Neuropharmacology,EcoleNormaleSupe ´rieure,75230ParisCedex05,France,2InstitutNationaldelaSante ´etdelaRechercheMe ´dicale,Unit796,
3InstitutNationaldelaSante ´etdelaRechercheMe ´dicale,Unit679,NeurologyandExperimentalTherapeutics,HospitalofPitie ´-Salpe ˆtrie `reandUniversity
birth to reach a stable 38 and 23% reduction in the SN and VTA, respectively, and that neuronal loss can be antagonized by En2
tant adult physiological functions. They also suggest that mice lacking only one En1 allele could provide a novel model for the study of
Homeoproteins, a class of transcription factors encoded by ho-
meogenes, have been studied primarily during development but
also are expressed at late developmental stages and in the adult.
Some adult functions probably are related to the persistence of
plastic events in the mature nervous system (Morgan, 2006), for
ical changes accompanying physiological activity. However, an-
new functions unrelated to development.
During development the mesencephalic dopaminergic (DA)
neurons of the substantia nigra (SN) and the ventral tegmental
area (VTA) are determined by the sequential expression of
growth and transcription factors. LIM homeobox transcription
factor 1? (Lmxb1) and nuclear receptor-related subfamily 1
(Nurr1) transcription factor act very early in the chain of events
leading to DA cell determination (Prakash and Wurst, 2006). In
contrast, pentraxin-related gene 3 (Ptx3) and Engrailed1 (En1)
homeogenes intervene at later stages of DA cell development.
Indeed, Ptx3 plays a role in the early morphological differentia-
tion of DA neurons (Smidt et al., 2004b), and En1, possibly also
2001; Sgado et al., 2006).
During development En1 is expressed first at the one-somite
stage, and its invalidation results in nearly complete midbrain
absence (Wurst et al., 1994). En2 expression is detected at the
five-somite stage, and its deletion gives rise primarily to a cere-
bellar phenotype (Joyner et al., 1991; Gerlai et al., 1996). Knock-
ing En2 into En1 allows for midbrain development and mouse
survival, suggesting that the two proteins are equivalent in the
midbrain, at least at this level of analysis (Hanks et al., 1995).
In the adult En2 is expressed in the tectum, SN, and VTA, but
its strongest site of expression is the granule cell layer in the
cerebellum. In contrast, En1 expression is predominant in the
che Me ´dicale, and Ecole Normale Supe ´rieure. Specific support also was obtained from the Association Franc ¸aise
Correspondence should be addressed to Alain Prochiantz, Centre National de la Recherche Scientifique, Unite ´
TheJournalofNeuroscience,January31,2007 • 27(5):1063–1071 • 1063
of En1/2 is DA cell survival. Indeed, the deletion of the two En1
and one/two En2 alleles leads to the prenatal loss of DA neurons
in the VTA and SN, and the extent of DA loss, probably via
deleted (Simon et al., 2001). Furthermore, the deletion of one
En1 and two En2 alleles (three En alleles of four) leads to massive
DA cell death in the SN of young adult mice, but not in the VTA
(Sgado et al., 2006).
In the present work we have used a mutant mouse line in
which only one En1 allele has been deleted in an En2 wild-type
context that preserves cerebellar integrity. We show that, in this
DA neurons is reduced specifically and progressively with age
from 8 to 24 weeks postnatal and then remains stable. DA loss
affects both the SN and VTA, gives rise to a strong hypo-DA
phenotype, and can be antagonized by the local infusion of En2
France) were crossed with En1LacZ?/?, En2?/? (referred to in this
homologous recombination (Hanks et al., 1995) maintained on a WT
identified by PCR with the use of LacZ primers. Male WT and En1?/?
littermates were housed in groups of 6–10 per cage (45 cm deep, 30 cm
wide, 20 cm high), except for the saccharin consumption test in which
ambient temperature of the room was maintained at 21.5 ? 1.5°C. For
immunohistochemistry and stereological count of DA cells in SN and
VTA five different batches of mice were used corresponding to 3, 8, 16,
24, and 48 d postnatal. Groups of n ? 3–4 mice per genotype were
analyzed. Those animals were not tested previously for behavior. For
striatal monoamine level evaluation one batch of mice at 55 weeks post-
natal was used (n ? 8 for WT; n ? 10 for En1?/?). These animals were
the open-field test at the age of 24 weeks. Batch 2 (WT, n ? 9; En1?/?,
n ? 10) was subjected to elevated plus maze and locomotor activity tests
at the age of 24 weeks, social interaction test at the age of 25 weeks, the
to amphetamine administration and horizontal locomotor activity eval-
uation at 32–33 weeks postnatal (WT, n ? 12; En1?/?, n ? 12).
Immunohistochemistry. Animals were anesthetized with pentobarbital
formed as described previously (Hirsch et al., 1988) on free-floating
cryomicrotome-cut sections (20 ?m in thickness) encompassing the en-
tire midbrain. After incubation in 3% H2O2/20% methanol, followed by
0.2% Triton X-100 and by 4% bovine serum albumin in 0.1 PBS, the
sections were incubated overnight at 4°C with a polyclonal antibody
AR). Sections then were treated with secondary antibodies (Vectastain,
Vector Laboratories, Burlingame, CA) and subsequently were incubated
with avidin-biotinylated horseradish peroxidase complex. Peroxidase
was revealed by incubation with 0.05% 3,3?-diaminobenzidine tetrahy-
drochloride containing 0.008% hydrogen peroxide. Sections were ana-
lyzed on a Nikon Optiphot 2 microscope with ExploraNova VisioScan
T4.18 software (La Rochelle, France) as previously described (Hoglinger
et al., 2003). Cell counts were quantified stereologically on regularly
anatomical landmarks (Paxinos mouse brain atlas).
cut sections (20 ?m in thickness) were incubated for 1 h in PBS, 0.2%
Triton X-100, and 10% fetal calf serum (FCS). Primary antibodies
against TH (1:1000; Pel-Freez Biologicals) and neuronal marker NeuN
(1:1000; Chemicon, Temecula, CA) were added overnight at 4°C before
the addition of anti-mouse FITC (1:1000; Jackson ImmunoResearch,
West Grove, PA) and anti-rabbit cyanine 3 (1:2000; Jackson Immuno-
Research) antibodies. Sections were analyzed by confocal microscopy.
TH staining was used to delimit the SN, and NeuN-positive cells were
counted in the delimited area by using NIH ImageJ software (Bethesda,
Statistical differences were assessed by two-way ANOVAs (genotype
status ? cell counts), followed by post hoc Tukey honest significant dif-
age and genotype).
Engrailed recombinant protein infusion. En2 recombinant protein was
produced in bacteria as described previously (Foucher et al., 2003) and
infused for 14 d above the SN (anterior, 0.27; lateral, 0.8; ?, 2.75 relative
to ?) with micro-osmotic pumps (0.3 ?l/h; model 102 ALZET, Durect,
saline (0.09% NaCl), and three WT mice and three En1?/? mice were
infused with En2 recombinant protein (150 ?g/ml En2 and 10 ?g/?l
were killed at 9 weeks by a 4% PFA infusion; their brains were removed,
postfixed for 24 h at 4°C in 4% PFA, cryoprotected with 10% sucrose,
frozen in isopentane at –30°C, and stored at –80°C.
Sections were immunostained for Engrailed to analyze the location of
the exogenous En2 protein. Free-floating cryomicrotome-cut sections
(20 ?m in thickness) encompassing the entire midbrain were heated in
1? PBS overnight at 65°C to inhibit endogenous alkaline phosphatase
(AP) activity, blocked for 30 min at room temperature (RT) in 1? PBS,
in the laboratory; 1:2000). Sections then were treated with anti-rabbit
biotin-IgGs (1:2000) for 2 h at RT and AP-streptavidin (1:2000) for 30
min at RT. AP activity was revealed by using nitroblue tetrazolium/5-
bromo-4-chloro-3-indolyl phosphate (NBT/BCIP) overnight at RT and
stopped with PBS. Sections were post-fixed for 30 min in 4% PFA.
the stereological protocol described above. Statistical differences were
assessed by two-way ANOVA, followed by Fisher least significant differ-
ences (LSD) post hoc test.
Striatal monoamine levels. Animals were decapitated at 55 weeks, and
total striae were dissected and stored at –30°C until use. Tissues were
homogenized (Potter Teflon) in 500 ?l of 0.1N perchloric acid (for cat-
sodium metabisulfites. After centrifugation (25,000 ? g for 30 min at
to a mass spectrometer (Finnigan LCQ, Thermo Fisher Scientific,
in inverse phases. The eluted amines were monitored by using electros-
in striatal monoamine levels were compared by using the Student’s t test
(n ? 8 for WT; n ? 10 for En1?/?).
Open field. Activity was measured in a 15 min session in four Plexiglas
transparent open fields (44.5 cm long, 44.5 cm wide, 40 cm high). The
distance traveled (in centimeters), the number of rearings, and the time
spent in the central zone of the open field (22.5 ? 22.5 cm2) were re-
corded in 5 min periods with infrared photobeam detection systems
(ActiTrack, LSI LETICA, Panlab, Barcelona, Spain). Two-way ANOVAs
(genotype status ? time period) with repeated measures on time period
were used to test for significant differences between WT and En1?/?
Amphetamine treatments. The effect of administrations of amphet-
amine (sulfate salt; Sigma) on open-field motor activity was studied in
one 60 min session at a dose of 5 mg/kg and in three 60 min sessions at
1064 • J.Neurosci.,January31,2007 • 27(5):1063–1071Sonnieretal.•MouseModelofProgressiveDopaminergicCellLoss
48 h intervals at a dose of 2 mg/kg. Animals in each group were divided
randomly in two subgroups and received either saline or amphetamine
traveled (in centimeters) was recorded during 30 min periods before
(data not shown) and 30 min after the injection. Two-way ANOVA for
repeated measures was used to determine the influence of treatment,
Student’s t test when appropriate.
Accelerating rotarod. Balance and motor coordination were tested by
using a constant-speed rotarod treadmill (Model 7600 with a 3 cm in
diameter rod; Ugo Basile, Comerio, Italy). The time spent by the mouse
on the rotarod was recorded up to a maximum of 120 s, with speeds
varying from 16 to 32 rpm at 4 rpm incremental steps. Each animal was
trained over 3 d (three trials per speed with an interval of at least 4 h
differences between WT and En1?/? mice.
using the forced swimming test (FST) (Porsolt et al., 1977). This test has
been performed as described by Ventura and colleagues (2002) with
minor modifications. Mice were placed individually into glass cylinders
(height, 25 cm; diameter, 55 cm) containing 20 cm of water at 25–26°C
during two sessions 24 h apart. The first session lasted 10 min and the
unaware of animal status measured immobility time. A mouse was
judged to be immobile when it floated in an upright position and made
only small movements to keep its head above water.
Two-way ANOVAs (genotype status ? time period) with repeated
measures on time period were used to test for significant differences
between WT and En1?/? mice in FST. When significant differences
that end, animals tested in FST were evaluated previously in the open
Bioseb, Chaville, France) for a 10 min period. The individual compart-
ments (40 ? 40 ? 30 cm) were in a dimly lit quiet room. The obtained
locomotor activity data were analyzed by a separate two-way ANOVA
Social interaction test. The general design was adapted from a pub-
lished protocol (Sams-Dodd, 1997). The test was performed in an open
arena (40 ? 40 cm; 30 cm high) that was placed in a dimly lit room and
cleaned after each test. Two unfamiliar mice of identical genotype that
had approximately the same weight were placed simultaneously in the
opposite corners of the arena. The track of each mouse was recorded for
10 min period of the test the locomotor activity and time spent in the
central zone also were recorded. The behavior of animals was followed
with a video tracking system (SMART, Bioseb). Social interaction data
found in social interaction and motor activity, ANCOVAs were con-
ducted to determine phenotype independences.
Saccharin preference. The mice placed in individual cages 2 weeks be-
fore the test then were submitted to a water versus saccharin two-bottle
preference test (saccharin sodium salt hydrate; Sigma). Increasing sac-
charin solutions were used (0.0125, 0.05, 0.2, 0.8, and 3.2%). Each con-
centration was presented in consecutive 2 d blocks. The solutions were
available 23 h/d, and the bottles were weighed and refilled during the
remaining 1 h. The left–right positions of the saccharin and water were
alternated for each concentration (to control for the preference of some
mice for a particular side). Total intakes were averaged and expressed in
“expected consumption” if there was no preference (total ingested flu-
id/2) from the actual amount of saccharin-containing liquid consumed
(Meliska et al., 1995). Significance was calculated by two-way ANOVAs
way ANOVAs and post hoc analysis (Fisher LSD) when appropriate.
Dopaminergicphenotypeof En1?/? mice
Illustrations of TH staining for DA cells and NeuN staining for
total neuronal cells are provided in Figures 1A and 2C for WT
and heterozygote mice (En1?/?). We evaluated the number of
TH-positive cells in the SN and VTA of En1?/? animals and
their WT littermates at different time points (for SN, 3, 8, 16, 24,
and 48 weeks postnatal; for VTA, 3 and 48 weeks postnatal). In
the SN and VTA the overall ANOVAs revealed a significant ge-
notype ? age effect [for SN, F(4,23)? 10 ( p ? 0.001); for VTA,
tests. At 3 weeks of age there is no difference in numbers of
TH-expressing neurons between WT and En1?/? mice both in
the SN and in the VTA (Fig. 1B). However, the percentage of
TH-positive cells decreases progressively in the mutant, and a
8 weeks ( p ? 0.01) in the SN. At 24 weeks the number of TH-
positive cells is reduced significantly by 38% in the mutant ( p ?
0.001), and we could measure no additional loss between 24 and
48 weeks (Fig. 1B). The loss in TH-positive neurons at 48 weeks
in the SN was twice that measured in the VTA (23%; p ? 0.01)
using the pan-neuronal marker NeuN at 3 and 48 weeks postna-
tal. The overall ANOVAs revealed a significant phenotype ? age
sive loss of DA neurons in SN with age in En1?/? mice (3, 8, 16, 24, 48 weeks postnatal).
at www.jneurosci.org as supplemental material). C, Loss of DA neurons in VTA of En1?/?
Sonnieretal.•MouseModelofProgressiveDopaminergicCellLossJ.Neurosci.,January31,2007 • 27(5):1063–1071 • 1065
Tukey HSD test demonstrated that, at 48
weeks, the mutant SN had lost 38% of its
neurons, compared with WT ( p ? 0.001)
(Fig. 2D). Thus the numbers of TH-
positive cells and NeuN-positive cells in
the SN decreased in parallel, demonstrat-
ing that the decrease in TH-positive cells
reflected cell death and not impaired TH
expression. Compared with WT, the size
of the SN was not modified with age in
En1?/? (Fig. 2A,B).
Evaluation of DA contents in the stria-
tum by liquid chromatography and mass
spectrometry revealed a significant de-
crease of 37% in DA levels in the striatum
of En1?/? mice ( p ? 0.01) (Fig. 2E).
This measure of striatal DA content takes
into account DA originating from the SN
and the VTA terminals. Interestingly, the
(DOPAC) on DA, an index of DA turn-
over, was increased by 23% in En1?/?
mice ( p ? 0.001) (Fig. 2F). En1?/? ani-
by increasing DA turnover. It is notewor-
ing that in the striatum the mutation pri-
marily affected the DA pathway.
In several previous studies we have dem-
onstrated that Engrailed is a messenger
homeoprotein and is internalized (and
secreted) by live cells thanks to the pres-
ence of two highly specific sequences
(Prochiantz and Joliot, 2003; Joliot and
Prochiantz, 2004).We used this property
to develop a gain of function approach
aimed at verifying that En2 and En1 are
biochemically equivalent in the midbrain
(Hanks et al., 1995) and that En1/2 are
ical activity after internalization previously had been demon-
3B, the protein was infused at a concentration of 150 ?g/ml be-
tween weeks 6 and 8, and the animals were killed at week 9. This
handled more easily than younger pups. Infusions were unilat-
eral; En2 protein could diffuse through the entire midbrain (Fig.
3A), but the intensity and radius of diffusion of the protein were
variable among animals. We then evaluated the number of TH-
positive cells in the SN ipsilateral to the infusion site. Two-way
ANOVA revealed a significant treatment ? genotype effect
[F(1,11)? 14.2 ( p ? 0.01)]. The Fisher LSD post hoc test showed
that the infusion had no effect on WT [p ? 0.4; not significant
fused with saline) have significantly less TH-positive cells than
demonstrated that En2 protein can antagonize DA cell loss fully
in the SN of En1?/? animals ( p ? 0.01) (Fig. 3C). These data
suggest that the DA cell death phenotype truly reflects a survival
function of En1/2.
Spontaneous locomotor activity in a novel environment (open
field) was investigated in WT and En1?/? mice at 24 weeks of
age (Fig. 4A–C). Compared with WT, En1?/? mice exhibited a
reduction in distance traveled (Fig. 4A) [F(1,36)? 7.66 ( p ?
no difference between the groups (period ? genotype interac-
w.jneurosci.org as supplemental material). E, Striatal DA contents in WT (n ? 8) and En1?/? (n ? 10) mice at 55 weeks
1066 • J.Neurosci.,January31,2007 • 27(5):1063–1071Sonnieretal.•MouseModelofProgressiveDopaminergicCellLoss
[F(1,36)? 12.3 ( p ? 0.01)] whatever the period tested (period ?
En1?/? and WT mice spent similar amounts of time in the
central area of the open field (F(1,36)? 0.3; NS). This could be
interpreted as an absence of difference in a behavior associated
with anxiety. This interpretation was confirmed by the elevated
able at www.jneurosci.org as supplemental material).
Amphetamine-induced activity (2 mg/kg) was investigated in
WT and En1?/? mice at 32–33 weeks of age. During the 30 min
preceding the injection of amphetamine or saline En1?/? mice
exhibited a lower motor activity than their WT littermates (data
not shown). Post hoc analysis showed that in the three sessions,
compared with WT, activity was decreased in En1?/? animals
ity during the three sessions ( p ? 0.9; NS) but induced an in-
not significant in session S1 but was significant in sessions S2
( p ? 0.05) and S3 ( p ? 0.01) (Fig. 4D). These results show that,
zation in En1?/? mice, but not in WT mice. Hyperactivity in
(data not shown). En1?/? mice were analyzed additionally in a
rotarod test, which evaluates motor coordination and sensori-
motor learning (Crawley, 1999) (Fig. 4E). En1?/? mice tended
to fall earlier than WT mice [F(1,17)? 2.7
notype ? speed; F(4,68)? 0.4; NS). How-
ever, this trend was not significant. The
performance of both groups varied with
speed [F(4,68)? 2.9 ( p ? 0.05)], with sta-
ble performances from 16 to 24 rpm, fol-
lowed by a decrease in stability when the
speed was increased to 28 and 32 rpm (at
rpm). Although the differences between
WT and En1?/? are not statistically sig-
utable to the exceptionally good perfor-
mance of one of the mutants (of 10
animals in the group) and that removing
ference of p ? 0.01 between the two
Forced swimming test
Figure 5, A and B, illustrates the perfor-
mances of 24-week-old mice in the FST, a
widely used paradigm to assess antide-
pressant activity (Porsolt et al., 1977;
Cryan and Mombereau, 2004). With the
use of the protocol of Cryan and Mombe-
reau (2004), during the first exposure to
the test situation (Fig. 5A) both WT and
En1?/? mice show a typical pattern: an
initial behavioral activation that is substi-
tuted progressively by increasing episodes
of rigid immobility as revealed by a signif-
icant time period effect [F(1,17)? 11 ( p ?
0.001)]. In addition, the analysis also re-
a significant genotype ? time period interaction [F(1,17)? 108
( p ? 0.01)]. A post hoc test revealed that En1?/? mice exhibit
min of the test ( p ? 0.01), whereas during the last 5 min the
immobility time is similar between WT and En1?/? mice.
During the second exposure (Fig. 5B) En1?/? mice exhibit a
greater immobility time than WT mice [F(1,17)? 6.9 ( p ?
time period ? genotype interaction (F(1,17)? 0.9; NS). In
addition, levels of immobility remained constant across peri-
ods (F(1,17)? 0.4; NS).
Animals tested in FST have been evaluated previously in an
open field to assess their locomotor activity (Fig. 5C). Compared
with WT, En1?/? mice exhibit a reduction in distance traveled
in an open field [F(1,17)? 6.6 ( p ? 0.05)]. In both groups loco-
no difference between the groups (F(1,17)? 1.2; NS). ANCOVA
has been used to determine whether reduction of locomotor ac-
tivity could be responsible for the increase in immobility time
seen in the first 5 min in En1?/? mice. The ANCOVA revealed
that immobility time (during the first day) is still increased sig-
nificantly even when hypoactivity has been accounted for
[F(1,16)? 10.9 ( p ? 0.01)]. Thus the FST test revealed a
observed in En1?/? mice.
amount in a region encompassing the SN. Anterior sections are to the left and posterior to the right. B, Progressive loss of DA
the WT (***p ? 0.001). No significant difference in TH-positive cell numbers is found between WT infused with saline or WT
Gain of function by intraparenchymal infusion of En2 protein. A, After 14 d of infusion, En2 is detected in a large
Sonnieretal.•MouseModelofProgressiveDopaminergicCellLoss J.Neurosci.,January31,2007 • 27(5):1063–1071 • 1067
Saccharin consumption test
Saccharine intake was measured in 35-
Preference Index for saccharine. The
charin concentration on the Preference
Index [F(4,64)? 41.7 ( p ? 0.001)] (Fig.
6A). There was also a significant concen-
tration ? genotype interaction effect
difference of Preference Index between
En1?/? and WT mice was not observed
at all saccharin concentrations. Separate
ANOVAs for each saccharin concentra-
erence Index for saccharin is significantly
higher in WT than in En1?/? mice at 0.2
and 0.8% and (2) the profile of saccharin
intake differs between WT and En1?/?
revealed a peak of saccharin intake at 0.2
and 0.8%, whereas no preference for sac-
charine was observed in En1?/? mice at
any concentration. Nevertheless, identical
aversion was observed in WT and
En1?/? at the highest saccharin concen-
tration (3.2%). Together, these results
Total fluid intake. As illustrated in Fig-
ure 6B, En1?/? mice consumed signifi-
cantly less fluid (saccharin plus water)
whatever the saccharin concentration
effect [F(4,64)? 2 ( p ? 0.05)], suggesting
an adipsic behavior in En1?/? mice. An
increase in total volume intake at 0.2 and
also was observed in WT mice.
Two unfamiliar mice (28-week-old mice)
with identical genotypes that had approx-
imately the same weight were placed si-
multaneously in the opposite corners of
the arena to evaluate social behavior.
Compared with WT, En1?/? mice spent
significantly less time in social interaction
[236.9 ? 35.6 vs 124.4 ? 16.2 s; t7? 3.1
( p ? 0.05)]. En1?/? mice also showed a
significant reduction in the number of
contacts [41.7 ? 3.5 vs 23.6 ? 1.7; t7?
in locomotor activity during the 10 min
significant difference was observed in this
and En1?/? mice (117.9 ? 11.6 vs
149.1 ? 30.6 s, respectively), confirming
the absence of differences in anxiety-
0.01. D, Effect of amphetamine (2 mg/kg) on the distance traveled in the open field (in centimeters; mean ? SD) by WT and
shown). E, Rotarod performances of 27-week-old En1?/? and WT mice expressed as the mean ? SD of latency (s) to fall
(average of 3 trials). Although the same trend (decreased performance for En1?/? mice) was observed at all speeds, the
as the mean ? SD of immobility time during the first (A) and the second (B) test exposure; *p ? 0.05 and **p ? 0.01. C,
1068 • J.Neurosci.,January31,2007 • 27(5):1063–1071Sonnieretal.•MouseModelofProgressiveDopaminergicCellLoss
reduction of locomotor activity could be responsible for the so-
cial interaction deficit found in En1?/? mice. The ANCOVA
revealed that both decrease of time spent in contact and number
for [F(1,15)? 12.6 ( p ? 0.01) and F(1,15)? 30 ( p ? 0.001),
The aim of this study was to investigate the participation of En-
control the development of several body structures. En1-
invalidated mice develop without midbrain and die before birth
(Wurst et al., 1994). The phenotype of the En2 mutant is milder,
with nonlethal cerebellar defects (Joyner et al., 1991). This is
for the absence of the latter. The reverse compensation is dem-
onstrated by the almost-normal brain phenotype of animals in
which En2 was knocked into En1 (Hanks et al., 1995).
In the search for adult homeogene functions En1/2 are inter-
esting because of their expression in structures of extreme phys-
iological importance. Indeed, En2 is expressed strongly in the
cerebellum, whereas En1 is abundant in the SN and VTA, two
interest for these two structures is substantial because of their
involvement in the regulation of motor behaviors and in the
physiology of mood and reward (Graybiel et al., 1994).
Simon and colleagues (2001) showed that, approximately at
birth, the number of surviving DA neurons in the midbrain is
proportional to the number of En1/2 alleles expressed. More re-
cently, Sgado and colleagues (2006) reported that, in a mouse
mutant with one En1 allele in a En2-null context (three alleles
missing), the number of DA neurons decreases rapidly in the SN
number of DA neurons at all stages. This result differs from our
finding that the latter genotype leads to DA cells death in the SN
genetic backgrounds (C57/BL6 vs Swiss).
Our study establishes that, at least until 3 weeks postnatal,
En1?/? mice and WT animals from the same litter have the
same number of TH-positive neurons both in the SN and VTA.
seen in the SN of the heterozygotes, and the number of TH-
positive neurons declines progressively to reach a plateau after 6
months. Counting neurons in the SN with the help of NeuN, a
pan-neural marker, demonstrates a parallel decrease in TH-
positive and NeuN-positive cells. We thus propose that the dele-
tion of one Engrailed allele only induces the progressive death of
DA neurons in the SN and, by extension, in the VTA.
The role of Engrailed proteins as survival factors was con-
firmed by infusion experiments based on the ability of extracel-
lular Engrailed to gain access to the cell cytoplasm and nucleus
(Prochiantz and Joliot, 2003; Joliot and Prochiantz, 2004). En2
infused above the SN diffuses into the ventral mesencephalon
and blocks the death of SN neurons between weeks 6 and 9. This
En2 gain of function experiment demonstrates that En2 and En1
are equivalent to promote DA cell survival without eliminating
possible effects on TH expression.
additional cell death. This suggests that DA neurons in the SN
and VTA are heterogeneous in terms of sensitivity to En dosage.
Another possibility is that En1/2 also act as growth factors. In-
deed, as with many other homeoproteins, they are secreted and
internalized and therefore can act as signaling molecules, in par-
in the ventral midbrain would be shared by a smaller number of
DA neurons and would reach a “survival threshold.” This possi-
a therapeutic protein.
En1?/? mice exhibit abnormal spontaneous motor activity
tor learning was observed in the rotarod, although significance
could be attained by removing one remarkably agile En1?/?
panying cell loss. Indeed, the nigrostriatal DA pathway is in-
volved in initiating and coordinating locomotion. In addition,
En1?/? mice show a strong sensitization to amphetamine after
only one injection, suggesting an impaired activity of DA path-
ways implicated in behavioral sensitization (Kalivas and Stewart,
1991). A similar response to amphetamine was observed in adult
mice that develop a hypo-DA phenotype after orthodenticle ho-
molog 2 (Otx2) homeogene invalidation (Borgkvist et al., 2006).
the second and third sessions, possibly because we used only 2
total fluid intake (B) in 35-week-old En1?/? and WT mice expressed as the mean ? SD
Sonnieretal.•MouseModelofProgressiveDopaminergicCellLoss J.Neurosci.,January31,2007 • 27(5):1063–1071 • 1069
in striatal DA probably induces postsynaptic hypersensitivity
(Borgkvist et al., 2006).
DA death in the VTA led us also to investigate nonmotor
behavior deficits. The FST revealed a depressive-like behavior
marked by an increase in immobility attributable to a reduced
sensitivity of the mesolimbic dopamine system (Ventura et al.,
2002). Anhedonia and adipsia observed in the saccharin con-
sumption are unlikely to reflect, although we remain careful, a
deficit in odorant and gustatory capacities, because mutant and
WT mice were inhibited equally by the highest (3.2%) saccharin
concentration. Accordingly, an implication of the nigrostriatal
pathway in adipsic behavior has been suggested (Casas et al.,
2000; Pal et al., 2001). Finally, the mutants showed a strong de-
crease in social interactions, suggesting social anhedonia.
Anhedonic-like behaviors observed in En1?/? animals can be
inputs from the VTA in drug and natural rewards (Kelley and
Berridge, 2002; Wise, 2002).
Other genetic models showing the importance of develop-
In addition to En1/2 and Otx2, previously mentioned, Lmx1b,
Nurr1, and Ptx3 (Eells, 2003; Nunes et al., 2003; Wallen and
Perlmann, 2003; Smidt et al., 2004a) could be of particular inter-
est. Nurr1 heterozygous and Pitx3 aphakia mice have a reduced
number of nigrostriatal DA neurons. In Pitx3 mutants the DA
loss is present at birth (van den Munckhof et al., 2003) and re-
stricted to the SN (Smidt et al., 2004a). The Nurr1?/? mouse
shows a reduced number of TH-positive neurons at 60 weeks in
the SN, with an associated deficit in motor behavior (Jiang et al.,
2005). A difference with the En1 model is that the number of DA
Jiang et al., 2005), and that the hypo-dopaminergic phenotype
appears much later and seemingly reflects a downregulation of
DA markers rather than cell loss.
This suggests that the En1?/? mutant might represent an-
other model for diseases, neurological or psychiatric-associated
with DA deficits. In particular, some traits could make it a useful
tool in Parkinson’s disease research (Fleming et al., 2005).
Among these are the progressive loss of DA neurons and the
Damier et al., 1999). Also, the association of motor deficits with
haviors is reminiscent of traits observed in Parkinson’s disease
(Isella et al., 2003; Lemke et al., 2004; Lieberman, 2006). How-
such a claim. Our view is that most models that have been devel-
oped, genetic and pharmacological, are useful, each with its
strengths and limitations.
In fact, our main message is that developmental genes ex-
pressed in the adult have important physiological functions. A
lesson from the present study is that the deletion of one allele of
This probably means that fluctuations, even minor, in En1/2 ex-
behaviors, even in the absence of dramatic phenotypes. We thus
speculate that the identification of the genetic pathways in which
iology. It is our hope that this new knowledge will permit the
identification of innovative therapeutic targets and strategies.
AlberiL,SgadoP,SimonHH (2004) Engrailedgenesarecell-autonomously
required to prevent apoptosis in mesencephalic dopaminergic neurons.
Borgkvist A, Puelles E, Carta M, Acampora D, Ang SL, Wurst W, Goiny M,
Fisone G, Simeone A, Usiello A (2006) Altered dopaminergic innerva-
tion and amphetamine response in adult Otx2 conditional mutant mice.
Mol Cell Neurosci 31:293–302.
A, Holt C (2005) The transcription factor Engrailed-2 guides retinal ax-
ons. Nature 438:94–98.
Casas M, Prat G, Robledo P, Barbanoj M, Kulisevsky J, Jane F (2000) Meth-
col Biochem Behav 66:257–263.
Crawley JN (1999) Behavioral phenotyping of transgenic and knockout
tions, motor abilities, and specific behavioral tests. Brain Res 835:18–26.
Cryan JF, Mombereau C (2004) In search of a depressed mouse: utility of
models for studying depression-related behavior in genetically modified
mice. Mol Psychiatry 9:326–357.
Damier P, Hirsch EC, Agid Y, Graybiel AM (1999) The substantia nigra of
the human brain. I. Nigrosomes and the nigral matrix, a compartmental
organization based on calbindin D28Kimmunohistochemistry. Brain
EellsJB (2003) Thecontrolofdopamineneurondevelopment,functionand
survival: insights from transgenic mice and the relevance to human dis-
ease. Curr Med Chem 10:857–870.
Eells JB, Lipska BK, Yeung SK, Misler JA, Nikodem VM (2002) Nurr1-null
levels and increased stress-induced locomotor activity. Behav Brain Res
Fleming SM, Fernagut PO, Chesselet MF (2005) Genetic mouse models of
parkinsonism: strengths and limitations. NeuroRx 2:495–503.
is the result of a highly conserved mechanism for direct interaction of
Gerlai R, Millen KJ, Herrup K, Fabien K, Joyner AL, Roder J (1996) Im-
paired motor learning performance in cerebellar En-2 mutant mice. Be-
hav Neurosci 110:126–133.
Graybiel AM, Aosaki T, Flaherty AW, Kimura M (1994) The basal ganglia
and adaptive motor control. Science 265:1826–1831.
Hanks M, Wurst W, Anson Cartwright L, Auerbach AB, Joyner AL (1995)
Rescue of the En-1 mutant phenotype by replacement of En-1 with En-2.
Hirsch E, Graybiel AM, Agid YA (1988) Melanized dopaminergic neurons
are differentially susceptible to degeneration in Parkinson’s disease. Na-
Oertel WH, Hirsch EC (2003) Chronic systemic complex I inhibition
induces a hypokinetic multisystem degeneration in rats. J Neurochem
Isella V, Iurlaro S, Piolti R, Ferrarese C, Frattola L, Appollonio I, Melzi P,
Grimaldi M (2003) Physical anhedonia in Parkinson’s disease. J Neurol
Neurosurg Psychiatry 74:1308–1311.
Jiang C, Wan X, He Y, Pan T, Jankovic J, Le W (2005) Age-dependent do-
paminergic dysfunction in Nurr1 knockout mice. Exp Neurol
Joliot A, Prochiantz A (2004) Transduction peptides: from technology to
physiology. Nat Cell Biol 6:189–196.
Joyner AL, Herrup K, Auerbach BA, Davis CA, Rossant J (1991) Subtle cer-
homeobox. Science 251:1239–1243.
Kalivas PW, Stewart J (1991) Dopamine transmission in the initiation and
expression of drug- and stress-induced sensitization of motor activity.
Brain Res Brain Res Rev 16:223–244.
1070 • J.Neurosci.,January31,2007 • 27(5):1063–1071Sonnieretal.•MouseModelofProgressiveDopaminergicCellLoss
Kelley AE, Berridge KC (2002) The neuroscience of natural rewards: rele- Download full-text
vance to addictive drugs. J Neurosci 22:3306–3311.
Lemke MR, Fuchs G, Gemende I, Herting B, Oehlwein C, Reichmann H,
Rieke J, Volkmann J (2004) Depression and Parkinson’s disease. J Neu-
rol 251[Suppl 6]:VI/24–27.
Lieberman A (2006) Depression in Parkinson’s disease—a review. Acta
Neurol Scand 113:1–8.
Meliska CJ, Bartke A, McGlacken G, Jensen RA (1995) Ethanol, nicotine,
amphetamine, and aspartame consumption and preferences in C57BL/6
and DBA/2 mice. Pharmacol Biochem Behav 50:619–626.
Morgan R (2006) Hox genes: a continuation of embryonic patterning?
Trends Genet 22:67–69.
Nunes I, Tovmasian LT, Silva RM, Burke RE, Goff SP (2003) Pitx3 is re-
quired for development of substantia nigra dopaminergic neurons. Proc
Natl Acad Sci USA 100:4245–4250.
PalGK,PalP,RajSS,MohanM (2001) Modulationoffeedinganddrinking
behaviour by catecholamines injected into nucleus caudatus in rats. In-
dian J Physiol Pharmacol 45:172–180.
Pellow S, Chopin P, File SE, Briley M (1985) Validation of open/closed arm
entries in an elevated plus-maze as a measure of anxiety in the rat. J Neu-
rosci Methods 14:149–167.
Porsolt RD, Le Pichon M, Jalfre M (1977) Depression: a new animal model
sensitive to antidepressant treatments. Nature 266:730–732.
Prakash N, Wurst W (2006) Development of dopaminergic neurons in the
mammalian brain. Cell Mol Life Sci 63:187–206.
ProchiantzA,JoliotA (2003) Cantranscriptionfactorsfunctionascell–cell
signaling molecules? Nat Rev Mol Cell Biol 4:814–819.
Sams-Dodd F (1997) Effect of novel antipsychotic drugs on phencyclidine-
induced stereotyped behaviour and social isolation in the rat social inter-
action test. Behav Pharmacol 8:196–215.
MP, Dyck RH, Simon HH (2006) Slow progressive degeneration of ni-
Acad Sci USA 103:15242–15247.
SimonHH,SaueressigH,WurstW,GouldingMD,O’LearyDD (2001) Fate
of midbrain dopaminergic neurons controlled by the Engrailed genes.
J Neurosci 21:3126–3134.
Smidt MP, Smits SM, Bouwmeester H, Hamers FP, van der Linden AJ, Hel-
lemons AJ, Graw J, Burbach JP (2004a) Early developmental failure of
substantia nigra dopamine neurons in mice lacking the homeodomain
gene Pitx3. Development 131:1145–1155.
SmidtMP,SmitsSM,BurbachJP (2004b) HomeoboxgenePitx3anditsrole
in the development of dopamine neurons of the substantia nigra. Cell
Tissue Res 318:35–43.
van den Munckhof P, Luk KC, Ste-Marie L, Montgomery J, Blanchet PJ,
Sadikot AF, Drouin J (2003) Pitx3 is required for motor activity and for
survival of a subset of midbrain dopaminergic neurons. Development
VenturaR,CabibS,Puglisi-AllegraS (2002) Geneticsusceptibilityofmeso-
cortical dopamine to stress determines liability to inhibition of mesoac-
cumbens dopamine and to behavioral “despair” in a mouse model of
depression. Neuroscience 115:999–1007.
Wallen A, Perlmann T (2003) Transcriptional control of dopamine neuron
development. Ann NY Acad Sci 991:48–60.
Wise RA (2002) Brain reward circuitry: insights from unsensed incentives.
Wurst W, Auerbach AB, Joyner AL (1994) Multiple developmental defects
in Engrailed-1 mutant mice: an early mid-hindbrain deletion and pat-
terning defects in forelimbs and sternum. Development 120:2065–2075.
Sonnieretal.•MouseModelofProgressiveDopaminergicCellLossJ.Neurosci.,January31,2007 • 27(5):1063–1071 • 1071