Journal of Chemical Neuroanatomy 39 (2010) 262–271

Dopaminergic neurons expressing Fos during waking and paradoxical sleepin the rat

Lucienne Leger a,b,*, Emilie Sapin a,b, Romain Goutagny a,b,1, Christelle Peyron a,b, Denise Salvert a,b,Patrice Fort a,b, Pierre-Herve Luppi a,b

a Universite de Lyon, Lyon, universite Lyon 1, CNRS, UMR5167, Physiologie et physiopathologie du cycle veille-sommeil, F-69008, Franceb Institut Federatif des Neurosciences, IFNL19, Lyon, France

A R T I C L E I N F O

Article history:

Received 29 September 2009

Received in revised form 27 January 2010

Accepted 1 March 2010

Available online 6 March 2010

Keywords:

Dopamine

Ventral midbrain

Hypothalamus

Immunohistochemistry

Double-labeling

Novel environment

A B S T R A C T

Formerly believed to contribute to behavioural waking (W) alone, dopaminergic (DA) neurons are now also

known to participate in the regulation of paradoxical sleep (PS or REM) in mammals. Indeed, stimulation of

postsynaptic DA1 receptors with agonists induces a reduction in the daily amount of PS. DA neurons in the

ventral tegmental area were recently shown to fire in bursts during PS, but nothing is known about the

activity of the other DA cell groups in relation to waking or PS. To fulfil this gap, we used a protocol in which

rats were maintained in continuous W for 3 h in a novel environment, or specifically deprived of PS for 3

days with some of them allowed to recover from this deprivation. A double immunohistochemical labeling

with Fos and tyrosine hydroxylase was then performed. DA neurons in the substantia nigra (A9) and ventral

tegmental area (A10), and its dorsocaudal extension in the periaqueductal gray (A10dc), almost never

showed a Fos-immunoreactive nucleus, regardless of the experimental condition. The caudal hypothalamic

(A11) group showed a moderate activation after PS deprivation and novel environment. During PS-

recovery, the zona incerta (A13) group contained a significant number and percentage of double-labeled

neurons. These results suggest that some DA neurons (A11) could participate in waking and/or the

inhibition of PS during PS deprivation whereas others (A13) would be involved in the control of PS.

� 2010 Elsevier B.V. All rights reserved.

Contents lists available at ScienceDirect

Journal of Chemical Neuroanatomy

journal homepage: www.e lsev ier .com/ locate / jchemneu

Abbreviations: A8, A8 dopaminergic group (retrorubral field); A9, A9 dopaminergic

group (substantia nigra); A10, A10 dopaminergic group (ventral tegmental area);

A10dc, dorsocaudal A10 (ventral periaqueductal gray); A10vr, ventrorostral A10

(supramammillary nucleus); A11, A11 dopaminergic group (caudal hypothalamus);

A12, A12 dopaminergic group (arcuate nucleus); A13, A13 dopaminergic group

(medial zona incerta); A14, A14 dopaminergic group (rostral third ventricle); AH,

anterior hypothalamic area; Cp, cerebral peduncle; DLG, dorsal lateral geniculate

nucleus; DM, dorsomedial hypothalamic nucleus; DpMe, deep mesencephalic

nucleus; f, fornix; ic, internal capsule; IP, interpeduncular nucleus; LH, lateral

hypothalamic area; Ml, medial lemniscus; MM, medial mammillary nucleus; NEv,

novel environment; Opt, optic tract; Ox, optic chiasm; Pa, paraventricular

hypothalamic nucleus; PAG, periaqueductal gray; Pc, posteriror commissure;

PVP, paraventricular thalamic nucleus posterior part; PH, posterior hypothalamic

area; PnO, pontine reticular nucleus oral part; PS, paradoxical sleep; PSC,

paradoxical sleep control condition; PSD, paradoxical sleep deprivation condition;

PSR, paradoxical sleep recovery condition; SC, superior colliculus; SI, substantia

innominata smstria medullaris thalami; SuM, supramammillary nucleus; VLG,

ventral lateral geniculate nucleus; VMH, ventromedial nucleus; VTA, ventral

tegmental area; Xscp, decussation of the superior cerebellar peduncle; ZI, zona

incerta.

* Corresponding author at: CNRS UMR 5167, Faculte de Medecine Laennec, 7 rue

Guillaume Paradin, 69373 Lyon Cedex 08, France. Tel.: +33 4 78 77 10 41;

fax: +33 4 78 77 10 22.

E-mail address: leger@sommeil.univ-lyon1.fr (L. Leger).1 Present address: Douglas Institute Research Center, McGill University, 6875,

boul. LaSalle, Verdun, Montreal, Quebec H4H 1R3, Canada.

0891-0618/$ – see front matter � 2010 Elsevier B.V. All rights reserved.

doi:10.1016/j.jchemneu.2010.03.001

1. Introduction

In mammals, dopaminergic (DA) neurons are contained inseveral cellular groups distributed from the caudal mesencephalonto the rostralmost levels of the brain (Hokfelt et al., 1984; Lindvalland Bjorklund, 1983). These neurons participate in physiologicalprocesses as diverse as locomotor activity, sensorimotor integration,motivation and reward, and sexual function (Giuliano and Allard,2001; Le Moal, 1995). Since the experimental lesioning made byJones et al. (1973) in the substantia nigra and ventral tegmental area(VTA), several attempts have been made to disclose the exact role ofdopamine in the regulation of sleep and waking (for review seeMonti and Monti, 2007). An overwhelming number of data indicatethat dopamine is a key transmitter in the modulation of behaviouralarousal, but its role in the regulation of sleep is less clear. Forexample, depending on the dose, agonists and antagonists of the DAreceptors either decrease or increase slow-wave sleep and para-doxical (REM) sleep (PS) (Monti and Monti, 2007). Interestingly, ithas been suggested that dopamine could participate in dreamingoccurring during PS (Solms, 2000). Early electrophysiologicalrecordings showed that DA neurons in the substantia nigra andVTA do not change their mean firing rate across vigilance stages(Trulson et al., 1981; Trulson and Preussler, 1984). However, recent

L. Leger et al. / Journal of Chemical Neuroanatomy 39 (2010) 262–271 263

findings indicate that DA cells in the VTA switch to a prominentbursting pattern of discharge during PS, similarly to the activity theydisplay during rewarding processes (Dahan et al., 2007). Moreover,dopamine concentrations, as measured by microdialysis in theprefrontal cortex, are higher during PS than during slow-wave sleep(Lena et al., 2005). In the VTA, DA neurons express Fos during therecovery period following a selective PS deprivation (Maloney et al.,2002). In addition, there are indications that the DA neurons locatedin the caudal periaqueductal gray express Fos during sleepdeprivation (Lu et al., 2006). The activity of the DA neurons in thehypothalamic cell groups in relation to the sleep-waking cycle isunknown. The present work was undertaken in order to re-examine,in the ventral mesencephalic groups, and study in the hypothalamicgroups the distribution of the DA neurons activated when enhancingwaking or suppressing PS.

To achieve this goal, we took advantage of the elevatedexpression of the immediate-early gene c-fos in neurons after astimulation (Dragunow and Faull, 1989; Kovacs, 1998; Morgan andCurran, 1991), a method widely used to map neurons activated indifferent experimental situations. This method has been success-fully used in the rat by our and other groups to localize, with theimmunohistochemical detection of the protein Fos, the brainstemneurons active during PS and/or W (Leger et al., 2009b; Maloneyet al., 1999, 2000; Sapin et al., 2009; Shiromani et al., 1995; Verretet al., 2003, 2005, 2006). In the present work, rats either remainedawake and active by placing them in a novel environment (NEv) orwere specifically deprived of PS, with some of them allowed torecover from this deprivation. The expression of the protein Fos inDA neurons was visualized by dual immunostaining combining Fosand tyrosine hydroxylase (TH) immunohistochemistry. The DA A8to A14 groups were examined. This work already appeared inabstract form (Leger et al., 2006, 2007, 2009a) and is complemen-tary of that describing the expression of Fos in the noradrenergicneurons in the same experimental conditions (Leger et al., 2009b).

2. Experimental procedures

2.1. Animals

OFA adult male rats (250–350 g) (Charles River France, L’Arbresle, France) were

used in this study. Before and during the experiments they were housed with a 12-h

light dark cycle (lights on between 7.00 h and 19.00 h). Four series of three animals

were submitted to the protocol consisting in a selective deprivation/recovery of PS.

Four additional animals were submitted to the NEv protocol to maintain them

awake. All experiments were performed according to the European Community

Council Directive (86/609/EEC). The protocol was approved by the Institutional

Animal Care and Use Committee of our University (BH 2006-09).

2.2. Novel environment and PS deprivation-recovery protocols

The NEv protocol was used to induce a continuous period of alert waking at a

time when rats are spending most of their time asleep. The animals were placed

two-by-two and free to move during 3 h (13.00 h to 16.00 h) in a square

1.5 m � 1.5 m � 0.5 m black open-field with numerous coloured plastic toys. The

light intensity above the open field was the same as that of the animal room. The

animals had free access to water and food. They were continuously observed by the

experimenter and the toys were moved when the animals seemed to get sleepy.

In the deprivation/recovery protocol, each series consisted in one control rat

remaining on a bed of woodchips in its individual container during the whole

experiment (PSC condition), a second rat deprived of PS during 75 h with the

flowerpot technique on a platform (6.5 cm in diameter) surrounded by water, and

sacrificed at the end of the deprivation period (PSD condition), and a third rat

deprived of PS during 72 h, but allowed to recover during 3 h on a dry bed of

woodchips (PSR condition). During the deprivation period, the three rats were

placed next to each other, which reduced the stress associated to social isolation.

They were regularly watched over through an eyepiece. During the daily cleaning of

the deprivation container (starting at 10 a.m.), the rats were transferred for 30 min

to a clean container with dry woodchips.

2.3. Histological and immunohistochemical procedures

Rats were perfused at the same time of the day, i.e. 16.00 h to avoid interference

of the nycthemeral rhythm with the results. They were deeply anesthetized with

pentobarbital and perfused with a Ringer-lactate solution containing 0.1% heparin,

followed by 400 mL of a fixative composed of 4% paraformaldehyde in 0.1 M

phosphate buffer (PB; pH 7.4). The brains were postfixed overnight in the same

fixative, stored for at least 2 days in 0.1 M PB with 30% sucrose and frozen with

expanded CO2 gas. Coronal 25-mm-thick sections were obtained with a cryostat

(HM550 Microm) and stored free-floating at 4 8C in 0.02 M PB containing 0.9% NaCl,

0.3% Triton X-100 (PBST) and 0.1% sodium azide (PBST-Az).

When Fos and TH immunostaining were combined, Fos immunostaining was

performed first in order to obtain black nuclei and TH second to obtain brown

cytoplasms. Sections from one animal in each series (PSC, PSD, PSR and NEv) were

run simultaneously. The sequence of incubations was: (i) a rabbit antiserum to Fos

(1: 10,000; Ab-5, Calbiochem) in PBST-Az for 3 days at 4 8C; (ii) a biotinylated goat

anti-rabbit IgG (1: 2,000; Vector) overnight at 4 8C and (iii) an ABC-HRP solution (1:

1,000; Elite kit, Vector) for 90 min at room temperature. The reaction was

developed in 0.05 M Tris–HCl buffer (pH 7.6) containing 0.025% 3,30-diamino-

benzidine-4HCl (DAB; Sigma), 0.003% H2O2 and 0.6% nickel ammonium sulphate.

The sections were kept not less than one night in PBST-Az at 4 8C. They were then

incubated in a rabbit antiserum to TH (1:5000; Institut J. Boy, France) in PBST-Az for

3 days at 4 8C, followed by the same sequence of antibodies and reagents as before,

except that DAB was prepared without nickel ammonium sulphate. The sections

were mounted on gelatin-coated slides, dried, dehydrated and coverslipped with

DePeX.

Controls with omission of the primary or secondary antibodies were run to check

for the absence of nonspecific staining and cross-reactions in the dual

immunostainings performed with two primary antibodies made in rabbit. The

clear-cut presence of three populations of immunoreactive (IR)-neurons, i.e. single

Fos, single TH and double-labeled Fos–TH neurons, together with a perfect match of

the first two populations with the descriptions reported in the literature (Chan-

Palay et al., 1984; Hokfelt et al., 1984; Lindvall and Bjorklund, 1983; Sapin et al.,

2009; Verret et al., 2003, 2005, 2006) strongly argued against any cross-reaction.

The antiserum to Fos was made against a synthetic peptide corresponding to the N-

terminal part (residues 4–17) of human Fos. This part of the protein displays 100%

homology between human and rat and no homology with Fos-related antigens such

as Fos B, Jun B, Fra-1 and Fra-2 (Blast 2 sequences, NCBI). The antiserum to TH was

made against enzyme isolated and purified from rat pheochromocytoma (Arluison

et al., 1984).

2.4. Analysis and quantification

The DA neurons form a continuum of cells from the ventral midbrain to the

preoptic area and extend dorsally up in the periaqueductal gray. To delineate the

different groups we followed the nomenclature of Dahlstrom and Fuxe (1964),

Hokfelt et al. (1984) and Lindvall and Bjorklund (1983), updated by German and

Manaye (1993) and Ikemoto (2007) for the ventral midbrain. The A8 group lies in

the retrorubral field, occupying the ventral part of the caudal mesencephalic

reticular formation. The A9 group encompasses the substantia nigra pars compacta,

pars reticulata and pars lateralis in the ventral mesencephalon. The A10 group

covers the whole VTA and adjacent nuclei on the midline, namely the interfascicular

and rostral linear nuclei (Ikemoto, 2007). The DA cells distributed in the ventral

periaqueductal gray, from the dorsal raphe nucleus to the rostralmost periaque-

ductal gray, were denominated A10 dorsocaudal (A10dc), according to Hokfelt et al.

(1984). The DA neurons in the central linear nucleus and rostral to it were included

in this DA cell group. The small round DA neurons located in the supramammillary

nucleus were counted separately (Shepard et al., 1988) and denominated

ventrorostral A10 (A10vr). The A11 group extends from the rostralmost

periaqueductal gray to the dorsal and caudal hypothalamus, medially to the

mammillothalamic tract. The caudally located A11 neurons were differentiated

from the rostral A10dc neurons by their larger size and higher immunoreactivity,

both parameters being similar to the A11 neurons in the hypothalamus. The A12

group is located in the hypothalamic arcuate nucleus. The A13 group is restricted to

the medial part of the zona incerta, at mid-rostrocaudal level of this nucleus. The

A14 group is located in the rostral hypothalamus, along the third ventricle and

extends dorsally into the posterior part of the paraventricular nucleus.

Sections, taken at 600 mm intervals across the DA groups on both sides of the

brain, were entirely drawn and plotted in one animal from each experimental

condition. In the other 3 animals, only the DA cell groups were drawn and plotted.

The selection of sections was as follows, according to Paxinos and Watson (1997):

A8 (2 sections at�7.2 and�6.6 from Bregma,); A9 and A10 (4 sections at�6.6,�6.0,

�5.4 and �4.8); A10vr (2 sections �4.8 and �4.2); A10dc (5 sections at �7.2, �6.6,

�6.0, �5.4 and �4.8); A11 (3 sections at �4.8, �3.9 and �3.3); A12 (3 sections at

�3.9, �3.3 and �2.7); A13 (1 section at �2.7) and A14 (2 sections at �2.1 and 1.5).

The delineation of each DA group was made by tracing a broken line enclosing all

the TH-IR cell bodies and passing 50–100 mm apart from them, usually across their

dendritic fields. The delineation and naming of all the other nuclei or areas were

made according to Paxinos and Watson (Paxinos and Watson, 1997) (Fig. 1).

Drawings of double immunostained sections were made with an image analysis

system (Mercator, ExploraNova, La Rochelle, France) coupled to a Zeiss Axioskop

microscope equipped with a motorized X–Y sensitive stage. The three categories of

neurons (single Fos, single TH and double Fos/TH) were plotted and, once plotted,

were automatically counted by the software Mercator. Fos-IR nuclei were

Fig. 1. Schematic distribution of the Fos-immunoreactive (small black dots), TH-immunoreactive (gray dots) and Fos–TH double-labeled neurons (red dots) in sections taken

at 600 mm intervals through the dopaminergic cell groups in a representative animal for control (PSC), PS-deprived (PSD), PS-recovery (PSR) and novel environment (NEv)

conditions. Abbreviations for the names of the structures are in the main list.

L. Leger et al. / Journal of Chemical Neuroanatomy 39 (2010) 262–271264

Fig. 1. (Continued )

L. Leger et al. / Journal of Chemical Neuroanatomy 39 (2010) 262–271 265

Fig. 1. (Continued ).

L. Leger et al. / Journal of Chemical Neuroanatomy 39 (2010) 262–271266

considered as positive when the DAB-Ni precipitate was of a deep blue-black color

(see Figs. 2 and 3). The single TH-IR cell bodies were counted when a pale area

corresponding to the nucleus, or part of the nucleus, was clearly identified. The

numbers counted here were proportional to those estimated by other authors for

the different DA groups (German and Manaye, 1993; Halliday and Tork, 1986;

Reymond et al., 1984; Stratford and Wirtshafter, 1990; van den Pol et al., 1984). For

the 16 animals used in this study, the averages were 286 � 39 TH-IR neurons for A8,

1269 � 124 neurons for A9, 988 � 74 neurons for A10, 60 � 15 neurons for A10vr,

254 � 24 neurons for A10dc, 56 � 6 neurons for A11, 148 � 26 neurons for A12,

186 � 26 neurons for A13 and 111 � 12 neurons for A14.

The digital images were taken with the CCD Color 10-bit QiCam camera used to

plot the labeled neurons. They were imported into Adobe Photoshop 7.0, digitally

adjusted for brightness and contrast, and were assembled into plates at a resolution

of 300 dpi.

2.5. Statistical analysis

To compare the data obtained in the 4 independent experimental conditions, we

performed non-parametric tests and reported median and quartiles values. We first

performed a Kruskal–Wallis test. Secondly, in order to uncover differences between 2

conditions, we used a new generation of non-parametric tests based on permutation,

with the StatXact software (StatXcat1 8 from Cytel Inc, Cambridge, USA).

3. Results

The distribution of the 3 categories of labeled neurons (Fos-IR,TH-IR and Fos/TH-IR) is illustrated in Figs. 1–3, from caudalmesencephalon to rostral hypothalamus. The localization of theFos-IR neurons in the PSC, PSD and PSR groups was similar to thatdescribed in our previous studies using the same deprivation/

recovery protocol (Sapin et al., 2009; Verret et al., 2003). Inaddition, in our study devoted to the noradrenergic neurons (Legeret al., 2009a) and using the same animals as here, we showed thatthe number of Fos-IR neurons was not significantly different fromthat obtained by Verret et al. (2005) in seven areas of thebrainstem. These areas were the lateral and dorsal paragiganto-cellular nuclei, the ventral part of the gigantocellular nucleus, thelocus coeruleus, the sublaterodorsal nucleus, the dorsal raphe andthe ventrolateral periaqueductal gray. If the localization andnumber of Fos-IR neurons do not vary significantly amongindependent series of animals, the figures obtained in each newseries can therefore be considered confidently.

3.1. A8 and A9 groups (retrorubral field and substantia nigra)

The A8 and A9 groups contained relatively few Fos-IR neurons(n = 7–61), except A9 in the NEv condition (126 neurons). Thesenumbers were significantly different from control in A9 in both thePSR and the NEv conditions. Among these Fos-IR neurons anextremely low number was TH-IR (2 neurons in NEv whichrepresents 0.1% of the TH-IR neurons).

3.2. A10 group (ventral tegmental area)

The number of Fos-IR neurons increased from PSC to PSD, fromPSD to PSR and from PSR to NEv conditions. This number was

Fig. 2. Photomicrographs of the ventral tegmental area (A10 in A), the supramammillary nucleus (A10vr in B) and the periaqueductal group (A10dc in C and D) groups in the

PS-deprived (PSD) (A and C) and novel environment (NEv) (B and D) conditions. In the A10 group (A) few double-labeled (arrows) and singly labeled Fos (short arrows) are

observed in the PSD condition. In A10vr (B) numerous Fos-IR nuclei (short arrows) are observed in the NEv condition and several neurons are double-labeled. Note the low

number of double-labeled neurons in the A10dc group in both conditions (arrows in C and D), although numerous singly labeled Fos-IR (short arrows) are intermingled with

the TH neurons. In A midline is toward the left. Bar = 100 mm.

L. Leger et al. / Journal of Chemical Neuroanatomy 39 (2010) 262–271 267

significantly different from control in PSD, PSR and NEv conditions.The number of Fos–TH-IR neurons was low in both PSD and PSRconditions (5 and 4 neurons, respectively). It was slightly higher inthe NEv condition (14 neurons or 1.3% of the TH neurons) and wassignificantly different from both control and PSR conditions. InPSD, PSR and NEv conditions, the double-labeled neurons tended tobe located in the rostral and ventral VTA, above the interfascicularnucleus (Fig. 2A) or above the anterior part of the interpeduncularnucleus. They usually possessed small round somata and fewdendrites.

3.3. Ventrorostral A10 (supramammillary nucleus)

Like in the A9 and A10 groups, the number of Fos-IR neuronsincreased from PSC to PSD, from PSD to PSR and further from PSRto NEv. In each condition, it was close to that counted in A10 andwas statistically different from control in PSR and NEv. Thenumber of double-labeled neurons was similar to that counted inA10 in the PSD condition (5 neurons). It was lower than in A10 inthe PSR and NEv conditions. The percentage of these double-labeled neurons versus the TH neurons reached or was close to12% in both PSD and NEv conditions (Fig. 2B) but with lowstatistical significance.

3.4. A10dc group (periaqueductal gray)

The number of Fos-IR neurons was significantly different fromcontrol in the three other conditions. It was the same in the PSDand PSR conditions (105 and 107 neurons, respectively). It wasapproximately doubled in the NEv condition (191 neurons), whichyielded a statistical significance versus the PSD condition. Thenumber of double-labeled neurons was statistically higher thancontrol only in the PSD and NEv conditions (9 and 6.5 neurons,

respectively). These numbers represented less than 5% of the TH-IRneurons (4.5% in PSD and 2.7% in NEv).

3.5. A11 group (caudal hypothalamus)

In this group, the number of Fos-IR neurons was significantlyhigher than control (32 neurons) in both the PSD (84.5 neurons)and NEv (123 neurons) conditions. In addition, in NEv the numberof Fos neurons was also significantly higher than PSR (68.5neurons). The number of double-labeled neurons followed thesame pattern but with much lower figures. In PSD and NEv thesedouble-labeled neurons accounted for 13.2 and 9.6% of the TH-IRneurons, respectively (Fig. 3A and B).

3.6. A12 group (arcuate nucleus)

Only the PSD and NEv conditions yielded numbers of Fos-IRneurons (100 and 89.5, respectively) significantly higher thancontrol (13 neurons). The number of Fos–TH-IR was low in eachcondition and did not show statistical difference across conditions.

3.7. A13 group (medial zona incerta)

As for both the Fos-IR neurons and the double-labeled neurons,this group reacted differently from the other groups (Fig. 3C–E).Indeed, it displayed a high number of Fos neurons in the PSRcondition (127.5 neurons). This number was significantly higherthan in PSC (33 neurons) and PSD (75 neurons). In the same PSRcondition, the number of Fos–TH-IR neurons was the highestamong all DA groups (26 neurons) and was significantly differentfrom the NEv condition. In addition, the percentage versus the THneurons (20.6%) was also significantly different from PSD.Interestingly in this PSR condition, the double-labeled neurons

Fig. 3. Photomicrographs of the A11 (A and B), A13 (C, D and E) and A14 (F) dopaminergic groups in the control (PSC), PS-deprived (PSD), PS-recovery (PSR) and novel

environment (NEv) conditions. Note the double-labeled neurons in the A11 group in the PSD and NEv conditions (A and B). In the A13 group, the double-labeled neurons are

located laterally in the PSD condition (arrows in D), whereas they are distributed over the whole mediolateral extent of the group in the PSR condition (arrows in E). In the A14

group, some double-labeled neurons are observed in the PSR condition (arrow in F). The midline is on the left in all photos. Bar = 100 mm.

L. Leger et al. / Journal of Chemical Neuroanatomy 39 (2010) 262–271268

were homogeneously distributed in the cellular group (Fig. 3E),whereas in the PSD condition the double-labeled cells tended tosegregate in the lateral half of the DA group (Fig. 3D).

3.8. A14 group (rostral third ventricle)

The number of Fos-IR neurons was significantly higher thancontrol in PSD, PSR and NEv (107.5, 55.5 and 157 neurons,respectively) and higher than PSR in NEv. The number of double-labeled neurons was low, except in the PSR condition (11.5neurons) (Fig. 3F). This number represented 9.3% of the TH-IRneurons, a figure significantly different from PSC, PSD and NEv, likein A13.

4. Discussion

In the present experimental conditions increasing or decreasingthe amount of PS or waking, Fos was induced in a substantialnumber of neurons in the areas occupied by DA neurons from thecaudal mesencephalon to the rostral hypothalamus. Whatever theDA group and experimental condition considered, a maximum of

20% of the DA neurons were Fos-IR. The percentage was higherthan 10% in 3 areas and 3 conditions, namely the ventrorostralextension of the VTA (A10vr) in the PSD and NEv conditions, thecaudal hypothalamic group (A11) in the PSD condition, and themedial zona incerta (A13) group in the PSR condition.

4.1. Methodological considerations

In the present study, the changes in Fos expression areinterpreted as reflecting changes in neuronal activity associatedwith the different experimental conditions. It should be remem-bered that the induction of Fos requires synaptic receptoractivation and increased concentration of intracellular calciumbut that increased spike activity does not necessarily induce Fos(Luckman et al., 1994). For example, the short high-frequencybursts of activity recorded in the hypothalamic oxytocin neuronsduring lactation in the female rat are not a stimulus for Fosinduction (Poulain and Wakerley, 1982). Conversely, Fos expres-sion can occur independently of neuronal discharge (Morgan andCurran, 1991). Fos may be regarded as a link between externalcellular signals and metabolic/genomic changes, which still

L. Leger et al. / Journal of Chemical Neuroanatomy 39 (2010) 262–271 269

renders Fos immunohistochemistry a useful tool to map activatedneurons (Kovacs, 2008).

Both the PSD and NEv periods generate stress, semi-chronic forthe first condition and acute for the second condition. To ourknowledge, no data is available on Fos expression in DA neurons ofchronically stressed animals. In addition, there is generalagreement that chronic stress is associated with a progressivedecrease in the amount of Fos-IR neurons in the stress-relatedareas of the brain (Chen and Herbert, 1995; Senba and Ueyama,1997). It is not excluded that acute stress, likely generated byexposure to the NEv, might participate in the induction of Fos insome DA neurons. For example, DA neurons in the A10 group wereshown to express Fos after acute exposure to restraint or a predatorodor (Deutch et al., 1991; Redmond et al., 2002).

4.2. DA neurons expressing Fos in low number across experimental

conditions

4.2.1. A8–A9 groups (retrorubral field and substantia nigra)

An extremely low number of double-labeled Fos–TH neuronswere observed in the A8 and A9 groups in our experimentalconditions. This paucity cannot be attributed to a deficiency in Fosstaining since clearly visible, sometimes numerous, e.g. in the NEv

Table 1Numbers and percentages of Fos- and Fos–TH-immunoreactive neurons in the dopami

s PSC PSD

Numbers of Fos

A8 2 8.5 (6.5; 20.2) 25.0 (19.0; 51

A9 4 7.0 (2.5; 17.5) 33.5 (20.5; 54

A10 4 16.0 (6.7; 30.5) 67.0 (38.2; 12

A10vr 2 19.5 (9.7; 27.7) 64.5 (34.5; 10

A10dc 5 41.0 (18.5; 53.0) 105.0 (98.2; 15

A11 3 32.5 (18.7; 48.5) 84.5 (59.7; 12

A12 3 13 (3.7; 23.0) 100.0 (28.0; 10

A13 1 33.0 (27.5; 52.0) 75.0 (31.5; 89

A14 2 33.0 (21.5; 38.7) 107.5 (77.5; 13

Numbers of Fos/TH

A8 2 0.0 (0.0; 0.0) 0.0 (0.0; 1.5)

A9 4 0.0 (0.0; 0.0) 0.0 (0.0; 0.7)

A10 4 0.0 (0.0; 0.7) 5.0 (0.7; 18.5

A10vr 2 1.0 (0.2; 3.2) 5.5 (4.2; 8.2)

A10dc 5 2.0 (0.5; 2.0) 9.0 (5.7; 16.0

A11 3 0.5 (0.0; 1.7) 6.0 (2.7; 8.5)

A12 3 0.5 (0.0; 1.7) 1.0 (0.0; 2.0)

A13 1 4.0 (1.2; 9.0) 14.5 (9.2; 22.0

A14 2 0.5 (0.0; 4.7) 2.5 (2.0; 9.0)

% Fos–TH vs Fos

A8 2 0.0 (0.0; 0.0) 0.0 (0.0; 0.7)

A9 4 0.0 (0.0; 0.0) 0.0 (0.0; 0.0)

A10 4 0.0 (0.0; 0.1) 1.4 (0.1; 8.2)

A10vr 2 3.5 (0.3; 11.0) 9.7 (7.1; 12.7

A10dc 5 0.7 (0.1; 4.3) 5.9 (4.1; 9.4)

A11 3 1.0 (0.0; 3.9) 10.5 (4.3; 13.9

A12 3 0.5 (0.0; 6.3) 0.8 (0.0; 1.6)

A13 1 3.7 (1.9; 14.0) 13.8 (7.3; 29.8

A14 2 0.3 (0.0; 4.5) 2.7 (1.6; 5.6)

%Fos–TH vs TH

A8 2 0.0 (0.0; 0.0) 0.0 (0.0; 0.0)

A9 4 0.0 (0.0; 0.0) 0.0 (0.0; 0.0)

A10 4 0.0 (0.0; 0.0) 0.5 (0.2; 1.0)

A10vr 2 2.1 (1.0; 3.2) 12 (10.7; 12.3

A10dc 5 0.6 (0.4; 0.7) 4.5 (3.3; 5.6)

A11 3 1 (0.0; 2.4) 13.2 (9.8; 15.3

A12 3 0.5 (0.0; 1.0) 0.3 (0.0; 0.8)

A13 1 2.2 (1.4; 3.0) 7.9 (6.0; 10.4

A14 2 0.4 (0.0; 1.8) 2.4 (1.8; 3.8)

Numbers and percentages (median and Q1; Q3 quartiles) of Fos-immunoreactive and do

groups in control (PSC, n = 4), PS-deprived (PSD, n = 4), PS-recovery (PSR, n = 4) and novel e

percentages of neurons counted bilaterally every 600 mm through the dopaminergic g#P�0.05 vs PSD; 1P�0.05 vs PSR; *P�0.05 vs NEv.

condition, single-labeled Fos neurons were encountered in thevicinity of the TH-IR neurons. According to authors (Dragunow andFaull, 1989; Ma et al., 1993; Sgambato et al., 1997), TH neurons inthe A8 and A9 groups rarely express detectable levels of Fos,whatever the experimental situation. This may lead to falsenegative result when an activation of these neurons is suspected,but is coherent with the absence of change in their mean firing rateacross the sleep-waking cycle (Trulson et al., 1981; Trulson andPreussler, 1984) (Table 1).

4.2.2. A10 group (ventral tegmental area)

A slightly higher number of TH-IR neurons were Fos-IR in the A10group (VTA) as compared to the A8 and A9 groups, especially in theNEv condition. In this condition, the activation of TH neurons couldreflect increased alertness and rewards that might be expected bythe rats during exploration of the NEv (Kakade and Dayan, 2002) andmight not be directly related to the continuous waking occurringduring this period. In the present work, the number of double-labeled neurons was very similar in the PSD and PSR conditions (5and 4 neurons, respectively). This result differs from that of Maloneyet al. (2002) who, using a protocol similar to ours (53 h in PSD insteadof 75 h), found that the number of double-labeled neurons wasincreased in the PSR condition (12 neurons) compared to the PSD

nergic groups.

PSR NEv

.2) 36.0 (33.0; 70.5)* 22.5 (12.75; 37.5)

.7) 61.0 (24.7; 92.7)* 126.0 (64.0; 154.2)*

8.7)* 84.5 (51.2; 143.2)* 188.5 (147.2; 241.0)*

6.5) 78.0 (61.2; 151.0)*,# 137.5 (70.0; 321.2)*

1.2)* 107.0 (67.2; 269.0)* 191.0 (174.0; 214.0)*,#

5.7)* 68.5 (38.2; 88.2) 123.0 (110.7; 172.0)*,1

0.0) 67.5 (15.7; 132.0) 89.5 (77.0; 209.5)*

.7) 127.5 (113.5; 207.5)*,# 79.5 (52.0; 157.2)

8.2)* 55.5 (51.2; 83.0)* 157.0 (124.7; 185.5)*,1

0.0 (0.0; 0.0) 0.0 (0.0; 0.0)

0.5 (0.0; 1.7) 2.0 (0.2; 3.7)

) 4.0 (2.2; 5.7)* 14.0 (10.2; 26.7)*,1

2.0 (0.2; 6.7) 5.5 (2.5; 28.0)

)* 5.0 (1.0; 12.7) 6.5 (5.2; 7.0)*

*,1 1.0 (0.0; 2.0) 5.5 (4.2; 12.7)*,1

1.5 (0.2; 3.5) 4.0 (0.2; 19.7)

) 26.0 (20.7; 44.0)*,* 10.5 (5.5; 12.5)

11.5 (7.7; 13.0)* 2.5 (1.0; 7.0)

0.0 (0.0; 0.0) 0.0 (0.0; 0.0)

0.1 (0.0; 1.1) 0.3 (0.0; 1.9)

2.6 (0.3; 3.9) 3.7 (1.2; 7.8)*

)1 3.2 (0.3; 7.0) 7.5 (2.7; 12.8)

*,* 3.6 (0.6; 4.7) 3.1 (2.4; 3.2)

)# 1.1 (0.0; 2.5) 6.4 (4.2; 11.6)1

1.5 (0.3; 2.4) 2.7 (0.1; 9.2)

) 19.8 (14.8; 26.5)* 7.0 (3.9; 10.2)

13.4 (8.9; 18.8)#,* 1.6 (0.8; 4.4)

0.0 (0.0; 0.0) 0.0 (0.0; 0.0)

0.0 (0.0; 0.1) 0.1 (0.0; 0.2)

0.3 (0.1; 0.5) 1.3 (1.0; 1.7)*

)* 4.7 (1.4; 8.1) 11.3 (7.1; 22.6)

* 2.5 (1.8; 3.2) 2.7 (2.1; 3.1)*

)*,1 1.3 (0.0; 2.8) 9.6 (6.4; 15.9)*,1

0.7 (0.0; 1.2) 2.3 (0.3; 7.3)

)* 20.6 (13.5; 28.9)*,#,* 4.2 (3.2; 5.2)

9.3 (8.1; 11.5)*,#,* 2.3 (1.0; 4.6)

uble-labeled Fos–TH-immunoreactive neurons counted in the rat dopaminergic cell

nvironment (NEv), n = 4) conditions. The values displayed represent the numbers or

roups, on one to five sections (column s). Significance values are *P�0.05 vs PSC;

L. Leger et al. / Journal of Chemical Neuroanatomy 39 (2010) 262–271270

condition (8 neurons). However, it should be realized that, in bothstudies, these numbers represent 1% or less of the TH-IR neurons.The bursty mode of discharge, recently shown to occur during PS inthe A10 group (Dahan et al., 2007) would not be sufficient to triggerthe induction of Fos in a large number of DA neurons.

Among the non-DA neurons exhibiting a Fos-IR nucleus in thePSR condition might reside a large percentage of GABA neurons, asobserved by Maloney et al. (2002) and, more significantly, by Sapinet al. (2009) who used the same deprivation protocol as here and acombination of in situ hybridization and immunohistochemistry todetect the activated GABA neurons.

4.3. Dopaminergic neurons expressing Fos in paradoxical sleep

deprivation and NEv: A10dc (periaqueductal gray) and A11 (caudal

hypothalamic) groups

In both A10dc and A11 groups, the number of activated THneurons is higher in PSD and NEv conditions than in PSC condition,but this number represents a substantial proportion of the THneurons only in the A11 group. Infusion of dopamine in the peri-LCalpha, a pontine nucleus critically involved in the generation of PSin the cat, yields decreases in the amounts of PS (Crochet and Sakai,1999). In addition, in the same species the DA axons present in thisnucleus are partly issued from the A11 group (Sakai, 1991).Therefore, it is not excluded that the DA neurons in the A11 groupcontribute to the inhibition of PS. The activation of DA neurons inthe A11 group could also partly reflect the possible modificationsin autonomic and sensory functions occurring at the end of thedeprivation period (Rechtschaffen et al., 1989). Indeed, it has beensuggested that this cell group might constitute both a sym-pathoexcitatory and endogenous pain inhibitory system throughits well-documented projection to the spinal cord (Skagerberg andLindvall, 1985). Finally, it should not be forgotten that the A11group also contains double-labeled neurons in the NEv condition,which suggests that the DA neurons of this group might play a rolein waking, since this physiological state is common between PSDand NEv.

Contrary to the results of Lu et al. (2006) we did not observe asubstantial increase in the number of double-labeled cells in theA10dc group at the end of the time spent in the NEv. These authorsfound that 41% of the TH neurons in this DA group were Fos-IR after2 h of total sleep deprivation, compared to less than 3% here after3 h of NEv. The sleep deprivation in Lu et al. (2006) consisted inremoving the cover of the cage of the rats when signs of slow-wavesleep appeared, as seen on the EEG monitor. The antibody used isthe same (Ab5 from Oncogene) in both studies. In our mind, thedifference is not likely due to a poor Fos immunoreactivity in thepresent work, since a high number of Fos-IR non-TH neurons werepresent in the area of the A10dc group (compare our Fig. 2D withFig. 1 by Lu et al., 2006). Even in the caudal half of the A10dc group,as sampled by Lu et al. (2006), we did not obtain a high number ofdouble-labeled cells. One possible explanation is that theautoinhibition of Fos occurs sooner in this neuronal group thanin the other DA groups. It might also be that this DA group is notinvolved in waking per se, this state being present in both thegentle deprivation and the NEv, but in certain behaviours displayedduring the protocol used by Lu et al. (2006). To resolve thisdiscrepancy, it would be necessary to record the unitary activity ofthe DA neurons in this group across the sleep-waking cycle.

4.4. Dopaminergic neurons expressing Fos in paradoxical sleep

recovery: A13 (medial zona incerta) and A14 groups (rostral third

ventricle)

Unexpectedly and not mentioned by other authors, the A13group and less markedly the A14 group contained a significant

percentage of double-labeled neurons (20.6 and 9.3%, respec-tively), specifically during the PS-recovery phase. Interestingly, werecently showed that this condition induces a large population ofFos-IR neurons in the zona incerta, one-third of which express theneuropeptide melanin-concentrating hormone (Verret et al.,2003). DA and melanin-concentrating hormone neurons representseparate neuronal populations in the zona incerta (Sita et al., 2003),which means that the neurotransmitter of approximately half ofthe Fos-IR neurons, likely participating in the regulation of PS inthis area, has been identified.

Both the A13 and A14 groups are part of the incerto-hypothalamic neuronal system (Lookingland and Moore, 2005)which sends axons to the paraventricular and supraoptic nuclei,the anterior hypothalamus, the basal forebrain and the lateralpreoptic area (Cheung et al., 1998; van Vulpen et al., 1999; Wagneret al., 1995). Dopamine, when released in both the paraventricularnucleus and preoptic area, would be a key transmitter in thecontrol of sexual behaviour and, particularly, penile erections(Giuliano and Allard, 2001). Interestingly, it was shown thatneurotoxic lesions of the lateral preoptic area reduce the number ofpenile erections occurring during PS without affecting thoseoccurring during waking (Schmidt et al., 2000). Further studies areneeded to determine whether the activation of the TH-IR neuronsin the A13 and A14 DA groups, as observed here, is related to thischaracteristic phenomenon of PS.

5. Conclusion

In experimental conditions manipulating the quantities ofwaking or PS in the rat, we observed that DA neurons as a whole arenot strongly activated, as seen by the expression of Fos. However, itis suggested that the caudal hypothalamic group (A11) participatesin waking and/or the inhibition of PS and that the zona incertagroup (A13) participates in the regulation of PS.

Acknowledgements

This work was supported by CNRS (CNRS UMR5167) andUniversite de Lyon, Universite Lyon 1.

References

Arluison, M., Dietl, M., Thibault, J., 1984. Ultrastructural morphology of dopami-nergic nerve terminals and synapses in the striatum of the rat using tyrosinehydroxylase immunocytochemistry: a topographical study. Brain Res. Bull. 13,269–285.

Chan-Palay, V., Zaborszky, L., Kohler, C., Goldstein, M., Palay, S.L., 1984. Distributionof tyrosine-hydroxylase-immunoreactive neurons in the hypothalamus of rats.J. Comp. Neurol. 227, 467–496.

Chen, X., Herbert, J., 1995. Regional changes in c-fos expression in the basalforebrain and brainstem during adaptation to repeated stress: correlationswith cardiovascular, hypothermic and endocrine responses. Neuroscience 64,675–685.

Cheung, S., Ballew, J.R., Moore, K.E., Lookingland, K.J., 1998. Contribution of dopa-mine neurons in the medial zona incerta to the innervation of the centralnucleus of the amygdala, horizontal diagonal band of Broca and hypothalamicparaventricular nucleus. Brain Res. 808, 174–181.

Crochet, S., Sakai, K., 1999. Effects of microdialysis application of monoamines onthe EEG and behavioural states in the cat mesopontine tegmentum. Eur. J.Neurosci. 11, 3738–3752.

Dahan, L., Astier, B., Vautrelle, N., Urbain, N., Kocsis, B., Chouvet, G., 2007. Prominentburst firing of dopaminergic neurons in the ventral tegmental area duringparadoxical sleep. Neuropsychopharmacology 32, 1232–1241.

Dahlstrom, A., Fuxe, K., 1964. Evidence for the existence of monoamine-containingneurons in the central nervous system. I. Demonstration of monoamines in thecell bodies of brain stem neurons. Acta Physiol. Scand. (Suppl.) 232, 1–55.

Deutch, A.Y., Lee, M.C., Gillham, M.H., Cameron, D.A., Goldstein, M., Iadarola, M.J.,1991. Stress selectively increases Fos protein in dopamine neurons innervatingthe prefrontal cortex. Cereb. Cortex 1, 273–292.

Dragunow, M., Faull, R., 1989. The use of c-fos as a metabolic marker in neuronalpathway tracing. J. Neurosci. Methods 29, 261–265.

L. Leger et al. / Journal of Chemical Neuroanatomy 39 (2010) 262–271 271

German, D.C., Manaye, K.F., 1993. Midbrain dopaminergic neurons (nuclei A8, A9,and A10): three-dimensional reconstruction in the rat. J. Comp. Neurol. 331,297–309.

Giuliano, F., Allard, J., 2001. Dopamine and male sexual function. Eur. Urol. 40, 601–608.

Halliday, G.M., Tork, I., 1986. Comparative anatomy of the ventromedial mesence-phalic tegmentum in the rat, cat, monkey and human. J. Comp. Neurol. 252,423–445.

Hokfelt, T., Martensson, R., Bjorklund, A., Kleinau, S., Goldstein, M., 1984. Distribu-tion maps of tyrosine-hydroxylase-immunoreactive neurons in the rat brain. In:A Bjorklund, H.T. (Ed.), Handbook of Chemical Neuroanatomy, vol. 2. ElsevierScience Publishers BV, pp. 277–287.

Ikemoto, S., 2007. Dopamine reward circuitry: two projection systems from theventral midbrain to the nucleus accumbens–olfactory tubercle complex. BrainRes. Rev. 56, 27–78.

Jones, B.E., Bobillier, P., Pin, C., Jouvet, M., 1973. The effect of lesions of catecho-lamine-containing neurons upon monoamine content of the brain and EEG andbehavioral waking in the cat. Brain Res. 58, 157–177.

Kakade, S., Dayan, P., 2002. Dopamine: generalization and bonuses. Neural Netw.15, 549–559.

Kovacs, K.J., 1998. c-Fos as a transcription factor: a stressful (re)view from afunctional map. Neurochem. Int. 33, 287–297.

Kovacs, K.J., 2008. Measurement of immediate-early gene activation – c-fos andbeyond. J. Neuroendocrinol. 20, 665–672.

Le Moal, M., 1995. Mesocorticolimbic dopaminergic neurons: functional and reg-ulatory roles. In: Bloom, F.E., Kupfer, D.J. (Eds.), Psychopharmacology: TheFourth Generation of Progress. Raven Press, New York, pp. 283–294.

Leger, L., Goutagny, R., Sapin, E., Salvert, D., Gervasoni, D., Fort, P., Luppi, P., 2006.Comparison of Fos expression in catecholaminergic neurons after paradoxicalsleep deprivation/recovery and forced wakefulness in the rat. SfN (Abstract).

Leger, L., Goutagny, R., Sapin, E., Salvert, D., Fort, P., Luppi, P., 2007. Comparison ofFos expression in catecholaminergic neurons after paradoxical sleep depriva-tion/recovery and forced wakefulness in the rat. Soc. Neurosci. (Abstract).

Leger, L., Goutagny, R., E. Sapin, D. Salvert, P. Fort, P.-H. Luppi, 2009a. Comparison ofFos expression in catecholaminergic neurons after paradoxical sleep depriva-tion/recovery and novel environment in the rat. In: 50th Anniversary of Para-doxical Sleep Discovery, Basic and Clinical Perspectives, a Joint ESRS/WASMInternational Symposium, vol., ed., Lyon, France.

Leger, L., Goutagny, R., Sapin, E., Salvert, D., Fort, P., Luppi, P.H., 2009b. Noradre-nergic neurons expressing Fos during waking and paradoxical sleep deprivationin the rat. J. Chem. Neuroanat. 37, 149–157.

Lena, I., Parrot, S., Deschaux, O., Muffat-Joly, S., Sauvinet, V., Renaud, B., Suaud-Chagny, M.F., Gottesmann, C., 2005. Variations in extracellular levels of dopa-mine, noradrenaline, glutamate, and aspartate across the sleep–wake cycle inthe medial prefrontal cortex and nucleus accumbens of freely moving rats. J.Neurosci. Res. 81, 891–899.

Lindvall, O., Bjorklund, A., 1983. Dopamine- and norepinephrine-containing neuronsystems: their anatomy in the rat brain. In: Emson, P.C. (Ed.), Chemical Neu-roanatomy. Raven Press, New York, pp. 229–255.

Lookingland, K.J., Moore, K.E., 2005. Functional neuroanatomy of hypothalamicdopaminergic neuroendocrine systems. In: S.B. Dunnett (Eds.), Handbook ofChemical Neuroanatomy. Vol. 21: Dopamine. pp. 435–502.

Lu, J., Jhou, T.C., Saper, C.B., 2006. Identification of wake-active dopaminergicneurons in the ventral periaqueductal gray matter. J. Neurosci. 26, 193–202.

Luckman, S.M., Dyball, R.E., Leng, G., 1994. Induction of c-fos expression inhypothalamic magnocellular neurons requires synaptic activation and notsimply increased spike activity. J. Neurosci. 14, 4825–4830.

Ma, Q.P., Zhou, Y., Han, J.S., 1993. Noxious stimulation accelerated the expression ofc-fos protooncogene in cholecystokininergic and dopaminergic neurons in theventral tegmental area. Peptides 14, 561–566.

Maloney, K.J., Mainville, L., Jones, B.E., 1999. Differential c-Fos expression incholinergic, monoaminergic, and GABAergic cell groups of the pontomesence-phalic tegmentum after paradoxical sleep deprivation and recovery. J. Neurosci.19, 3057–3072.

Maloney, K.J., Mainville, L., Jones, B.E., 2000. c-Fos expression in GABAergic, seroto-nergic, and other neurons of the pontomedullary reticular formation and rapheafter paradoxical sleep deprivation and recovery. J. Neurosci. 20, 4669–4679.

Maloney, K.J., Mainville, L., Jones, B.E., 2002. c-Fos expression in dopaminergic andGABAergic neurons of the ventral mesencephalic tegmentum after paradoxicalsleep deprivation and recovery. Eur. J. Neurosci. 15, 774–778.

Monti, J.M., Monti, D., 2007. The involvement of dopamine in the modulation ofsleep and waking. Sleep Med. Rev. 11, 113–133.

Morgan, J.I., Curran, T., 1991. Stimulus-transcription coupling in the nervoussystem: involvement of the inducible proto-oncogenes fos and jun. Annu.Rev. Neurosci. 14, 421–451.

Paxinos, G., Watson, C., 1997. The Rat Brain in Stereotaxic Coordinates. AcademicPress, Sydney; Orlando.

Poulain, D.A., Wakerley, J.B., 1982. Electrophysiology of hypothalamic magnocel-lular neurones secreting oxytocin and vasopressin. Neuroscience 7, 773–808.

Rechtschaffen, A., Bergmann, B.M., Everson, C.A., Kushida, C.A., Gilliland, M.A., 1989.Sleep deprivation in the rat: X. Integration and discussion of the findings. Sleep12, 68–87.

Redmond, A.J., Morrow, B.A., Elsworth, J.D., Roth, R.H., 2002. Selective activation ofthe A10, but not A9, dopamine neurons in the rat by the predator odor, 2,5-dihydro-2,4,5-trimethylthiazoline. Neurosci. Lett. 328, 209–212.

Reymond, M.J., Arita, J., Dudley, C.A., Moss, R.L., Porter, J.C., 1984. Dopaminergicneurons in the mediobasal hypothalamus of old rats: evidence for decreasedaffinity of tyrosine hydroxylase for substrate and cofactor. Brain Res. 304, 215–223.

Sakai, K., 1991. Physiological properties and afferent connections of the locuscoeruleus and adjacent tegmental neurons involved in the generation of para-doxical sleep in the cat. Prog. Brain Res. 88, 31–45.

Sapin, E., Lapray, D., Berod, A., Goutagny, R., Leger, L., Ravassard, P., Clement, O.,Hanriot, L., Fort, P., Luppi, P.H., 2009. Localization of the brainstem GABAergicneurons controlling paradoxical (REM) sleep. PLoS ONE 4, e4272.

Schmidt, M.H., Valatx, J.L., Sakai, K., Fort, P., Jouvet, M., 2000. Role of the lateralpreoptic area in sleep-related erectile mechanisms and sleep generation in therat. J. Neurosci. 20, 6640–6647.

Senba, E., Ueyama, T., 1997. Stress-induced expression of immediate early genes inthe brain and peripheral organs of the rat. Neurosci Res. 29, 183–207.

Sgambato, V., Abo, V., Rogard, M., Besson, M.J., Deniau, J.M., 1997. Effect of electricalstimulation of the cerebral cortex on the expression of the Fos protein in thebasal ganglia. Neuroscience 81, 93–112.

Shepard, P.D., Mihailoff, G.A., German, D.C., 1988. Anatomical and electrophysio-logical characterization of presumed dopamine-containing neurons within thesupramammillary region of the rat. Brain Res. Bull. 20, 307–314.

Shiromani, P.J., Malik, M., Winston, S., McCarley, R.W., 1995. Time course of Fos-likeimmunoreactivity associated with cholinergically induced REM sleep. J. Neu-rosci. 15, 3500–3508.

Sita, L.V., Elias, C.F., Bittencourt, J.C., 2003. Dopamine and melanin-concentratinghormone neurons are distinct populations in the rat rostromedial zona incerta.Brain Res. 970, 232–237.

Skagerberg, G., Lindvall, O., 1985. Organization of diencephalic dopamine neuronesprojecting to the spinal cord in the rat. Brain Res. 342, 340–351.

Solms, M., 2000. Dreaming and REM sleep are controlled by different brain mechan-isms. Behav Brain Sci. 23, 843–850.

Stratford, T.R., Wirtshafter, D., 1990. Ascending dopaminergic projections from thedorsal raphe nucleus in the rat. Brain Res. 511, 173–176.

Trulson, M.E., Preussler, D.W., 1984. Dopamine-containing ventral tegmental areaneurons in freely moving cats: activity during the sleep–waking cycle andeffects of stress. Exp. Neurol. 83, 367–377.

Trulson, M.E., Preussler, D.W., Howell, G.A., 1981. Activity of substantia nigra unitsacross the sleep-waking cycle in freely moving cats. Neurosci. Lett. 26, 183–188.

van den Pol, A.N., Herbst, R.S., Powell, J.F., 1984. Tyrosine hydroxylase-immunor-eactive neurons of the hypothalamus: a light and electron microscopic study.Neuroscience 13, 1117–1156.

van Vulpen, E.H., Yang, C.R., Nissen, R., Renaud, L.P., 1999. Hypothalamic A14 andA15 catecholamine cells provide the dopaminergic innervation to the supraop-tic nucleus in rat: a combined retrograde tracer and immunohistochemicalstudy. Neuroscience 93, 675–680.

Verret, L., Goutagny, R., Fort, P., Cagnon, L., Salvert, D., Leger, L., Boissard, R., Salin,P., Peyron, C., Luppi, P.H., 2003. A role of melanin-concentrating hormoneproducing neurons in the central regulation of paradoxical sleep. BMC Neu-rosci. 4, 19.

Verret, L., Leger, L., Fort, P., Luppi, P.H., 2005. Cholinergic and noncholinergicbrainstem neurons expressing Fos after paradoxical (REM) sleep deprivationand recovery. Eur. J. Neurosci. 21, 2488–2504.

Verret, L., Fort, P., Gervasoni, D., Leger, L., Luppi, P.H., 2006. Localization of theneurons active during paradoxical (REM) sleep and projecting to the locuscoeruleus noradrenergic neurons in the rat. J. Comp. Neurol. 495, 573–586.

Wagner, C.K., Eaton, M.J., Moore, K.E., Lookingland, K.J., 1995. Efferent projectionsfrom the region of the medial zona incerta containing A13 dopaminergicneurons: a PHA-L anterograde tract-tracing study in the rat. Brain Res. 677,229–237.