Neuroscience 228 (2013) 36–46

LITHIUM AUGMENTATION OF THE EFFECTS OF DESIPRAMINE IN A

MOUSE MODEL OF TREATMENT-RESISTANT DEPRESSION: A ROLE
FOR HIPPOCAMPAL CELL PROLIFERATION
O. F. O’LEARY, a* S. ZANDY, a T. G. DINAN b,c AND Key words: treatment-resistant depression, lithium, antide-
J. F. CRYAN a,b* pressant, augmentation, hippocampal neurogenesis, behav-
a
iour.
Department of Anatomy and Neuroscience, University College
Cork, Ireland
b
Alimentary Pharmabiotic Centre, University College Cork, Ireland
c
INTRODUCTION
Department of Psychiatry, University College Cork, Ireland
Approximately 50% of patients with a major depressive
episode fail to achieve remission with first-line
Abstract—Approximately 50% of patients with a major antidepressant treatment (Trivedi et al., 2006). Second-
depressive episode fail to achieve remission with first-line
line treatment strategies for such patients include: dose
antidepressant treatments. Second-line treatment strategies
for such patients include lithium augmentation of antide-
optimisation; switching to another antidepressant; or
pressants, particularly with tricyclic antidepressants. The augmentation with an antipsychotic drug, lithium or
neurobiological mechanisms underlying the therapeutic thyroid hormone (Carvalho et al., 2007). Most of the
effects of lithium augmentation are not yet fully understood. supporting evidence for the clinical efficacy of
Unravelling these mechanisms could aid the development of augmentation strategies comes from lithium studies
more effective antidepressant drugs. In the present study, (Bauer et al., 2010). The neurobiological mechanisms
we investigated whether chronic treatment with a combina- underlying the therapeutic effects of lithium
tion of lithium and the tricyclic antidepressant, desipramine, augmentation are not yet fully understood although
could produce antidepressant-like behaviour in a mouse there is evidence that serotonin may play a role
strain (BALB/cOLaHsd) that exhibited reduced sensitivity
(Haddjeri et al., 2000; Wegener et al., 2003).
to the behavioural effects of chronic desipramine treatment
in the novelty-induced hypophagia test. Since chronic treat-
Unravelling the precise mechanisms underlying the
ment with antidepressant drugs increases the proliferation efficacy of lithium augmentation could prove fruitful for
of newly-born cells in the hippocampus, and hippocampal the development of new more effective antidepressant
cell proliferation is required for the behavioural effects of drugs thus also avoiding complications associated with
at least some antidepressants in neohypophagia tests, the polypharmacy.
present study also investigated whether lithium plus desi- Animal models are powerful tools for investigating
pramine increased cell proliferation in the hippocampus. such mechanisms but yet, few studies have investigated
Chronic treatment with lithium plus desipramine but neither whether lithium augments the behavioural effects of
drug alone, induced antidepressant-like behaviour and antidepressants in animal models (Nixon et al., 1994;
increased hippocampal cell proliferation, thus suggesting
Redrobe et al., 1998). To the best of our knowledge,
that increased hippocampal cell proliferation may be a
mechanism underlying lithium augmentation of antidepres- animal studies investigating the behavioural effects of
sants. Moreover, since BALB/cOLaHsd mice respond to lith- chronic lithium augmentation of chronic antidepressant
ium plus desipramine but not to either drug alone, they may treatment have not yet been conducted. Therefore, we
become useful in the development of a mouse model of investigated whether chronic treatment with a
treatment-refractory depression for which there is an unmet combination of lithium and the tricyclic antidepressant,
need. Ó 2012 IBRO. Published by Elsevier Ltd. All rights desipramine, could produce antidepressant-like
reserved. behaviour in a mouse strain (BALB/cOLaHsd) that
exhibits reduced sensitivity to the behavioural effects of
chronic desipramine treatment in the novelty-induced
hypophagia test (NIH). The NIH test is a behavioural
*Corresponding authors. Address: Department of Anatomy and test of anxiety that is sensitive to chronic but not acute
Neuroscience, University College Cork, Ireland. Tel: +353-21-420- antidepressant treatment and thus is used to assess
5480; fax: +353-21-427-3518 (O. F. O’Leary), tel: +353-21-420- behavioural changes following chronic antidepressant
5426; fax: +353-21-427-3518 (J. F. Cryan).
E-mail addresses: o.oleary@ucc.ie (O. F. O’Leary), j.cryan@ucc.ie
treatment (Dulawa and Hen, 2005). Since the
(J. F. Cryan). behavioural effects of chronic antidepressant treatment
Abbreviations: BDNF, brain-derived neurotrophic factor; GCL, granule in hyponeophgia tests can be dependent upon
cell layer; GSK3, glycogen synthase kinase 3; NGS, normal goat hippocampal cell proliferation in some mouse strains
serum; NIH, novelty-induced hypophagia; qRT-PCR, Quantitative Real
Time-PCR; SGZ, subgranular zone; VEGF, vascular endothelial (Santarelli et al., 2003; Wang et al., 2008; David et al.,
growth factor. 2009; Petrik et al., 2012), and chronic antidepressant

0306-4522/12 $36.00 Ó 2012 IBRO. Published by Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.neuroscience.2012.09.072
36
 O. F. O’Leary et al. / Neuroscience 228 (2013) 36–46 37

Fig. 1. A combination of lithium plus desipramine but neither drug alone increases antidepressant-like behaviour and hippocampal cell proliferation
in the BALB/cOLaHsd mouse strain. (A) Experimental design: two separate cohorts of mice were used for the behavioural (n = 9–10 per treatment
group) and hippocampal cell proliferation experiments (n = 6 per treatment group). (B) In the novelty-induced hypohagia (NIH) test, a combination
of lithium plus desipramine but neither drug alone decreased the latency (latency in novel cage minus latency in home cage) to drink the milk, thus
indicating an antidepressant effect. (C) A combination of lithium plus desipramine but neither drug alone decreased defecation in the novel cage,
thus indicating a reduction in an autonomic stress response. (D and E) A combination of lithium plus desipramine but neither drug alone increased
cell proliferation in the dentate gyrus of the hippocampus with this effect occurring specifically in the subgranular zone (SGZ) and granule cell layer
(GCL) but not in the hilus. Arrows indicate examples of cells or clusters of cells that are BrdU-positive. ⁄Significantly different to the control diet-
vehicle group according to Fisher’s PLSD ⁄p < 0.05, ⁄⁄p < 0.01.

treatment can increase hippocampal cell proliferation and Experimental design and drug treatment
neurogenesis in depressed humans (Boldrini et al., 2009;
Lucassen et al., 2010), we also examined the effects of Three separate cohorts of male 9-week-old BALB/cOLaHsd mice
combination treatment on hippocampal cell proliferation were fed a lithium-supplemented diet (0.2% LiCl; Teklad, Harlan)
or control diet as previously described (O’Leary et al., 2010) and
in these mice.
received daily injections of either desipramine (10 mg/kg
(calculated as base weight), 10 ml/kg, i.p.) or saline for 21 days
(Fig. 1A). Similar doses of desipramine have previously been
reported to have antidepressant-like behavioural effects and
EXPERIMENTAL PROCEDURES increase hippocampal cell proliferation in the mouse brain
(Crowley et al., 2004; Gur et al., 2007; O’Leary et al., 2007).
Subjects The dose of lithium was chosen based on previous reports that
similar treatment regimens achieved serum lithium levels of
Nine-week-old male BALB/cOLaHsd mice (Harlan, UK) were 0.75–0.97 mM in mice (Chen et al., 2000; Riadh et al.,
housed under controlled conditions (temperature 20–21 °C, 2011) which is within the therapeutic concentration range
55–60% humidity) on a 12 h light/dark cycle and provided with (0.6–1.2 mM) in humans. Mice were also given free access to a
chow and water ad libitum. Mice were acclimated to housing second drinking bottle containing 0.89% NaCl to prevent
conditions for one week prior to experimental treatment. lithium-induced ion imbalance (Gould et al., 2008).
Experiments were conducted in accordance with the European One day following the last drug treatment, the behaviour of
Directive 86/609/EEC and the Recommendation 2007/526/65/ one cohort of mice was examined using the novelty-induced
EC, and were approved by the Animal Experimentation Ethics hypophagia test (Expt. 1; Fig. 1A), another cohort of mice were
Committee of the University College Cork. All efforts were injected with BrdU (75 mg/kg, i.p) every 2 h over an 8-h period
made to minimise the number of animals used and their to label newly-born cells as previously described (Wu and
suffering. Castren, 2009; O’Leary et al., 2012) (Expt. 2; Fig. 1A), and a
38 O. F. O’Leary et al. / Neuroscience 228 (2013) 36–46

third cohort of mice were culled for the removal of both (90 min at RT; VectastainÒ Elite ABC kit, Vector Laboratories)
hippocampi for the analysis of vascular endothelial growth followed by incubation in ABC reagent (VectastainÒ Elite ABC
factor (VEGF) mRNA, brain-derived neurotrophic factor (BDNF) kit, Vector Laboratories). 3,30 -Diaminobenzidine (0.02%, Sigma)
mRNA, phospho-glycogen synthase kinase (GSK)3a(ser21), was used as a chromogen with 0.0075% H2O2 in PBS.
phospho-GSK3b(ser9), total GSK3a and total GSK3b protein BrdU-positive cells were counted by an investigator blind to
levels. experimental treatment. Cells were counted in every 6th section
of the hippocampus from AP:0.94 to AP:3.80. For each
animal, analysed sections were in approximately the same A–P
Novelty-induced hypophagia test (NIH; Expt. 1)
plane, thus limiting the influence of differences in area. On
average, 12–13 sections were analysed per mouse (n = 6 per
The NIH test is a behavioural test of anxiety sensitive to chronic group), and the number was multiplied by six to estimate the
but not acute antidepressant treatment (Dulawa and Hen, 2005). total cell number per brain.
This test measures hyponeophagia, the inhibition of feeding by
exposure to a novel and anxiety-provoking environment. The
test was conducted essentially as described by Dulawa and Quantitative Real Time-PCR (qRT-PCR) of BDNF and
Hen (2005) whereby mice are trained to drink a highly VEGF mRNA levels in the hippocampus
palatable solution of diluted sweetened condensed milk (1:3,
milk:water) in their home cage for 30 min per day over 3 days
(days 18, 19 and 20 of Expt. 1). On the fourth day (day 21 of Various growth factors including BDNF and VEGF regulate the
Expt. 1), mice are brought to a quiet and dimly-lit (40 lux) room proliferation and survival of newly-born cells in the adult
where they remain undisturbed for 1.5 h prior to assessing their hippocampus and are increased in the brain by antidepressant
latency to drink the milk in their home cage. Mice rapidly treatments as well as lithium (Warner-Schmidt and Duman,
approach and drink the milk in their home cage under these 2007; O’Leary and Castren, 2010). Therefore, we investigated
conditions. However, 1 day later (day 22 of Expt. 1) when both whether lithium plus desipramine produced synergistic
increases in hippocampal mRNA expression of BDNF and
the milk and the mouse are placed in a new brightly-lit and
VEGFA, the predominant form of VEGF in the nervous system.
unfamiliar environment (clean cage without bedding, 1200 lux),
One day following the last drug treatment, the right
they exhibit a longer latency to drink the milk. Chronic but not
hippocampus was rapidly removed and immediately placed in
acute treatment with antidepressant drugs can reduce the
RNAlaterÒ (Applied Biosystems) for 24 h at 4 °C prior to
latency to drink the milk in the novel cage (Dulawa and Hen,
2005). To account for potential inter-individual differences in storage at 80 °C until further processing. Total RNA was
palatability or learning to drink the milk, the latency to drink the isolated using the miRNA isolation kit as per manufacturer’s
milk in the home cage is subtracted from the latency to drink in instructions (Ambion) and isolated RNA was stored at 80 °C
the novel cage (Dulawa and Hen, 2005). It has been reported until use. RNA concentrations were quantified using the ND-
that the effects of chronic antidepressant treatment in similar 1000 spectrophotometer (NanoDropÒ) and the quality was
tests of hyponeophagia can be dependent upon hippocampal assessed using the Agilent 2100 Bioanalyzer. Only samples
cell proliferation and neurogenesis in at least some mouse with an RIN of >8 were used.
strains (Santarelli et al., 2003; Wang et al., 2008; David et al., RNA was reverse transcribed to cDNA using random primers
2009) and thus behavioural changes in this test in response to on Applied Biosystem’s GeneAmp PCR System 9700. qRT-PCR
antidepressant treatments may be correlated with changes in of cDNA was conducted using probes (6 carboxy fluorescein –
the proliferation and survival of newly-born cells in the FAM) made by Applied Biosystems for VEGFA
hippocampus. Given the role of hippocampal cell proliferation (Mm01281449_m1) and total BDNF (exon IX; (Aid et al., 2007);
and neurogenesis in antidepressant activity in neohypophagia FAM probe: CCTGCAGCCTTCCTTG; Forward: GACCAAGTG
TAATCCCATGGGTTA; Reverse: GTTCCAGTGCCTTTTGTC
(Santarelli et al., 2003) and the responsivity of the NIH to
TATGC). RT-PCR was conducted using the ABI7300 Real
chronic but not acute antidepressant drug treatment (Dulawa
Time PCR machine (Applied Biosystems, Warrington, UK).
and Hen, 2005), we deemed the NIH test the most appropriate
Samples were heated to 95 °C for 10 min, and then subjected
model to test our hypothesis that antidepressant-like behaviour
to 40 cycles of amplification by melting at 95 °C and annealing
induced by chronic co-administration of lithium plus
desipramine would be correlated with increased cell at 60 °C for 1 min. Experimental samples were run in triplicate
proliferation in the adult hippocampus. with 1 ll cDNA per reaction. Data were normalised using
b-actin and transformed using the 2DDCt method followed by
comparison to the control-vehicle group (Livak and Schmittgen,
Assessment of cell proliferation in the hippocampus 2001).
(Expt. 2)

A separate cohort of mice were injected with BrdU on day 22 and Western blotting of phospho-GSK3a(ser21) and
transcardially perfused 24 h following the last BrdU injection. phospho-GSK3b(ser9) in the hippocampus
BrdU is a thymidine analogue that becomes incorporated into
the DNA during the S-phase of the cell cycle and is used as a Glycogen synthase kinase 3 (GSK3) is a constitutively active
marker of mitosis and cell proliferation. Cell proliferation was serine/threonine protein kinase that is indirectly inhibited by
analysed using BrdU immunohistochemistry of free-floating phosphorylation at serine-9(ser9) of its beta isoform, or serine-
30-lm sections as described previously (O’Leary et al., 2012; 21(ser21) of its alpha isoform (Li and Jope, 2010). Inhibition of
Felice et al., 2012). Briefly, sections were washed in PBS (pH GSK-3 has been suggested to be an important mechanism of
7.4) and heated in 0.01 M sodium citrate buffer in a waterbath action of lithium (Klein and Melton, 1996; O’Brien et al., 2011).
set at 85 °C (20 min). Sections were cooled to room In addition, inhibition of this enzyme has antidepressant-like
temperature followed by DNA denaturation in 2 N HCl (30 min, behavioural effects in rodents (Gould et al., 2004; Kaidanovich-
37 °C). Sections were neutralized in 0.1 M borate buffer, pH 8.5 Beilin et al., 2009). Furthermore, lithium and some
(2  5 min) and pre-treated with 0.75% H2O2 to block antidepressant drugs have been reported to increase phospho-
endogenous peroxidase activity. Non-specific binding was GSK3a(ser21) and phospho-GSK3b(ser9) levels in the
blocked (10% normal goat serum (NGS), 0.1% Triton-X 100, in hippocampus (Li et al., 2004; Eom and Jope, 2009). Moreover,
PBS, 1 h at RT) and sections were incubated in anti-Rat BrdU it was recently reported that phosphorylation of these serine
(Abcam; 1:100 in 1% NGS, 0.1% Triton-X 100, PBS; overnight; residues is required for lithium plus fluoxetine-induced
4 °C). Sections were washed, incubated in biotinylated anti-rat increases in hippocampal cell proliferation in adult mice
 O. F. O’Leary et al. / Neuroscience 228 (2013) 36–46 39

(Eom and Jope, 2009). Thus, increased concentrations of lithium nor desipramine treatment alone altered latency to
phospho-GSK3a(ser21) and phospho-GSK3b(ser9) in the drink (p = 0.79; p = 0.42, respectively), but a
hippocampus may represent important mechanisms through
combination of both drug treatments significantly reduced
which lithium plus desipramine induced hippocampal cell
proliferation and behavioural effects in the NIH. To test this latency to drink the milk when compared to control mice
hypothesis, we employed western blotting to measure changes (p = 0.02) thus suggesting an antidepressant-like
in phospho-GSK3a(ser21) and phospho-GSK3b(ser9) concen- behavioural effect. In addition, a combination of both
trations in the hippocampus. drugs but neither drug alone significantly reduced stress-
One day following the last drug treatment, the left induced defecation in the novel cage (Fig. 1C).
hippocampus was rapidly removed on an ice-cold petri dish, Specifically, two-way ANOVA revealed a significant main
frozen on dry ice, and stored at 80 °C until use. Using a
hand-held motorised homogeniser, each hippocampus was
effect for lithium (F(1, 33) = 6.55, p = 0.015) but no
homogenised in 300 ll of NP-40 lysis buffer (137 mM NaCl, effect of desipramine or a lithium  desipramine
20 mM Tris–HCl, pH 8.0, 1% NP-40 (v/v), 10% glycerol (v/v), interaction. However, post hoc analysis revealed that
48 mM NaF, H2O, 1 complete protease inhibitor mix (Roche) neither lithium nor desipramine treatment alone affected
and 2 mM Na3VO4), incubated on ice for 15 min and stress-induced defecation (p = 0.24; p = 0.90,
centrifuged at 12,000g for 15 min at 4 °C. The supernatants respectively), but a combination of both drug treatments
were removed and their total protein content was measured
significantly reduced defecation in the novel cage
using the QubitÒ Flourometer system (Invitrogen). Protein
extracts were stored at 80 °C until use. (p = 0.03; n = 9–10 per group).
Protein extracts were diluted with 4 sample buffer, heated
to 95 °C for 5 min and cooled on ice. Thirty micrograms of total Lithium in combination with desipramine but neither
protein from each sample was loaded onto an 8% SDS–PAGE
drug alone increases hippocampal cell proliferation
gels with a 5% stacking gel. Separated proteins were blotted
from gel onto polyvinylidene difluoride (PVDF) membranes. The effects of lithium plus desipramine on hippocampal
Membranes were blocked with 3% bovine serum albumin in
cell proliferation are illustrated in Fig. 1D and E. In the
TBS with 0.1% Tween-20 (TBST) for 2 h at room temperature
and then incubated with either rabbit anti-pGSK3a(ser21) subgranular zone (SGZ), two-way ANOVA revealed a
(1:300; cell signalling) or rabbit anti-pGSK3b(ser9) (1:100; cell significant main effect for desipramine (F(1, 20) = 5.405,
signalling) in 5% BSA–TBST, overnight at 4 °C with gentle p = 0.03) but not for lithium (F(1, 20) = 2.389,
agitation. Membranes were washed with TBST (4  15 min), p = 0.14) or lithium  desipramine (F(1, 20) = 1.565,
incubated with HRP-conjugated anti-rabbit IgG (1:10,000) in 3% p = 0.225). However, post hoc analysis revealed that
BSA–TBST for 1 h at room temperature. Protein bands were neither desipramine (p = 0.84) nor lithium alone
visualised using an ECL reagent (Pierce, Rockford, IL, USA)
(p = 0.46) altered cell proliferation, while a combination
and then exposed and photographed in a luminescent image
analyser (Fujifilm, Bedford, UK). Membranes were stripped, of both drugs significantly increased cell proliferation in
blocked and incubated with rabbit anti-GSK3a/b (1:1000, the SGZ (p = 0.013).
Upstate) in 3% BSA, overnight at 4 °C with gentle agitation. In the granule cell layer (GCL), two-way ANOVA
Following visualisation of protein bands as described above, revealed a significant main effect for desipramine
membranes were stripped, blocked and incubated with mouse (F(1, 20) = 8.219, p = 0.01) but not for lithium
anti-b-actin (1:5000, Sigma) for 2 h at room temperature.
(F(1, 20) = 2.119, p = 0.161) or lithium  desipramine
Following washes in TBST, membranes were incubated with
HRP-conjugated anti-mouse IgG (1:10,000) in 3% BSA–TBST (F(1, 20) = 0.427, p = 0.521). However, post hoc
for 1 h at room temperature. Optical densities of protein bands analysis revealed that neither desipramine (p = 0.58)
were measured using Multi Gauge software (Fujifilm). nor lithium alone (p = 0.13) altered cell proliferation but
Measurements of phospho-GSK3 isoforms were normalised a combination of both drugs significantly increased cell
against the total amount of the respective GSK3 isoform proliferation in the GCL (p = 0.006).
followed by normalisation with b-actin which was used as a
In the hilus, neither desipramine (F(1, 20) = 0.49,
loading control. Data are expressed as % change from control-
vehicle group. p = 0.49), lithium (F(1, 20) = 0.31, p = 0.583) nor
lithium plus desipramine (F(1, 20) = 2.49, p = 0.13)
affected cell proliferation.
Statistical analysis In the total dentate gyrus (SGZ plus GCL plus hilus),
two-way ANOVA revealed a significant main effect for
Statistical significance was determined using two-way ANOVA desipramine (F(1, 20) = 6.50, p = 0.019) but not for
followed by Fisher’s PLSD post hoc test using SPSS software.
lithium (F(1, 20) = 2.848, p = 0.11) or lithium 
For all comparisons, p < 0.05 was the criterion for statistical
significance. desipramine (F(1, 20) = 1.378, p = 0.254). However,
post hoc analysis revealed that neither desipramine
(p = 0.72) nor lithium (p = 0.34) treatment alone
RESULTS altered cell proliferation but a combination of both drugs
Lithium in combination with desipramine but neither significantly increased cell proliferation in the total
drug alone reduces neohypophagia and stress- dentate gyrus (p = 0.007).
induced defecation in the NIH test
Co-administration of lithium and desipramine does
In the NIH test (Fig. 1B), two-way ANOVA revealed a
not increase mRNA expression of the neurotrophins
significant main effect for desipramine (F(1, 33) = 4.32,
VEGF and BDNF in the hippocampus
p < 0.05) but no effect of lithium or a
lithium  desipramine interaction on latency to drink the VEGF and BDNF are important mediators of
milk. However, post hoc analysis revealed that neither antidepressant drug action and play a role in the
40 O. F. O’Leary et al. / Neuroscience 228 (2013) 36–46

proliferation and survival of newly-born cells in the proliferation. The effects of lithium plus desipramine on
hippocampus (O’Leary and Castren, 2012). Therefore, mRNA expression of VEGF and BDNF in the
we investigated whether a combination of lithium and hippocampus are illustrated in Fig. 2.
desipramine would produce synergistic increases in the Neither desipramine alone (F(1, 24) = 0.283,
expression of these neurotrophins and thus could p = 0.599), lithium alone (F(1, 24) = 3.683, p = 0.069)
account for the observed increase in hippocampal cell nor a combination of lithium with desipramine

Fig. 2. The effects of lithium, desipramine and co-administration of lithium plus desipramine on VEGF mRNA, BDNF mRNA, phospho-
GSK3a(ser21) and phospho-GSK3b(ser9) in the hippocampus. (A) Neither lithium, desipramine nor co-administration of lithium plus desipramine
altered VEGF mRNA levels in the hippocampus (n = 7 per treatment group). (B) BDNF mRNA expression in the hippocampus was decreased by
desipramine treatment and the co-administration of lithium plus desipramine; lithium treatment did not alter BDNF mRNA expression in the
hippocampus (n = 7–9 per treatment group). (C) Neither lithium, desipramine nor co-administration of lithium plus desipramine altered phospho-
GSK3a protein expression in the hippocampus (n = 7–10 per treatment group). (D) Neither lithium, desipramine nor co-administration of lithium
plus desipramine altered phospho-GSK3b protein expression in the hippocampus (n = 8–10 per treatment group). ⁄⁄Significantly different to the
control diet-vehicle group according to Fisher’s PLSD, p < 0.01. ##Significantly different to the lithium diet-vehicle group according to Fisher’s
PLSD, p < 0.01.
 O. F. O’Leary et al. / Neuroscience 228 (2013) 36–46 41

(F(1, 24) = 0.41, p = 0.528) significantly altered the DISCUSSION

expression of VEGF mRNA in the hippocampus
(Fig. 2A). In the present study, a combination of lithium plus
Two-way ANOVA revealed that lithium alone had no desipramine but neither drug alone induced
effect on BDNF mRNA expression in the hippocampus antidepressant-like behaviour and increased
(F(1, 28) = 0.02, p = 0.888; Fig. 2B). Desipramine hippocampal cell proliferation in BALB/cOLaHsd mice.
treatment alone significantly decreased BDNF mRNA To our knowledge, this is the first study to demonstrate
expression in the hippocampus (F(1, 28) = 20.144, that lithium can augment both the behavioural and
p < 0.001) but the addition of lithium did not alter this cellular effects of chronic treatment with an
effect (no lithium  desipramine interaction F(1, 28) = antidepressant drug in an animal model. In agreement
0.036, p = 0.85). with our findings, it was recently reported that chronic
treatment with both lithium and the antidepressant
imipramine, but neither drug alone prevented ACTH-
induced reductions in cell proliferation in the rat
Co-administration of lithium and desipramine does
hippocampus (Kitamura et al., 2011), however whether
not produce synergistic increases in phospho-
these changes in hippocampal cell proliferation
GSK3a(ser21) or phospho-GSK3b(ser9) in the
correlated with antidepressant-like behaviour was not
hippocampus
assessed. Other treatment strategies for treatment-
Lithium and some antidepressant drugs have been resistant depression have also been reported to
reported to increase inhibitory phosphorylation of increase hippocampal cell proliferation and/or
GSK3a and GSK3b at ser21 and ser9, respectively hippocampal neurogenesis including deep-brain
(Li et al., 2004; Eom and Jope, 2009). Moreover, it was stimulation (Encinas et al., 2011), vagus nerve
recently reported that phosphorylation of these serine stimulation (Revesz et al., 2008) and electroconvulsive
residues is required for lithium plus fluoxetine-induced therapy (Segi-Nishida et al., 2008). Taken together,
increases in hippocampal cell proliferation in adult mice these data suggest that enhanced hippocampal cell
(Eom and Jope, 2009). In addition, inhibition of GSK-3 proliferation and/or hippocampal neurogenesis may play
has antidepressant-like behavioural effects (Gould a role in the antidepressant-enhancing effects of lithium
et al., 2004) and GSK-3 mediates some of the as well as other treatment options for treatment-
behavioural effects of lithium in rodents (O’Brien et al., resistant depression.
2011). Therefore, we investigated whether a combi- Mechanisms other than increased hippocampal cell
nation of lithium and desipramine would produce proliferation have also been suggested to play a role in
synergistic increases in inhibitory phosphorylation of the antidepressant-enhancing effects of lithium. Both
GSK3a and GSK3b which might then account for the preclinical and clinical studies have suggested that the
observed increase in hippocampal cell proliferation and serotonergic system may play a role. Electrophy-
reduced neohypohagia in the NIH. The effects of siological studies in rats have reported that short-term
lithium plus desipramine on protein expression of lithium treatment (3 days) in combination with chronic
phospho-GSK3a(ser21) and phospho-GSK3b(ser9) in antidepressant treatment increases tonic activation of
the hippocampus are illustrated in Fig. 2. forebrain 5-HT1A receptors (Haddjeri et al., 2000). In
Lithium treatment alone produced a small increase vivo microdialysis studies in rodents have reported that
(35%) in phospho-GSK3a(ser21) relative to total lithium increases both basal and antidepressant-induced
GSK3a protein in the hippocampus, however, this effect increases in extracellular serotonin levels in both the
was not statistically significant (F(1, 31) = 1.089, hippocampus (Wegener et al., 2003) and the medial
p = 0.305). Desipramine treatment alone did not alter prefrontal cortex (Muraki et al., 2001; Kitaichi et al.,
phospho-GSK3a(ser21) relative to total GSK3a protein 2005) of the rat. However, whether lithium-induced
(F(1, 31) = 0.45, p = 0.833). Similarly, the addition of increases in serotonin can mediate lithium’s enhancing
lithium to desipramine did not result in any further effects on antidepressant-induced behaviour has yet to
increases in phospho-GSK3a(ser21) relative to any be determined using serotonin depletion methods.
other experimental group (no interaction effect: Nevertheless, a recent pharmacogenetic study
F(1, 31) = 1.43, p = 0.241) thus suggesting that indirect suggested that a genetic variant of the serotonin
inhibition of GSK-3a via phosphorylation at its ser21 transporter is associated with the clinical response to
residue does not mediate the effects of lithium plus lithium augmentation. Specifically, patients homozygous
desipramine on behaviour in the NIH and hippocampal for the short (s) allele of the serotonin transporter gene-
cell proliferation (Fig. 2C). linked polymorphic region
Similarly, neither lithium alone (F(1, 32) = 0.157, (5-HTTLPR) which has been reported to confer lower
p = 0.694), desipramine (F(1, 32) = 1.688, p = 0.206) transcriptional activity of the serotonin transporter
nor lithium plus desipramine (F(1, 32) = 1.886, relative to the long (l) allele (Lesch et al., 1996), had a
p = 0.179) significantly altered phospho-GSK3b(ser9) more favourable response to lithium augmentation than l
relative to total GSK3b protein thus suggesting that allele carriers (Stamm et al., 2008). Several clinical
indirect inhibition of GSK-3b via phosphorylation at its studies have also reported that the addition of lithium to
ser9 residue does not mediate the effects of lithium plus tricyclic antidepressants enhances serotonergic function
desipramine on behaviour in the NIH and hippocampal (Cowen et al., 1991), however, these effects were not
cell proliferation (Fig. 2D). correlated with the clinical response. More studies using
42 O. F. O’Leary et al. / Neuroscience 228 (2013) 36–46

serotonin depletion methods are required to determine at least some antidepressants induce a biphasic
whether such increases in serotonin are directly regulation of BDNF gene expression. For example,
responsible for the antidepressant-enhancing effects of while 4 days of fluoxetine treatment decreased BDNF
lithium (Lucki and O’Leary, 2004). mRNA levels in the hippocampus, 14 days of treatment
The molecular mechanisms underlying the effects of increased expression in the same brain area (De
lithium plus desipramine treatment on hippocampal cell Foubert et al., 2004). Similarly, it was reported that
proliferation and antidepressant-like behaviour in the chronic fluoxetine treatment increased BDNF mRNA
present study have yet to be determined. Adult levels in the dentate gyrus of the rat hippocampus when
hippocampal cell proliferation can be regulated by it was measured 24 h following the last injection, but
serotonin (Benninghoff et al., 2010) and since lithium- decreased expression when measured 4 h following the
induced increases in serotonin have been reported (see last injection (Coppell et al., 2003). It is possible that
above), the effects observed in the present study may desipramine might also induce a similar biphasic
be a result of increased serotonin in the hippocampus. regulation of BDNF gene expression and that this might
Another possible candidate is glycogen synthase kinase account for discrepancies in the literature. Finally, the
3 (GSK3), a constitutively active serine/threonine protein gene encoding BDNF has a complex structure in both
kinase that is inhibited by phosphorylation at serine-9 in rodents and humans. Both the mouse and rat bdnf
its beta isoform, or serine-21 in its alpha isoform. genes consist of at least eight 50 untranslated exons
Recently, it was reported that phosphorylation of these that are alternatively spliced to the protein coding 30
serine residues is required for the effects of lithium and exon (Aid et al., 2007). It has been reported that
the antidepressant, fluoxetine on hippocampal cell different antidepressants increase BDNF expression in
proliferation in mice (Eom and Jope, 2009). Moreover, the hippocampus and frontal cortex using different
pharmacogenetic studies have reported that carriers of combinations of these promoters (Dias et al., 2003;
the C allele of a 50T/C SNP in the promoter region of Dwivedi et al., 2006; Khundakar and Zetterstrom, 2006;
GSK3, exhibit a better response to lithium augmentation Tsankova et al., 2006). Indeed, it has been suggested
than patients of the TT genotype (Adli et al., 2007). that the biphasic change in bdnf gene expression
Using western blotting, we investigated whether following treatment with antidepressants including
desipramine plus lithium could produce additive effects desipramine might be explained by differential mRNA
on inhibitory phosphorylation of GSK3 in the transcript regulation (Khundakar and Zetterstrom, 2006).
hippocampus, however, this drug combination did not While it could be concluded that neither BDNF, VEGF
increase phospho-GSK3b(ser9) or phospho-GSK3a or phospho-GSK3a/b mediated the increase in
(ser21) levels any further when compared with lithium hippocampal cell proliferation, it should be noted that
treatment alone. homogenates from whole hippocampi were employed.
Various growth factors including BDNF and VEGF Therefore, increased expression within specific
regulate the proliferation or survival of newly-born cells subregions of the hippocampus cannot be ruled out. For
in the hippocampus and are increased by example, it has been reported that chronic treatment
antidepressant treatments as well as lithium (O’Leary with lithium alone (Hammonds and Shim, 2009) or
and Castren, 2010). Using RT-PCR, we did not observe desipramine alone (Jacobsen and Mork, 2004)
synergistic increases in mRNA levels of either growth increased BDNF mRNA and protein selectively in the
factor in the hippocampus. In fact, chronic treatment dentate gyrus, although negative findings have also
with desipramine either alone or in combination with been reported (Jacobsen and Mork, 2004; Martinez-
lithium significantly decreased BDNF mRNA in whole Turrillas et al., 2005; Hanson et al., 2011). Given that
hippocampus homogenates. This finding may seem hippocampal cell proliferation occurs within the dentate
counterintuitive given other reports that antidepressants gyrus of the hippocampus, future studies using in situ
including desipramine increase BDNF mRNA in the rat hybridisation or immunohistochemistry should determine
hippocampus (Nibuya et al., 1995; Jacobsen and Mork, whether co-administration of lithium and desipramine
2004; Martinez-Turrillas et al., 2005; Dwivedi et al., selectively increases the expression of these molecular
2006). However, negative findings have also been substrates within the dentate gyrus relative to other
reported both at the mRNA level (Coppell et al., 2003; hippocampal subregions. Moreover, future studies using
Vinet et al., 2004; Torregrossa et al., 2005; Bravo et al., gene array technology would aid the identification of
2009) and at the protein level (Altar et al., 2003; targets underlying the augmenting effects of lithium on
Jacobsen and Mork, 2004; Balu et al., 2008, 2009). desipramine in antidepressant-like behaviour and
Moreover, desipramine-induced decreases have also hippocampal cell proliferation.
been reported by others (Torregrossa et al., 2005). The In the present study, desipramine alone did not alter
reasons for these conflicting results are unclear but may behaviour in the NIH test and did not increase
be a function of species (most studies have been hippocampal cell proliferation in BALB/cOLaHsd mice.
conducted in rats rather than mice), strain, differences in This finding may initially appear surprising given the
dosing regimens and/or in the methods used to quantify many reports that chronic treatment with antidepressant
BDNF mRNA levels as discussed below (for review see drugs increases hippocampal cell proliferation in rats
(O’Leary and Castren, 2010)). Moreover, the length of and mice. The reasons underlying the lack of effect of
treatment and the time of sacrifice post-treatment may desipramine alone in the present study are unclear but
be important factors. Indeed, it has been suggested that may be a function of strain differences or drug dose.
 O. F. O’Leary et al. / Neuroscience 228 (2013) 36–46 43

Although there are some reports that chronic desipramine hippocampal cell proliferation and hippocampal
increases hippocampal cell proliferation in the mouse neurogenesis (Santarelli et al., 2003; Wang et al., 2008;
(Gur et al., 2007; Balu et al., 2009; Pechnick et al., David et al., 2009), the NIH test was deemed the most
2011), negative findings have also been reported (Balu appropriate model to test the hypothesis that
et al., 2009; Stevenson et al., 2009; Onksen et al., antidepressant-like behaviour induced by the chronic
2011). None of the above studies were conducted in co-administration of lithium plus desipramine would be
BALB/c mice so it is possible that the lack of effect correlated with increased hippocampal cell proliferation.
observed in the present study may be due to strain Nevertheless, it will be important to determine whether
differences. Indeed, strain differences in antidepressant- the effects observed in the NIH extend to other tests of
induced increases in hippocampal cell proliferation have antidepressant drug-like activity. Commonly-used tests
already been reported in the mouse with the of antidepressant drug-like activity include the tail
antidepressants, fluoxetine and desipramine, although suspension test (O’Leary and Cryan, 2009) and the
these studies did not examine the BALB/c mouse strain forced swim test (Cryan and Holmes, 2005) but these
(Miller et al., 2008; Balu et al., 2009). However, it has tests respond to acute antidepressant treatment thus
already been reported that chronic treatment with making it difficult to dissociate the effects of chronic
another antidepressant, fluoxetine, does not increase treatment from acute treatment. The unpredictable
hippocampal cell proliferation in the BALB/cj mouse chronic mild stress (UCMS) model of depression has
strain under basal conditions (Holick et al., 2008; Huang been used with success in neurogenesis studies (Surget
et al., 2008), thus raising the intriguing possibility that et al., 2008), however such models are notoriously
some BALB/c mouse strains may be insensitive to difficult to reliably reproduce between laboratories
antidepressant-induced increases in hippocampal cell (Cryan and Mombereau, 2004) and thus were not
proliferation. Nevertheless, both positive (Gur et al., employed in the present study.
2007; Pechnick et al., 2011), and negative (Balu et al.,
2009; Stevenson et al., 2009; Onksen et al., 2011)
findings with desipramine have been reported in mice CONCLUSION
on a C57/BL6 background, thus suggesting that In conclusion, the present study demonstrated that a
variables other than strain may also play a role. In the combination of lithium plus desipramine but neither drug
present study, the effects of a single dose of alone induced antidepressant-like behaviour and
desipramine were examined, thus we cannot exclude increased hippocampal cell proliferation in a mouse
the possibility that other doses of desipramine could strain with reduced sensitivity to chronic desipramine
increase hippocampal cell proliferation and induce treatment. To our knowledge, this is the first study to
antidepressant-like behavioural effects in BALB/ demonstrate that lithium can augment both the
cOLaHsd mice. However, we have also observed a lack behavioural and cytogenic effects of chronic treatment
of effect of chronic treatment with a higher dose of with an antidepressant drug in an animal study. These
desipramine in the forced swim test in this mouse strain data together with other studies suggest that enhanced
(O’Leary and Cryan, unpublished). Taken together, the hippocampal cell proliferation may be a plausible
lack of effect of desipramine on hippocampal cell mechanism underlying the antidepressant-enhancing
proliferation and in the NIH test, raise the intriguing effects of lithium as well as other treatment options for
possibility that the BALB/cOLaHsd mouse strain may be treatment-resistant depression. However, this
useful in the development of an animal model of hypothesis has yet to be directly tested by inhibiting
treatment-refractory depression (Samuels et al., 2011). hippocampal cell proliferation, e.g., by irradiation.
The present study has several limitations that should Similarly, the question remains as to how quickly this
be considered when interpreting the results. First, BrdU- drug combination can induce its antidepressant-like
positive cells were not counted using a stereological behavioural effects and whether the time course for
approach. However as outlined in the methods section, lithium plus desipramine to produce a behavioural
several measures were adopted to reduce any response directly correlates with increased cell
disadvantage of this limitation. In addition, some proliferation in the hippocampus. In addition, whether
suggest that stereology offers little advantage over surviving newly-born cells differentiate into neurons has
absolute counting of BrdU-positive cells because yet to be determined. Finally, since BALB/cOLaHsd
counting frames may miss clusters of BrdU cells which mice demonstrate behavioural responses to lithium plus
typically have a heterogeneous distribution; and the desipramine but not to either drug alone, they may be
volume of the hippocampus increases along its useful in the development of an animal model of
rostrocaudal plane (Crews et al., 2004; Noori and treatment-refractory depression and thus be a useful
Fornal, 2011). Second, the behavioural effects of lithium tool for the discovery of new more effective
and desipramine were examined using a single test, the antidepressants where there is an unmet medical need
NIH test. While the NIH test is sensitive to chronic but (Samuels et al., 2011).
not acute antidepressant treatment, it is important to
note that reduced neohypophagia is also an index of Acknowledgments—The authors gratefully acknowledge the
reduced anxiety. However, since the effects of at least technical assistance of Ms. Elaine O’Mahony, Mr. Seamus Boyle,
some antidepressant drugs in some mouse strains in Dr. Richard O’Connor and Dr. Sue Grenham. This work was
this test have been shown to be dependent upon funded by a Career Development Award from the Health
44 O. F. O’Leary et al. / Neuroscience 228 (2013) 36–46

Research Board Ireland (PD/2008/26; O.F.O.), Science Foundation David DJ, Samuels BA, Rainer Q, Wang JW, Marsteller D, Mendez I,
Ireland (02/CE/B12 and 07/CE/B1368; J.F.C.) and the European Drew M, Craig DA, Guiard BP, Guilloux JP, Artymyshyn RP,
Community’s Seventh Framework Programme (FP7/2007-2013, Gardier AM, Gerald C, Antonijevic IA, Leonardo ED, Hen R
Grant Agreement 201714; J.F.C.). (2009) Neurogenesis-dependent and -independent effects of
fluoxetine in an animal model of anxiety/depression. Neuron
62:479–493.
De Foubert G, Carney SL, Robinson CS, Destexhe EJ, Tomlinson R,
REFERENCES Hicks CA, Murray TK, Gaillard JP, Deville C, Xhenseval V,
Thomas CE, O’Neill MJ, Zetterstrom TS (2004) Fluoxetine-
Adli M, Hollinde DL, Stamm T, Wiethoff K, Tsahuridu M, Kirchheiner induced change in rat brain expression of brain-derived
J, Heinz A, Bauer M (2007) Response to lithium augmentation in neurotrophic factor varies depending on length of treatment.
depression is associated with the glycogen synthase kinase 3- Neuroscience 128:597–604.
beta 50T/C single nucleotide polymorphism. Biol Psychiatry Dias BG, Banerjee SB, Duman RS, Vaidya VA (2003) Differential
62:1295–1302. regulation of brain derived neurotrophic factor transcripts by
Aid T, Kazantseva A, Piirsoo M, Palm K, Timmusk T (2007) Mouse antidepressant treatments in the adult rat brain.
and rat BDNF gene structure and expression revisited. J Neurosci Neuropharmacology 45:553–563.
Res 85:525–535. Dulawa SC, Hen R (2005) Recent advances in animal models of
Altar CA, Whitehead RE, Chen R, Wortwein G, Madsen TM (2003) chronic antidepressant effects: the novelty-induced hypophagia
Effects of electroconvulsive seizures and antidepressant drugs on test. Neurosci Biobehav Rev 29:771–783.
brain-derived neurotrophic factor protein in rat brain. Biol Dwivedi Y, Rizavi HS, Pandey GN (2006) Antidepressants reverse
Psychiatry 54:703–709. corticosterone-mediated decrease in brain-derived neurotrophic
Balu DT, Hodes GE, Anderson BT, Lucki I (2009) Enhanced factor expression: differential regulation of specific exons
sensitivity of the MRL/MpJ mouse to the neuroplastic and by antidepressants and corticosterone. Neuroscience 139:
behavioral effects of chronic antidepressant treatments. 1017–1029.
Neuropsychopharmacology 34:1764–1773. Encinas JM, Hamani C, Lozano AM, Enikolopov G (2011) Neurogenic
Balu DT, Hoshaw BA, Malberg JE, Rosenzweig-Lipson S, Schechter hippocampal targets of deep brain stimulation. J Comp Neurol
LE, Lucki I (2008) Differential regulation of central BDNF protein 519:6–20.
levels by antidepressant and non-antidepressant drug treatments. Eom TY, Jope RS (2009) Blocked inhibitory serine-phosphorylation of
Brain Res 1211:37–43. glycogen synthase kinase-3alpha/beta impairs in vivo neural
Bauer M, Adli M, Bschor T, Pilhatsch M, Pfennig A, Sasse J, Schmid precursor cell proliferation. Biol Psychiatry 66:494–502.
R, Lewitzka U (2010) Lithium’s emerging role in the treatment of Felice D, O’Leary OF, Pizzo RC, Cryan JF (2012) Blockade of the
refractory major depressive episodes: augmentation of GABA(B) receptor increases neurogenesis in the ventral but not
antidepressants. Neuropsychobiology 62:36–42. dorsal adult hippocampus: Relevance to antidepressant action.
Benninghoff J, Gritti A, Rizzi M, Lamorte G, Schloesser RJ, Schmitt Neuropharmacology 63:1380–1388.
A, Robel S, Genius J, Moessner R, Riederer P, Manji HK, Grunze Gould TD, Einat H, Bhat R, Manji HK (2004) AR-A014418, a selective
H, Rujescu D, Moeller HJ, Lesch KP, Vescovi AL (2010) GSK-3 inhibitor, produces antidepressant-like effects in the forced
Serotonin depletion hampers survival and proliferation in swim test. Int J Neuropsychopharmacol 7:387–390.
neurospheres derived from adult neural stem cells. Gould TD, O’Donnell KC, Dow ER, Du J, Chen G, Manji HK (2008)
Neuropsychopharmacology 35:893–903. Involvement of AMPA receptors in the antidepressant-like effects
Boldrini M, Underwood MD, Hen R, Rosoklija GB, Dwork AJ, John of lithium in the mouse tail suspension test and forced swim test.
Mann J, Arango V (2009) Antidepressants increase neural Neuropharmacology 54:577–587.
progenitor cells in the human hippocampus. Gur TL, Conti AC, Holden J, Bechtholt AJ, Hill TE, Lucki I, Malberg
Neuropsychopharmacology 34:2376–2389. JE, Blendy JA (2007) CAMP response element-binding protein
Bravo JA, Diaz-Veliz G, Mora S, Ulloa JL, Berthoud VM, Morales P, deficiency allows for increased neurogenesis and a rapid onset of
Arancibia S, Fiedler JL (2009) Desipramine prevents stress- antidepressant response. J Neurosci 27:7860–7868.
induced changes in depressive-like behavior and hippocampal Haddjeri N, Szabo ST, de Montigny C, Blier P (2000) Increased tonic
markers of neuroprotection. Behav Pharmacol 20:273–285. activation of rat forebrain 5-HT(1A) receptors by lithium addition to
Carvalho AF, Cavalcante JL, Castelo MS, Lima MC (2007) antidepressant treatments. Neuropsychopharmacology
Augmentation strategies for treatment-resistant depression: a 22:346–356.
literature review. J Clin Pharm Ther 32:415–428. Hammonds MD, Shim SS (2009) Effects of 4-week treatment with
Chen G, Rajkowska G, Du F, Seraji-Bozorgzad N, Manji HK (2000) lithium and olanzapine on levels of brain-derived neurotrophic
Enhancement of hippocampal neurogenesis by lithium. J factor, B-cell CLL/lymphoma 2 and phosphorylated cyclic
Neurochem 75:1729–1734. adenosine monophosphate response element-binding protein in
Coppell AL, Pei Q, Zetterstrom TS (2003) Bi-phasic change in BDNF the sub-regions of the hippocampus. Basic Clin Pharmacol
gene expression following antidepressant drug treatment. Toxicol 105:113–119.
Neuropharmacology 44:903–910. Hanson ND, Nemeroff CB, Owens MJ (2011) Lithium, but not
Cowen PJ, McCance SL, Ware CJ, Cohen PR, Chalmers JS, Julier fluoxetine or the corticotropin-releasing factor receptor 1 receptor
DL (1991) Lithium in tricyclic-resistant depression. Correlation of antagonist R121919, increases cell proliferation in the adult
increased brain 5-HT function with clinical outcome. Br J dentate gyrus. J Pharmacol Exp Ther 337:180–186.
Psychiatry 159:341–346. Holick KA, Lee DC, Hen R, Dulawa SC (2008) Behavioral effects of
Crews FT, Nixon K, Wilkie ME (2004) Exercise reverses ethanol chronic fluoxetine in BALB/cJ mice do not require adult
inhibition of neural stem cell proliferation. Alcohol 33:63–71. hippocampal neurogenesis or the serotonin 1A receptor.
Crowley JJ, Jones MD, O’Leary OF, Lucki I (2004) Automated tests Neuropsychopharmacology 33:406–417.
for measuring the effects of antidepressants in mice. Pharmacol Huang GJ, Bannerman D, Flint J (2008) Chronic fluoxetine treatment
Biochem Behav 78:269–274. alters behavior, but not adult hippocampal neurogenesis, in
Cryan JF, Holmes A (2005) The ascent of mouse: advances in BALB/cJ mice. Mol Psychiatry 13:119–121.
modelling human depression and anxiety. Nat Rev Drug Discov Jacobsen JP, Mork A (2004) The effect of escitalopram, desipramine,
4:775–790. electroconvulsive seizures and lithium on brain-derived
Cryan JF, Mombereau C (2004) In search of a depressed mouse: neurotrophic factor mRNA and protein expression in the rat
utility of models for studying depression-related behavior in brain and the correlation to 5-HT and 5-HIAA levels. Brain Res
genetically modified mice. Mol Psychiatry 9:326–357. 1024:183–192.
 O. F. O’Leary et al. / Neuroscience 228 (2013) 36–46 45

Kaidanovich-Beilin O, Lipina TV, Takao K, van Eede M, Hattori S, O’Leary OF, Cryan JF (2009) The tail suspension test: a model for
Laliberte C, Khan M, Okamoto K, Chambers JW, Fletcher PJ, characterizing antidepressant activity in mice. In: Gould TD,
MacAulay K, Doble BW, Henkelman M, Miyakawa T, Roder J, editor. Neuromethods, vol. 42. USA: Springer. p. 119–137.
Woodgett JR (2009) Abnormalities in brain structure and behavior O’Leary OF, O’Connor RM, Cryan JF (2012) Lithium-induced effects
in GSK-3alpha mutant mice. Mol Brain 2:35. on adult hippocampal neurogenesis are topographically
Khundakar AA, Zetterstrom TS (2006) Biphasic change in BDNF gene segregated along the dorso-ventral axis of stressed mice.
expression following antidepressant drug treatment explained by Neuropharmacology 62:247–255.
differential transcript regulation. Brain Res 1106:12–20. O’Leary OF, Castren E (2010) Neurotrophic factors and
Kitaichi Y, Inoue T, Nakagawa S, Izumi T, Koyama T (2005) Effect of antidepressant action: recent advances. In: Cryan JF, Leonard
milnacipran on extracellular monoamine concentrations in the BE, editors. Depression: from psychopathology to
medial prefrontal cortex of rats pre-treated with lithium. Eur J pharmacotherapy, Vol. 27. Basel: Karger. p. 99–223.
Pharmacol 516:219–226. Onksen JL, Brown EJ, Blendy JA (2011) Selective deletion of a cell
Kitamura Y, Doi M, Kuwatsuka K, Onoue Y, Miyazaki I, Shinomiya K, cycle checkpoint kinase (ATR) reduces neurogenesis and alters
Koyama T, Sendo T, Kawasaki H, Asanuma M, Gomita Y (2011) responses in rodent models of behavioral affect.
Chronic treatment with imipramine and lithium increases cell Neuropsychopharmacology 36:960–969.
proliferation in the hippocampus in adrenocorticotropic hormone- Pechnick RN, Zonis S, Wawrowsky K, Cosgayon R, Farrokhi C,
treated rats. Biol Pharm Bull 34:77–81. Lacayo L, Chesnokova V (2011) Antidepressants stimulate
Klein PS, Melton DA (1996) A molecular mechanism for the effect of hippocampal neurogenesis by inhibiting p21 expression in the
lithium on development. Proc Natl Acad Sci U S A 93:8455–8459. subgranular zone of the hipppocampus. PLoS One 6:e27290.
Lesch KP, Bengel D, Heils A, Sabol SZ, Greenberg BD, Petri S, Petrik D, Lagace DC, Eisch AJ (2012) The neurogenesis hypothesis
Benjamin J, Muller CR, Hamer DH, Murphy DL (1996) Association of affective and anxiety disorders: are we mistaking the
of anxiety-related traits with a polymorphism in the serotonin scaffolding for the building? Neuropharmacology 62:21–34.
transporter gene regulatory region. Science 274:1527–1531. Redrobe JP, Bourin M, Colombel MC, Baker GB (1998) Dose-
Li X, Jope RS (2010) Is glycogen synthase kinase-3 a central dependent noradrenergic and serotonergic properties of
modulator in mood regulation? Neuropsychopharmacology venlafaxine in animal models indicative of antidepressant
35:2143–2154. activity. Psychopharmacology (Berl) 138:1–8.
Li X, Zhu W, Roh MS, Friedman AB, Rosborough K, Jope RS (2004) Revesz D, Tjernstrom M, Ben-Menachem E, Thorlin T (2008) Effects
In vivo regulation of glycogen synthase kinase-3beta (GSK3beta) of vagus nerve stimulation on rat hippocampal progenitor
by serotonergic activity in mouse brain. proliferation. Exp Neurol 214:259–265.
Neuropsychopharmacology 29:1426–1431. Riadh N, Allagui MS, Bourogaa E, Vincent C, Croute F, Elfeki A
Livak KJ, Schmittgen TD (2001) Analysis of relative gene expression (2011) Neuroprotective and neurotrophic effects of long term
data using real-time quantitative PCR and the 2(Delta Delta lithium treatment in mouse brain. Biometals 24:747–757.
C(T)) method. Methods 25:402–408. Samuels BA, Leonardo ED, Gadient R, Williams A, Zhou J, David DJ,
Lucassen PJ, Meerlo P, Naylor AS, van Dam AM, Dayer AG, Fuchs Gardier AM, Wong EH, Hen R (2011) Modeling treatment-
E, Oomen CA, Czeh B (2010) Regulation of adult neurogenesis resistant depression. Neuropharmacology 61:408–413.
by stress, sleep disruption, exercise and inflammation: Santarelli L, Saxe M, Gross C, Surget A, Battaglia F, Dulawa S,
implications for depression and antidepressant action. Eur Weisstaub N, Lee J, Duman R, Arancio O, Belzung C, Hen R
Neuropsychopharmacol 20:1–17. (2003) Requirement of hippocampal neurogenesis for the
Lucki I, O’Leary OF (2004) Distinguishing roles for norepinephrine behavioral effects of antidepressants. Science 301:805–809.
and serotonin in the behavioral effects of antidepressant drugs. J Segi-Nishida E, Warner-Schmidt JL, Duman RS (2008)
Clin Psychiatry 65(Suppl. 4):11–24. Electroconvulsive seizure and VEGF increase the proliferation
Martinez-Turrillas R, Del Rio J, Frechilla D (2005) Sequential of neural stem-like cells in rat hippocampus. Proc Natl Acad Sci U
changes in BDNF mRNA expression and synaptic levels of S A 105:11352–11357.
AMPA receptor subunits in rat hippocampus after chronic Stamm TJ, Adli M, Kirchheiner J, Smolka MN, Kaiser R, Tremblay
antidepressant treatment. Neuropharmacology 49:1178–1188. PB, Bauer M (2008) Serotonin transporter gene and response to
Miller BH, Schultz LE, Gulati A, Cameron MD, Pletcher MT (2008) lithium augmentation in depression. Psychiatr Genet 18:92–97.
Genetic regulation of behavioral and neuronal responses to Stevenson JR, Schroeder JP, Nixon K, Besheer J, Crews FT, Hodge
fluoxetine. Neuropsychopharmacology 33:1312–1322. CW (2009) Abstinence following alcohol drinking produces
Muraki I, Inoue T, Hashimoto S, Izumi T, Ito K, Koyama T (2001) depression-like behavior and reduced hippocampal
Effect of subchronic lithium treatment on citalopram-induced neurogenesis in mice. Neuropsychopharmacology
increases in extracellular concentrations of serotonin in the 34:1209–1222.
medial prefrontal cortex. J Neurochem 76:490–497. Surget A, Saxe M, Leman S, Ibarguen-Vargas Y, Chalon S, Griebel
Nibuya M, Morinobu S, Duman RS (1995) Regulation of BDNF and G, Hen R, Belzung C (2008) Drug-dependent requirement of
trkB mRNA in rat brain by chronic electroconvulsive seizure and hippocampal neurogenesis in a model of depression and of
antidepressant drug treatments. J Neurosci 15:7539–7547. antidepressant reversal. Biol Psychiatry 64:293–301.
Nixon MK, Hascoet M, Bourin M, Colombel MC (1994) Additive Torregrossa MM, Folk JE, Rice KC, Watson SJ, Woods JH (2005)
effects of lithium and antidepressants in the forced swimming test: Chronic administration of the delta opioid receptor agonist
further evidence for involvement of the serotoninergic system. (+)BW373U86 and antidepressants on behavior in the forced
Psychopharmacology (Berl) 115:59–64. swim test and BDNF mRNA expression in rats.
Noori HR, Fornal CA (2011) The appropriateness of unbiased optical Psychopharmacology (Berl) 183:31–40.
fractionators to assess cell proliferation in the adult hippocampus. Trivedi MH, Rush AJ, Wisniewski SR, Nierenberg AA, Warden D, Ritz
Front Neurosci 5:140. L, Norquist G, Howland RH, Lebowitz B, McGrath PJ, Shores-
O’Brien WT, Huang J, Buccafusca R, Garskof J, Valvezan AJ, Berry Wilson K, Biggs MM, Balasubramani GK, Fava M (2006)
GT, Klein PS (2011) Glycogen synthase kinase-3 is essential for Evaluation of outcomes with citalopram for depression using
beta-arrestin-2 complex formation and lithium-sensitive behaviors measurement-based care in STAR⁄D: implications for clinical
in mice. J Clin Invest 121:3756–3762. practice. Am J Psychiatry 163:28–40.
O’Leary OF, Bechtholt AJ, Crowley JJ, Valentino RJ, Lucki I (2007) Tsankova NM, Berton O, Renthal W, Kumar A, Neve RL, Nestler EJ
The role of noradrenergic tone in the dorsal raphe nucleus of the (2006) Sustained hippocampal chromatin regulation in a mouse
mouse in the acute behavioral effects of antidepressant drugs. model of depression and antidepressant action. Nat Neurosci
Eur Neuropsychopharmacol 17:215–226. 9:519–525.
46 O. F. O’Leary et al. / Neuroscience 228 (2013) 36–46

Vinet J, Carra S, Blom JM, Brunello N, Barden N, Tascedda F (2004) Wegener G, Bandpey Z, Heiberg IL, Mork A, Rosenberg R (2003)
Chronic treatment with desipramine and fluoxetine modulate BDNF, Increased extracellular serotonin level in rat hippocampus
CaMKKalpha and CaMKKbeta mRNA levels in the hippocampus of induced by chronic citalopram is augmented by subchronic
transgenic mice expressing antisense RNA against the lithium: neurochemical and behavioural studies in the rat.
glucocorticoid receptor. Neuropharmacology 47:1062–1069. Psychopharmacology (Berl) 166:188–194.
Wang JW, David DJ, Monckton JE, Battaglia F, Hen R (2008) Chronic Wu X, Castren E (2009) Co-treatment with diazepam prevents
fluoxetine stimulates maturation and synaptic plasticity of adult- the effects of fluoxetine on the proliferation and survival of
born hippocampal granule cells. J Neurosci 28:1374–1384. hippocampal dentate granule cells. Biol Psychiatry 66:
Warner-Schmidt JL, Duman RS (2007) VEGF is an essential 5–8.
mediator of the neurogenic and behavioral actions of
antidepressants. Proc Natl Acad Sci U S A 104:4647–4652.

(Accepted 29 September 2012)

(Available online 13 October 2012)