Cendes F (1995) Autosomal dom-
Shallice T, Burgess PW (1991) Deficits in strategy application following
frontal lobe damage in man. Brain 114:727–741.
4072 • J.Neurosci.,March16,2011 • 31(11):4063–4073Kimetal.•ThalamicFeedbackMediatesPrefrontalDysfunction
Sharp FR, Bernaudin M (2004) HIF1 and oxygen sensing in the brain. Nat
Rev Neurosci 5:437–448.
Sherman SM (2001) Tonic and burst firing: dual modes of thalamocortical
relay. Trends Neurosci 24:122–126.
ShueKL,DouglasVI (1992) Attentiondeficithyperactivitydisorderandthe
frontal lobe syndrome. Brain Cogn 20:104–124.
Sitnikova E, Dikanev T, Smirnov D, Bezruchko B, van Luijtelaar G (2008)
Granger causality: cortico-thalamic interdependencies during absence
seizures in WAG/Rij rats. J Neurosci Methods 170:245–254.
Steriade M (2006) Grouping of brain rhythms in corticothalamic systems.
Steriade M, Contreras D (1995) Relations between cortical and thalamic
cellular events during transition from sleep patterns to paroxysmal activ-
ity. J Neurosci 15:623–642.
Steriade M, Contreras D (1998) Spike-wave complexes and fast compo-
nents of cortically generated seizures. I. Role of neocortex and thalamus.
J Neurophysiol 80:1439–1455.
Steriade M, McCormick DA, Sejnowski TJ (1993) Thalamocortical oscilla-
tions in the sleeping and aroused brain. Science 262:679–685.
Stores G, Zaiwalla Z, Bergel N (1991) Frontal lobe complex partial seizures
Child Neurol 33:998–1009.
Stuss DT, Gow CA (1992) “Frontal dysfunction” after traumatic brain in-
jury. Cogn Behav Neurol 5:272–282.
Tan M, Appleton R (2005) Attention deficit and hyperactivity disorder,
methylphenidate, and epilepsy. Arch Dis Child 90:57–59.
Walton NY, Treiman DM (1988) Experimental secondarily generalized
convulsive status epilepticus induced by d,l-homocysteine thiolactone.
Epilepsy Res 2:79–86.
Wienbruch C, Moratti S, Elbert T, Vogel U, Fehr T, Kissler J, Schiller A,
Rockstroh B (2003) Source distribution of neuromagnetic slow wave
activity in schizophrenic and depressive patients. Clin Neurophysiol
Ragland JD, Carter CS (2008) Association of dorsolateral prefrontal
cortex dysfunction with disrupted coordinated brain activity in schizo-
phrenia: relationship with impaired cognition, behavioral disorganiza-
tion, and global function. Am J Psychiatry 165:1006–1014.
ZikopoulosB,BarbasH (2006) Prefrontalprojectionstothethalamicretic-
Kimetal.•ThalamicFeedbackMediatesPrefrontalDysfunctionJ.Neurosci.,March16,2011 • 31(11):4063–4073 • 4073