Adult hippocampal neuroplasticity triggers susceptibility to recurrent depression

Abstract

Depression is a highly prevalent and recurrent neuropsychiatric disorder associated with alterations in emotional and cognitive domains. Neuroplastic phenomena are increasingly considered central to the etiopathogenesis of and recovery from depression. Nevertheless, a high number of remitted patients experience recurrent episodes of depression, remaining unclear how previous episodes impact on behavior and neuroplasticity and/or whether modulation of neuroplasticity is important to prevent recurrent depression. Through re-exposure to an unpredictable chronic mild stress protocol in rats, we observed the re-appearance of emotional and cognitive deficits. Furthermore, treatment with the antidepressants fluoxetine and imipramine was effective to promote sustained reversion of a depressive-like phenotype; however, their differential impact on adult hippocampal neuroplasticity triggered a distinct response to stress re-exposure: while imipramine re-established hippocampal neurogenesis and neuronal dendritic arborization contributing to resilience to recurrent depressive-like behavior, stress re-exposure in fluoxetine-treated animals resulted in an overproduction of adult-born neurons along with neuronal atrophy of granule neurons, accounting for an increased susceptibility to recurrent behavioral changes typical of depression. Strikingly, cell proliferation arrest compromised the behavior resilience induced by imipramine and buffered the susceptibility to recurrent behavioral changes promoted by fluoxetine. This study shows that previous exposure to a depressive-like episode impacts on the behavioral and neuroanatomical changes triggered by subsequent re-exposure to similar experimental conditions and reveals that the proper control of adult hippocampal neuroplasticity triggered by antidepressants is essential to counteract recurrent depressive-like episodes.

Figure 1

Anhedonic- and anxiety-like behaviors induced by recurrent stress exposure are prevented by imipramine treatment, but not fluoxetine. (a) Schematic representation of the experimental timeline used, including behavioral assessments throughout the protocol and the respective treatments. (b–d) Longitudinal assessment of anhedonic-like behavior by the SPT, revealed an increased susceptibility after re-exposure to stress driven by fluoxetine. (e) Assessment of behavioral despair, at the end of the experimental protocol, by the FST test revealed that recurrent stress exposure induced a significant increase of immobility time, only prevented in animals treated with imipramine. (f–h) Anxiety-like behavior was continuously tested throughout the experimental protocol, at week 6 by the EPM (f) and the NSF at weeks 10 (g) and 16 (h), evidencing the efficacy of imipramine in preventing anxiety-like behavior after stress re-exposure, contrarily to fluoxetine. (i) Non-treated and particularly, fluoxetine-treated animals, subjected to repeated uCMS exposure presented elevated corticosterone levels in the serum. Basal corticosterone levels were measured in the serum of rats collected between 0800 and 0900 at the end of the protocol. See also Supplementary Figures 1 and 2. *Denotes the effect of uCMS analyzed by Student's t-test; ^#Denotes the effect of ADs, by comparison of treatment and SAL animals; and ^‡denotes differences between ADs, analyzed by one-way analysis of variance (ANOVA). Data are represented as mean±s.e.m. ^#P⩽0.05, ^{**, ##}P⩽0.01, ^{***, ###, ‡‡‡}P⩽0.001; n=6–8 animals per group. uCMS, unpredictable chronic mild stress protocol (uCMS*, slightly modified version. See Supplementary Information). AD, antidepressant; CTRL, non-stressed animals; EPM, elevated-plus maze; FLX, animals repeatedly exposed to uCMS and treated with fluoxetine; FST, forced-swimming test; IMIP, animals repeatedly exposed to uCMS and treated with imipramine; MWM, Morris water maze; NOR, novel object recognition; NSF, novelty-suppressed feeding; SAL, animals repeatedly exposed to uCMS and non-treated; SDT, sweet-drive test; SPT, sucrose preference test; TP, time point.

Figure 2

Evaluation of cognitive function throughout the experimental protocol revealed that imipramine is able to prevent cognitive deficits induced by recurrent stress exposure while fluoxetine specifically prevents alterations in behavior flexibility. (a–c) Continuous assessment of long-term memory using the novel object recognition (NOR) test revealed that recurrent stress induces cognitive deficits, which are also observed in fluoxetine-treated animals but not with imipramine treatment. (d–g) At the end of the protocol, the MWM test was used to evaluate cognitive performances, including working (d) and reference memory (e), reversal learning (f) and working memory by a probe trial (g). *Denotes the effect of uCMS analyzed by Student's t-test; ^#Denotes the effect of ADs, by comparison of treatment and SAL animals; and ^‡denotes differences between ADs analyzed by one-way analysis of variance (ANOVA). ANOVA repeated measures was used to analyze cognitive learning tasks performance. Data are represented as mean±s.e.m. ^{*, #}P⩽0.05, ^{**, ##, ‡‡}P⩽0.01, ^{***, ###, ‡‡‡}P⩽0.001; n=6–8 animals per group. AD, antidepressant; AUC, area under the curve; CTRL, non-stressed animals; FLX, animals repeatedly exposed to uCMS and treated with fluoxetine; IMIP, animals repeatedly exposed to uCMS and treated with imipramine; MWM, Morris water maze; NOR, novel object recognition; SAL, animals repeatedly exposed to uCMS and non-treated; TP, time point; uCMS, unpredictable chronic mild stress protocol.

Figure 3

Treatment with fluoxetine boosts the generation and survival of newborn hippocampal neurons even after recurrent exposure to stress. (a and b) Quantification of the number of BrdU⁺ and BrdU⁺NeuN⁺ cells, per dentate gyrus (DG) area, revealed an increased production and survival of newborn neurons evoked by fluoxetine treatment. (c) Representative coronal section of the DG stained for BrdU (in red), NeuN (in green) and DAPI (in blue). (d) Density of newborn astroglial cells, identified as BrdU⁺GFAP⁺ cells, revealed no major effect of stress or ADs treatment. (e) Representative staining for BrdU (in red), GFAP (in green) and DAPI (in blue) in hippocampal DG. (f) Analysis of relative expression levels of STAT3 (upper panel) and BMP4 (lower panel) in the macrodissected DG corroborated the absence of major effects on gliogenesis in consequence of the recurrent stress exposure or AD treatment. (g and h) Quantitative analysis of Ki-67⁺ cells and amplifying progenitors, identified as Sox-2⁺Ki-67⁺ cells, revealed that fluoxetine leads to a depletion of DG progenitor cells. (i) Representative confocal image of Sox-2 (green) and Ki-67 (red) immunostaining. (j) Analysis of the number of DCX⁺Ki-67⁺ cells, representative of the neuroblasts population in the DG showed that animals subjected to repeated stress exposure, non-treated and treated with fluoxetine, presented a decreased number of neuroblasts. (k) Coronal section of the DG stained for DCX (in green), Ki-67 (in red) and DAPI (in blue). (l) Relative mRNA expression levels of NEUROD1 (upper panel) and DCX (lower panel) in the macrodissected DG revealed a decrease in the expression levels as a consequence of repeated stress exposure. In general, treatment with ADs restored the expression levels of both makers of neuronal maturation. Scale bars represent 30 μm. *Denotes the effect of unpredictable chronic mild stress (uCMS) analyzed by Student's t-test; ^#denotes the effect of ADs, by comparison of treatment and SAL animals, analyzed by one-way analysis of variance (ANOVA). Data represented as mean±s.e.m. ^{*, #}P⩽0.05; n=± 4 animals per group. AD, antidepressant; CTRL, non-stressed animals; FLX, animals repeatedly exposed to uCMS and treated with fluoxetine; IMIP, animals repeatedly exposed to uCMS and treated with imipramine; SAL, animals repeatedly exposed to uCMS and non-treated.

Figure 4

Imipramine enhances dentate gyrus (DG) neuronal arborization, whereas fluoxetine promotes an atrophy of granule neurons after recurrent stress. Dendritic length analysis (a) and neuronal organization (b) of DG granule neurons showed a dendritic shrinkage promoted by fluoxetine treatment, whereas imipramine induced an enlargement of the neuronal arborization after stress re-exposure. (c) Representative three-dimensional (3D) morphometric reconstruction of DG granule neurons of each experimental group. (d and e) The relative gene expression levels of remodeling genes, NCAM and SYN1, corroborates the neuronal remodeling promoted by imipramine treatment. Scale bars represent 50 μm. *Denotes the effect of unpredictable chronic mild stress (uCMS) analyzed by Student's t-test. ^#Denotes the effect of ADs, by comparison of treatment and SAL animals; and ^‡denotes differences between ADs, analyzed by one-way analysis of variance (ANOVA). ANOVA repeated measures was used to analyze Sholl analysis. Data are represented as mean±s.e.m. ^{#, ‡}P⩽0.05, ^{**, ‡‡}P⩽0.01, ^{***, ###, ‡‡‡}P⩽0.001; n=± 4 animals per group. AD, antidepressant; CTRL, non-stressed animals; FLX, animals repeatedly exposed to uCMS and treated with fluoxetine; IMIP, animals repeatedly exposed to uCMS and treated with imipramine; SAL, animals repeatedly exposed to uCMS and non-treated.

Figure 5

Ablation of adult hippocampal neurogenesis by MAM rescues behavioral deficits induced by fluoxetine in recurrent stress while inhibiting the protective effects promoted by imipramine. (a) Representative scheme of the experimental timeline used, including drug treatments with ADs and MAM. (b) Quantification of the number of BrdU⁺NeuN⁺ cells in the hippocampal dentate gyrus (DG), representing the population of newborn neurons denotes that proliferation arrest promotes a normalization of adult neurogenesis in consequence of treatment with fluoxetine, whereas a depletion in the levels of neurogenesis in animals treated with imipramine occurred. (c) The assessment of anhedonic-like behavior by the SDT, at the end of the protocol, revealed that the arrestment of adult neurogenesis rescues the anhedonic phenotype presented by fluoxetine-treated animals and inhibits the effect of imipramine to prevent anhedonic alterations. (d) Anxiety-like behavior, tested by the EPM, demonstrated that ablation of adult hippocampal neurogenesis has a preventive effect on anxiety-induced fluoxetine while leading to an anxious-like state when MAM was co-treated with imipramine. (e and f) MWM performed to assess cognitive performance on working memory revealed no major effects of MAM treatment in this behavioral dimension. (g) Schematic representation of the impact of adult neuroplasticity alterations promoted by unpredictable chronic mild stress (uCMS) and ADs. ^#Denotes the effect of ADs, by comparison of treatment and SAL animals and ^‡denotes differences between ADs, analyzed by one-way analysis of variance (ANOVA). ANOVA repeated measures was used to analyze cognitive learning tasks performance. Data are represented as mean±s.e.m. ^{#, ‡}P⩽0.05, ^{##, ‡‡}P⩽0.01, ^{###, ‡‡‡}P⩽0.001; n=6–8 animals per group. AD, antidepressant; CTRL, non-stressed animals; EPM, elevated-plus maze; FLX, animals repeatedly exposed to uCMS and treated with fluoxetine; FLX+MAM, animals repeatedly exposed to uCMS and treated with fluoxetine and methylazoxymethanol; IMIP, animals repeatedly exposed to uCMS and treated with imipramine; IMIP+MAM, animals repeatedly exposed to uCMS and treated with imipramine and methylazoxymethanol; MAM, methylazoxymethanol; MWM, Morris water maze; SAL, animals repeatedly exposed to uCMS and non-treated; SDT, sweet-drive test.