Gender-Dependent Regulation of G-protein-Gated

Inwardly Rectifying Potassium Current in Dorsal

Raphe Neurons in Knock-Out Mice Devoid of the
5-Hydroxytryptamine Transporter

Alexandre Julien Châu Loucif,1 Patricia Bonnavion,1 Béatrice Macri,1,2

Jean-Louis Golmard,3 Claudette Boni,1 Maxette Melfort,1 Grégoire Leonard,1
Klaus-Peter Lesch,4 Joëlle Adrien,1 Thierry Didier Jacquin1,*
1
UMR 677, INSERM/UPMC, NeuroPsychoPharmacologie, Faculté de Médecine Pierre et Marie Curie,
Site Pitié-Salpêtrière, 91 Boulevard de l’Hôpital, 75634 Paris Cedex 13, France
2
Department of Animal Physiology and Biophysics, Faculty of Biology, University of Bucharest,
91–95, Splaiul Independentei, Bucharest 050095, Romania
3
UPMC, Unité Fonctionnelle de Biostatistiques, Faculté de Médecine Pierre et Marie Curie,
Site Pitié-Salpêtrière, 91 Boulevard de l’Hôpital, 75634 Paris Cedex 13, France
4
Molecular and Clinical Psychobiology, Department of Psychiatry and Psychotherapy,
University of Würzburg, 97080 Würzburg, Germany

Received 9 March 2006; accepted 20 June 2006

ABSTRACT: Agonists at G-protein-coupled recep-
tors in neurons of the dorsal raphe nucleus (DRN) of knock-
out mice devoid of the serotonin transporter (5-HTT2/2)
exhibit lower efficacy to inhibit cellular discharge than in
wild-type counterparts. Using patch-clamp whole-cell
recordings, we found that a G-protein-gated inwardly recti-
fying potassium (GIRK) current is involved in the inhibi-
tion of spike discharge induced by 5-HT1A agonists (5-car-
boxamidotryptamine (5-CT) and (6)-2-dipropylamino-8-
hydroxy-1,2,3,4-tetrahydronaphthalene hydrobromide
(8-OH-DPAT); 50 nM–30 lM) in both wild-type and
5-HTT2/2 female and male mice. These effects
were mimicked by 50 -guanylyl-imido-diphosphate
(Gpp(NH)p; 400 lM) dialysis into cells with differences
between genders. The 5-HTT2/2 knock-out mutation
reduced the current density induced by Gpp(NH)p in
females but not in males. These data suggest that the
decreased response of 5-HT1A receptors to agonists in 5-
HTT2/2 mutants reflects notably alteration in the cou-
pling between G-proteins and GIRK channels in females
but not in males. Accordingly, gender differences in cen-
tral 5-HT neurotransmission appear to depend—at least
in part—on sex-related variations in corresponding re-
ceptor-G protein signaling mechanisms.
' 2006 Wiley Periodicals, Inc. J Neurobiol 66: 1475–1488, 2006
Keywords: 5-HT1A; 5-CT; 8-OH-DPAT; Gpp(NH)p;
functional desensitization

Correspondence to: T.D. Jacquin (tjacquin@ext.jussieu.fr). Contract grant sponsor: Deutsche Forschungsgemeinschaft; con-
Contract grant sponsor: INSERM and Université Pierre et Marie tract grant number: SFB 581.
Curie. ' 2006 Wiley Periodicals, Inc.
Contract grant sponsor: European Community; contract grant Published online 29 September 2006 in Wiley InterScience (www.
number: LSHM-CT-2004-503474. interscience.wiley.com).
DOI 10.1002/neu.20321

1475
1476 Loucif et al.
INTRODUCTION
Functional desensitization of G-protein-coupled recep-
tors, reflected by a reduced potency of agonists to inhibit
neuronal activity, has been observed in neurons of the
dorsal raphe nucleus (DRN) in knockout mice for the
gene encoding the Na+/Cl
/K+-dependent 5-HT trans-
porter (5-HTT
/
mice), therefore devoided of 5-HT
reuptake capacity. This has been observed with 5-HT1A
autoreceptors (Gobbi et al., 2001; Mannoury la Cour
et al., 2001; Bouali et al., 2003), interestingly in a man-
ner more pronounced in females than in males (Bouali
et al., 2003), and also with GABAB receptors (Man-
noury la Cour et al., 2004), supporting the idea that both
5-HT1A and GABAB receptors share common transduc-
tion pathways. Indeed, mediation by G proteins has
clearly been shown for both receptor types, in raphe-
(Innis et al., 1988) and hippocampus-neurons (Andrade
et al., 1986) in the rat. Long-term administration of
selective serotonin reuptake inhibitors (SSRIs), which
act on the 5-HT transporter (Graham et al., 1989; Lesch,
1997), also induces 5-HT1A autoreceptor desensitization
(Blier et al., 1998; Le Poul et al., 1995, 2000; Czachura
and Rasmussen, 2000; El Yacoubi et al., 2003), but their
effects on the function of GABAB receptors have not
yet been studied. This adaptive change at G-protein-
coupled receptors is believed to be a key mechanism
underlying the action of SSRIs to treat psycho-affective
disorders. Furthermore, depression occurs with a two-
fold higher frequency in women than in men (Kornstein
et al., 2000b; Piccinelli and Wilkinson, 2000) and the
response to SSRI treatment in depressed patients also
exhibits some gender-dependency (Kornstein et al.,
2000a). Because 5-HTT
/
mutants can be considered
as a model of whole-life treatment with SSRIs, analysis
of the mechanisms underlying G-protein-coupled recep-
tor adaptive changes in the DRN of these mutants, with
care to differences between females and males, should
be of interest to better understand the neurobiological
bases of the therapeutic action of SSRIs as the most fre-
quently used antidepressants.
The functional desensitization of 5-HT1A autore-
ceptors in DRN neurons of 5-HTT
/
mice has been
attributed to a downregulation of 5-HT1A receptors
and related G-proteins (Fabre et al., 2000; Li et al.,
2000; Bouali et al., 2003). Similarly, the concomitant
functional desensitization of GABAB receptors has
also been attributed to a downregulation of related G-
proteins (Mannoury la Cour et al., 2004). However,
no studies to date have assessed the possible implica-
tion of downstream mechanisms in the transduction
system such as the coupling between G-proteins and
related currents in mice. In the rat, 5-HT1A and
GABAB receptors control a G-protein-gated inwardly
Journal of Neurobiology. DOI 10.1002/neu
rectifying potassium (GIRK) current on the same neu-
rons of the DRN to induce membrane hyperpolariza-
tion and decrease spontaneous spike discharge (Innis
et al., 1988; Williams et al., 1988; Penington et al.,
1993). In the present work, we compared quantita-
tively the coupling between G-proteins and the related
inwardly rectifying current in DRN neurons of
5-HTT
/
mutants and paired wild-type mice of the
same genetic background, in both females and males.
For this purpose, patch-clamp techniques were used
under in vitro conditions applied to brain stem slices.
METHODS
Animals
Experiments were performed using females and males of
wild-type 5-HTT+/+ and homozygous 5-HTT
/
genotype
with C57Bl/6 genetic background obtained from hetero-
zygous and homozygous breeding. Genotyping was per-
formed as described already (Bengel et al., 1998). After
weaning and sexing, males and females were housed sepa-
rately in groups of 6 per cage and maintained under stand-
ard laboratory conditions (22 6 18C, 60% relative humid-
ity, 12–12 h light-dark cycle, food and water ad libitum).
Procedures involving animals and their care conformed to
institutional guidelines and complied with national and
international laws and policies (French Council Directive
no 87-848, 19 October 1987 of the Ministère de l’Agri-
culture et de la Forêt, Service Vétérinaire de la Santé et de
la Protection Animale, permissions 75-125 to J.A. and 75-
1153 to T.D.J., and European Community Council Direc-
tive of 24 November 1986; 86/609/EEC).
Preparation of Brainstem Slices
and Neuron Recording
Animals were used at 2 months of age. Because previous
studies showed that 5-HT1A receptor-mediated regulation
of DRN firing does not vary along the estrus cycle (Bouali
et al., 2003), females underwent no monitoring of the cycle.
Mice were decapitated and the brains were rapidly removed
and immersed in an ice-cold (&48C) artificial cerebrospinal
fluid (ACSF) of the following composition (in mM): NaCl,
128; KCl, 3.5; NaH2PO4, 1.2; MgCl2, 1.3; CaCl2, 2;
NaHCO3, 25; D-glucose, 11; sucrose, 17, maintained at pH
7.4 by continuous bubbling with O2/CO2 mixture (95%/
5%). A block of tissue containing the DRN was affixed to
the stage of a vibratome (TPI, series 1000) with cyanoacry-
late glue, submerged in ice-cold and bubbled ACSF and cut
into coronal sections of 250
m thickness. Brain stem slices
were collected in gassed ACSF (continuous bubbling with
95% O2/5% CO2 mixture, pH 7.4) and maintained at room
temperature (22–258C). A slice was then transferred into a
recording chamber (2 mL) and covered with a nylon mesh,
while continuously superfused with gassed ACSF (22–
258C) at a constant flow rate of 3 mL/min. In some experi-

ments, KCl was increased from 3.5 to 23.5 mM and osmo-
larity was maintained by reducing the concentration of
NaCl correspondingly. Recording electrodes (4–6 MO)
were pulled from borosilicate glass tubes (Phymep, France)
on a PP-830 micropipette puller (Narishige, Japan), filled
with internal solution containing (in mM): KH2PO4, 108;
CaCl2, 1; MgCl2, 1; EGTA, 11; Hepes, 10; NaCl, 5; Mg-
ATP, 4; Na-GTP, 0.4, adjusted to pH 7.4 with 36 mM KOH
(final K+ concentration 144 mM), and kept in ice until use.
Na in the internal solution (Mechaly et al., 2005) allowed
firing of DRN neurons, otherwise usually silent in vitro in
current clamp recordings (9/10 cells recorded without NaCl
in the pipette were spontaneously silent in our preparation,
see also VanderMaelen and Aghajanian, 1983). Phenyleph-
rine also induced spiking in the DRN (5/5 cells exhibited
spike discharge in the presence of 3
M phenylephrine, see
VanderMaelen and Aghajanian, 1983) by regulating sodium
and potassium currents (Pan et al., 1994; Brown et al.,
2002), but it was not used because unknown interactions
with the mechanisms under study may occur, in particular
in 5-HTT
/
mice. When indicated, Na-GTP of internal
solution was either omitted without adjusting ionic strength,
either replaced by the nonhydrolyzable GTP analog,
Gpp(NH)p, also at 0.4 mM. Electrodes were advanced
towards DRN neurons under visual guidance, using an
upright microscope equipped with Normarski optics (Nikon
E600FN, Japan). Pipette voltage offset was neutralized
prior to the formation of a GO seal and was not further cor-
rected. Capacitance was neutralized before breaking the
seal. Patch-clamp recordings were performed in whole-cell
configuration and amplified with an Axopatch 200B (Axon
Instruments, Union City, CA) using current clamp fast- and
voltage clamp-modes. Signals were low-pass filtered at 2
kHz, digitized at 15 kHz, and analyzed with a computer
using a Digidata 1200 interface coupled to the pCLAMP 9
software (Axon Instruments). Input resistance (Rm), series
resistance (Rs), and membrane capacitance (Cm) were deter-
mined from current transients elicited by a 5 mV depolariz-
ing step from a holding potential of
60 mV. Criteria for
cell inclusion in the study were as follows: Rm (>100 MO)
and stable Rs (20 MO), Cm (30–55 pF), and spike ampli-
tude (55 mV). The neurons were considered healthy when
maintaining the same cellular characteristics, except drug-
induced modifications, during the whole recording period.
Liquid junction potential error (Verheugen et al., 1999) was
calculated using pCLAMP 9.2 (
13.9 mV) and corrected.
In current clamp mode, spike discharge frequency (f: Hz)
was regularly monitored and determined for at least 2 min
at various epochs of recordings. In voltage clamp mode,
currents were evoked by applying 5 mV increment voltage
steps to potentials from
30 to
145 mV from a holding
potential of –60 mV, and for 1.2–1.4 s duration. Steady-
state current activation curves were plotted against voltage
(I/V curves). To compare and quantify data between cells,
current density (d: pA/pF) was measured during steady-
state current induced by voltage steps down to
125 mV.
When testing a 5-HT1A agonist, f and d were compared
before (f0 and d0) and after superfusion of 35 mL agonist-
containing ACSF (fagonist and dagonist). In the case of free
GIRK Current in DRN of 5-HTT
/
Mice 1477
Gpp(NH)p dialysis into cells, f and d were compared by tak-
ing values within the first 5 min of recordings after breaking
the seal (baseline values: f0 and d0), and after *20–30 min
Gpp(NH)p dialysis into cells (fGpp(NH)p and dGpp(NH)p). In
particular, the variations of f (Dfagonist ¼ fagonist
f0
or DfGpp(NH)p ¼ fGpp(NH)p
f0, Hz, respectively) and d
(Ddagonist ¼ dagonist
d0 or DdGpp(NH)p ¼ dGpp(NH)p
d0,
pA/pF, respectively) were calculated. Neurons were consid-
ered as inhibited when the following criteria were all met:
(1) decrease of f, (2) hyperpolarization, (3) increased ampli-
tude of currents evoked by hyperpolarizing voltage steps,
and (4) reversibility of these effects upon washing or WAY
100635 or barium superfusion after 5-HT1A agonist applica-
tion, or barium superfusion after Gpp(NH)p dialysis.
Thresholds for \Gpp(NH)p-induced inhibition" were con-
sidered when ratios after *15–30 min Gpp(NH)p dialysis
into cells (a reasonable amount of time to observe
Gpp(NH)p effects, see Ikeda and Schofield, 1989) to base-
line values were as follows: fGpp(NH)p/f0  0.33 and
dGpp(NH)p/d0  1.5.
Statistical Analyses
Descriptive statistics displayed frequencies for qualitative
data and means, standard deviations, medians and Q1–Q3
(interquartile) intervals for quantitative ones. The relation-
ships between qualitative variables were tested using logis-
tic regression analyses when comparing a binary criterion
and several potential factors, and Fisher’s exact tests
(Fleiss, 1981) when comparing two variables. With samples
of small size and consequently, normality hypotheses diffi-
cult to assess, we adopted a nonparametric approach. There-
fore, the relationships between quantitative variables and
qualitative ones were tested using Kruskal–Wallis tests for
three or four-valued qualitative variables and Wilcoxon two-
sample tests for binary ones (comparisons of two groups).
No correction for multiple comparisons was performed. All
the tests were two-tailed, and p-values lower than 0.05 were
considered as statistically significant. Computations were
performed using the SAS V8 statistical software.
Chemicals
The following drugs were aliquoted, stored at –208C, and
dissolved in ACSF immediately before their use in superfu-
sion experiments: 5-carboxamidotryptamine (5-CT; 50 nM–
20
M) from Research Biochemicals, Natick, MA; (6)-2-
dipropylamino-8-hydroxy-1,2,3,4-tetrahydronaphthalene hy-
drobromide (8-OH-DPAT; 20 nM
30
M), phenylephrine
(3
M) and barium chloride (2 mM) from Sigma-Aldrich
Chimie, St Quentin Fallavier, France; 4-(N-ethyl-N-phenyl-
amino)-1,2-dimethyl-6-(methylamino) pyrimidium chloride (ZD
7288; 0.1 mM) from Tocris, Illkirch, France; and N-[2-[4-(2-
methoxyphenyl)-1-piperazinyl]ethyl]-N-(2-pyridinyl)-cyclo-
hexane carboxamide (WAY 100635, 2–20
M) from
Wyeth-Ayerst, Princeton, NJ.
Journal of Neurobiology. DOI 10.1002/neu
1478 Loucif et al.

RESULTS

Recordings in the DRN

Whole-cell recordings in the DRN were obtained
from 79 cells of wild-type mice (46 in females and 33
in males) and from 71 cells of 5-HTT
/
mice (39 in
females and 32 in males). Cells were located mostly
in the dorsal subdivision of the DRN (84–100% of the
total number of recorded cells in each group of mice),
throughout the rostro-caudal extent of the nucleus
(c.f. plates 66–72; Franklin and Paxinos, 1997, see
[Fig. 1(b,c)]).

Electroresponsive Properties of Three

Types of DRN Neurons in Wild-Type
and 5-HTT2/2 Mice
A very large majority of neurons recorded in the
DRN of wild-type (n ¼ 63/64) and 5-HTT
/
(n ¼
70/71) mice exhibited regular firing in current clamp
recordings, with 3.5 mM KCl in ACSF and 5 mM
NaCl and 0.4 mM GTP in the electrode solution. The
two remaining neurons were silent. Electrophysiolog-
ical characteristics of active neurons of the DRN are
provided in females and males of wild-type and 5-
HTT
/
mice (see [Figs. 1(b,c)] and Table 1). These
characteristics were not significantly different
between the four groups.

Figure 1 Characteristics of the recorded neurons in the

Effects of 5-HT1A Agonists on Spike
Discharge and Inwardly Rectifying
Current
In wild-type (Fig. 2) and mutant (Fig. 3) mice, most
DRN neurons (n ¼ 39 out of a total of 63 recorded
cells) were reversibly inhibited by the 5-HT1A ago-
nists, 5-CT (50 nM–3
M) and 8-OH-DPAT (300
nM–30
M, see Table 2 and the selection criteria in
methods). In current clamp recordings, the regular fir-
ing of neurons with 5 mM internal sodium was fol-
lowed by a clear decrease of spike discharge fre-
quency and a hyperpolarization when perfusing with
5-HT1A agonists [Figs. 2(a) and 3(a)]. Note that in
wild type animals, a small 5-CT concentration
(300 nM) induced strong inhibition of firing, with
complete arrest, while a large 5-CT concentration
(1
M) in 5-HTT
/
mutants could not totally stop
firing. In voltage clamp recordings, application of the
agonists produced an increase of the inward current
evoked by hyperpolarizing steps of 5 mV increment
and 1.2 s duration from a holding potential close to
the resting potential (Vholding ¼
60 mV; [Figs. 2(b)
and 3(b)]). Current–voltage relations were examined
Journal of Neurobiology. DOI 10.1002/neu
DRN. (a) The hatched area shows the localization of the re-
corded neurons, mostly in the dorsal subdivision of the
DRN. The limits of the DRN, indicated by the dotted line,
were easily identified in the slice preparation from its charac-
teristic location and aspect. Aq: aqueduct of Sylvius; plate
no. 69 from the atlas of Franklin and Paxinos (1997). (b,c)
Measurements of electrophysiological characteristics in the
same neuron such as, in panel b: resting membrane potential
(arrowhead on the left indicates
55 mV membrane poten-
tial), duration of afterhyperpolarization (DAHP) and thresh-
old of action potential (T, defined as the membrane potential
at which the velocity of depolarization exhibits a rapid
increase), and in panel c: amplitude (AAP), duration (DAP),
and threshold (T, arrowhead on the left indicates
47 mV
membrane potential) of action potential and amplitude of
afterhyperpolarization (AAHP). Top time calibration bar
refers to panel b and bottom calibration bar refers to panel c.
in the absence and the presence of the agonists in volt-
age clamp recordings [Figs. 2(c) and 3(c)] and
changes in currents elicited by the agonists were
determined as the \agonist difference current" [Figs.
2(d) and 3(d)]. The 5-HT1A agonist difference current
reversed around
90 mV at 3.5 mM of external K+
 GIRK Current in DRN of 5-HTT
/
Mice 1479

Table 1 Cellular Characteristics of DRN Neurons in Females and Males of Wild-Type and 5-HTT
/
Genotype
5-HTT+/+ 5-HTT
/

# (26) $ (38) # (32) $ (39)
Resting membrane
55.0 6 6.4
57.3 6 8.1
52.9 6 7.5
54.3 6 7.6
potential (mV)
54.0 [
59.0;
52.0]
55.0 [
62.0;
52.0]
51.0 [
59.0;
48.0]
54.0 [
59.0;
48.0]
Input resistance (MO) 647.0 6 248.5 481.6 6 298.4 472.2 6 350.0 435.9 6 257.1
355.4 [239.9; 753.0] 460.0 [229.0; 623.6] 349.0 [238.6; 618.0] 382.5 [226.9; 625.1]
Tau (ms) 40.6 6 20.3 39.5 6 19.1 41.1 6 18.6 42.1 6 29.1
36.7 [27.0; 56.0] 38.8 [28.3; 49.0] 40.2 [28.3; 53.5] 32.1 [16.6; 60.0]
AP threshold (mV)
45.0 6 5.3
45.3 6 6.4
42.3 6 6.7
44.0 6 6.6

44.0 [
47.0;
42.0]
45.2 [
50.5;
40.0]
41.6 [
46.0;
38.0]
43.1 [
49.0;
39.0]
AP amplitude (mV) 70.8 6 16.8 73.6 6 11.6 67.7 6 9.5 67.0 6 11.3
71.0 [57.5; 80.0] 73.0 [68.0; 80.6] 70.0 [62.0; 75.0] 67.7 [57.8; 75.0]
AP duration (ms) 4.2 6 1.9 4.0 6 1.7 5.0 6 1.7 4.4 6 1.2
3.9 [3.2; 5.2] 3.7 [2.8; 5.2] 5.2 [3.5; 5.9] 4.4 [3.5; 5.1]
AHP amplitude (mV) 20.9 6 5.0 21.8 6 6.3 19.0 6 4.6 19.7 6 4.2
21.5 [17.5; 24.0] 22.0 [18.1; 25.7] 19.3 [16.4; 22.7] 20.2 [17.0; 22.5]
AHP duration (ms) 105.0 6 85.0 119.2 6 88.8 122.5 6 88.9 147.0 6 124.2
75.9 [60.0; 113.6] 82.4 [66.0; 150.0] 106.0 [61.5; 156.9] 115.4 [65.0; 181.4]
AHP tau (ms) 1.5 6 1.3 1.5 6 1.8 3.1 6 8.0 1.4 6 1.6
1.0 [0.7; 2.1] 0.9 [0.5; 1.7] 1.1 [0.7; 2.5] 0.8 [0.7; 1.3]
Current density (pA/pF) 3.7 6 1.2 4.2 6 2.1 4.6 6 2.9 4.1 6 2.1
3.7 [2.7; 4.9] 3.6 [3.0; 5.5] 3.6 [2.3; 5.8] 3.5 [2.4; 5.6]
Spike discharge 12.1 6 10.6 9.7 6 6.5 8.3 6 5.0 8.8 6 6.0
frequency (Hz) 9.3 [4.1; 15.0] 8.4 [5.0; 12.2] 7.5 [4.8; 10.2] 7.5 [5.5; 10.5]
Values indicate mean 6 SD, and Median values and Q1–Q3 intervals; numbers in parentheses indicate total number of cells studied. Note
that values are not significantly different between groups. AP: action potential, AHP: afterhyperpolarization.

[Figs. 2(c) and 3(c)] and around
40 mV with 23.5
mM of external K+ [Fig. 4(b)], compatible with the
reversal potential for K+ according to the Nernst
equation:
93 and
45 mV, respectively. Current
density, measured during steady-state current induced
by hyperpolarizing voltage steps to
125 mV,
increased during 5-CT superfusion from about 3 to 7
pA/pF [Figs. 2(e) and 3(e)]. The current was more
pronounced in 23.5 mM versus 3.5 mM of external
K+ (Fig. 4), as expected from K+-dependent kinetic
properties of the GIRK current (Isomoto et al., 1997).
The 5-HT1A agonist difference current was voltage
sensitive, exhibiting an inward rectification at depo-
larized potentials, as depicted by I/V plots for either
3.5 mM [Figs. 2(d) and 3(d)] or 23.5 mM [Fig. 4(c)]
external K+. The specific 5-HT1A receptor antagonist,
WAY 100635 (1–10
M; n ¼ 7 during 5-CT superfu-
sion; n ¼ 3 during 8-OH-DPAT superfusion; Fig. 2)
and barium, a potassium channel blocker (2 mM, n ¼
4; not shown) reversed spike discharge frequency and
current density, and could block the 5-HT1A agonist
difference current.
When pooling together the results with 5-CT and
8-OH-DPAT, the proportions of 5-HT1A agonist-
sensitive cells exhibited no significant differences
between females and males and between wild-type
and mutant mice (logistic regression analysis, see
Table 2).
In other DRN neurons (n ¼ 24 out of the 63
recorded cells), the hyperpolarizing voltage steps-
evoked current was unchanged during superfusion
with more than 300 nM of 5-CT or 8-OH-DPAT,
although firing was reduced or even abolished in
some cells (n ¼ 5). Possibly, 5-HT1A receptors con-
trol a distinct current in these cells, such as a transient
calcium current (Chen and Penington, 1996), or the
reduction in activity depends on an unidentified net-
work interaction.
In no cases was the GIRK current observed during
superfusion of 5-HT1A agonists, when GTP was omit-
ted in the pipette.
Gpp(NH)p-Induced Inhibition of DRN
Neurons in Females and Males of
Wild-Type and 5-HTT2/2 Genotype
To test for G-protein-dependent activation of the
inwardly rectifying current, GTP in the pipette was
replaced by the nonhydrolyzable GTP analog,
Journal of Neurobiology. DOI 10.1002/neu
1480 Loucif et al.
Figure 2 5-CT-induced inhibition of a DRN neuron in a
5-HTT+/+ female. (a) Neuronal activity in control condi-
tions, as well as during 5-CT perfusion and addition of
WAY 100635 in a brain stem slice in current clamp record-
ings. 5-CT (300 nM) induces hyperpolarization and cessa-
tion of spike discharge. WAY 100635 (1
M, added to 5-
CT at 300 nM) reverses 5-CT-induced inhibition. The
arrowhead on the left indicates
55 mV membrane poten-
tial. (b) Inward current evoked by applying 5 mV increment
voltage steps to potentials from
45 mV to
145 mV, from
a holding potential of
60 mV, during control, perfusion of
5-CT (300 nM), and WAY 100635 (1
M, added to 5-CT at
300 nM). The 5-CT-induced increase of current amplitude
is reversed by WAY 100635. Same neuron as in panel a.
Current traces at each voltage step are an average of 3
sweeps. (c) Activation curves of steady state current
recorded in panel b are plotted against voltage. Note an
inward rectification at relatively depolarized voltages. (d)
The 5-CT difference current obtained by subtracting the
control- from the 5-CT (300 nM)-induced I–V curve exhib-
its inward rectification in the depolarized voltage range,
compared with the dotted line, and reverses polarity at
95
mV. (e) Spike discharge frequency (f; top) and current den-
sity (d; bottom) as a function of 5-CT and WAY 100635
concentrations, for recordings shown in panels a and b,
respectively. d was measured during steady-state current
induced by hyperpolarizing voltage steps to
125 mV, as in
the following figures. [K+]ext ¼ 3.5 mM .
Journal of Neurobiology. DOI 10.1002/neu
Gpp(NH)p (400
M). In a proportion of neurons of 5-
HTT+/+ (5/25 in females and 11/19 in males, Fig. 5)
and 5-HTT
/
(8/18 in females and 7/14 in males,
Fig. 6) mice, inhibition similar to that induced by 5-
HT1A agonist superfusion was observed within 15
min after the start of whole cell recordings. As in
the case of 5-HT1A agonists, the proportions of
Figure 3 5-CT-induced inhibition of a DRN neuron in a
5-HTT
/
female. (a) Neuronal activity in control, during
5-CT and wash, in current clamp recordings. 5-CT (1
M)
inhibits cellular activity with hyperpolarization and
decrease of spike discharge frequency. Partial recovery is
shown during wash. The arrowhead on the left indicates

55 mV membrane potential. (b) Inward current evoked by
5 mV increment voltage steps to potentials from
45 mV
to
145 mV, from a holding potential of
60 mV, during
control, 5-CT (1
M) and wash. Same neuron as in panel a.
Current traces at each voltage step are an average of 3
sweeps. (c) Activation curves of steady state current
recorded in panel b are plotted against voltage. Note an
inward rectification at relatively depolarized voltages. (d)
The 5-CT difference current obtained by subtracting the
control- from the 5-CT (1
M)-induced I–V curve exhibits
inward rectification in the depolarized voltage range and
reverses polarity at
95 mV. (e) Spike discharge frequency
(top) and current density (bottom) as a function of 5-CT
concentration, for recordings shown in panels a and b,
respectively. [K+]ext ¼ 3.5 mM .
 GIRK Current in DRN of 5-HTT
/
Mice 1481

Table 2 Proportions of Inhibited Neurons in the with the emergence of either a slow-, or a fast-relaxa-
DRN for Females and Males of Wild-Type and tion, according to neurons. The amplitude of the slow
5-HTT2/2 Genotype during Gpp(NH)p Dialysis relaxation [Fig. 5(c,f)] increased at hyperpolarized
or 5-HT1A Agonist Superfusion potentials, consistent with current activation. In neu-
5-HTT+/+ 5-HTT
/
rons with the slow relaxation, the time course was
best described by a fit of two exponentials, with aver-
$ # $ #
age time constants s1 ¼ 17.0 6 2.1 ms and s2 ¼ 165.1
Gpp(NH)p 5/25 11/19 8/18 7/14 6 11.7 ms (n ¼ 11) at
125 mV hyperpolarizing
% 20* 58 44 50 voltage steps after 30 min Gpp(NH)p dialysis into
5-CT 6/9 3/7 3/6 5/8 cells. In the other neurons, the fast relaxation was par-
8-OH-DPAT 7/7 1/4 7/9 7/13 tially masked by the settling time of the clamp.
Total 13/16 4/11 10/15 12/21 In all cells sensitive to Gpp(NH)p dialysis, current
% 81 36 66 57
density increased irreversibly as a function of time to
Top: Gpp(NH)p was applied at 400
M. Bottom: 5-CT and 8- reach a maximum [Figs. 5(e) and 6(e)]. However, this
OH-DPAT were applied at 50 nM–3
M and 300 nM – 30
M, was blocked by barium (2 mM), a GIRK channel
respectively. The highest concentrations were used to ascertain the
absence of neuronal sensitivity. Numbers are indicated versus the
blocker, both in mice of 5-HTT+/+ (Fig. 5) and 5-
total number of recorded cells in each group of mice. HTT
/
(Fig. 6) genotypes. Barium could depolarize
*p < 0.001, different from 5-HT1A agonist-sensitive cells in the cells with recovery of spike discharge (in current
wild-type females. clamp recordings) and decrease of Gpp(NH)p-
induced current (in voltage clamp recordings), ruling
out run-up artifacts during Gpp(NH)p dialysis into
cells. Interestingly, barium also could block a compo-
Gpp(NH)p-inhibited neurons were not significantly
different between females and males and between
wild-type and mutant mice (logistic regression analy-
sis, see Table 2).

Gpp(NH)p-Sensitive Cells. Gpp(NH)p produced a

strong inhibition, with hyperpolarization and marked
decrease or arrest of spontaneous firing [Figs. 5(a)
and 6(a)], using 3.5 mM of external K+ and current
clamp mode. Furthermore, a basal outward current
appeared and the inward current elicited by hyper-
polarizing voltage steps increased in amplitude
[Figs. 5(b) and 6(b)], in voltage clamp mode.
The current elicited by Gpp(NH)p, called the
\Gpp(NH)p-difference current", was obtained by
subtracting the amount of current elicited by hyperpo-
larizing steps at the beginning of the whole cell
recordings (baseline value) from the amount of cur- Figure 4 [K+]ext and 5-CT-mediated increase in GIRK
rent elicited after *20–30 min of Gpp(NH)p dialysis current in a DRN neuron of a 5-HTT+/+ female. (a) Increas-
into cells [Figs. 5(c) and 6(c)]. At steady state during ing [K+]ext from 3.5 mM (l) to 23.5 mM (^) enhances
hyperpolarizing voltage steps at 3.5 mM external K+, inward current evoked by 5 mV increment voltage steps to
the Gpp(NH)p difference current reversed around potentials from
45 to
125 mV, from a holding potential

95 mV as shown in I/V plots [Figs. 5(d) and 6(d)]. of
60 mV. Current increase is even more pronounced in
This is compatible with the reversal potential for K+ the presence of 100 nM of 5-CT (~). This effect is reversi-
according to the Nernst equation,
93 mV, and sug- ble after 5-CT removal from the superfusing medium (n).
(b) Current/voltage curves at steady state of currents shown
gests that the Gpp(NH)p difference current was exclu-
in panel a. Current traces at each voltage step are an aver-
sively carried by potassium ions. The Gpp(NH)p dif-
age of 3 sweeps. (c) The 5-CT (100 nM) difference current
ference current also exhibited voltage dependency in 23.5 mM K+ is obtained by subtracting the control K+
with inward rectification at depolarized voltages as 23.5 mM curve from the 5-CT 100 nM curve in 23.5 mM
shown in I/V plots [Figs. 5(d) and 6(d)]. Furthermore, K+. The 5-CT difference current exhibits inward rectifi-
two types of time dependent inward rectification were cation in the depolarized voltage range and reverses at
distinguished in the Gpp(NH)p difference current,
40 mV.
Journal of Neurobiology. DOI 10.1002/neu
1482 Loucif et al.
nent of the control current [Fig. 5(b,c)]. On the con-
trary, ZD 7288 (0.1 mM), an organic Ih blocker
(BoSmith et al., 1993), had no effect on the
Gpp(NH)p-induced inhibition (n ¼ 3, not shown).
Therefore, reversal potential close to EK (
93 mV),
similar voltage- and time-dependency of inward
relaxation, and barium sensitivity suggest that the
Gpp(NH)p difference current observed in DRN neu-
rons is compatible with activation of a G-protein-
gated inwardly rectifying current (Isomoto et al.,
1997; Yamada et al., 1998).
In similar recordings for 1 h duration, but with
GTP rather than Gpp(NH)p in the pipette, action
potential discharge and current density remained sta-
ble as did resting membrane potential, membrane re-
sistance, and action potential amplitude (n ¼ 5, not
shown).
Gpp(NH)p-Induced Changes of Spike Discharge
Frequency and Current Density with Gender and
Genotype. 6 criteria (f0, fGpp(NH)p, DfGpp(NH)p, d0,
dGpp(NH)p and DdGpp(NH)p) have been compared
between sex by genotype and between genotype by
sex using Wilcoxon two-sample tests. Results are
Journal of Neurobiology. DOI 10.1002/neu
displayed in Table 3. f0 and d0 were not significantly
different between all groups of mice (*10 Hz and
*5 pA/pF in females and males of wild-type and 5-
HTT
/
genotype, respectively) but fGpp(NH)p was
significantly larger in 5-HTT
/
females than in
wild-type counterparts (p ¼ 0.017), while there were
no significant differences between 5-HTT
/
and
wild-type males. In addition, dGpp(NH)p and
DdGpp(NH)p (pA/pF) were significantly smaller in 5-
HTT
/
females than in wild-type counterparts (p ¼
0.040 and p ¼ 0.019, respectively). In contrast, these
Figure 5 Gpp(NH)p increases GIRK current in a DRN
neuron of a 5-HTT+/+ female. (a) Gpp(NH)p (400
M)
induces a progressive inhibition with hyperpolarization and
cessation of spike firing in current clamp mode. Barium
(2 mM) reverses Gpp(NH)p-induced changes in membrane
potential and firing of the cell. The arrowhead on the left
indicates
55 mV membrane potential. (b) Gpp(NH)p
(400
M) increases progressively the amplitude of the cur-
rent evoked by 5 mV increment voltage steps between
potentials from
45 to
145 mV, from a holding potential
of
60 mV, as shown at the beginning (l) and after 12 ()
and 17 (~) min of the recording shown in panel a. Note
that barium (n, 2 mM) suppresses the Gpp(NH)p effect.
Current traces at each voltage step are an average of 3
sweeps. (c) The barium-sensitive component of the control
current (D) is obtained by subtracting the traces during bar-
ium superfusion from traces obtained at the beginning of
the recording, as shown in panel b. The Gpp(NH)p differ-
ence current ($) is obtained by subtracting the traces at the
beginning of the recording from traces obtained after
17 min of Gpp(NH)p application. Note that inward relaxa-
tion at the beginning of the voltage steps occurs at poten-
tials more hyperpolarized than
90 mV and increases with
hyperpolarization (see arrow). Kinetics are best fitted to
two exponentials with time constants for an initial phase
(s1; 1–120 ms interval) and a late phase (s2; 120–1400 ms
interval) as indicated. During a voltage step to
125 mV,
s1 ¼ 18.8 ms and s2 ¼ 148.2 ms for the Gpp(NH)p differ-
ence current in this cell. Same calibration as in panel b. (d).
Steady state current/voltage relationship of the barium sen-
sitive component of the control current (D) and the
Gpp(NH)p difference current ($), as shown in panel c.
Note that the Gpp(NH)p difference current exhibits inward
rectification in the depolarized voltage range and reverses
polarity at
95 mV. (e) Spike discharge frequency (top)
and current density (bottom) as a function of time for the
recordings shown in panels a and b, respectively. Note the
blocking effect of barium. Such barium effect was reproduced
in 5 cells in females and 7 cells in males. (f). Time constants
s1 (!) and s2 (^) of inward relaxation as a function of volt-
age for the Gpp(NH)p difference current shown in panel c.
Note the increase of time constants with hyperpolarization.
Gpp(NH)p was present in the pipette since the beginning of
the recording, as in Fig. 6. [K+]ext ¼ 3.5 mM.

Figure 6 Gpp(NH)p increases GIRK current in a DRN
neuron of a 5-HTT
/
female. (a) Control and Gpp(NH)p-
induced inhibition after 20 min of dialysis, in current clamp
mode. Gpp(NH)p (400
M) produces hyperpolarization and
cessation of spike firing. Barium (2 mM) reverses
Gpp(NH)p-induced changes in membrane potential and fir-
ing of the cell. The arrowhead on the left indicates
55 mV
membrane potential. (b). Gpp(NH)p (400
M) increases the
amplitude of the current evoked by 5 mV increment voltage
steps between potentials from
45 to
145 mV, from a
holding potential of
60 mV, as shown at the beginning
(l) and after 20 min (~) of the recording shown in panel
a. Note that barium (2 mM, n) suppresses the Gpp(NH)p
effect. Current traces at each voltage step are an average of
3 sweeps. (c) Activation curves of steady state current
recorded in panel b are plotted against voltage. Note an
inward rectification at relatively depolarized voltages. (d).
Steady state current/voltage relationship of the Gpp(NH)p
difference current ($), as shown in panel c. Note that the
Gpp(NH)p difference current exhibits inward rectification
in the depolarized voltage range and reverses polarity at

92 mV. (e) Spike discharge frequency (top) and current
density (bottom) as a function of time for the recordings
shown in panels a and b, respectively. Note the blocking
effect of barium. Such barium effect was reproduced in 3
cells in females and 2 cells in males.
variables were not significantly different in males
between 5-HTT
/
and wild-type genotypes. Thus,
Gpp(NH)p effects decreased in females and did not
change in males with the 5-HTT
/
mutation. Fur-
GIRK Current in DRN of 5-HTT
/
Mice 1483
ther comparison between females and males indicates
that fGpp(NH)p and DfGpp(NH)p (Hz) are significantly
different in mutant mice, with fGpp(NH)p larger in
females than in males (p ¼ 0.020) and DfGpp(NH)p
smaller in females than in males (p ¼ 0.011).
Comparison between the Proportions of Gpp(NH)
p- and 5-HT1A Agonist-Sensitive Cells. In a logistic
regression analysis including three factors (drug type:
Gpp(NH)p or 5-HT1A agonists, with the proportions
of 5-CT- and 8-OH-DPAT-sensitive cells pooled to-
gether; gender; and genotype), the only significant
factor was the type of drug (p ¼ 0.013). When consid-
ering interactions, the interaction between drug and
gender was almost significant (p ¼ 0.051). Hence in
wild-type females, the percentage of Gpp(NH)p-
sensitive cells was significantly smaller than that of 5-
HT1A agonist-sensitive cells (p ¼ 0.0002, see Table
2). On the contrary, no significant differences
between Gpp(NH)p- and 5-HT1A agonist-sensitive
cells were observed in other groups (mutant females
and wild-type and mutant males).
Comparison between the Effects of Gpp(NH)p and
5-CT on GIRK Current. The mean values of current
density variation for 5-CT (2–3
M, Dd5-CT ¼ d5-CT

d0 pA/pF) and Gpp(NH)p were similar for each
group of mice (females and males of wild-type and 5-
HTT
/
genotype). Dd5-CT was 3.9 6 4.6 pA/pF (n
¼ 6) and 1.2 6 1.9 pA/pF (n ¼ 3) in females of wild-
type and 5-HTT
/
genotype, respectively, and 4.8
6 2.7 pA/pF (n ¼ 3) and 1.1 6 2.1 pA/pF (n ¼ 5) in
males of wild-type and 5-HTT
/
genotype, respec-
tively, not different from DdGpp(NH)p respective values
(see Table 3).
Gpp(NH)p NonSensitive Cells. Some cells in 5-
HTT+/+ (20/25 in females and 8/19 in males) and 5-
HTT
/
(10/18 in females and 7/14 in males) mice,
exhibited little or no change in spike discharge fre-
quency or the current elicited by hyperpolarizing volt-
age steps after *20–30 min Gpp(NH)p dialysis.
These neurons were considered as Gpp(NH)p nonsen-
sitive cells. Their spike discharge frequency and cur-
rent density did not differ significantly from corre-
sponding baseline values for Gpp(NH)p-sensitive
cells in all groups of mice (*10 Hz and *5 pA/pF in
females and males of wild-type and 5-HTT
/
geno-
type, respectively). Furthermore, superfusion of 5-CT
(30
M) did not change spike discharge frequency
and the current elicited by hyperpolarizing voltage
steps in neurons exhibiting no changes in these pa-
rameters after *20–30 min Gpp(NH)p dialysis
Journal of Neurobiology. DOI 10.1002/neu
1484 Loucif et al.

(7 cells in 5-HTT+/+ and 4 cells in 5-HTT
/
mice,

5.41 [4.64; 5.62]

2.08 [1.38; 3.48]

in frequency and increase in current density in all groups of mice after Gpp(NH)p dialysis. For each column, significant differences between pairs, with the Wilcoxon two-sample tests, are highlighted
Mean 6 SD, Median values and Q1–Q3 (interquatile intervals) of spike discharge frequency (f, Hz) and current density (d, pA/pF) at the beginning of the recordings (f0 and d0, respectively), after
*20–30 min Gpp(NH)p dialysis into cells (fGpp(NH)p and dGpp(NH)p, respectively), and their differences (DfGpp(NH)p and DdGpp(NH)p, respectively) are indicated for each group of mice. Note the decrease
3.93 [2.26; 5.32]

3.93 [1.50; 4.28]

DdGpp(NH)p

2.4 6 1.2a
not shown).

5.0 6 2.1

4.3 6 2.9

3.4 6 1.6
Table 3 Spike Discharge Frequency and Current Density in Gpp(NH)p-Inhibited Neurons in Females and Males of Wild-Type and 5-HTT
/
Genotype

DISCUSSION

5-HT1A agonists activated a GIRK current, hence in-

12.26 [8.85; 13.44]

6.63 [4.50; 9.62]

7.33 [5.96; 8.80]

7.03 [6.00; 9.94]

hibiting spike discharge, in neurons of the DRN
6.9 6 2.9a
11.6 6 3.5

8.3 6 3.7

7.5 6 2.6
dGpp(NH)p

recorded in brain stem slices from adult females

and males of wild-type and 5-HTT
/
genotype.
Gpp(NH)p mimicked these effects, but exhibited
lower efficacy in 5-HTT
/
females than in wild-
type counterparts, while no differences were observed
between males of both genotypes. These data suggest
alterations of the coupling between G-proteins and
6.64 [3.44; 8.70]

4.61 [2.70; 6.20]

3.70 [2.68; 5.00]

4.40 [2.50; 5.78]

GIRK channels, specifically in 5-HTT
/
females,

6.7 6 3.7

4.5 6 2.0

4.0 6 1.3

4.1 6 2.0

consistent with the existence of gender-related differ-

d0

ences in the efficacy of G-protein-dependent neuro-

transmission in the DRN during chronic 5-HT re-
uptake blockade.

11.00 [
15.00;
9.50]

8.25 [
10.00;
4.50]

7.75 [
9.25;
5.50]

7.5 [
11.5;
4.0]

5-HT1A-Receptor-Activated GIRK Current

7.3 6 2.4b

9.4 6 7.3

8.8 6 4.4

12.7 6 5.1
DfGpp(NH)p

The outward current underlying the 5-HT1A (8-OH-

DPAT or 5-CT)-agonist-induced hyperpolarization of
DRN neurons in females and males of wild-type and
5-HTT
/
genotype was identified as a GIRK cur-
rent because (1) it was dependent on intracellular
GTP concentration, (2) it was also activated by intra-
cellular Gpp(NH)p in the absence of agonists, exhibit-
1.75 [0.25; 3.75]

ing time dependency, with slow and fast inward relax-

0.0 [0.0; 1.50]
0.0 [0.0; 0.0]

0.0 [0.0; 0.0]

2.1 6 2.0a,b
fGpp(NH)p

ation, (3) it presented a reversal potential close to EK,

0.8 6 1.0

0.1 6 0.4
060

(4) it exhibited voltage sensitivity, as shown by acti-

vation at hyperpolarized potentials and inward rectifi-
cation at depolarized voltages, (5) it was blocked by
external barium (2 mM), a GIRK channel blocker, as
p < 0.05, different from wild-type mice of the same gender.
p < 0.05, different from male mice of the same genotype.

well as the Gpp(NH)p-difference current, but not ZD

11.00 [9.50; 15.00]
7.50 [4.00; 11.50]

9.75 [8.00; 10.75]

9.00 [6.20; 12.00]

7288, a Ih blocker (BoSmith et al., 1993; Isomoto

9.4 6 7.3

9.4 6 2.0

9.6 6 4.3

12.9 6 4.9

et al., 1997; Yamada et al., 1998). The 8-OH-DPAT-

or 5-CT-difference GIRK current was also blocked by
f0

WAY 100635, a specific 5-HT1A antagonist (Man-

noury la Cour et al., 2001). Such an increase in GIRK
current by 5-HT1A-receptor stimulation, described
here for the first time in the mouse, is consistent with
data obtained in the DRN of the rat (Aghajanian and
5
HTT
/
, n ¼ 8

5
HTT
/
, n ¼ 7
5
HTT+/+, n ¼ 11
,n¼5

Lakoski, 1984; Innis et al., 1988; Williams et al.,

1988; Penington et al., 1993; Beck et al., 2004).
+/+

In recordings for 1 h duration, but with GTP rather

5-HTT

than Gpp(NH)p in the pipette, action potential dis-

charge and current density remained stable. This sug-
gests firing decrease, hyperpolarization, and increase
by bold.

in current density observed during 5-HT1A agonists

b
a
#

superfusion and/or Gpp(NH)p dialysis into cells,

$

Journal of Neurobiology. DOI 10.1002/neu


together with their reversibility upon wash-out or
when barium or WAY100635 were added to the
superfusion medium, resulted from pharmacological
applications rather than dialysis of neurons.
A few recorded cells exhibited 5-HT1A agonist-
induced decrease in spike discharge frequency with-
out any changes in hyperpolarizing voltage steps-
evoked current density. 5-HT1A receptor activation
could inhibit these cells by other mechanisms, such as
a transient calcium current (Chen and Penington,
1996), or a network interaction with activity of the
recorded cells during agonist superfusion. By analogy
with the regulatory network implicating a medial pre-
frontal cortex loop projecting onto serotonergic neu-
rons located in the DRN (Celada et al., 2001), such a
mechanism might involve 5-HT1A agonist-induced
inhibition of excitatory interneurons directly imping-
ing onto the recorded cells that were probably 5-HT1A
agonist insensitive. Indeed, glutamatergic perikarya
(Clements et al., 1991) and 5-HT1A agonist-sensitive
nonserotonergic neurons (Liu et al., 2002; Kirby
et al., 2003) have already been described in the DRN
and could participate in this mechanism.
Gpp(NH)p-Induced Current
In Gpp(NH)p-inhibited neurons of the DRN, spike
discharge frequency and current density at the begin-
ning of the recordings were not significantly different
between all groups of mice (females and males of
wild-type and 5-HTT
/
genotype), nor from those
of Gpp(NH)p nonsensitive DRN neurons in these var-
ious groups of mice. This suggests that a large tonic
G-protein-, and hence GIRK current-activation,
which may have resulted from a large concentration
of extracellular serotonin caused by serotonin trans-
porter deficiency, does not occur under baseline
conditions in 5-HTT
/
mice. Indeed, extracellular
serotonin does not increase in DRN of chronically
SSRI-treated rats, another model of serotonin trans-
porter blockade (Bel and Artigas, 1993). Therefore,
Gpp(NH)p-induced effects observed in our prepara-
tion are very probably due to G-protein coupling to
GIRK channels in the plasma membrane of DRN neu-
rons, and not to increased basal serotonin binding to
receptors.
Surprisingly, no Gpp(NH)p-induced depolariza-
tion was observed, as would be expected from G-pro-
tein-coupled excitatory receptors in the DRN. This
can be explained by more powerful G-protein-medi-
ated inhibitions than excitations, even within a cell.
For instance, 5-HT2 receptor-mediated excitations are
detectable only after 5-HT1A receptor blockade in the
GIRK Current in DRN of 5-HTT
/
Mice 1485
DRN (Craven et al., 2001). Furthermore, in wild-type
females, the percentage of sensitive cells was lower
during Gpp(NH)p dialysis (20%) than during 5-HT1A
agonists superfusion (81%), although the only
described transduction mechanism in the literature for
5-HT1A receptor gating of the GIRK current involves
G proteins (Innis et al., 1988), while no differences
were observed in other groups of wild-type males and
mutant females and males. Such discrepancy in wild-
type females is puzzling, eventhough data obtained
here with 5-HT1A agonists are reported for the sake of
qualitative-, but not quantitative comparison with
Gpp(NH)p effects. This could result from the use of a
single dose of Gpp(NH)p, i.e. 0.4 mM, which may not
have been maximal to trigger GIRK current while 5-
HT1A agonists were used at maximal concentrations.
Also, several G-protein related mechanisms trigger-
ing GIRK current may be involved, including the 5-
HT1A receptor-dependent process. A large number of
accessory proteins involved in signal processing
through G proteins have been described (Sato et al.,
2006) and may be regulated differently during 5-
HT1A receptors activation. For example, in the hypo-
thalamus of female mice, it has been shown that estro-
gens can block a GIRK current through G-proteins
(Kelly et al., 2003). Hence, in neurons in the DRN of
female mice where estrogen receptors have been iden-
tified (Sheng et al., 2004), the estrogen transduction
pathway may counteract G protein-GIRK current cou-
pling triggered by Gpp(NH)p dialysis, more in wild-
type than in 5-HTT
/
mice (Bouali et al., 2003).
The slow and fast inward relaxations observed in
the Gpp(NH)p-difference current suggest involve-
ment of different GIRK subunits in DRN neurons,
consistent with the localization of several GIRK subu-
nits, namely GIRK 1, GIRK 2 and GIRK 3, but not
GIRK 4, subunits in the DRN of rats and mice
(Karschin et al., 1996; Chen et al., 1997; Murer et al.,
1997; Wickman et al., 2000). The slow relaxation
may reflect the involvement of a GIRK 1 subunit,
which possesses slow activation properties (Wisch-
meyer et al., 1997; Yamada et al., 1998). The barium-
blocked component of the control current observed in
our preparation may correspond to a basal activity of
GIRK channels (Rishal et al., 2005). Heterogeneity of
GIRK multimeric complexes could underlie involve-
ment of different GIRK currents triggered by various
physiological states of the animals. Identification of
these GIRK subunits in wild-type and 5-HTT
/

mice will need further experiments such as single cell
RT-PCR investigations.
We observed 58% (11 out of 19 recorded cells) of
Gpp(NH)p-sensitive neurons in male wild-type mice,
a result that can be compared to data in the literature
Journal of Neurobiology. DOI 10.1002/neu
1486 Loucif et al.
where most studies are performed in males. Knowing
that 60% of DRN neurons exhibit 5-HT1A autorecep-
tors in mice (Kirby et al., 2003) and rats (Sotelo et al.,
1990), and that a large majority (85%) of neurons
exhibiting serotonin transporter mRNA also coex-
press GABAB receptors (Serrats et al., 2003), it can
be hypothesized that in males, the pool of G-proteins
linked to 5-HT1A- and/or GABAB-receptors is similar
to the pool of G-proteins present within a cell and/or
within the DRN to trigger GIRK current. This is con-
firmed by the fact that 5-HT1A agonists had no addi-
tive effects on the Gpp(NH)p difference current and
Gpp(NH)p nonsensitive cells were also nonsensitive
to 5-HT1A agonists in our preparation.
G-Protein-Dependent Inhibition
of DRN Neurons Varies with Gender
and Genotype
The G-protein coupling to GIRK channels in the
plasma membrane of DRN neurons appeared less effi-
cient in females, but not in males, of 5-HTT
/
ge-
notype, than in wild-type counterparts. These data
suggest that a lower potency of neurotransmitters to
stimulate G-protein-related receptors in 5-HTT
/

mutants, as measured from changes in cell discharge
in response to activation of 5-HT1A (Mannoury la
Cour et al., 2001)- and GABAB (Mannoury la Cour
et al., 2004)-receptors, reflects alterations at all levels
of the transduction system in females, and only
upstream the G-protein-coupled GIRK current in
males. Actually, several factors can potentially con-
tribute to lower efficacy of these receptors’ signaling
in the DRN: (i) a 30–50% decrease in 5-HT1A recep-
tor density, as observed by autoradiography and
mRNA in situ hybridization (Fabre et al., 2000; Li
et al., 2000; Bouali et al., 2003; Mannoury la Cour
et al., 2004), (ii) a potential deficit in their coupling
with G-proteins, as evidenced by a decrease in 5-
HT1A- and GABAB-receptor agonist-evoked [35S]GTP-
-S binding (Fabre et al., 2000; Mannoury la Cour
et al., 2004), and (iii) qualitative and/or quantitative
changes in RGS proteins interacting with the functional
pool of G-proteins shared by 5-HT1A- and GABAB-
receptors and GIRK channels (De Vries et al., 2000).
In the present study, we provided evidence that altera-
tion in G protein-GIRK channel coupling does contrib-
ute in 5-HTT
/
female mice to adaptive functional
changes in 5-HT1A-mediated neurotransmission, specif-
ically.
Altogether, our data indicate that alterations in G-
protein-coupled receptors, and notably 5-HT1A- or
GABAB-receptors located (mainly if not exclusively;
see Sotelo et al., 1990) on serotonergic neurons, are
Journal of Neurobiology. DOI 10.1002/neu
more pronounced in females than in males of the 5-
HTT
/
genotype. Such data can be related to gender
differences previously observed in vivo, where 5-HT1A
autoreceptor stimulation could no longer inhibit DRN
neuronal firing in 5-HTT
/
females while it
remained partially effective in males (Bouali et al.,
2003). Downregulating factors at all levels of the 5-
HT1A-receptor transduction system in 5-HTT
/

females, and to a lesser extent in 5-HTT
/
males,
could explain the larger influence of the 5-HTT
/

knock-out mutation observed in females compared to
males.
This can be of importance knowing that depression
has been related to serotonergic neurotransmission
deficit possibly caused by inhibition of DRN neurons
through hypersensitive 5-HT1A autoreceptors (Blier
et al., 1998; El Yacoubi et al., 2003) as well as altera-
tions of GABA neurotransmission (Brambilla et al.,
2003; Tunnicliff and Malatynska, 2003). Hence, the
demonstration of gender-dependent changes in the
pool of Gpp(NH)p-inhibited DRN neurons can be par-
ticularly relevant since depression occurs with a two-
fold higher frequency in women than in men (Korn-
stein et al., 2000b; Piccinelli and Wilkinson 2000)
and that the response to SSRI treatment in depressed
patients also exhibits some gender-dependency
(Kornstein et al., 2000a).
In conclusion, the present in vitro electrophysio-
logical investigations demonstrated that the 5-HTT
/

knock-out mutation reduces the Gpp(NH)p-induced
coupling between G-proteins and GIRK channels in
females but not in males. These characteristics might
be relevant to gender-dependent molecular adap-
tive mechanisms affecting 5-HT neurotransmission
during chronic 5-HT reuptake blockade by SSRI anti-
depressants.
We are grateful to Dr. R. Miles for precious discussion
and comments on the manuscript, and to Dr. Lee Schechter
(Wyeth Res. Lab., Princeton) for generous gift of WAY
100635.
REFERENCES
Aghajanian GK, Lakoski JM. 1984. Hyperpolarization of
serotonergic neurons by serotonin and LSD: Studies in
brain slices showing increased K+ conductance. Brain
Res 305:181–185.
Andrade R, Malenka RC, Nicoll RA. 1986. A G protein
couples serotonin and GABAB receptors to the same
channels in hippocampus. Science 234:1261–1265.
Beck SG, Pan YZ, Akanwa C, Kirby L. 2004. Median and
dorsal raphe neurons are not electrophysiologically iden-
tical. J Neurophysiol 91:994–1005.

Bel N, Artigas F. 1993. Chronic treatment with fluvox-
amine increases extracellular serotonin in frontal cortex
but not in raphe nuclei. Synapse 15:243–245.
Bengel D, Murphy DL, Andrews AM, Wichems CH, Felt-
ner D, Heils A, Mossner R, et al. 1998. Altered brain
serotonin homeostasis and locomotor insensitivity to 3,
4-methylenedioxymethamphetamine (\ecstasy") in sero-
tonin transporter-deficient mice. Mol Pharmacol 53:649–
655.
Blier P, Pineyro G, El Mansari M, Bergeron R, De Mon-
tigny C. 1998. Role of somatodendritic 5-HT autorecep-
tors in modulating 5-HT neurotransmission. Ann NY
Acad Sci 861:204–216.
BoSmith RE, Briggs I, Sturgess NC. 1993. Inhibitory
actions of ZENECA ZD7288 on whole-cell hyperpolar-
ization activated inward current (If) in guinea-pig dissoci-
ated sinoatrial node cells. Br J Pharmacol 110:343–349.
Bouali S, Evrard A, Chastanet M, Lesch KP, Hamon M,
Adrien J. 2003. Sex hormone-dependent desensitization
of 5-HT1A autoreceptors in knock-out mice deficient in
the 5-HT transporter. Eur J Neurosci 18:2203–2212.
Brambilla P, Perez J, Barale F, Schettini G, Soares JC.
2003. GABAergic dysfunction in mood disorders. Mol
Psychiatry 8:721–737.
Brown RE, Sergeeva OA, Eriksson KS, Haas HL. 2002.
Convergent excitation of dorsal raphe serotonin neurons
by multiple arousal systems (orexin/hypocretin, hista-
mine and noradrenaline). J Neurosci 22:8850–8859.
Celada P, Puig MV, Casanovas JM, Guillazo G, Artigas F.
2001. Control of dorsal raphe serotonergic neurons by
the medial prefrontal cortex: Involvement of serotonin-
1A, GABAA, and glutamate receptors. J Neurosci 21:
9917–9929.
Chen SC, Ehrhard P, Goldowitz D, Smeyne RJ. 1997. De-
velopmental expression of the GIRK family of inwardly
rectifying potassium channels: Implications for abnor-
malities in the weaver mutant mouse. Brain Res 778:
251–264.
Chen Y, Penington NJ. 1996. Differential effects of protein
kinase C activation on 5-HT1A receptor coupling to Ca2+
and K+ currents in rat serotonergic neurons. J Physiol
(Lond) 496:129–137.
Clements JR, Toth DD, Highfield DA, Grant SJ. 1991. Glu-
tamate-like immunoreactivity is present within choliner-
gic neurons of the laterodorsal tegmental and pedunculo-
pontine nuclei. Adv Exp Med Biol 295:127–142.
Craven RM, Grahame-Smith DG, Newberry NR. 2001. 5-
HT1A and 5-HT2 receptors differentially regulate the
excitability of 5-HT-containing neurones of the guinea
pig dorsal raphe nucleus in vitro. Brain Res 899:159–
168.
Czachura JF, Rasmussen K. 2000. Effects of acute and
chronic administration of fluoxetine on the activity of
serotonergic neurons in the dorsal raphe nucleus of the
rat. Naunyn Schmiedeberg’s Arch Pharmacol 362:266–
275.
De Vries L, Zheng B, Fischer T, Elenko E, Farquhar MG.
2000. The regulator of G protein signalling family. Annu
Rev Pharmacol Toxicol 40:235–271.
GIRK Current in DRN of 5-HTT
/
Mice 1487
El Yacoubi M, Bouali S, Popa D, Naudon L, Leroux-Nicol-
let I, Hamon M, Costentin J, et al. 2003. Behavioral, neu-
rochemical and electrophysiological characterization of a
genetic mouse model of depression. Proc Natl Acad Sci
USA 100:6227–6232.
Fleiss JL. 1981. Statistical Methods for Rates and Propor-
tions. New York: Wiley. 24 pp.
Fabre V, Beaufour C, Evrard A, Rioux A, Hanoun N, Lesch
KP, Murphy DL, et al. 2000. Altered expression and
functions of serotonin 5-HT1A and 5-HT1B receptors in
knock-out mice lacking the 5-HT transporter. Eur J Neu-
rosci 12:2299–2310.
Franklin KBJ, Paxinos G. 1997. The Mouse Brain in Ste-
reotaxic Coordinates. San Diego: Academic Press.
Gobbi G, Murphy DL, Lesch KP, Blier P. 2001. Modifica-
tions of the serotoninergic system in mice lacking sero-
tonin transporters: An in vivo electrophysiological study.
J Pharmacol Exp Ther 296:987–995.
Graham D, Esnaud H, Habert E, Langer SZ. 1989. A
common binding site for tricyclic and nontricyclic 5-hy-
droxytryptamine uptake inhibitors at the substrate re-
cognition site of the neuronal sodium-dependent 5-hy-
droxytryptamine transporter. Biochem Pharmacol 38:
3819–3826.
Ikeda SR, Schofield GG. 1989. Somatostatin blocks a cal-
cium current in rat sympathetic ganglion neurones.
J Physiol (Lond) 409:221–240.
Innis RB, Nestler EJ, Aghajanian GK. 1988. Evidence for
G protein mediation of serotonin- and GABAB-induced
hyperpolarisation of rat dorsal raphe neurons. Brain Res
459:27–36.
Isomoto S, Kondo C, Kurachi Y. 1997. Inwardly rectifying
potassium channels: Their molecular heterogeneity and
function. Jpn J Physiol 47:11–39.
Karschin C, Dibmann E, Stühmer W, Karschin A. 1996.
IRK (1–3) and GIRK (1–4) inwardly rectifying K+ chan-
nel mRNAs are differentially expressed in the adult rat
brain. J Neurosci 16:3559–3570.
Kelly MJ, Qiu J, Ronnekleiv OK. 2003. Estrogen modula-
tion of G-protein-coupled receptor activation of potas-
sium channels in the central nervous system. Ann NY
Acad Sci 1007:6–16.
Kirby LG, Pernar L, Valentino RJ, Beck SG. 2003. Distin-
guishing characteristics of serotonin and non-serotonin-
containing cells in the dorsal raphe nucleus: Electrophys-
iological and immunohistochemical studies. Neuro-
science 116:669–683.
Kornstein SG, Schatzberg AF, Thase ME, Yonkers KA,
McCullough JP, Keitner GI, Gelenberg AJ, et al. 2000a.
Gender differences in treatment response to sertraline
versus imipramine in chronic depression. Am J Psychia-
try 157:1445–1452.
Kornstein SG, Schatzberg AF, Thase ME, Yonkers KA,
McCullough JP, Keitner GI, Gelenberg AJ, et al. 2000b.
Gender differences in chronic major and double depres-
sion. J Affect Disord 60:1–11.
Le Poul E, Boni C, Hanoun N, Laporte AM, Laaris N,
Chauveau J, Hamon M, et al. 2000. Differential adapta-
tion of brain 5-HT1A and 5-HT1B receptors and 5-HT
Journal of Neurobiology. DOI 10.1002/neu
1488 Loucif et al.
transporter in rats treated chronically with fluoxetine.
Neuropharmacology 39:110–122.
Le Poul E, Laaris N, Doucet E, Laporte AM, Hamon M,
Lanfumey L. 1995. Early desensitization of somato-den-
dritic 5-HT1A autoreceptors in rats treated with fluoxetine
or paroxetine. Naunyn Schmiedeberg’s Arch Pharmacol
352:141–148.
Lesch KP. 1997. Molecular biology, pharmacology, and
genetics of the serotonin transporter: Psychobiological
and clinical implications. In: Baumgarten HG, Göthert
M, editors. Serotoninergic Neurons and 5-HT Receptors
in the CNS (Handbook of Experimental Pharmacology,
Vol. 129), Berlin: Springer, pp 671–705.
Li Q, Wichems C, Heils A, Lesch KP. 2000. Reduction in
the density and expression, but not G-protein coupling,
of serotonin receptors (5-HT1A) in 5-HT transporter
knock-out mice: Gender and brain region differences.
J Neurosci 20:7888–7895.
Liu RJ, Van den Pol A, Aghajanian GK. 2002. Hypocretins
(orexins) regulate serotonin neurons in the dorsal raphe
nucleus by excitatory direct and inhibitory indirect
actions. J Neurosci 22:9453–9464.
Mannoury la Cour C, Boni C, Hanoun N, Lesch KP, Hamon
M, Lanfumey L. 2001. Functional consequences of 5-HT
transporter gene disruption on 5-HT1A receptor mediated-
regulation of dorsal raphe and hippocampal cell activity.
J Neurosci 21:2178–2185.
Mannoury la Cour C, Hanoun N, Melfort M, Hen R, Lesch
KP, Hamon M, Lanfumey L. 2004. GABAB receptors in
5-HT transporter- and 5-HT1A receptor-knock-out mice:
Further evidence of a transduction pathway shared with
5-HT1A receptors. J Neurochem 89:886–896.
Mechaly I, Scamps F, Chabbert C, Sans A, Valmier J.
2005. Molecular diversity of voltage-gated sodium chan-
nel  subunits expressed in neuronal and non-neuronal
excitable cells. Neuroscience 130:389–396.
Murer G, Adelbrecht C, Lauritzen I, Lesage F, Lazdunski
M, Agid Y, Raisman-Vozari R. 1997. An immunocyto-
chemical study on the distribution of two G protein-gated
inward rectifier potassium K+ channels (GIRK2 and
GIRK4) in the adult rat brain. Neuroscience 80:345–357.
Pan ZZ, Grudt TJ, Williams JT. 1994.  1-adrenoceptors in
rat dorsal raphe neurons: Regulation of two potassium
conductances. J Physiol (Lond) 478:437–447.
Penington NJ, Kelly JS, Fox AP. 1993. Whole-cell record-
ing of inwardly rectifying K+ currents activated by 5-
HT1A receptors on dorsal raphe neurons of the adult rat.
J Physiol (Lond) 469:387–405.
Journal of Neurobiology. DOI 10.1002/neu
Piccinelli M, Wilkinson G. 2000. Gender differences in
depression. Critical review. Br J Psychiatry 177:486–
492.
Rishal I, Porozov Y, Yakubovich D, Varon D, Dascal N.
2005. G-dependent and G-independent basal activ-
ity of G protein-activated K+ channels. J Biol Chem
280:16685–16694.
Sato M, Blumer JB, Simon V, Lanier SM. 2006. Accessory
proteins for G proteins: Partners in signalling. Annu Rev
Pharmacol Toxicol 46:151–187.
Serrats J, Artigas F, Mengod G, Cortes R. 2003. GABAB re-
ceptor mRNA in the raphe nuclei: Co-expression with se-
rotonin transporter and glutamic acid decarboxylase.
J Neurochem 84:743–752.
Sheng Z, Kawano J, Yanai A, Fujinaga R, Tanaka M, Wata-
nabe Y, Shinoda K. 2004. Expression of estrogen recep-
tors (, ) and androgen receptor in serotonin neurons of
the rat and mouse dorsal raphe nuclei; sex and species
differences. Neurosci Res 49:185–196.
Sotelo C, Cholley B, El Mestikawy S, Gozlan H, Hamon
M. 1990. Direct immunohistochemical evidence of the
existence of 5-HT1A autoreceptors on serotoninergic neu-
rons in the midbrain raphe nuclei. Eur J Neurosci 2:
1144–1154.
Tunnicliff G, Malatynska E. 2003. Central GABAergic sys-
tems and depressive illness. Neurochem Res 28:965–976.
VanderMaelen CP, Aghajanian GK. 1983. Electrophysio-
logical and pharmacological characterization of seroto-
nergic dorsal raphe neurons recorded extracellularly and
intracellularly in rat brain slices. Brain Res 289:109–119.
Verheugen JAH, Fricker D, Miles R. 1999. Noninvasive
measurements of the membrane potential and GABAer-
gic action in hippocampal interneurons. J Neurosci 19:
2546–2555.
Wickman K, Karschin C, Karschin A, Picciotto MR, Clap-
ham DE. 2000. Brain localization and behavioral impact
of the G-protein-gated K+ channel subunit GIRK4.
J Neurosci 20:5608–5615.
Williams JT, Colmers WF, Pan ZZ. 1988. Voltage and
ligand-activated inwardly rectifying currents in dorsal
raphe neurons in vitro. J Neurosci 89:3499–3506.
Wischmeyer E, Döring F, Wischmeyer E, Spauschus A,
Thomizig A, Veh R, Karschin A. 1997. Subunit interac-
tions in the assembly of neuronal Kir3.0 inwardly rectify-
ing K+ channels. Mol Cell Neurosci 9:194–206.
Yamada M, Inanobe A, Kurachi YG. 1998. Protein regula-
tion of potassium ion channels. Pharmacol Rev 50:723–
760.