Uncoupling protein 2/3 immunoreactivity and the ascending dopaminergic and noradrenergic neuronal systems: relevance for volume transmission.

Abstract

Uncoupling proteins in the inner mitochondrial membrane uncouples oxidative phosphorylation from ATP synthesis. It has been suggested that these proteins are involved in thermogenesis as well as in the regulation of reactive oxygen species production in the mitochondria. The present work was conducted to investigate the localization of the uncoupling protein 2-like immunoreactivity (uncoupling protein 2/3 immunoreactivity) in the main catecholaminergic projection fields in the rat brain as well as in the areas of the dopaminergic and noradrenergic nerve cell groups. In particular, the relationships of tyrosine hydroxylase, dopamine beta-hydroxylase and uncoupling protein 2/3 immunoreactivity were assessed by double immunolabeling and confocal laser microscopy analysis associated with computer-assisted image analysis. Uncoupling protein 2/3 immunoreactivity was observed in discrete dopaminergic terminals in the nucleus accumbens and in the cerebral cortex whereas it was found in scattered noradrenergic terminals in the caudate putamen and Islands of Calleja Magna. One interesting finding was that uncoupling protein 2/3 immunoreactivity together with tyrosine hydroxylase immunoreactivity in the shell of nucleus accumbens was observed surrounding the previously characterized D1 receptor rich nerve cell column system characterized by a relative lack of tyrosine hydroxylase immunoreactivity. Moreover, in animal models of dopaminergic pathway degeneration, plastic changes in uncoupling protein 2/3 terminals have been shown in the cerebral cortex and striatum as seen from the increased size and intensity of uncoupling protein 2/3 immunoreactivity of their varicosities. Taken together, these findings open up the possibility that uncoupling protein 2/3 could play an important role modulating the dopaminergic and noradrenergic neurotransmission within discrete brain regions.

Full text

Page 1

UNCOUPLING PROTEIN 2/3 IMMUNOREACTIVITY AND THE
ASCENDING DOPAMINERGIC AND NORADRENERGIC NEURONAL
SYSTEMS: RELEVANCE FOR VOLUME TRANSMISSION
A. RIVERA,a* L.F. AGNATI,cT.L. HORVATH,d
J.J. VALDERRAMA,aA. DE LA CALLEaAND K. FUXEb
aDepartment of Cell Biology, School of Science, University of Málaga,
Campus de Teatinos s/n, 29071 Málaga, Spain
bDepartment of Neuroscience, Division of Cellular and Molecular Neu-
rochemistry, Karolinska Institute, 171 77 Stockholm, Sweden
cDepartment of Biomedical Science, Section of Physiology, University
of Modena, 41 100 Modena, Italy
dDepartment of Obstetrics and Gynecology, Yale Universty School of
Medicine, New Haven, CT 06520, USA
Abstract—Uncoupling proteins in the inner mitochondrial
membrane uncouples oxidative phosphorylation from ATP syn-
thesis. It has been suggested that these proteins are involved in
thermogenesis as well as in the regulation of reactive oxygen
species production in the mitochondria. The present work was
conducted to investigate the localization of the uncoupling pro-
tein 2-like immunoreactivity (uncoupling protein 2/3 immunore-
activity) in the main catecholaminergic projection fields in the
rat brain as well as in the areas of the dopaminergic and nor-
adrenergic nerve cell groups. In particular, the relationships of
tyrosine hydroxylase, dopamine ?-hydroxylase and uncoupling
protein 2/3 immunoreactivity were assessed by double immu-
nolabeling and confocal laser microscopy analysis associated
with computer-assisted image analysis. Uncoupling protein 2/3
immunoreactivity was observed in discrete dopaminergic ter-
minals in the nucleus accumbens and in the cerebral cortex
whereas it was found in scattered noradrenergic terminals in
the caudate putamen and Islands of Calleja Magna. One inter-
estingfindingwasthatuncouplingprotein2/3immunoreactivity
together with tyrosine hydroxylase immunoreactivity in the
shell of nucleus accumbens was observed surrounding the
previously characterized D1receptor rich nerve cell column
system characterized by a relative lack of tyrosine hydroxylase
immunoreactivity. Moreover, in animal models of dopaminergic
pathway degeneration, plastic changes in uncoupling protein
2/3 terminals have been shown in the cerebral cortex and stri-
atum as seen from the increased size and intensity of uncou-
pling protein 2/3 immunoreactivity of their varicosities. Taken
together, these findings open up the possibility that uncoupling
protein 2/3 could play an important role modulating the dopa-
minergic and noradrenergic neurotransmission within discrete
brain regions. © 2005 Published by Elsevier Ltd on behalf of
IBRO.
Key words: uncoupling protein, intercellular communication,
catecholamines, 6-OHDA, rat brain, UCP2/3 overexpressing
mice.
In the inner membrane of the mitochondria, the uncoupling
proteins (UCPs) mediate the passive transport of hydrogen
protons from the intermembrane space to the matrix com-
partment. Through this process, UCPs dissipate the proton
gradient across the inner membrane diminishing the ATP
synthesis and generating an important amount of heat
(Bouillaud et al., 1985; Horvath et al., 2003a; Nicholls and
Locke, 1984; Palou et al., 1998; Richard et al., 2001).
UCPs have also been demonstrated to reduce the produc-
tion of superoxides and increase the calcium efflux from
the mitochondria with the reduction of the mitochondrial
membrane potential (Arsenijevic et al., 2000; Horvath et
al., 2003a; Negre-Salvayre et al., 1997).
UCP proteins are expressed in a variety of tissues,
including the brown adipose tissue where an important role
in the processes of thermogenesis has been assigned to
them (Bouillaud et al., 1985; Nicholls and Locke, 1984).
The proteins UCP2 and UCP3 are members of this protein
family and at least UCP2 is expressed in the CNS (Diano
et al., 2000; Fleury et al., 1997; Horvath et al., 1999;
Lengacher et al., 2004; Richard et al., 1998). The role
played by UCP2 in the brain has yet to be fully described,
but it has been suggested that UCP2 could contribute to
the control of neurotransmission (Horvath et al., 1999,
2003a) and exert neuroprotective actions (Bechmann et
al., 2002; Diano et al., 2003; Horvath et al., 2003a,b;
Paradis et al., 2003).
It is now generally accepted that brain communication
operates not only by synaptic transmission but also by
volume transmission involving diffusion and convection of
the transmitters in the extracellular fluid (ECF) and in the
cerebrospinal fluid (CSF) (Agnati et al., 1994, 2000, 2005;
Fuxe and Agnati, 1991). The present study was initiated to
test our hypothesis that UCP2/3 may be involved in volume
transmission in the brain. In view of the previous work
indicating that the ascending dopaminergic (DAergic) and
noradrenergic (NAergic) systems operate mainly by vol-
ume transmission (Jansson et al., 2002), an immunohisto-
chemical analysis was performed to characterize the rela-
tionship of the UCP2/3 and tyrosine hydroxylase (TH) and
dopamine ?-hydroxylase (DBH) immunoreactivity (IR) in
cell bodies, dendrites and nerve terminals using double
immunolabeling procedures in combination with confocal
laser microscopy. Of special interest was the nucleus ac-
*Corresponding author. Tel: ?34-952131935; fax: ?34-952132000.
E-mail address: arivera@uma.es (A. Rivera).
Abbreviations: CA, catecholamine; CAergic, catecholaminergic; CSF,
cerebrospinal fluid; DA, dopamine; DAergic, dopaminergic; DBH, dopa-
mine ?-hydroxylase; ECF, extracellular fluid; FA, field area; IR, immuno-
reactivity; LC, locus coeruleus; MFB, medial forebrain bundle; NA, nor-
adrenaline; NAergic, noradrenergic; OD, optical density; PB, phosphate
buffer; PBS, phosphate-buffered saline; PBS-TX, phosphate-buffered sa-
line containing 0.2% Triton X-100; SN, substantia nigra; SNc, substantia
nigra pars compacta; SNr, substantia nigra pars reticulata; TH, tyrosine
hydroxylase; UCP, uncoupling protein; VTA, ventral tegmental area;
6-OHDA, 6-hydroxydopamine.
Neuroscience 137 (2006) 1447–1461
0306-4522/06$30.00?0.00 © 2005 Published by Elsevier Ltd on behalf of IBRO.
doi:10.1016/j.neuroscience.2005.05.051
1447

Page 2

cumbens where clear dopamine (DA) terminal/DA receptor
mismatches have been observed (Jansson et al., 1999).
We also explored the functional role of UCP2/3 by using a
rat model of Parkinson’s disease with 6-hydroxydopamine
(6-OHDA) induced degeneration of the ascending DA neu-
rons on one side, where compensatory mechanisms have
previously been demonstrated increasing DA release and
promoting volume transmission (Zoli et al., 1998, 1999).
Finally, we used UCP2/3 overexpressing mice and their
littermates to test the specificity of the UCP2/3 antibody
used in the present work.
EXPERIMENTAL PROCEDURES
Animals
Male Sprague–Dawley rats weighing 250 g (n?10) and male
UCP2/3 overexpressing mice (n?4) and their wild-type littermates
(n?4) (Fuller et al., 2000) were used. The animals were kept
under a standard 12-h light/dark cycle and constant room temper-
ature (23 °C), and had free access to tap water and food pellets.
Animal care and use followed the directives of the Council of
European Communities (86/609/EEC). All the experiments were
approved by the appropriate animal care committee at the Uni-
versity of Málaga, and all efforts were made to minimize the
number of animals used and their suffering.
Lesion of the ascending DA pathways
Ascending DA axon bundles to the forebrain were lesioned with
local administration of the catecholamine (CA) neurotoxin
6-OHDA into the medial forebrain bundle (MFB). The rats (n?4)
were pretreated 30 min prior to surgery with desipramine
(25 mg/kg i.p. in saline solution) to protect NA projections (Breese
and Traylor, 1970; Jacks et al., 1972) and then anesthetized with
a mixture of ketamine (75 mg/kg i.p.) and medetomidine
(0.5 mg/kg i.p.). Rats were placed in a stereotaxic frame and
unilateral injection of 6-OHDA (14 ?g in 4 ?l of saline containing
0.02% of ascorbic acid) was infused for 2 min in the MFB using a
26-gauge cannula at the coordinates AP ?3.7 mm, L ?1.4 mm
and V ?8.6 mm from Bregma (Paxinos and Watson, 1986). Two
weeks after the surgery, the animals were perfused and the brains
processed for immunohistochemistry (as described below).
Tissue processing
The animals were deeply anesthetized with sodium pentobarbital
(60 mg/kg i.p.) and perfused transcardially with 0.1 M phosphate-
buffered saline, pH 7.4 (PBS) followed by 4% paraformaldehyde
(w/v) in 0.1 M phosphate buffer, pH 7.4 (PB). The brains were
dissected out, post-fixed in the same fixative for 2 h, immersed in
30% sucrose in PBS for at least 48 h and frozen in dry ice. Coronal
free-floating sections (20 ?m thick) were obtained and stored in PBS
with 0.02% sodium azide at 4 °C until use for immunohistochemistry.
Antibodies
Affinity-purified rabbit polyclonal UCP2 antiserum developed by
Horvath et al. (1999) was used at a 1:500 (in the double immu-
nofluorescence stainings) or 1:1000 dilution (in the avidin–biotin
immunohistochemistry experiments). The UCP2 antibody has
been previously characterized and used to visualize the UCP2
protein distribution in the brain especially the hypothalamus and
entorhinal cortex (Bechmann et al., 2002; Diano et al., 2000;
Horvath et al., 1999). This antiserum also recognizes UCP3, an
UCP that we recently observed to be present in certain brain areas
(Horvath et al., unpublished observations). Thus, we refer to
UCP2/3 IR throughout the manuscript. TH was detected using a
mouse monoclonal antibody (DiaSorin, SpA, Saluggia, Italy) di-
luted at 1:1000 or 1:10,000. DBH was detected with a mouse
monoclonal antibody (Chemicon International Inc., Temecula, CA,
USA) diluted at 1:1000 or 1:10,000. High dilutions (1:10,000) were
used for TH and DBH antibodies at the substantia nigra (SN) and
the locus coeruleus (LC) level due to the high expression of TH
and DBH in these areas. Low dilutions (1:1000) of TH and DBH
antibodies were used to analyze the other brain regions, where a
higher sensitivity is required to label varicose terminal-like fibers.
All antibodies were diluted in phosphate-buffered saline contain-
ing 0.2% Triton X-100 (PBS-TX) and 0.1% sodium azide.
Immunohistochemistry
Coronal sections from different bregma levels according to the
atlas of Paxinos and Watson (1986) were processed for either
single or double immunolabeling procedures. The bregma levels
used were: ?1.70 mm to ?0.70 mm (striatum, nucleus accum-
bens, Islands of Calleja, olfactory tubercle and cingulate cortex);
?4.8 mm to ?5.8 mm (SN and ventral tegmental area, VTA); and
?9.6 mm to ?10.0 mm (LC).
Single antigen immunohistochemistry was performed to com-
pare UCP2/3 IR on the unoperated vs the DA denervated side in
rat and UCP2/3 IR in the wild-type vs UCP2/3 overexpressing
mice. Endogenous peroxidase activity was quenched by incuba-
tion for 10 min with 3% hydrogen peroxide in PBS. After washing
with PBS, sections were incubated for 48 h at 4 °C with UCP2/3
antibody. After washing with PBS sections were incubated for 1 h
in biotin-conjugated goat anti-rabbit IgG diluted 1:500 in PBS-TX
(Vector Laboratories, Burlingame, CA, USA). The sections were
washed again and incubated for 1 h in a streptavidin–peroxidase
complex (Sigma, St. Louis, MO, USA) diluted 1:2000 in PBS-TX.
Peroxidase activity was visualized with 0.05% 3,3=-diaminobenzi-
dine (DAB, Sigma) and 0.002% H2O2, and staining was intensified
with 0.8% nickel ammonium sulfate. The sections were mounted
on gelatin-coated slides, air dried, dehydrated in xylene and cov-
erslipped with DPX mounting medium.
Double immunohistochemistry for UCP2/3 and TH or DBH IR
was performed in rat sections using green and red fluorescent
staining, respectively. Non-specific binding sites were blocked for
30 min with 5% bovine serum albumin in PBS-TX. Sections were
washed in PBS and incubated for 48 h at 4 °C with the UCP2/3
antibody followed by incubation for 24 h at 4 °C with TH or DBH
antibodies. After washing in PBS, the sections were incubated for
1 h at room temperature with goat biotinylated anti-rabbit IgG
(1:500, Vector Laboratories), washed again and incubated in a
mixture of streptavidin conjugated with Alexa 488 (1:2000, Molec-
ular Probes, Carlsbad, CA, USA) and goat anti-mouse IgG con-
jugated with Cy3 (1:200, Jackson Immunoresearch, West Grove,
PA, USA). Finally, the sections were rinsed in PBS and mounted
on gelatin-coated slides with an antifading mounting medium
(Dako Corporation, Carpinteria, USA).
Image analysis
Quantitative analysis of both size and intensity of UCP2/3 IR
varicose terminal-like fibers was performed using the image ana-
lyzing system NIH Image (http://rsb.info.nih.gov/nih-image/). Pho-
tographs of UCP2/3 IR varicose terminal-like fibers were taken
with a digital camera (Coolpix 4500, Nikon, Tokyo, Japan) using a
?100 objective. Measurements of size and UCP2/3 IR intensity of
varicosities were only performed from those that were in focus.
The size of each varicosity was determined of the average value
of four measures of diameter according to different sampling of the
same varicosity. The intensity of UCP2/3 IR was expressed as the
mean optical density (OD) and the value was corrected with the
OD from an immunonegative area. Ten fibers (with three to eight
varicosities) were measured from each region on the unoperated
A. Rivera et al. / Neuroscience 137 (2006) 1447–14611448

Page 3

and 6-OHDA-lesioned side as well as in the wild-type and UCP2/3
overexpressing mice. The statistical analysis was performed by
Mann Whitney U test.
Specific immunofluorescence signals were visualized by excita-
tion at blue (488 nm) for Alexa 488 (absorption: 490; emission: 519)
and green (545 nm) for Cy3 (absorption: 550; emission: 580) using a
confocal laser microscope (Leica TCS-NT, Wetzlar, Germany; laser
ArKr) or a Nikon Microphot-Fx (Japan; Hg-lamp). In the confocal
microscope, the emission signal was separated through a short-pass
filter (RSP580), and the short-wavelength signal corresponding to
Alexa 488 was band-pass filtered (BP530/30) and collected in the
green channel. The long-wavelength signal was long-pass filtered
Fig. 1. Fluorescence photomicrographs from the rat shell part of the nucleus accumbens showing UCP2/3 IR (A, C) and TH IR (B, D) in the same
section. Arrowheads indicate patches displaying low UCP2/3 IR (A, C) where also weak TH IR is shown (B, D). The major part of shell of nucleus
accumbens shows strong UCP2/3 IR matched by strong TH IR. Asterisks in C–D show an example of a strongly UCP2/3-immunolabeled patch with
weak TH IR. Crossed arrows indicate dorsal and lateral direction in the four panels. AcbSh, shell part of the nucleus accumbens; ICjM, Islands of
Calleja Magna. Scale bar?100 ?m in A and B and in C and D 200 ?m.
A. Rivera et al. / Neuroscience 137 (2006) 1447–14611449

Page 4

(LP590) and collected in the red channel. The scanning was made
sequentially to avoid crosstalk. In the figures, UCP2/3 IR is shown in
green and TH and DBH IR in red.
The analysis of the overlap between UCP2/3 and TH IR was
made using the image analyzer KS 400 (Zeiss Kontron, Zeiss,
Arese, Italy). Area measurements were obtained from the shell
nucleus accumbens from four different rats and expressed as
means?SEM. Field areas (FA) in pixels were determined for TH-
and UCP2/3 immunolabelings and were evaluated at a low and
high threshold of discrimination. The two thresholds were selected
by measuring with an automatic interactive procedure the mean
gray values and the standard deviations within the area of interest
(X, S.D.) and then using as a “low threshold of discrimination” the
value equal to X and as “high threshold of discrimination” the value
equal to X?2SD.
The Boolean operator “AND” was applied to evaluate the
overlap between TH IR and UCP2/3 IR at the two different thresh-
olds of discrimination. The FA of these overlaps was determined
and expressed in percent of the corresponding TH IR FA.
RESULTS
UCP2/3 IR in catecholaminergic (CAergic) projection
fields in the brain
In line with previous results (Diano et al., 2000; Horvath et
al., 1999) UCP2/3 IR terminal-like fibers were observed in
many brain regions, including CAergic projection fields.
Intensely stained UCP2/3 IR terminal-like fibers were
found in the ventral striatum (nucleus accumbens and
olfactory tubercle) as well as in the lateral septum and the
diagonal band of Broca. Less UCP2/3 IR was observed in
the dorsal striatum and in the cerebral cortex. A detailed
description on the UCP2/3 IR in the main CAergic projec-
tion fields and its relationship to the TH and DBH IR nerve
terminal plexa will be given.
Nucleus accumbens
In the nucleus accumbens prominent UCP2/3 IR terminals
were found in the shell and core part but with different distri-
bution patterns. The shell nucleus accumbens showed a
heterogeneous pattern with both strongly and weakly
UCP2/3 IR patches (Fig. 1A). This pattern was found to be in
good register with the TH IR patches identified in the same
section (Fig. 1A, B). Nevertheless, certain strong UCP2/3
IR patches were not associated with high levels of TH IR
but with low levels (Fig. 1C, D). Based on image analysis,
the overlap between UCP2/3 IR and TH IR was about 78%
and 96% at the lower and higher threshold of discrimina-
tion, respectively when evaluating the UCP2/3 IR and TH
IR patches with strong IR (Table 1).
Scattered UCP2/3 IR varicose terminal-like fibers dis-
playing strong intensity of immunostaining were observed
in both strongly and weakly UCP2/3 IR patches in the shell
nucleus accumbens. A detailed analysis using laser con-
focal microscopy showed that these UCP2/3 IR varicose
terminals like fibers were also TH IR (data not shown) and
DBH IR (Fig. 2). However, the vast majority of TH- and DBH
IR terminal-like fibers did not express UCP2/3 IR (Fig. 2).
As seen in Fig. 3A, the core part of the nucleus ac-
cumbens showed a prominent UCP2/3 IR patch, presum-
ably built up of a nerve terminal plexus surrounding the
anterior commissure, where a large number of TH IR ter-
minal-like fibers were also found (Fig. 3B). Analysis with
confocal laser microscopy demonstrated a few intensely
UCP2/3 IR varicose terminal-like fibers in close proximity
Fig. 2. Confocal photomicrographs from the rat shell part of the nu-
cleus accumbens showing double immunolabeling with anti-UCP2/3
(green) and anti-DBH (red) antibodies. One intense UCP2/3-immuno-
labeled terminal-like fiber is illustrated co-expressing DBH IR resulting
in yellowish color (arrowhead) in the weak UCP2/3 patch (asterisk) (A).
The majority of the DBH axonal processes did not show UCP2/3 IR
(arrows). The panel B shows a high magnification of a UCP2/3-DBH
double-immunolabeled terminal-like fiber (yellowish, arrowhead) and
DBH alone IR-labeled terminal-like fibers (arrows). Crossed arrows
indicate dorsal and lateral direction in all the panels. Scale bar?20 ?m
in A and in B.
Table 1. Overlap between TH- and UCP2/3 IR terminal-like fibers in
the shell part of the nucleus accumbens
FA of IR (pixels)
Low thresholdHigh threshold
TH
UCP2/3
TH?UCP2/3
TH?UCP2/3 overlap (%)
100?4
102?7
78?3
78?11
24?6
28?2
23?8
96?33
The FA (pixels) values represent the measurement of area of TH-
and UCP2/3 IR at two threshold levels (low and high). The FA is
expressed as means?SEM (n?4). The TH?UCP2/3 field areas over-
laps were also evaluated at the two different thresholds of discrimina-
tion and expressed as percentages of the respective FA of the TH IR
staining.
A. Rivera et al. / Neuroscience 137 (2006) 1447–1461 1450

Page 5

to the anterior commissure (Fig. 3C, E). These UCP2/3 IR
profiles were found to be immunoreactive also for TH (Fig.
3C, D) but seemed to lack DBH IR (Fig. 3E, F).
Olfactory tubercle and Islands of Calleja
A low to high density of UCP2/3 IR terminal-like fibers was
found in the olfactory tubercle (Fig. 4A). High density re-
gions in the inner layer were characterized by a densely
punctate of UCP2/3 immunofluorescence overlapping with
TH IR terminal-like fibers also of a high density. This
overlap did not exist in the outer layer. It was not possible
to determine if UCP2/3 IR and TH IR were co-located in the
same terminals. On the other hand, few scattered UCP2/3
IR varicose terminal-like fibers in the outer layer were
Fig. 3. Confocal photomicrographs from the rat core part of the nucleus accumbens showing double immunolabeling using UCP2/3 and TH or
DBH antibodies. A strongly UCP2/3 IR area surrounds the anterior commissure displaying intense TH IR (arrows in A–B). In high magnification
the confocal laser microscopy analysis demonstrated intense large UCP2/3 IR varicose terminal-like fibers displaying also TH IR (arrowheads
in C–D), but not DBH IR (E, F). Weakly stained TH IR fibers did not display UCP2/3 IR (arrows in C–D). Crossed arrows indicate dorsal and
lateral direction in all the panels. aca, anterior commissure; AcbC, core part of the nucleus accumbens. Scale bar?200 ?m in A and B and in
C, D, E and F 25 ?m.
A. Rivera et al. / Neuroscience 137 (2006) 1447–14611451

Page 6

found to contain TH IR (Fig. 4C, D) as shown with confocal
laser microscopy.
The core of the Islands of Calleja displayed a moderate
degree of UCP2/3 IR punctate while the outer part lacked
UCP2/3 IR (Fig. 4A). Large number of TH IR terminal-like
fibers surrounded these islands but only few were found
inside (Fig. 4B). In the Islands of Calleja Magna, as indi-
cated in Fig. 5A and B, low to moderate numbers of TH and
DBH IR varicose terminal-like fibers were observed. In
contrast only few UCP2/3 IR varicose terminal-like fibers
were found in the Islands of Calleja Magna. Double immuno-
labeling experiments and detailed analysis of optical sections
using confocal laser microscopy, revealed that TH- but not
DBH IR co-localize with UCP2/3 IR in their varicosities (insets
in Fig. 5A and B).
Caudate putamen
Scattered strongly UCP2/3 IR cell bodies and varicose
terminal-like fibers appeared in the caudate putamen
(Fig. 6A, B). They did not seem to have a selective
localization within the different compartments (strio-
somes vs matrix) and topographic regions in the cau-
date putamen. As seen in Fig. 6C and D, the UCP2/3 IR
varicose terminal-like fibers were also immunoreactive
for TH and for DBH.
Fig. 4. Confocal photomicrographs from the rat olfactory tubercle with the Islands of Calleja showing double immunolabeling using UCP2/3 and TH
antibodies. UCP2/3 IR is shown in low magnification in A where this IR is mainly found in the inner layers of the olfactory tubercle (arrowheads) and
within the Island of Calleja (arrowheads). In contrast, strong TH IR is also found in the outer layers of the olfactory tubercle but almost absent in the
Island of Calleja (arrowheads) (B). Few scattered UCP2/3 IR varicose terminal-like fibers were found in the outer layer of the olfactory tubercle (arrow
in A). In high magnification it is possible to see single UCP2/3 IR varicose terminal-like fibers co-expressing TH IR (arrows in C–D). Crossed arrows
indicate dorsal and medial direction in all the panels. ICj, Islands of Calleja; Tu, olfactory tubercle. Scale bar?100 ?m in A and B and in C and D 20 ?m.
A. Rivera et al. / Neuroscience 137 (2006) 1447–14611452

Page 7

Cerebral cortex
In cortical areas, such as cingulate, frontal, parietal, ento-
rhinal and piriform cortex, scattered UCP2/3 IR cell bodies
and varicose terminal-like fibers were observed (Fig. 7A).
Most of them were located in the supragranular layers (I, II
and III), although they could also be found in the infra-
granular layers. The UCP2/3 IR nerve cells displayed mor-
phological features of cortical interneurons. TH IR could be
demonstrated in all of the UCP2/3 IR varicose terminal-like
fibers analyzed characterized by large varicosities and
intense IR (Fig. 7B, C). Frequently fine TH IR terminal-like
fibers with small varicosities and with less intensity of
staining were observed to lack UCP2/3 IR. In optical sec-
tions using laser confocal microscopy, the relationship be-
tween DBH IR and UCP2/3 IR was further studied. As seen
in Fig. 7D and E, UCP2/3 varicose terminal-like fibers were
found to lack DBH IR but with DBH IR terminal-like fibers
present in the surround.
UCP2/3 expression in DAergic and NAergic nerve
cell groups
The existence of UCP2/3 IR was analyzed in the DAergic
nerve cell bodies located in the SN and the VTA and in the
NAergic nerve cell bodies in the LC.
SN
In the substantia nigra pars compacta (SNc) and VTA
prominent UCP2/3 IR was demonstrated in both cell bod-
ies and dendritic branches of DAergic neurons immuno-
stained in the same section with anti-TH antibodies (Fig.
8). Within the substantia nigra pars reticulata (SNr) abun-
dant UCP2/3 IR processes and cell bodies were also
present (Fig. 8A, B, D). Double immunolabeling for TH and
UCP2/3 IR revealed that TH IR cell bodies and dendrites
also contain UCP2/3 IR. However, a large number of
Fig. 5. Confocal photomicrographs from the rat Island of Calleja Magna showing double immunolabeling with anti-UCP2/3 (green) and anti-TH (red)
(A) or anti-DBH (red) (B) antibodies. A low to moderate density of TH and DBH IR varicose terminal-like fibers was found in these islands as well as
a few of UCP2/3 IR alone varicose terminal-like fibers (arrows in A and B). Insets in A and B show a high magnification of a UCP2/3 IR varicose
terminal-like fiber co-expressing TH but not DBH IR. AcbSh, shell part of the nucleus accumbens; ICjM, Islands of Calleja Magna. Scale bar?40 ?m
in A and B and in the insets 5 ?m.
Fig. 6. Fluorescence photomicrographs of rat coronal sections of the
caudate putamen illustrating the localization of UCP2/3 IR terminal-like
fibers(arrow)withstronglylabeledmoderatetolargevaricositiesandtheir
relationship with TH (red) or DBH (red) IR varicose terminal-like fibers
(A–D). In A it was also possible to see UCP2/3 IR in scattered cell bodies
(arrowheads). The high magnification in panel B shows strongly UCP2/3
IR varicose terminal-like fibers where the varicosities appear to have
different sizes. Nevertheless this may be due to the fact that the varicos-
ities are in different optical planes. Panels C and D show UCP2/3-TH and
UCP2/3-DBH double-immunolabeled terminal-like fibers with strongly co-
labeled varicosities (yellow), respectively. Crossed arrows indicate dorsal
and medial direction of all the pictures. CPu, caudate putamen. Scale
bar?100 ?m in A; in B 50 ?m; and in C and D 20 ?m.
A. Rivera et al. / Neuroscience 137 (2006) 1447–1461 1453

Page 8

perikarya and dendritic profiles in SNr were found to ex-
press exclusively UCP2/3 IR (Fig. 8D).
LC
In line with the previous results of Horvath et al. (1999),
a moderate degree of UCP2/3 IR was found in the LC.
The UCP2/3 IR was either found in the cell bodies and
their dendritic processes or in terminal-like fibers (Fig.
9A). UCP2/3 IR appeared as punctate in the cytoplasm
of NAergic cell bodies, identified in the same section
with anti-DBH antibodies (Fig. 9). This analysis was
performed by confocal laser microscopy that allows
through optical sectioning the localization of the punc-
tate UCP2/3 IR to the cytoplasm of the cell bodies. Most
of the DBH IR nerve cells displayed high degrees of
UCP2/3 IR, although moderate–low degrees were also
observed in some cells. UCP2/3 IR was also demon-
strated in varicose terminal-like fibers in close proximity
to the NAergic nerve cells (Fig. 9).
Effects of 6-OHDA induced lesions of the ascending
DA pathways to the forebrain on UCP2/3 expression
Remaining striatal DA nerve terminals are known to be-
come hypertrophic after 6-OHDA lesions, probably as a
result of a compensatory activation of surviving DAergic
neurons (Zoli et al., 1998, 1999). To test whether UCP2/3
expression is affected by the 6-OHDA-induced lesion of
the ascending DA pathways we analyzed the size of
UCP2/3 IR varicosities as well as the intensity of UCP2/3
Fig. 7. Confocal microphotograph from the rat frontal cortex showing double immunolabeling using UCP2/3, TH and DBH antibodies. In panel A
scattered strongly UCP2/3 IR varicose terminal-like fibers (arrows) and one cell body (arrowhead) were found in the supragranular layers. In panels
B and C large varicose terminal-like fibers were found to contain both UCP2/3 and TH IR (arrowhead) while the fine varicose TH IR terminal-like fibers
lacked UCP2/3 IR (arrows). In panels D and E the UCP2/3 IR varicose terminal-like fibers are shown to lack DBH IR. Scale bar?100 ?m in A and
in B–D 40 ?m.
A. Rivera et al. / Neuroscience 137 (2006) 1447–14611454

Page 9

IR in both cerebral cortex and caudate putamen from the
unoperated and 6-OHDA-lesioned side.
The analysis showed a marked increase in the size
of UCP2/3 IR varicose terminal-like fibers as well as in
their intensity of UCP2/3 IR on the lesioned side as
compared with the unoperated side (Fig. 10A, B). The
diameter of the UCP2/3 IR varicosities increased by
95% in the cerebral cortex and by 97% in the striatum
(Fig. 10C). The OD as a measure of intensity of UCP2/3
IR increased by 37% and 16% in the cerebral cortex and
striatum, respectively (Fig. 10D). A detailed analysis of
the percentage of varicosities belonging to different cat-
egories of size (Fig. 10E, F) and OD (Fig. 10G, H) in
both the cerebral cortex and the caudate putamen
showed a displacement toward larger categories in
those from the DA-lesioned side. It should be noted that
the scattered UCP2/3 IR terminal-like fibers remain with
a similar distribution pattern after the 6-OHDA-induced
lesion. No differences in the OD were observed in the
hypothalamus (Fig. 10D).
UCP2/3 expression in UCP2/3 overexpressing mice
UCP2/3 IR was analyzed in UCP2/3 overexpressing mice.
No difference was observed in the distribution pattern of
UCP2/3 IR in mice compared with that of rat. The areas
selected were the cerebral cortex, the caudate putamen
and the hypothalamus, where UCP2/3 IR terminal-like fi-
bers were observed.
In the UCP2/3 overexpressing mice significant in-
creases of the intensity of UCP2/3 IR were observed in
terminals of the cerebral cortex (by 44%), the caudate
putamen (by 23%) and the hypothalamus (by 28%) (Fig.
11A). UCP2/3 IR varicose terminal-like fibers of cerebral
Fig. 8. Confocal photomicrographs from the rat SN showing double immunolabeling using the UCP2/3 and TH antibodies (A–D). Strong UCP2/3
IR was found in many TH IR nerve cell bodies especially in the pars compacta but also in the pars reticulata (arrow in A). It was possible to
demonstrate UCP2/3 IR in the TH IR dendritic like processes (open arrow in D). Many UCP2/3 IR cell bodies in the pars reticulata did not show
TH IR (arrowheads in A and D). Confocal photomicrographs from the VTA showing double immunolabeling using UCP2/3 and TH antibodies
(yellow) (E). Strong UCP2/3 IR was found in the majority of the TH IR cell bodies. Scale bar?100 ?m in A, in B–C is 200 ?m, D 200 ?m and
in E 150 ?m.
A. Rivera et al. / Neuroscience 137 (2006) 1447–14611455

Page 10

cortex and caudate putamen also showed an increase (by
21% and 23%, respectively) in their size as compared with
wild-type (Fig. 11B).
DISCUSSION
Strikingly almost 100% overlap exists between UCP2
mRNA expression and UCP2 IR in the brain of both rats
(Horvath et al., 1999) and non-human primates (Diano et
al., 2000). Additionally, the use of commercially available
antisera, which in Western blot analysis of tissues from
UCP2 knock mice to demonstrated the lack of the UCP2
protein (Horvath et al., 2002), resulted in immunostaining
that overlaps UCP2 mRNA expression in mice (Diano et
al., 2003). The same was true for immunolabeling obtained
by the affinity purified polyclonal antiserum against human
UCP2/3 in both mice and rats (Diano et al., 2000, 2003;
Horvath et al., 1999), the antiserum of choice in the
present study. In further support of our view that the im-
munolabeling in CAergic cells depicts UCP expression, we
recently observed that mitochondrial uncoupling is ele-
vated in SN samples of UCP2/3 overexpressing animals
and diminished in UCP2 knockout animals (Andrews et al.,
2005). The increase of UCP2/3 IR observed in the UCP2/3
overexpressing mice support the specificity of the anti-
serum used in the present work. Nevertheless, the possi-
bility exists that the antisera used in the present study may
recognize epitopes of other peptides, particularly other
UCPs such as UCP3. Thus, throughout the text of this
manuscript, we refer to the immunolabeling as UCP2/3.
Previous work by Horvath’s group (Bechmann et al.,
2002; Diano et al., 2000; Horvath et al., 1999, 2003b) and
others (Richard et al., 1998) has demonstrated the exis-
tence of UCP2 producing neurons in the mammalian brain
especially in the hypothalamus, where it is found in various
types of neuropeptidergic neurons. Previous work also
indicates that UCP2 can be both a neuromodulator and a
neuroprotector by producing mitochondrial uncoupling with
generation of heat leading to modulation of synaptic trans-
mission and reduction of the production of superoxides
(Bechmann et al., 2002; Diano et al., 2003; Horvath et al.,
1999, 2003a). Immunohistochemistry and in situ hybridiza-
tion studies showing high levels of UCP2 mRNA and UCP2
IR especially in the hypothalamus indicated an increased
role of UCP2 in homeostatic centers involved in the regu-
lation of autonomic, endocrine and metabolic processes
(Diano et al., 2000; Horvath et al., 1999; Richard et al.,
1998).
The present results show coexistence of UCP2/3 and
TH or DBH IR in many cell bodies and discrete terminal-
like fibers of the ascending DA and NA neuron systems.
Fig. 9. Confocal photomicrograph from the rat LC showing double immunolabeling using UCP2/3 (A) and DBH (B) antibodies. Co-localization is
shown in C. Strong punctate UCP2/3 IR (arrowheads) was demonstrated in the DBH IR cell bodies (C, D). Cells with weak UCP2/3 IR are indicated
with open arrows (C, E). UCP2/3 IR was also present in DBH IR dendrites and terminal-like fibers (arrows) (C, E) in close proximity to the DBH IR
nerve cell bodies. Scale bar?100 ?m in A–C and in D–E is 25 ?m.
A. Rivera et al. / Neuroscience 137 (2006) 1447–14611456

Page 11

These observations open up the possibility that UCP2/3
also can play a major role in modulation of CA volume
transmission in the brain by producing temperature gradi-
ents (see Horvath et al., 1999, 2003a) causing convective
fluid movements. In this way UCP2/3 may enhance migra-
tion of neurotransmitters released from discrete forebrain
CAergic nerve terminals and from large numbers of CA cell
bodies and their dendrites.
In the shell of nucleus accumbens there exist special
nerve cell systems that contain a high density of D1recep-
tors but very few DA terminals that are surrounded by a
high density of DA but by only sparse NA nerve terminals
(Jansson et al., 1999). The existence of neuronal cell
clusters and neurochemical compartments in the rat nu-
cleus accumbens was demonstrated by Herkenham et al.
(1984). In the Jansson et al. (1999) studies it was pro-
posed that DA can diffuse from the surrounding DA termi-
nals to the high numbers of high affinity D1DA receptors
found in these special accumbens cell compartments. In
the present studies strongly UCP2/3 IR terminals were
Fig. 10. (A, B) Representative photomicrographs showing the increase in size and intensity of UCP2/3 IR of striatal nerve terminals on the
6-OHDA-lesioned side (B) compared with the unoperated side (A). Scale bar?20 ?m in A–B. (C, D) Comparison of the diameter (in ?m) (C) and OD
(D) of UCP2/3 IR varicose nerve terminals on the unoperated (white bars) and 6-OHDA-lesioned side (black bars) of the cerebral cortex, the caudate
putamen and the hypothalamus. Values are expressed as means?S.E.M. (n?4). Statistical analysis of the data was performed with Mann-Whitney
U test, ** P?0.01. (E–H) Graphical representations of percentage of varicosities found in each category of size (E and F) and OD (G and H) in the
cerebral cortex and caudate putamen. White bars represent values from control side and black bars from 6-OHDA-lesioned side.
A. Rivera et al. / Neuroscience 137 (2006) 1447–1461 1457

Page 12

observed surrounding these accumbens nerve cell com-
partments characterized by a D1/TH IR mismatches as
observed in the present images (Fig. 1). These two termi-
nal systems overlap in the distribution but the co-existence
of these two IR in the DA terminals remains to be investi-
gated using ultrastructural analysis. Nevertheless, the
present findings again open up the possibility that if UCP2/
3-generated temperature gradients (Horvath et al., 1999,
2003a) exist in the UCP2/3 IR-enriched patches, they may
enhance release of DA from DA terminals and DA migra-
tion with the convective fluid movements. In this way DA
may reach the D1receptor rich cell system more rapidly
and in biologically relevant concentrations, enhancing the
DA communication via volume transmission.
In the core of the nucleus accumbens it was however
possible to demonstrate co-localization of UCP2/3 and TH
IR in the same varicose nerve terminals, but not of UCP2/3
and DBH IR. These results indicate that UCP2/3 can exist
in DAergic terminals where by its postulated generation of
heat (Horvath et al., 1999, 2003a) it may facilitate release
and migration of DA in the surrounding ECF by the result-
ing temperature gradient. It should be noted that UCP2/
3-TH co-storing nerve terminals have large varicosities
while the TH IR fibers lacking UCP2/3 IR had other mor-
phological features with only small varicosities and with
low amount of TH IR. It is possible to speculate that the
UCP2/3-TH co-storing terminals may represent terminals
specialized for volume transmission. Large varicosities
with strong TH IR probably will have the ability to synthe-
size, store and release large amounts of CA into the ex-
tracellular space. Thus, with high amounts of CA released
the sphere of migrating CA in the ECF reaching the high
affinity CA receptors in biologically relevant concentrations
will become larger than with small varicosities with low TH
IR. It can be speculated that this migrating process may be
enhanced by co-stored UCP2/3 IR via heat generation
producing a temperature gradient. However, synaptic spe-
cialization at these large varicosities cannot be excluded
but it must be remembered that the majority of CA recep-
tors are located extrajunctionally (see Zoli et al., 1998,
1999) and most CA terminals lack junctional specializa-
tions (Descarries and Mechawar, 2000).
A partial lesion of the DAergic mesostriatal pathway
induces an increase in the size and in the content in TH IR
of the DA-varicosities, which is interpreted as a compen-
satory mechanism after the lesion of the DA pathway to
increase extrasynaptic DA release and to promote volume
transmission (Zoli et al., 1998, 1999). Similar results were
observed in the present study on UCP2/3 IR terminals in all
cortical regions where only the few strongly TH IR termi-
nals with large varicosities, probably representing DA ter-
minals, were found to contain UCP2/3 IR. The DA charac-
ter of the UCP2/3-TH co-storing terminals was indicated by
the absence of the UCP2/3-DBH co-storage in the cerebral
cortex. Thus, the increase in the size and intensity of the
UCP2/3 IR terminals on the 6-OHDA-lesioned side may
indicate a compensatory increase in their activity in re-
sponse to the partial forebrain DA denervation with a
switch toward the VT mode of communication (see above).
In contrast the vast majorities of the cortical TH IR termi-
nals characterized by small varicosities with weak IR lack
UCP2/3 IR.
0
5
10
15
20
25
30
2060 100 140 180 220
Optical density x10-3
260 300 340 380 420460 500
0
5
10
15
20
25
30
2060 100 140 180 220 260 300 340 380 420 460 500
Cortex
G
Optical density x10-3
H
% varicosities
% varicosities
Caudate Putamen
Fig. 10. (Continued).
A. Rivera et al. / Neuroscience 137 (2006) 1447–14611458

Page 13

In line with the findings of Horvath et al. (2003b), UCP2/3
IR could in the present studies be demonstrated in the large
number of TH IR nerve cells in the SNc and also in the SNr.
It is of substantial interest that UCP2/3 IR also could be
demonstrated in the TH IR dendritic-like processes in the
SNr. This opens up the possibility that UCP2/3 may also
facilitate migration of DA upon its release from dendrites in
addition to enhancing the dendritic DA release process
(Ruffieux and Schultz, 1980). It is of particular interest that
Horvath et al. (2003b) have found that the UCP2/3 IR DA
cellsarelessaffectedbyneurotoxicinjury.However,itshould
be considered that the UCP2/3 IR in the DA cell bodies can
alsofacilitatethemigrationofDAfromthecellbodiestoreach
e.g. close-by D2 autoreceptors (Sesack et al., 1994). The
present evidence indicates that these UCP2/3 mechanisms
exist also in the VTA DA cell bodies.
In the case of the NA nerve terminal networks in the
forebrain it was also possible to demonstrate in certain
regions the co-localization of UCP2/3 and DBH IR. In the
D1/TH mismatch regions of the shell of the nucleus accum-
bens poor in UCP2/3 IR it was possible to demonstrate that
the few DBH IR nerve terminals present in this area con-
tain UCP2/3 IR. Like in the case of the DA nerve terminals
with UCP2/3 IR the UCP2/3-DBH IR terminals had large
varicosities with strong UCP2/3 IR. This was even more
clearly demonstrated in the caudate putamen, where all
the scattered UCP2/3-DBH IR terminals were strongly
UCP2/3-DBH IR with large varicosities. As expected, these
terminals were also strongly TH IR. These results also
underline the interpretation that nerve terminals with large
varicosities with high amount of TH and DBH IR i.e. large
NA terminals with high NA synthesis containing UCP2/3 IR
are specialized for volume transmission, capable of gen-
erating temperature gradients by means of UCP2 resulting
in increased migration of NA in the ECF from the release
sites (see above).
The scattered UCP2/3 IR NAergic terminals in the
striatum are protected against 6-OHDA-induced lesions by
pretreatment with desipramine (Breese and Traylor, 1970;
Jacks et al., 1972). However, on the 6-OHDA-lesioned
side we observed an increase of the size and intensity of
the scattered striatal UCP2/3 IR varicose-like terminals,
similar to those observed in the cerebral cortex. The ob-
servation that in the hypothalamus there were no changes
in the UCP2/3 IR intensity supports the view of a specific
CA terminal response developed after the lesion of the
ascending DAergic pathway. In the caudate putamen the
morphological analysis indicates that the compensatory
responses occurred in the scattered NA-UCP2/3 terminal
system. It has been shown that NA exhibits a high affinity
for the D4DA receptors (Lanau et al., 1997; Newman-
Tancredi et al., 1997). An increased release of NA from the
scattered NA-UCP2/3 terminal system could more effec-
tively activate the DA D4receptors and partially compen-
sate for the loss of DA. An increase of NA migration may
occur by increasing the expression of UCP2/3 with an
increase of postulated heat generation.
The confocal analysis elegantly demonstrated the ex-
istence of punctate UCP2/3 IR in the cytoplasm of the LC
NA cells probably representing aggregates of UCP2/3 IR
mitochondria. Like in the case of the DA cell bodies these
UCP2/3 IR mitochondria may play a role inter alia in facil-
itating NA release and migration in the surrounding ECF to
reach e.g. ?2 autorreceptors (Aoki et al., 1994). This
mechanism may be in operation also in the NA dendrites
which also possessed UCP2/3-DBH IR. These results are
consistent with the finding of Callado and Stamford (2000)
suggesting interactions between NA and ?2 autorrecep-
tors or NA transporters as far as 10 ?m from the NA
release sites.
It is clear that despite the fact that large numbers of DA
and NA cell bodies contain cytoplasmic UCP2/3 IR only a
distinct minority of DA and NA nerve terminals contains
UCP2/3 IR and is characterized by large varicosities and
high amounts of TH and/or DBH IR. It may be that only
certain types of DA and NA nerve cells have the capability
to transport the UCP2/3 IR mitochondria into the nerve
terminal networks by axonal flow, at least in amounts
detectable by immunohistochemistry techniques. Under all
circumstances there must exist substantial differences in
the ability of DAergic and NAergic cells to transport
UCP2/3 IR mitochondria to the DAergic and NAergic net-
0
0,5
1
B
diameter (µm)
**
**
Cx CPu
0
50
100
150
200
250
WT OE
**
**
A
**
optical density x10-3
CxCPu
Hyp
WT OEWT OE
WT OE WT OE
Fig. 11. Mean OD values (A) (?m) and diameter (B) of UCP2/3 IR
varicosities in UCP2/3 overexpressing mice. White bars represent
wild-type mice and black bars transgenic mice. Scatter plots show the
mean values from each individual animal together with the overall
mean value. Statistical analysis was made with Mann-Whitney U test
(** P?0.01).
A. Rivera et al. / Neuroscience 137 (2006) 1447–14611459

Page 14

works. It is of particular interest that UCP2/3 IR DA and NA
nerve terminals have large varicosities and large amounts
of the CA synthesizing enzymes. Such terminals may be
highly specialized for volume transmission with the ability
to synthesize and release large amounts of CA and to
facilitate CA migration in the ECF by the UCP2/3-induced
temperature gradients.
Acknowledgments—This work has been supported by a grant
from the Swedish Research Council (14X-715) (to K.F.), a grant
from the Spanish Research Council (BFI2002-00587), RED CIEN
(Instituto de Salud Carlos III) and Junta de Andalucía (CTS-0161)
(to A.C.) and an NIH grant (NS-041725) (to T.H.). The authors
wish to express their thanks to Antonio Peñafiel for his assistance
with the confocal microscopy.
REFERENCES
Agnati LF, Cortelli P, Biagini G, Bjelke B, Fuxe K (1994) Different
classes of volume transmission signals exist in the central nervous
system and are affected by metabolic signals, temperature gradi-
ents and pressure waves. NeuroReport 6:9–12.
Agnati LF, Fuxe K, Nicholson C, Sykova ˇ E (2000) Volume transmis-
sion revisited: progress in brain research. Vol. 125. Amsterdam:
Elsevier Science BV.
Agnati LF, Genedani S, Lenzi PL, Leo G, Mora F, Ferré S, Fuxe K (2005)
Energy gradients for the homeostatic control of the ECF composition
and for VT signal migration. J Neural Transm 112:45–65.
Andrews ZB, Horvath B, Barnstable CJ, Elseworth J, Yang L, Beal MF,
Roth RH, Matthews RT, Horvath TL (2005) Uncoupling protein-2 is
critical for nigral dopamine cell survival in a mouse model of
Parkinson’s disease. J Neurosci 25:184–191.
Aoki C, Go CG, Venkatesan C, Kurose H (1994) Perikaryal and
synaptic localization of alpha 2A-adrenergic receptor-like immuno-
reactivity. Brain Res 650:181–204.
Arsenijevic D, Onuma H, Pecqueur C, Raimbault S, Manning BS,
Miroux B, Couplan E, Alves-Guerra MC, Goubern M, Surwit R,
Bouillaud F, Richard D, Collins S, Ricquier D (2000) Disruption
of the uncoupling protein-2 gene in mice reveals a role in
immunity and reactive oxygen species production. Nat Genet
26:435–439.
Bechmann I, Diano S, Warden CH, Bartfai T, Nitsch R, Horvath TL (2002)
Brain mitochondrial uncoupling protein 2 (UCP2): a protective stress
signal in neuronal injury. Biochem Pharmacol 64:363–367.
Bouillaud F, Ricquier D, Thibault J, Weissenbach J (1985) Molecular
approach to thermogenesis in brown adipose tissue: cDNA cloning
of the mitochondrial uncoupling protein. Proc Natl Acad Sci U S A
82:445–448.
Breese GR, Traylor TD (1970) Effect of 6-hydroxydopamine on brain
norepinephrine and dopamine evidence for selective degeneration
of catecholamine neurons. J Pharmacol Exp Ther 174:413–420.
Callado LF, Stamford JA (2000) Spatiotemporal interaction of alpha(2)
autoreceptors and noradrenaline transporters in the rat locus
coeruleus: implications for volume transmission. J Neurochem
74:2350–2358.
Descarries L, Mechawar N (2000) Ultrastructural evidence for diffuse
transmission by monoamine and acetylcholine neurons of the cen-
tral nervous system. Prog Brain Res 125:27–47.
Diano S, Urbanski HF, Horvath B, Bechmann I, Kagiya A, Nemeth G,
Naftolin F, Warden CH, Horvath TL (2000) Mitochondrial uncou-
pling protein 2 (UCP2) in the nonhuman primate brain and pituitary.
Endocrinology 141:4226–4238.
Diano S, Matthews RT, Patrylo P, Yang L, Beal MF, Barnstable CJ,
Horvath TL (2003) Uncoupling protein 2 prevents neuronal death
including that occurring during seizures: a mechanism for precon-
ditioning. Endocrinology 144:5014–5021.
Fleury C, Neverova M, Collins S, Raimbault S, Champigny O, Levi-
Meyrueis C, Bouillaud F, Seldin MF, Surwit RS, Ricquier D, War-
den CH (1997) Uncoupling protein-2: a novel gene linked to obesity
and hyperinsulinemia. Nat Genet 15:269–272.
Fuller PM, Warden CH, Barry SJ, Fuller CA (2000) Effects of 2-G
exposure on temperature regulation, circadian rhythms, and adi-
posity in UCP2/3 transgenic mice. J Appl Physiol 89:1491–1498.
Fuxe K, Agnati L (1991) Two principal modes of electrochemical
communication in the brain: Volume versus wiring transmission. In:
Volume transmission in the brain, novel mechanisms for neural
transmission, Vol. 1 (Fuxe K, Agnati L, eds), pp 1–9. New York:
Raven Press.
Herkenham M, Edley SM, Stuart J (1984) Cell clusters in the nucleus
accumbens of the rat, and the mosaic relationship of opiate recep-
tors, acetylcholinesterase and subcortical afferent terminations.
Neuroscience 11:561–593.
Horvath B, Spies C, Horvath G, Kox WJ, Miyamoto S, Barry S, Warden
CH, Bechmann I, Diano S, Heemskerk J, Horvath TL (2002) Un-
coupling protein 2 (UCP2) lowers alcohol sensitivity and pain
threshold. Biochem Pharmacol 64:369–374.
Horvath TL, Warden CH, Hajos M, Lombardi A, Goglia F, Diano S
(1999) Brain uncoupling protein 2: uncoupled neuronal mitochon-
dria predict thermal synapses in homeostatic centers. J Neurosci
19:10417–10427.
Horvath TL, Diano S, Barnstable C (2003a) Mitochondrial uncoupling
protein 2 in the central nervous system: neuromodulator and neu-
roprotector. Biochem Pharmacol 65:1917–1921.
Horvath TL, Diano S, Leranth C, Garcia-Segura LM, Cowley MA,
Shanabrough M, Elsworth JD, Sotonyi P, Roth RH, Dietrich EH,
Matthews RT, Barnstable CJ, Redmond DE Jr (2003b) Coenzyme
Q induces nigral mitochondrial uncoupling and prevents dopamine
cell loss in a primate model of Parkinson’s disease. Endocrinology
144:2757–2760.
Jacks BR, de Champlain J, Cordeau JP (1972) Effects of 6-hydroxy-
dopamine on putative transmitter substances in the central ner-
vous system. Eur J Pharmacol 18:353–360.
Jansson A, Goldstein M, Tinner B, Zoli M, Meador-Woodruff JH, Lew
JY, Levey AI, Watson S, Agnati LF, Fuxe K (1999) On the distri-
bution patterns of D1, D2, tyrosine hydroxylase and dopamine
transporter immunoreactivities in the ventral striatum of the rat.
Neuroscience 89:473–489.
Jansson A, Descarries L, Cornea-Hérbert V, Riad M, Vergé D, Bancila
M, Agnati LF, Fuxe K (2002) Transmitter-receptor mismatches in
central dopamine, serotonin, and neuropeptide systems. Further
evidence for volume transmission. In: The neuronal environment:
brain homeostasis in health and disease (Walz W, ed), pp 83–101.
Totowa, NJ: Humana Press Inc.
Lanau F, Zenner MT, Civelli O, Hartman DS (1997) Epinephrine and
norepinephrine act as potent agonists at the recombinant human
dopamine D4 receptor. J Neurochem 68:804–812.
Lengacher S, Magistretti PJ, Pellerin L (2004) Quantitative RT-PCR
analysis of uncoupling protein isoforms in mouse brain cortex:
methodological optimization and comparison of expression with
brown adipose tissue and skeletal muscle. J Cereb Blood Flow
Metab 24:780–788.
Negre-Salvayre A, Hirtz C, Carrera G, Cazenave R, Troly M, Salvayre
R, Penicaud L, Casteilla L (1997) A role for uncoupling protein-2 as
a regulator of mitochondrial hydrogen peroxide generation. FASEB
J 11:809–815.
Newman-Tancredi A, Audinot-Bouchez V, Gobert A, Millan MJ (1997)
Noradrenaline and adrenaline are high affinity agonists at dopa-
mine D4 receptors. Eur J Pharmacol 319:379–383.
Nicholls DG, Locke RM (1984) Thermogenic mechanisms in brown fat.
Physiol Rev 64:1–64.
Palou A, Pico C, Bonet ML, Oliver P (1998) The uncoupling protein,
thermogenin. Int J Biochem Cell Biol 30:7–11.
A. Rivera et al. / Neuroscience 137 (2006) 1447–14611460

Page 15

Paradis E, Clavel S, Bouillaud F, Ricquier D, Richard D (2003) Un-
coupling protein 2: a novel player in neuroprotection. Trends Mol
Med 9:522–525.
Paxinos G, Watson C (1986) The rat brain in stereotaxic coordinates.
Sydney: Academic Press.
Richard D, Rivest R, Huang Q, Bouillaud F, Sanchis D, Champigny O,
Ricquier D (1998) Distribution of the uncoupling protein 2 mRNA in
the mouse brain. J Comp Neurol 397:549–560.
Richard D, Clavel S, Huang Q, Sanchis D, Ricquier D (2001) Uncou-
pling protein 2 in the brain: distribution and function. Biochem Soc
Trans 29:812–817.
Ruffieux A, Schultz W (1980) Dopaminergic activation of reti-
culata neurones in the substantia nigra. Nature 285:240–241.
Sesack SR, Aoki C, Pickel VM (1994) Ultrastructural localization of
D2 receptor-like immunoreactivity in midbrain dopamine neu-
rons and their striatal targets. J Neurosci 14:88–106.
Zoli M, Torri C, Ferrari R, Jansson A, Zini I, Fuxe K, Agnati LF (1998)
The emergence of the volume transmission concept. Brain Res
Rev 26:136–147.
Zoli M, Jansson A, Sykova E, Agnati LF, Fuxe K (1999) Volume
transmission in the CNS and its relevance for neuropsychophar-
macology. Trends Pharmacol Sci 20:142–150.
(Accepted 5 May 2005)
(Available online 4 January 2006)
A. Rivera et al. / Neuroscience 137 (2006) 1447–14611461