Health Articles

 

Stress.cl

Neuroscience 169 (2010) 98 –108 CHRONIC FLUOXETINE TREATMENT INDUCES STRUCTURAL
PLASTICITY AND SELECTIVE CHANGES IN GLUTAMATE RECEPTOR
SUBUNITS IN THE RAT CEREBRAL CORTEX
E. AMPUERO,a F. J. RUBIO,a R. FALCON,a
The selective serotonin reuptake inhibitor fluoxetine (flx) is M. SANDOVAL,a G. DIAZ-VELIZ,b R. E. GONZALEZ,a
one of the most commonly prescribed antidepressant drugs.
N. EARLE,a A. DAGNINO-SUBIABRE,c F. ABOITIZ,d
The mechanisms underlying its antidepressant action, how- F. ORREGOa AND U. WYNEKENa*
ever, are still unclear. Although serotonin levels rise rapidly aLaboratorio de Neurociencias, Universidad de los Andes, Santiago, after acute administration, several weeks are required before therapeutic benefits are achieved. The delayed onset of an- bFacultad de Medicina, Universidad de Chile, Santiago, Chile tidepressant action suggests that plastic changes over cLaboratorio de Neurobiologia y Conducta, Facultad de Medicina, protracted periods of time might be causally related to the Universidad Catolica del Norte, Santiago, Chile therapeutic effect. Repetitive antidepressant administration dDepartamento de Psiquiatria, Pontificia Universidad Catolica de promotes plastic changes such as neurogenesis, synapto- Chile, Santiago, Chile genesis, neurotrophin signalling, and changes in chromatinstructure and gene expression Abstract—It has been postulated that chronic administration
Increasing evidence implicates glutamatergic circuit- of antidepressant drugs induces delayed structural and mo-
ries in these plastic changes lecular adaptations at glutamatergic forebrain synapses that
might underlie mood improvement. To gain further insight
into these changes in the cerebral cortex, rats were treated
We found that 2 weeks of flx treat- with fluoxetine (flx) for 4 weeks. These animals showed de-
ment induced brain-derived neurotrophic factor (BDNF) creased anxiety and learned helplessness. N-methyl-D-aspar-
signalling at excitatory forebrain synapses Glutamatergic synapses are situated on dendritic propionate (AMPA) receptor subunit levels (NR1, NR2A,
spines containing postsynaptic densities (PSDs), which NR2B, GluR1 and GluR2) were analysed in the forebrain by
allow glutamate receptors to anchor through interactions both western blot of homogenates and immunohistochemis-
with scaffolding proteins. Spines and PSDs are plastic try. Both methods demonstrated an upregulation of NR2A,
structures, and large spines are associated with more GluR1 and GluR2 that was especially significant in the retro-
efficient synapses The devel- splenial granular b cortex (RSGb). However, when analysing
subunit content in postsynaptic densities and synaptic mem-
opment of dendritic spines begins with the formation of branes, we found increases of NR2A and GluR2 but not
filopodia. After establishing contact with pre-synaptic ter- GluR1. Instead, GluR1 was augmented in a microsomal frac-
minals, they mature into spines acquiring thin, then stubby, tion containing intracellular membranes. NR1 and GluR2
and finally mushroom morphologies. The latter represent were co-immunoprecipitated from postsynaptic densities
the most mature and stable spines.
and synaptic membranes. In the immunoprecipitates, NR2A
Ionotropic glutamate receptors include the alpha-amino- was increased while GluR1 was decreased supporting a
3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA-R) change in receptor stoichiometry. The changes of subunit
and N-methyl-D-aspartate (NMDA-R) types. NMDA-Rs are levels were associated with an upregulation of dendritic
heterotetramers composed of NR1 and NR2 subunits that spine density and of large, mushroom-type spines. These
coassemble to form a functional channel. A single NR1 sub- molecular and structural adaptations might be involved in
neuronal
unit exists in eight splice isoforms, and there are four distinct treatment. 2010 IBRO. Published by Elsevier Ltd. All rights
NR2 subunits (NR2A–D). AMPA-Rs are homo- or heterotet- ramers composed mainly of GluR1 and 2/3 subunits in theadult forebrain. NMDA-R and AMPA-R subunit composition Key words: antidepressants, glutamate, ionotropic receptors,
is a major determinant of biophysical properties, association dendritic spines, cerebral cortex.
to protein complexes, downstream signalling, receptor traf-ficking and synaptic targeting Glutamate receptor availability de- *Corresponding author. Tel: ⫹56-2-4129353; fax: ⫹56-2-2141752.
termines spine structure, and NMDA-Rs, especially those E-mail address: (U. Wyneken).
Abbreviations: BDNF, brain-derived neurotrophic factor; EPM, ele- containing NR2A subunits, are present in large and stable vated plus maze; flx, fluoxetine; FST, forced swim test; IHC, immuno- histochemistry; NMDA, N-methyl-D-aspartate; NSF, novelty-sup- larly, GluR2 subunits promote the formation and growth of pressed feeding; PrL, prelimbic cortex; PSDs, postsynaptic densities; RSGb, retrosplenial granular b cortex; TST, tail suspension test; WB,western blots.
0306-4522/10 $ - see front matter 2010 IBRO. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.neuroscience.2010.04.035 E. Ampuero et al. / Neuroscience 169 (2010) 98 –108 The present study examined glutamate receptor levels with 2 cm of wooden bedding. Twenty-four hours before behav- following long-term flx treatment in the prelimbic (PrL), retro- ioural testing (immediately following the last drug injection), ani- splenial granular b (RSGb) and secondary motor (M2) corti- mals were deprived of food in the home cage. At the time oftesting, one food pellet was placed on a piece of round filter paper ces, as well as in subcellular fractions obtained from the (10 cm in diameter) positioned in the centre of the box. The animal forebrain. Spine morphology was analysed in the same re- was placed in a corner of the cage. The latency to begin feeding gions. Our results strongly suggest that flx induces plastic was recorded (maximum time, 15 min). Afterwards, the amount of changes in extensive forebrain networks that are consistent food consumed in 5 min in the home cage was measured.
with a functional stabilization of synaptic connections.
Spontaneous motor activity. Rats were individually placed in a plexiglass cage (30⫻30⫻35 cm3) located inside a soundproof chamber. The spontaneous motor activity and rearing was moni-tored during a period of 30 min and evaluated as described We used adult male Sprague–Dawley rats (250 – 400 g) in all of The apparatus consisted of a central the experiments. Procedures involving animals and their care platform (10⫻10 cm2), two opposed open arms (50⫻10 cm2) and were performed in accordance with the Universidad de los, Andes two opposed closed arms of the same size with 40-cm-high Bioethical Committee and the Guide for the Care and Use of opaque walls. The maze was elevated 83 cm above the ground.
Laboratory Animals from the National Institutes of Health. All Each animal was placed at the centre of the maze facing one of efforts were made to minimize animal suffering. Flx at doses of 0.7 the open arms. During a 5-min interval, the number of open and or 3.5 mg/kg (Ely-Lilly Co., Indianapolis, USA) or 0.9% NaCl (sal) closed arms entries, plus the time spent in the open and closed were administered daily by i.p. injection between 9:00 and 10:00 arms were measured in dim light.
AM for 28 days. Body weight was controlled daily and the percent-age of weight gain was calculated. In total, 224 rats were sacri- Tail suspension test. Rats were individually suspended by ficed 24 h following the last flx or sal injection to perform immu- the tail to a horizontal ring stand bar (distance from the table, 40 nohistochemistry (n⫽6 per group), Golgi staining (n⫽6 per group) cm) at 4 –5 cm from the beginning of the tail. A 6-min test session or subcellular fractionation (n⫽10 per group, in total 10 indepen- was recorded. The behavioural parameter measured was the dent preparations were performed). For immunohistochemistry or number of seconds spent in a completely immobile posture.
for Golgi staining, rats were sacrificed under ketamine (50 mg/kg)and xylazine (5 mg/kg) anaesthesia and then perfused intracardi- Forced swim test. The forced swim test was performed ally with sal followed by 300 ml of 4% paraformaldehyde in PBS.
according to a modification, suggested by Lucki of After perfusion, brains were removed immediately and processed the traditional method described by Porsolt for immunohistochemistry or for Golgi staining.
The following behavioural responses were recorded: escape orclimbing behaviour, which was defined as upward-directed move- ment of the forepaws along the side of the swim chamber; swim-ming behaviour, which was defined as movement throughout the All chemical reagents were purchased from Sigma (St Louis, MO, swim chamber; and immobility, which was recorded when the rat USA), unless otherwise stated. Protein G Sepharose was from made no further attempts to escape except the movement nec- Amersham Biosciences (Freiburg, Germany).
essary to keep its head above the water.
The primary antibodies used for immunohistochemistry (IHC), synaptic membrane staining and western blots (WB) were as Tissue preparation and subcellular fractionation
follows: guinea pig anti-ProSAP2/Shank3 (synaptic membranes,1:2000) (kindly donated by Eckart D. Gundelfinger at the Leibniz Rats were sacrificed by rapid decapitation and the cortices and Institute for Neurobiology, Magdeburg, Germany); anti-GluR1 hippocampi were immediately separated on ice and placed in (synaptic membranes, 1:10; IHC, 1:50; WB, 1:1000), anti-GluR2 homogenization buffer (0.32 M sucrose, 0.5 mM EGTA, 5 mM (synaptic membranes, 1:50; IHC, 1:400) and anti-NR2A (IHC, Hepes, pH 7.4) supplemented with a mixture of protease inhibitors 1:100; Chemicon International, Temecula, CA, USA); anti-GluR2 (Boehringer, Mannheim). Subcellular fractionation was performed (WB, 1:1000; BD Biosciences, Pharmingen, San Jose, CA, USA); following the method of Wyneken et al. anti-NR2A (synaptic membranes, 1:2000; WB, 1:1000; Millipore Synaptosomes were collected from the first sucrose gradient at Corporation, MA, USA); anti-NR1 (IHC, 1:250; Pharmingen, San the 1/1.2 M interphase and submitted to a hypo-osmotic shock to Diego, CA, USA); and anti-NR2B (IHC, 1:100; WB, 1:1000), anti- release intracellular organelles. Synaptic membranes were col- SAP102 (WB, 1:1000) and anti-PSD-95 (WB, 1:250; BD Trans- lected from the second sucrose gradient at the 1/1.2 M interphase duction Laboratories, San Jose, CA, USA).
and delipidated in 0.5% Triton to yield PSDs. The microsomalfraction (P3) resulted from the centrifugation of the supernatant (S2) at 100,000 g for 1 h. Protein concentrations were determined Rats were housed in groups of three in a 12-h (8:00 –20:00) using the BCA assay (Pierce, Rockford, IL, USA).
light/dark cycle at 22⫾1 °C with standard rodent pellet food andwater available ad libitum. Each animal behaviour was evaluated only once, 24 h after the final injection using the following se- Protein (250 ␮g) was solubilized for 2 h in 1 ml of solubilization quence: novelty-suppressed feeding (NSF), spontaneous motor buffer (50 mM Tris–Cl, 1% deoxycholate plus protease inhibitors, activity, elevated plus maze (EPM), tail suspension test (TST) and pH 9.0). The corresponding primary antibody or control IgG (2 ␮g) the pretest session of the forced swim test (FST). The test session was added to the supernatant and incubated overnight. Protein G of the FST was applied on the second day (i.e., 48 h after the last Sepharose (20 ␮l pre-washed with solubilization buffer and drug injection). Scores were generated from live observations byan experimenter blinded to the treatment condition, and video blocked with 0.2% BSA) was added and incubated for 1 h. The sequences were used for later reanalysis when necessary.
samples were centrifuged for 5 min at 1000 g and the superna-tants were discarded. The immunoprecipitates were washed three Novelty-suppressed feeding test. The testing apparatus times with solubilization buffer and were resuspended in 60 ␮l of consisted of a plastic cage (80⫻70⫻40 cm3) with its floor covered electrophoresis loading buffer.
E. Ampuero et al. / Neuroscience 169 (2010) 98 –108 Over 1200 spines from 18 to 23 neurons (four to five neurons per animal) were analysed per condition. In addition, images from Western blots. Equal amounts of protein (20 ␮g) were sep- each analysed dendritic segment were captured as visual support arated by 4 –20% SDS gradient gel electrophoresis, transferred to at 1 ␮m-spaced focal plane in the region of interest using a BX61 nitrocellulose membranes and immunoblotted with the indicated Olympus microscope (100⫻ objective, numeric aperture 1.3) at- primary and corresponding secondary antibodies.
tached to a Diagnostic Instruments 25.4/2 MP camera (final mag-nification of 4700⫻; see movie in supplementary material).
Synaptic membranes were immuno- stained following the procedure of Ciruela Synaptic membranes were plated on polylysine-coated slides,fixed in 4% paraformaldehyde in PBS containing 4% sucrose for Data were analysed using Graph Pad Prism 4.0 software and 10 min and washed with PBS containing 25 mM glycine. Mem- presented as mean⫾SEM. The statistical test used for behav- branes were permeabilized, blocked and incubated with the indi- ioural data was one-way analysis of variance (ANOVA) followed cated primary antibody. After incubation with the primary antibody, by Newman–Keuls post hoc test for multiple comparisons. For the membranes were washed and stained with the corresponding morphological and immunochemical staining, the Mann–Whitney secondary antibody (1:600; Alexa Fluor® 647 conjugated goat U-test was applied. In WBs, the relative optic densities of bands anti-Guinea-pig IgG, Alexa Fluor® 488 conjugated goat anti-rabbit were compared by two-tailed t-tests. A probability level of P⬍0.05 IgG or Alexa Fluor® 555 conjugated donkey anti-mouse IgG; (*) or P⬍0.01 (**) was accepted as significant.
Invitrogen, Eugene, OR, USA). Images of serial sections werecaptured with a Nikon Eclipse TE2000-U inverted epifluorescence microscope with a Plan Fluor 60⫻/1.25 numeric aperture oil-immersion objective attached to a cooled monochrome camera Twenty-eight days of 0.7 mg/kg fluoxetine reduced
DS-2MBWc (final magnification 3500⫻).
anxiety and learned helplessness in rats
After perfusion, brains were cryopre- To evaluate meaningful changes elicited by flx, we first served and cut serially in 30-␮m frozen coronal sections. Staining examined behaviour using two different flx doses, 0.7 was performed according to Ampuero and collaborators mg/kg (0.7 flx) or 3.5 mg/kg (3.5 flx), to select a dose with Immunoreactivity for the RSGb was quantified in coronalsections restricted to interaural 5.86 mm/bregma ⫺3.14 mm and optimal effects following 4 weeks of treatment. Although interaural 4.84 mm/ bregma ⫺4.16 mm. Quantification of immuno- selective serotonin reuptake inhibitors are known to induce reactivity in the PrL and M2 regions was restricted to interaural 13.20 anorectic effects in rats, there was no significant difference mm/bregma 4.20 mm and interaural 11.20/bregma 2.20 mm in body weight gain among the experimental groups ex- Layer V pyramidal neurons were examined amined in this study (sal group, 140.8⫾6.2%; 0.7 flx group, in the RSGb and PrL area. In M2, layer II/III pyramidal neurons were 126⫾4%; 3.5 flx group, 131.4⫾4.1%; compared to day 1 of analysed. Brain slices were visualized under a light microscope(AxiosKop, Zeiss, Germany; 10⫻ magnification; numeric aperture treatment). This may be due to the low doses of flx used in 0.3), and images of serial sections were captured with a digital the present study. Interestingly, treatment with 0.7 mg/kg camera (Nikon, CoolPIX 995) with a final magnification of 880⫻.
flx has already been shown to lead to plasma levels con- Digitized images were analysed with ImageJ program. The number sidered therapeutically effective in humans, and this dose of positive cells was counted in equal sample areas of 0.0289 mm2.
stimulates BDNF/TrkB signalling at excitatory synapses At least five sections per animal were analysed in control and treated animals (n⫽5 per group).
We first assessed anxiety-like behaviour in the NSF test left panel). The NSF test is a behavioural paradigm that is sensitive to chronic antidepressant treat- Brains were processed using the FDRapid GolgiStainTM kit (FD ments and acute treatments with anxiolytics (such as ben- Neuro Technologies, Baltimore, USA) and zodiazepines) but not subchronic antidepressant treat- analysed by an individual blind to the experimental conditions.
ments. In the NSF test, the latency to feed decreased from Eighteen to twenty-three randomly selected layer V pyramidal 396.9⫾39.8 s to 257.3⫾34.2 s in the 0.7 flx group and to neurons were examined per experimental condition in the RSGb 272.3⫾37.6 s in the 3.5 flx group (F and PrL area. In M2, layer II/III pyramidal neurons were analysed.
The selected neurons were required to have no breaks in staining groups P⬍0.05 compared to sal). After the test session, along the dendrites. Starting from the origin of the soma (for animals were returned to the home cage and allowed to primary dendrites) or the branch point (for secondary dendrites) eat. Next, we measured spontaneous motor activity be- and continuing away from the cell soma, spines were counted in cause a change in this activity might indirectly influence 8 ␮m segments along an 80 ␮m stretch of the dendrite. Dendritic behavioural tests. Although the flx-treated rats revealed a spines were classified under the microscope at different focal trend towards decreased motor activity, this was not sig- planes, by a single rater, blinded to the experimental groups, intothree shape categories filopodia/thin, stubby nificant when counted over a 30 min period (sal group, and mushroom/branched spines. Thin spines had a greater length 1482⫾119.7, 0.7 flx group, 1388⫾113.3, 3.5 flx group, than neck diameter and similar head and neck diameters. Stubby 1303⫾140.3; (F ⫽0.4963)). Because rearings are pos- spines had neck diameters that were similar to the total spine itively correlated with anxiety we also ex- length. For mushroom-shaped spines, the diameter of the head amined the number of rearings as an index of vertical was much greater than the diameter of the neck. Branched spines exploratory behaviour. We found significantly increased had more than one head emerging from a single neck originating rearing in the 3.5 flx group compared to the 0.7 flx group from the dendrite. If there was ever any doubt in a classification(less than 10%), spine and neck diameters were measured from (P⬍0.05) (sal group, 30⫾2.2, 0.7 flx group, 21.1⫾2.2, 3.5 the picture taken at the focal plane in which it appeared at its flx group, 37.6⫾7.3; F ⫽3.872; P⬍0.05). Although a maximal size.
trend towards decreased rearing with 0.7 flx treatment was

E. Ampuero et al. / Neuroscience 169 (2010) 98 –108 ses of structural and molecular effects at glutamatergic fore-brain synapses.
Long-term flx induces subunit-specific changes in
glutamate receptor subunit levels
Initially, we measured the levels of glutamate receptor sub-units in forebrain homogenates by WB We detecteda large increase in GluR1 (P⬍0.01), GluR2 (P⬍0.05), NR2Aand NR1 (P⬍0.01%), in the homogenates, but no changewas found for NR2B. PSD-95 and SAP102 are examples ofglutamate receptor scaffolding proteins, which have beenimplicated in the trafficking and anchoring of glutamate re-ceptor subunits. In the present study, we did not detect anychanges in PSD-95, but SAP102 increased by 60⫾6%. Tocheck these results in cerebrocortical subregions, we per- Fig. 1. Fluoxetine decreases depressive-like behaviours in rats. Rats
formed IHC in coronal sections of the forebrain. In were treated for 28 days with sal (white bars), 0.7 (black bars) or 3.5 (upper panel), two of the selected regions are shown: the PrL (grey bars) mg/kg of flx. (A) Anxiety was analysed using the NSF test andthe EPM test. Performances were video-recorded and analysed blindly. In cortex, a subregion of the prefrontal cortex thought to mod- the EPM, the time spent in the open arms divided by the number of entries ulate depressive symptoms into the open arms is shown. (B) Learned helplessness was assessed by and the RSGb, a limbic cortex on the posterior the TST and the FST. In both cases, immobility was scored (in sec- cingulate gyrus that has recently been implicated in the con- onds). In addition, swimming and climbing behaviours were measured trol of emotions separately in the FST. Results are presented as mean⫾SEM.
* P⬍0.05; ** P⬍0.01: compared to sal; # P⬍0.05: 3.5 flx compared (lower panel) shows representative stainings of NR2A and to 0.7 flx; ## P⬍0.01: 3.5 flx compared to 0.7 flx (one-way ANOVA GluR2 at two different magnifications. The quantification of followed by post hoc test).
layer V-immunopositive cells revealed an increase of NR2A, evident, the difference between the sal and 0.7 or 3.5 flx NR2B, GluR1 and GluR2 in the RSGb area, whereas no groups did not reach significance. Anxiety-like behaviour change was found in the PrL area was also evaluated in the EPM test, which has been widely To determine whether the upregulated receptors were validated to measure anxiety in rodents transported to the synapse, glutamate receptor subunits In the EPM test right were detected by WB in isolated PSDs We found panel), the time in the open arm per entry increased in the no change in NR2B even though NR1 and NR2A subunits rats treated with 0.7 flx (F ⫽5.227; P⬍0.05). The New- increased. In addition, the AMPA-R subunit GluR2 did not man–Keuls post hoc multiple comparisons revealed that change, but GluR1 was decreased by 12⫾5% and the the scores of the 3.5 flx group were not different from glutamate receptor scaffolding proteins PSD-95 and controls. In addition, the 0.7 flx group scores differed from SAP102 were increased by 23⫾7% and 30⫾4%, respec- those of both the sal and the 3.5 flx groups (P⬍0.05).
tively (not shown). Thus, treatment with flx appeared to These results confirm that the 0.7 mg/kg dose is more favour synapses with higher ratios of NR2A to NR2B and efficient in decreasing anxiety.
GluR2 to GluR1. A possible explanation of decreased Learned helplessness was assessed by the TST and the AMPA-R detection in PSDs is that GluR1 and GluR2 sub- FST In both tests, immobility decreased following units, although increased in homogenates, might be pref- 0.7 or 3.5 flx treatment. In the TST (F erentially localized intracellularly. To test this hypothesis, the post hoc Newman-Keuls test revealed a significant de- we used WB to determine that glutamate receptor subunits crease for both flx doses compared with sal (P⬍0.01). In the in the microsomal P3 fraction This fraction is a ⫽4.655, P⬍0.05), the post hoc test revealed a heterogeneous membrane compartment known to be en- significant decrease for both 0.7 and 3.5 flx (P⬍0.01 and riched in endoplasmic reticulum, Golgi network, endo- P⬍0.05, respectively). However, only the 0.7 flx dose caused somes, trafficking vesicles and synaptic vesicles, but not in a significant increase in swimming behaviour when com- PSDs (see Suppl. Fig. 1). Consistent with our previous pared to the sal and 3.5 flx doses (F data, we found increases of NR2A (P⬍0.05) whereas The post hoc test revealed a significant effect (P⬍0.01) for NR2B decreased (P⬍0.01). NR1 content was not modi- 0.7 flx compared to sal and 3.5 flx. In this same test, the fied. Significantly elevated GluR1 and GluR2 were found climbing behaviour was increased by 3.5 flx but not by 0.7 flx (P⬍0.05 in both cases) suggesting that these subunits are ⫽3.934, P⬍0.05, and P⬍0.05 in the Newman-Keuls present in intracellular membranes, which contain a variety test comparing sal with 3.5 flx). Although both flx doses of cellular components including trafficking compartments.
decreased immobility, the lower dose enhanced swimming Another possibility was that AMPA-Rs were lost during the behaviour, and the higher dose increased climbing behav- PSD purification procedure because they are loosely at- iour. Overall, these results established that 0.7 flx for 28 days tached to PSDs and a proportion of them is localized on consistently induced antidepressant-like behavioural effects their periphery. Therefore, we used immunodetection to (reducing both anxiety and learned helplessness). Therefore, determine AMPA-R subunit localization in synaptic mem- the 0.7 mg/kg dose of flx was chosen for subsequent analy- branes, which are the cellular fractions from which PSDs

E. Ampuero et al. / Neuroscience 169 (2010) 98 –108 are obtained by delipidation. Synaptic membranes corre-sponded to synaptosomes that had been subjected to anosmotic shock to release intracellular membranes (i.e.,their microsomal compartment). For increased sensitivityin this analysis compared to WBs, we stained synapticmembranes fixed on slides and quantified the levels ofreceptor subunits present exclusively in membrane com-partments containing the scaffolding protein ProSAP2/Shank3, a reliable marker of excitatory synapses As a positive control, we quantified the co-localization ofNR2A and ProSAP2. Consistent with the WBs and immu-nohistochemical analysis, we found a significant increaseof NR2A localization at excitatory synapses after flx treat-ment left panel, P⬍0.01). Interestingly, the size ofthe stained membranous structures, corresponding tospines that frequently retain their presynaptic terminal, wasaugmented right panel, P⬍0.05). This observa-tion was consistent with the idea that NR2A subunits arepresent on larger spines. A similar analysis for GluR1 andGluR2 reflected no change in GluR1/ProSAP2 co-localiza-tion but a significant increase in GluR2/ProSAP2 colocal-ization (P⬍0.05). In both cases, the synaptic membranearea increased (P⬍0.05). These results suggested thatAMPA-Rs present in the spine membrane preferentiallycontain GluR2 subunits and confirm that NMDA-Rs pref-erentially contain NR2A over NR2B. To test whether thesubunit composition of N-methyl-D-aspartate (NMDA) andAMPA-Rs changed following flx, we co-immunoprecipi-tated relevant subunits from PSDs, synaptic membranesand synaptosomes as starting material. showsrepresentative immunoprecipitations of NR1 and GluR2subunits from PSDs. Following flx, NR2A increased by4.50⫾0.88 times over control in NR1 immunoprecipitates(n⫽3, P⬍0.05) and no change was observed for NR2B.
GluR1 decreased in GluR2 immunoprecipitates to 0.62⫾0.11when compared to sal (n⫽4, P⬍0.05). This was confirmedby immunoprecipitations from synaptic membranes inwhich the decrease of GluR1 was found to be 0.59⫾0.13(n⫽3, P⬍0.05). However, when synaptosomes (i.e. syn-aptic membranes containing intracellular membranes)were taken as starting material for immunoprecipitation(n⫽4), no significant differences in AMPA-R stoichiometrywere found (1.07⫾0.14). These results indicate that at thelevel of the spine membrane, flx induced a switch towardsAMPA-Rs preferentially enriched in GluR2 subunits and of Fig. 2. Fluoxetine induces region-specific changes in glutamate receptor
levels. (A) WBs of glutamate receptor subunits (NR2A, NR2B, NR1,
NMDAR-Rs enriched in NR2A subunits. In contrast, GluR1 GluR1 and GluR2) and scaffolding proteins (PSD-95 and SAP102) in was increased but it might accumulate in intracellular lo- cortical homogenates are shown. The left panel shows representative cations. Taken together, these results confirm that GluR1 western blots for each protein, and the right panel shows the mean might not reach dendritic spine membranes or PSDs even change of control ⫾ SEM obtained by densitometric quantification of though immunohistochemistry and WBs demonstrated that WBs. The number of independent experiments is indicated. (B) Gluta-mate receptor subunits were immunodetected in coronal sections through it was increased.
the rat cortex in the indicated brain areas (after Paxinos and Watson,2008) (RSGb: retrosplenial granular cortex b; PrL: prelimbic cortex). Be- Changes in dendritic spine density and morphology
low, representative photomicrographs of NR2A and GluR2-stained sec- after 28 days of flx
tions are shown (scale bar: 25 ␮m). (C) Quantification of immunopositivecells expressed as change over control. Significant increases were found To determine whether changes in glutamate receptor sub- in the number of NR2A, NR2B, GluR1 and GluR2 positive cells in the unit composition were associated with changes in spine RSGb but not in the PrL. Results are presented as mean⫾SEM and were morphology, we performed Golgi stainings in the described determined from analysing 4 – 6 sections of five rats per group. Data werestatistically evaluated with Mann–Whitney U-test, * P⬍0.05; ** P⬍0.01.
cortical subregions and movie in SupplementaryMaterial). shows representative inverted images,

E. Ampuero et al. / Neuroscience 169 (2010) 98 –108 Fig. 3. Fluoxetine-induced changes on AMPA and NMDA-R subunit levels. (A) Western blots of glutamate receptor subunits in isolated postsynaptic
densities (PSDs) are shown. Left panel shows representative western blots for each protein, and the right panel shows the mean change of control ⫾ SEM
obtained by densitometric quantification of western blots. Note that GluR1 decreases in PSDs. The number of independent determinations is indicated. (B)
Quantification of glutamate receptor subunits detected by western blots in the microsomal fraction (P3), which is enriched in intracellular membranes but not
PSDs. (C) Immunocytochemical identification of NR2A (green) in the glutamatergic population of synaptic membranes (identified as ProSAP2/Shank3
immunoreactive; red). In the superimposed picture of the double immunocytochemical labelling (merge) of these representative fields, co-localization was
quantified (scale bar: 5 ␮m). (D) The left panel shows the quantification of co-localization of five different fields per slide from four independent experiments
using different membrane preparations and confirmed the augmented co-localization of NR2A and GluR2 but not of GluR1 with isolated glutamatergic
synapse compartments. In the right panel, the size of the double-positive membrane compartments was calculated. Note that the synaptic membranes
correspond to synaptosomes that were subjected to a hypo-osmotic shock in order to release intracellular membranes. Results are presented as
mean⫾SEM. Data were statistically evaluated with a paired t-test, * P⬍0.05; ** P⬍0.01. (E) Equal amounts of NR1 and GluR2 were immunoprecipitated from
PSDs to evaluate changes in receptor stoichiometry. WBs of the corresponding subunits are shown. In NR1 immunoprecipitates, a large increase in NR2A
(4.5⫾0.88; P⬍0.05) was found whereas a decrease of GluR1 (0.62⫾0.11; P⬍0.05) was found in GluR2-containing AMPA-Rs (n⫽3).
taken at three different focal planes, of a primary dendrite the RSGb, which was the same region that revealed sig- present in the RSGb of sal and flx treated animals and nificant increases in glutamate receptor subunits excluding indicates the main morphological types (filopodia/thin, NR2B. To evaluate whether this is a more general phe- stubby and mushroom). Spine density was measured from nomenon, secondary dendrites in these areas were exam- its emergence from the soma along an 80 ␮m stretch of the ined (not shown), and the analysis was extended to the primary apical dendrite that was subdivided into 10 seg- secondary motor cortex (M2, supplementary material), ments The number of spines per 8 ␮m segment which is not thought to be directly involved in the modula- increased beginning at the fifth segment in the RSGb and tion of depressive behaviours. For secondary dendrites, at the seventh segment in the PrL. This was accompanied shorter segments (from 32 to 80 ␮m) were measured with by a robust increase in the percentage of mushroom-type the premise that the selected segment had to be continu- spines and a decrease of thin spines in the RSGb ous with the primary dendrite and the cell soma. In the right panel, P⬍0.01). In the PrL, thin spines also de- case of the RSGb, total spine density (measured as the creased and stubby spines increased (P⬍0.05), but there number of spines per ␮m) increased from 0.70⫾0.04 to was no change in mushroom-type spines. Therefore, the 0.84⫾0.04 (P⬍0.05). Mushroom-type spines increased change in spine morphology was especially significant in from 32.9% to 43.7% (P⬍0.05) with a concomitant de-

E. Ampuero et al. / Neuroscience 169 (2010) 98 –108 Fig. 4. The effect of flx on spine density and morphology. Cortical pyramidal neurons were visualized by the rapid Golgi impregnation method. (A)
Representative images taken at three different focal planes in sal and flx-treated rats were inverted for better visualization. The original image, used
for quantification at 12 different focal planes per dendritic segment, is presented in the supplementary material. The three shape categories of spines
were shown as follows: filopodia/thin (T), stubby (S) and mushroom/branched (M). The numbers in parentheses indicate superimposed spines that
were resolved by observation at different focal planes: (2) represents a thin and a mushroom-type spine in both cases and (3) represents a thin, a
stubby and a mushroom-type spine. See experimental procedures and the movie in supplementary material for details. The scale bar was 1 ␮m. (B)
The left panels show spine number along primary dendrites in consecutive dendritic segments of 8 ␮m away from the soma in the RSGb and the PrL.
The right panels show the classification of dendritic spines, which revealed a significant increase of mushroom-type spines in the RSGb, but no
differences in the PrL, between sal and flx treated animals. All results are presented as mean⫾SEM determined from analysing 18 –23 cells/area
obtained from five rats per group, * P⬍0.05; ** P⬍0.01.
crease of thin and stubby-like spines (P⬍0.05). In the PrL, Taken together, the results indicated that flx induced no change in secondary spine density was found, and this morphological and molecular changes affecting large fore- was associated with no change in spine morphology. In- brain areas. Increased mushroom-type spines in selected terestingly, the changes in both spine morphology and regions were accompanied by global increases of NR2A- glutamate receptor subunit levels extended to layer II/III of containing NMDA-Rs and of GluR2-containing AMPA-Rs.
M2 (Suppl. Fig. 2).
GluR1 subunits, although upregulated, might be retained E. Ampuero et al. / Neuroscience 169 (2010) 98 –108 in the cell soma, or they might be located near, or even mance). In the future, behavioural studies of flx dosing within, spines. Surprisingly, such changes were highly sig- should be undertaken in animals in which depressive-like nificant in the RSGb but less prominent in the PrL cortex.
behaviour is induced (e.g., by exposure to chronic stress).
Glutamate receptors and antidepressants
A strong relationship between mood disorders and gluta- Our results were the first to demonstrate that long-term mate neurotransmission has already been postulated behaviourally relevant doses of flx induced an enrichment of specific glutamate receptor subunits in the rat forebrain.
like activity of NMDA-R antagonists and participation of Whereas GluR2 and NR2A increased in spine membranes AMPA-Rs, especially its GluR2 subunit and PSDs, GluR1, although increased in homogenates In agreement with our results, imipra- and in immunohistochemical analysis, did not reach these mine upregulated synaptic GluR2 and synaptic compartments. Changes in subunits levels were increased GluR1 in synaptosomes In- associated with an increase of mushroom-type dendritic creased GluR1 levels were found in a hippocampal syn- spines in a region-specific manner and were especially aptic fraction obtained from ovariectomized, but not con- significant in the RSGb. The molecular changes observed trol, rats treated with 5 mg/kg flx for 21 days here may underlie the restoration of plasticity previously ob- The difference in our results may be due to the flx served in the visual cortex following flx dose or the time course of the molecular adaptations. Flx These changes are compatible with a "maturation- might counteract the decrease of NR2A, but not NR2B, like" process leading to stabilization of synaptic connections observed in major depression that may be related to a functional recovery of glutamatergic or the decreases of NR1, NR2A and SAP102 forebrain networks observed in bipolar disorder Our behavioural data showed that the flx dose used in the long-term studies is an important variable to be con- NMDA and AMPA-Rs may be particularly important targets sidered. Consistent with our results, a similar low-dose of for the treatment of mood disorders, we postulated that flx (1 mg/kg) induced antidepressant-like effects and effec- subunit-specific strategies should be considered.
tively modulated neuronal firing rates Role of glutamate receptor subunits in structural and
Conflicting data re- garding mechanisms involved in antidepressant actionmight be due to the use of potentially harmful doses The importance of NMDA-Rs in synaptic plasticity and Increased fear has been reported memory are well described following 10 mg/kg flx treatment in the EPM However, the potential opposing contributions of NR2A Our results indicated that 3.5 mg/kg flx was not and NR2B subunits to LTP versus LTD have been highly effective in reducing anxiety-like behaviour in the EPM or controversial LTP itself might rearing but was effective in the NSF. Although both the induce an immediate switch favouring NR2A over NR2B- NSF and EPM are conflict tests to assess anxiety, they containing receptors recruit different neuronal circuits. NSF depends crucially The switch of NR2B to NR2A containing on appetite drive whereas the EPM depends on explor- NMDA-Rs during development is accompanied by synap- atory behaviour Consistent with our tic maturation, stabilization and growth findings, contrasting results using both tests have previ- In contrast to NR2B subunits, NR2A subunits are ously been reported, and it has been proposed that the predominantly present on large spines EPM may show a decreased sensitivity to chronic antide- In accordance with this, the pressant treatment as well as high drug doses synaptic enrichment of NR2A induced by flx was accom- A differential dose-dependent effect was confirmed panied by a higher proportion of mushroom-type spines, in the FST. Climbing, a behaviour considered to be depen- suggesting that flx might induce a maturation-like state.
dent on noradrenergic neurotransmission, was favoured NMDA-Rs regulate synaptic strength by controlling the over swimming after 3.5 flx but not 0.7 flx trafficking of AMPA-Rs in and out of postsynaptic sites In addition, both flx doses Synaptic strengthening in- caused opposing trends on rearing behaviour that were not volves activity-dependent addition of GluR subunits con- significant when compared to sal: treatment with 0.7 flx taining a long intracellular C-terminal tail (e.g. GluR1-con- induced a decrease in the amount of rearing, and 3.5 flx taining) to synapses, whereas short-tailed subunits (GluR2 treatment induced an increase resulting in a significant and GluR3) constitutively replace existing receptors.
difference between the 0.7 and 3.5 flx doses (P⬍0.05).
Therefore, inserted GluR1 subunits are later replaced by Increased rearing has been shown to correlate positively GluR2, although the requirement of this transient process with anxiety-like behaviour is currently debated A temporal Our results suggested that the 3.5 dose, in characterization of synaptic changes induced by flx would addition to modulating noradrenergic neurotransmission, be necessary to determine whether the switch towards might negatively affect anxiety-like behaviour (such as GluR2-containing receptors required a preceding GluR1 rearing) or have no effect (for instance, on EPM perfor- insertion. The fact that we did not find changes in GluR1 or E. Ampuero et al. / Neuroscience 169 (2010) 98 –108 NR1 content in PSDs 2 weeks after flx treatment indicates ies it was found that the antagonists acted in an excitatory that these adaptations, measured after 4 weeks, are slowly manner because the blocked NMDA-Rs were localized on GABAergic interneurons A highly positive correlation also exists between GluR2 levels and synaptic size, spine density and mEPSC fre- It is possible that the thera- quency, a relationship that has not been established for peutic effect of NMDA receptor antagonists, at the sub- anaesthetic doses currently under investigation for the In addition, GluR2 is more stably tethered to the treatment of depression, is in part due to their pro-excita- synapse and its incorporation is necessary for the long-term tory effect in the retrosplenial cortex expression of synaptic plasticity In line with these re- The increased levels of the scaffolding proteins sults, the gray matter in bipolar disorder was reduced in the PSD-95 and SAP102 in homogenates and PSDs are consis- posterior cingulate/retrosplenial cortex of unmedicated tent with their role in anchoring glutamate receptors and subjects relative to medicated patients delivering critical elements to growing spines.
The interesting possibility that the plastic changes It is likely that the flx-induced effects on spine morphol- induced by flx in this cerebrocortical region might be re- ogy and glutamate receptor content were mediated by lated with its positive effects on emotion and cognition neurotrophic factors Several growth should be investigated in the near future. In general, these factors, including BDNF and vascular endothelial growth findings underscore the importance of discriminating be- factor (VEGF) and their signalling pathways are necessary tween cortical subregions affected by antidepressants.
for a response to antidepressant drug treatment Several questions should be addressed in the future. For example, what are the cellular mechanisms that induce a specific contribution to the observed changes needs to be coordinated switch of both NMDA and AMPA-R subunits? addressed in the future.
Are plastic changes in the glutamatergic system causallyrelated to mood recovery? Can we identify brain circuits that Morphological changes in the forebrain
are specifically related to a subset of depressive behaviours,and are the RSGb and M2 part of such circuits? In the future, Animal models currently used to elicit depressive-like relevant brain circuitries might be preferentially targeted by symptoms lead to structural changes in neuronal networks, antidepressant treatments (e.g., by transcranial magnetic including dendritic length and complexity and spine mor- stimulation) in a personalized manner depending on individ- ual symptomatology Our findings might These changes can be reversed by antidepressant contribute to the search of new and faster-acting antidepres- sant interventions that target specific glutamate receptor sub- gesting that the flx-induced spine remodelling that we ob- units in concert as well as forebrain circuits critically involved served is involved in antidepressant action. Increases in in depressive symptoms.
spine density and mushroom-type spines in the RSGb and The delayed molecular adaptations in extensive corti- M2 might reflect enhanced basal neurotransmission. We cal networks reported here may underlie the therapeutic observed region-specific changes that correlated with action of antidepressants. Our findings that a widely pre- changes in glutamate receptor subunits. However, these scribed antidepressant in humans induces structural and effects probably extend to wide forebrain areas because molecular plasticity in the adult forebrain suggests a po- changes in glutamate receptor subunits could be detected tential clinical application for antidepressants in neurolog- in subcellular fractions obtained from whole forebrain ho- ical disorders in which synaptic function is compromised.
mogenates. Consistent with this idea, flx induced structuralplasticity in the rat somatosensory cortex Cognitive-emotional behaviours rely on complex Acknowledgments—This work was supported by Proyecto AnilloACT09-06 (U Wyneken). We are grateful to Soledad Sandoval for interactions of networks in several brain areas The contribution of specific areas to depressive-likebehaviours and antidepressant treatment is not completely understood. Besides mood disturbance, depression is ac-companied by sensorimotor disturbances, and manipulat- Adesnik H, Nicoll RA (2007) Conservation of glutamate receptor ing the motor system (e.g., by physical exercise) improves 2-containing AMPA receptors during long-term potentiation. J Neu-rosci 27:4598 – 4602.
mood It is therefore conceivable that the Al-Hallaq RA, Conrads TP, Veenstra TD, Wenthold RJ (2007) NMDA structural modulation of M2 has some therapeutic signifi- di-heteromeric receptor populations and associated proteins in rat cance. In turn, the RSGb is connected with cortices known hippocampus. J Neurosci 27:8334 – 8343.
to be involved in emotional control, such as the hippocam- Alvarez VA, Ridenour DA, Sabatini BL (2007) Distinct structural and pal formation and the midline limbic cortices ionotropic roles of NMDA receptors in controlling spine and syn- Activation of the RSGb by emotionally apse stability. J Neurosci 27:7365–7376.
Ampuero E, Dagnino-Subiabre A, Sandoval R, Zepeda-Carreno R, salient stimuli has already been shown Sandoval S, Viedma A, Aboitiz F, Orrego F, Wyneken U (2007) Interestingly, the retrosplenial cortex has been implicated Status epilepticus induces region-specific changes in dendritic in an experimental model of schizophrenia induced by spines, dendritic length and TrkB protein content of rat brain cortex.
NMDA receptor antagonists. Unexpectedly, in these stud- Brain Res 1150:225–238.
E. Ampuero et al. / Neuroscience 169 (2010) 98 –108 Antal M, Fukazawa Y, Eordogh M, Muszil D, Molnar E, Itakura M, Gray NA, Du J, Falke CS, Yuan P, Manji HK (2003) Lithium regulates Takahashi M, Shigemoto R (2008) Numbers, densities, and colo- total and synaptic expression of the AMPA glutamate receptor calization of AMPA- and NMDA-type glutamate receptors at indi- GluR2 in vitro and in vivo. Ann N Y Acad Sci 1003:402– 404.
vidual synapses in the superficial spinal dorsal horn of rats. J Neu- Grosshans DR, Clayton DA, Coultrap SJ, Browning MD (2002) LTP leads to rapid surface expression of NMDA but not AMPA recep- Bellone C, Nicoll RA (2007) Rapid bidirectional switching of synaptic tors in adult rat CA1. Nat Neurosci 5:27–33.
NMDA receptors. Neuron 55:779 –785.
Guirado R, Varea E, Castillo-Gomez E, Gomez-Climent MA, Rovira- Beneyto M, Meador-Woodruff JH (2008) Lamina-specific abnormali- Esteban L, Blasco-Ibanez JM, Crespo C, Martinez-Guijarro FJ, ties of NMDA receptor-associated postsynaptic protein transcripts Nacher J (2009) Effects of chronic fluoxetine treatment on the rat in the prefrontal cortex in schizophrenia and bipolar disorder. Neu- somatosensory cortex: activation and induction of neuronal struc- tural plasticity. Neurosci Lett 457:12–15.
Bessa JM, Ferreira D, Melo I, Marques F, Cerqueira JJ, Palha JA, Hajszan T, Szigeti-Buck K, Sallam NL, Bober J, Parducz A, Maclusky Almeida OF, Sousa N (2009) The mood-improving actions of an- NJ, Leranth C, Duman RS (2010) Effects of estradiol on learned tidepressants do not depend on neurogenesis but are associated helplessness and associated remodeling of hippocampal spine with neuronal remodeling. Mol Psychiatry 14:764 –773, 739.
synapses in female rats. Biol Psychiatry 67:168 –174.
Bleakman D, Alt A, Witkin JM (2007) AMPA receptors in the therapeu- Harris KM, Jensen FE, Tsao B (1992) Three-dimensional structure of tic management of depression. CNS Neurol Disord Drug Targets dendritic spines and synapses in rat hippocampus (CA1) at postnatal day 15 and adult ages: implications for the maturation of synaptic Borta A, Schwarting RK (2005) Inhibitory avoidance, pain reactivity, physiology and long-term potentiation. J Neurosci 12:2685–2705.
and plus-maze behavior in Wistar rats with high versus low rearing Holmes A, Kinney JW, Wrenn CC, Li Q, Yang RJ, Ma L, Vishwanath activity. Physiol Behav 84:387–396.
J, Saavedra MC, Innerfield CE, Jacoby AS, Shine J, Iismaa TP, Canbeyli R (2010) Sensorimotor modulation of mood and depression: Crawley JN (2003) Galanin GAL-R1 receptor null mutant mice an integrative review. Behav Brain Res 207:249 –264.
display increased anxiety-like behavior specific to the elevated Castren E (2005) Is mood chemistry? Nat Rev Neurosci 6:241–246.
Castren E, Voikar V, Rantamaki T (2007) Role of neurotrophic factors Holsboer F (2008) How can we realize the promise of personalized in depression. Curr Opin Pharmacol 7:18 –21.
antidepressant medicines? Nat Rev Neurosci 9:638 – 646.
Ciruela F, Casado V, Rodrigues RJ, Lujan R, Burgueno J, Canals M, Isaac JT, Ashby M, McBain CJ (2007) The role of the GluR2 subunit in Borycz J, Rebola N, Goldberg SR, Mallol J, Cortes A, Canela EI, AMPA receptor function and synaptic plasticity. Neuron 54:859 – 871.
Lopez-Gimenez JF, Milligan G, Lluis C, Cunha RA, Ferre S, Franco R Kessels HW, Malinow R (2009) Synaptic AMPA receptor plasticity and (2006) Presynaptic control of striatal glutamatergic neurotransmission behavior. Neuron 61:340 –350.
by adenosine A1-A2A receptor heteromers. J Neurosci 26:2080 – Kobayashi C, Aoki C, Kojima N, Yamazaki H, Shirao T (2007) Drebrin a content correlates with spine head size in the adult mouse Contreras CM, Rodriguez-Landa JF, Gutierrez-Garcia AG, Bernal- cerebral cortex. J Comp Neurol 503:618 – 626.
Morales B (2001) The lowest effective dose of fluoxetine in the Kopec C, Malinow R (2006) Neuroscience. Matters of size. Science forced swim test significantly affects the firing rate of lateral septal 314:1554 –1555.
nucleus neurones in the rat. J Psychopharmacol 15:231–236.
Krishnan V, Nestler EJ (2008) The molecular neurobiology of depres- Cryan JF, Markou A, Lucki I (2002) Assessing antidepressant activity sion. Nature 455:894 –902.
in rodents: recent developments and future needs. Trends Phar- Kumar P, Waiter G, Ahearn T, Milders M, Reid I, Steele JD (2008) macol Sci 23:238 –245.
Abnormal temporal difference reward-learning signals in major Cryan JF, O'Leary OF, Jin SH, Friedland JC, Ouyang M, Hirsch BR, depression. Brain 131:2084 –2093.
Page ME, Dalvi A, Thomas SA, Lucki I (2004) Norepinephrine- Lira A, Zhou M, Castanon N, Ansorge MS, Gordon JA, Francis JH, deficient mice lack responses to antidepressant drugs, including Bradley-Moore M, Lira J, Underwood MD, Arango V, Kung HF, selective serotonin reuptake inhibitors. Proc Natl Acad Sci U S A Hofer MA, Hen R, Gingrich JA (2003) Altered depression-related 101:8186 – 8191.
behaviors and functional changes in the dorsal raphe nucleus of Chen W, Prithviraj R, Mahnke AH, McGloin KE, Tan JW, Gooch AK, Inglis FM (2009) AMPA glutamate receptor subunits 1 and 2 reg- serotonin transporter-deficient mice. Biol Psychiatry 54:960 –971.
ulate dendrite complexity and spine motility in neurons of the Liston C, Miller MM, Goldwater DS, Radley JJ, Rocher AB, Hof PR, developing neocortex. Neuroscience 159:172–182.
Morrison JH, McEwen BS (2006) Stress-induced alterations in pre- Diaz-Veliz G, Mora S, Gomez P, Dossi MT, Montiel J, Arriagada C, frontal cortical dendritic morphology predict selective impairments in Aboitiz F, Segura-Aguilar J (2004) Behavioral effects of manganese perceptual attentional set-shifting. J Neurosci 26:7870 –7874.
injected in the rat substantia nigra are potentiated by dicumarol, a Lucki I (1997) The forced swimming test as a model for core and DT-diaphorase inhibitor. Pharmacol Biochem Behav 77:245–251.
component behavioral effects of antidepressant drugs. Behav Dickerson J, Sharp FR (2006) Atypical antipsychotics and a Src kinase inhibitor (PP1) prevent cortical injury produced by the psychomi- Maddock RJ (1999) The retrosplenial cortex and emotion: new insights metic, noncompetitive NMDA receptor antagonist MK-801. Neuro- from functional neuroimaging of the human brain. Trends Neurosci psychopharmacology 31:1420 –1430.
22:310 –316.
Du J, Creson TK, Wu LJ, Ren M, Gray NA, Falke C, Wei Y, Wang Y, Malenka RC, Bear MF (2004) LTP and LTD: an embarrassment of Blumenthal R, Machado-Vieira R, Yuan P, Chen G, Zhuo M, Manji riches. Neuron 44:5–21.
HK (2008) The role of hippocampal GluR1 and GluR2 receptors in Maya Vetencourt JF, Sale A, Viegi A, Baroncelli L, De Pasquale R, manic-like behavior. J Neurosci 28:68 –79.
O'Leary OF, Castren E, Maffei L (2008) The antidepressant fluox- Du J, Gray NA, Falke CA, Chen W, Yuan P, Szabo ST, Einat H, Manji etine restores plasticity in the adult visual cortex. Science HK (2004) Modulation of synaptic plasticity by antimanic agents: the role of AMPA glutamate receptor subunit 1 synaptic expres- McClung CA, Nestler EJ (2008) Neuroplasticity mediated by altered sion. J Neurosci 24:6578 – 6589.
gene expression. Neuropsychopharmacology 33:3–17.
Gould TD, O'Donnell KC, Dow ER, Du J, Chen G, Manji HK (2008) McCullumsmith RE, Kristiansen LV, Beneyto M, Scarr E, Dean B, Involvement of AMPA receptors in the antidepressant-like effects Meador-Woodruff JH (2007) Decreased NR1, NR2A, and SAP102 of lithium in the mouse tail suspension test and forced swim test.
transcript expression in the hippocampus in bipolar disorder. Brain Res 1127:108 –118.
E. Ampuero et al. / Neuroscience 169 (2010) 98 –108 McEwen BS (2007) Physiology and neurobiology of stress and adap- Passafaro M (2007) Extracellular interactions between GluR2 and tation: central role of the brain. Physiol Rev 87:873–904.
N-cadherin in spine regulation. Neuron 54:461– 477.
Medvedev NI, Rodriguez-Arellano JJ, Popov VI, Davies HA, Tigaret Sairanen M, O'Leary OF, Knuuttila JE, Castren E (2007) Chronic CM, Schoepfer R, Stewart MG (2008) The glutamate receptor 2 antidepressant treatment selectively increases expression of plas- subunit controls post-synaptic density complexity and spine shape ticity-related proteins in the hippocampus and medial prefrontal in the dentate gyrus. Eur J Neurosci 27:315–325.
cortex of the rat. Neuroscience 144:368 –374.
Monyer H, Burnashev N, Laurie DJ, Sakmann B, Seeburg PH (1994) Sanacora G, Zarate CA, Krystal JH, Manji HK (2008) Targeting the Developmental and regional expression in the rat brain and func- glutamatergic system to develop novel, improved therapeutics for tional properties of four NMDA receptors. Neuron 12:529 –540.
mood disorders. Nat Rev Drug Discov 7:426 – 437.
Morimoto T, Hashimoto K, Yasumatsu H, Tanaka H, Fujimura M, Schulz D, Buddenberg T, Huston JP (2007) Extinction-induced "despair" Kuriyama M, Kimura K, Takehara S, Yamagami K (2002) Neuro- in the water maze, exploratory behavior and fear: effects of chronic pharmacological profile of a novel potential atypical antipsychotic antidepressant treatment. Neurobiol Learn Mem 87:624 – 634.
drug Y-931 (8-fluoro-12-(4-methylpiperazin-1-yl)-6H-[1] benzothi- Shepherd JD, Huganir RL (2007) The cell biology of synaptic plasticity: AMPA receptor trafficking. Annu Rev Cell Dev Biol 23:613– 643.
eno[2, 3-b] [1,5] benzodiazepine maleate). Neuropsychopharma- Shinohara Y, Hirase H, Watanabe M, Itakura M, Takahashi M, Shige- cology 26:456 – 467.
moto R (2008) Left-right asymmetry of the hippocampal synapses Muigg P, Hetzenauer A, Hauer G, Hauschild M, Gaburro S, Frank E, with differential subunit allocation of glutamate receptors. Proc Natl Landgraf R, Singewald N (2008) Impaired extinction of learned fear Acad Sci U S A 105:19498 –19503.
in rats selectively bred for high anxiety— evidence of altered neu- Sousa N, Almeida OF, Wotjak CT (2006) A hitchhiker's guide to ronal processing in prefrontal-amygdala pathways. Eur J Neurosci behavioral analysis in laboratory rodents. Genes Brain Behav 5 28:2299 –2309.
(Suppl 2):5–24.
Nugent AC, Milham MP, Bain EE, Mah L, Cannon DM, Marrett S, Tanaka J, Horiike Y, Matsuzaki M, Miyazaki T, Ellis-Davies GC, Kasai Zarate CA, Pine DS, Price JL, Drevets WC (2006) Cortical abnor- H (2008) Protein synthesis and neurotrophin-dependent structural malities in bipolar disorder investigated with MRI and voxel-based plasticity of single dendritic spines. Science 319:1683–1687.
morphometry. Neuroimage 30:485– 497.
Vaisanen J, Linden AM, Lakso M, Wong G, Heinemann U, Castren E O'Leary OF, Wu X, Castren E (2009) Chronic fluoxetine treatment (1999) Excitatory actions of NMDA receptor antagonists in rat increases expression of synaptic proteins in the hippocampus of entorhinal cortex and cultured entorhinal cortical neurons. Neuro- the ovariectomized rat: role of BDNF signalling. Psychoneuroen- Van Groen T, Wyss JM (2003) Connections of the retrosplenial gran- Olney JW, Farber NB (1995) NMDA antagonists as neurotherapeutic ular b cortex in the rat. J Comp Neurol 463:249 –263.
drugs, psychotogens, neurotoxins, and research tools for studying Velisek L (2006) Prenatal exposure to betamethasone decreases anx- iety in developing rats: hippocampal neuropeptide y as a target Olney JW, Labruyere J, Wang G, Wozniak DF, Price MT, Sesma MA molecule. Neuropsychopharmacology 31:2140 –2149.
(1991) NMDA antagonist neurotoxicity: mechanism and preven- Vizi ES, Zsilla G, Caron MG, Kiss JP (2004) Uptake and release of tion. Science 254:1515–1518.
norepinephrine by serotonergic terminals in norepinephrine trans- Passafaro M, Nakagawa T, Sala C, Sheng M (2003) Induction of porter knock-out mice: implications for the action of selective se- dendritic spines by an extracellular domain of AMPA receptor rotonin reuptake inhibitors. J Neurosci 24:7888 –7894.
subunit GluR2. Nature 424:677– 681.
Warner-Schmidt JL, Duman RS (2008) VEGF as a potential target for Paxinos G, Watson C (1998) The Rat brain in stereotaxic coordinates.
therapeutic intervention in depression. Curr Opin Pharmacol San Diego: Academic Press.
Pellow S, Chopin P, File SE, Briley M (1985) Validation of open: closed Wyneken U, Sandoval M, Sandoval S, Jorquera F, Gonzalez I, arm entries in an elevated plus-maze as a measure of anxiety in Vargas F, Falcon R, Monari M, Orrego F (2006) Clinically rele- the rat. J Neurosci Methods 14:149 –167.
vant doses of fluoxetine and reboxetine induce changes in the Pessoa L (2008) On the relationship between emotion and cognition.
TrkB content of central excitatory synapses. Neuropsychophar- Nat Rev Neurosci 9:148 –158.
Pittenger C, Duman RS (2008) Stress, depression, and neuroplastic- Wyneken U, Smalla KH, Marengo JJ, Soto D, de la Cerda A, Tischmeyer ity: a convergence of mechanisms. Neuropsychopharmacology W, Grimm R, Boeckers TM, Wolf G, Orrego F, Gundelfinger ED 33:88 –109.
(2001) Kainate-induced seizures alter protein composition and N- Popoli M, Gennarelli M, Racagni G (2002) Modulation of synaptic plas- methyl-D-aspartate receptor function of rat forebrain postsynaptic ticity by stress and antidepressants. Bipolar Disord 4:166 –182.
densities. Neuroscience 102:65–74.
Porsolt RD, Anton G, Blavet N, Jalfre M (1978) Behavioural despair in Yao Y, Kelly MT, Sajikumar S, Serrano P, Tian D, Bergold PJ, Frey JU, rats: a new model sensitive to antidepressant treatments. Eur Sacktor TC (2008) PKM zeta maintains late long-term potentiation J Pharmacol 47:379 –391.
by N-ethylmaleimide-sensitive factor/GluR2-dependent trafficking Radley JJ, Rocher AB, Miller M, Janssen WG, Liston C, Hof PR, of postsynaptic AMPA receptors. J Neurosci 28:7820 –7827.
McEwen BS, Morrison JH (2006) Repeated stress induces den- Yashiro K, Philpot BD (2008) Regulation of NMDA receptor subunit expression and its implications for LTD, LTP, and metaplasticity.
dritic spine loss in the rat medial prefrontal cortex. Cereb Cortex Rodriguez-Landa JF, Contreras CM, Gutierrez-Garcia AG, Bernal-Mo- rales B (2003) Chronic, but not acute, clomipramine or fluoxetinetreatment reduces the spontaneous firing rate in the mesoaccumbens neurons of the rat. Neuropsychobiology 48:116 –123.
Saglietti L, Dequidt C, Kamieniarz K, Rousset MC, Valnegri P, Supplementary data associated with this article can be found, in Thoumine O, Beretta F, Fagni L, Choquet D, Sala C, Sheng M, the online version, at doi: (Accepted 16 April 2010) (Available online 22 April 2010)

Source: http://www.stress.cl/pdf/17.pdf