Antidepressants and Adolescent Brain Development

Emily Karanges; Iain S McGregor


Future Neurology. 2011;6(6):783-808. 

In This Article

Neural Effects of Antidepressants in Adolescent Animals

Effects on Major Neurotransmitter Systems

Serotonin System Interest in the role of 5-HT in depression and the antidepressant response began in the 1960s following the discovery of the serotonergic effects of the TCAs[110] and was heightened following the development of the SSRIs. While it is now generally accepted that there is no simple relationship between serotonergic dysfunction and depression,[111] 5-HT remains a major research interest due to its involvement in the regulation of mood, emotional processing, appetite and sleep, all of which are disrupted in depression,[111,112] and its importance in the mechanism of action of the SSRIs.

The effects of the SSRIs on the serotonergic system during adulthood are generally well characterized. 5-HTT, the primary target of the SSRIs, controls the intensity and duration of 5-HT signaling, being responsible for the reuptake of synaptic 5-HT into the presynaptic neuron. Upon administration, SSRIs bind 5-HTT with high affinity, inhibiting reuptake and increasing the synaptic concentration of 5-HT. However, increases in synaptic 5-HT are rapidly attenuated by homeostatic activation of 5-HT1A and 5-HT1B autoreceptors.[113] Therefore, lasting alterations in serotonergic tone may not occur until these receptors become desensitized 2–3 weeks later.[114]

In addition to 5-HT1A and 5-HT1B autoreceptor desensitization, chronic SSRI treatment in adults has frequently been associated with desensitization and/or downregulation of other receptor subtypes including 5-HT2A,[115,116] 5-HT2C[117] and 5-HT4 receptors.[116] Downregulation of 5-HTT is also often observed,[106,108] although some studies report no difference from controls.[118] The reduction in the activity of 5-HT receptors and 5-HTT has been used to explain the delayed therapeutic response to SSRIs,[119] although additional adaptive mechanisms are likely involved. Reductions in 5-HT turnover and in concentrations of the 5-HT metabolite 5-hydroxyindoleacetic acid also accompany chronic SSRI treatment in adults.[86,120]

Studies in adolescent animals suggest that some components of the developing serotonergic system respond to SSRI treatment in a similar fashion to the adult system. For example, chronic (over 22 days) paroxetine had similar effects on whole-tissue concentrations of 5-HT (unchanged), 5-hydroxyindoleacetic acid (reduced) and 5-HT turnover (reduced) in the striatum of adult and adolescent rats (see Figure 2[86]). However, the majority of studies describe age-specific effects of SSRIs on this system.

Figure 2.

Chronic paroxetine (10 mg/kg in drinking water) exerts differential neurochemical effects in the striatum of adult and adolescent rats (n = 8/group).
*Significant treatment effect compared with age-matched controls (p < 0.05).
**Significant age × treatment interaction effect (p < 0.05).
5-HIAA: 5-hydroxyindoleacetic acid; 5-HT: Serotonin; Ado: Adolescent; CON: Control; DA: Dopamine; DOPAC: 3,4-dihydroxyphenylacetic acid; HVA: Homovanillic acid; PRX: Paroxetine.
Adapted with permission from [86].

Two separate groups have investigated SSRI-induced changes in 5-HTT density in various regions of the adolescent rat brain following chronic SSRI administration. In contrast to the often-found decrease or null effect on 5-HTT binding density observed in adults, both studies report regional increases in 5-HTT binding in their younger cohort. An early study by Wegerer et al. reported increased levels of 5-HTT in the frontal cortex of early adolescent (P25) rats treated with fluoxetine for 14 days, with no alterations in binding density in the parietal cortex, occipital cortex, midbrain or hypothalamus.[105] Interestingly, no such effects were found in rats when treatment was started at P50, pointing to a sensitive period during early adolescent life. Karanges et al. reported similar findings, showing upregulation of 5-HTT in the amygdala, but not the hippocampus, in adolescent rats following chronic paroxetine treatment.[86] The frontal cortex and amygdala both receive serotonergic innervation from the raphe nucleus,[121] suggesting that these findings may reflect regional increases in serotonergic innervation and synaptic outgrowth rather than a direct increase in 5-HTT expression. Indeed, given the extensive remodeling and strengthening of connections occurring during adolescence and the role of 5-HT in synaptic outgrowth during development,[101,122] such an explanation is plausible.

Interestingly, Wegerer et al. provide evidence that the regional increases in 5-HTT density endure into adult life,[105] in contrast to the rapid recovery of SSRI-induced 5-HTT downregulation in adults.[119] Lasting changes in 5-HTT may explain the increases in anxiety-like behavior and sexual dysfunction observed in adult rats that have been treated with SSRIs during adolescence.[123]

Several studies exploring the effects of SSRIs on 5-HT receptor function in adolescents have employed neuroendocrine or behavioral drug challenge techniques. In a study investigating the effects of chronic sertraline on serotonergic function, treated adult rats displayed the usual suppression of prolactin release to fenfluramine challenge, suggesting desensitization of postsynaptic serotonergic receptors in response to sertraline.[85] By contrast, the response to fenfluramine challenge was not altered by sertraline in prepubertal or peripubertal rats, suggesting that receptor desensitization may not occur prior to adulthood. Similarly, fluoxetine appears to have different effects on hypothalamic 5-HT2A receptor function in adult and adolescent rats, as shown by the neuroendocrine response to DOI challenge.[124]

Two studies have investigated the effects of adolescent SSRI treatment on 5-HT1A receptors in adulthood. These studies report no changes in 5-HT1A receptor binding[125] or function[123] following extended wash-out periods (14–17 days and 14 weeks, respectively). Changes in 5-HT1A binding density do not appear to occur during SSRI treatment of adult animals,[126] although decreases in receptor function have been observed.[127] It is currently unknown how 5-HT1A function is affected during treatment in adolescent animals.

In adults, chronic SSRI treatment probably modulates the 5-HT2C receptor in two ways: the receptor becomes desensitized[117] and alterations in pre-mRNA editing modify the balance of different receptor isoforms.[128] The primary transcript of the 5-HT2C receptor is subject to post-transcriptional editing, producing various receptor isoforms that differ in their sensitivity for 5-HT and their ability to activate the receptor's associated G protein, Gαq.[129] Studies have demonstrated that chronic stress alters pre-mRNA editing, and treatment with SSRIs during adulthood reverses these effects.[128]

Although the effect of antidepressants on 5-HT2C receptor desensitization in adolescents is unknown, Bhansali et al. investigated the impact of 28 days of fluoxetine treatment on adult and adolescent BALB/c mice with or without a history of infant maternal separation (IMS), and reported differential effects of the antidepressant depending on prior stress experiences and age.[130] IMS increased editing, thus reducing the sensitivity of the receptor for 5-HT. In what appears to be a compensatory response for the reduced interaction of the receptor with its G protein, the level of Gαq was also increased in IMS mice. Fluoxetine treatment normalized these effects in adolescents, but only reversed the increase in Gαq in adults. In contrast, normal adolescent mice showed the opposite response to fluoxetine, with increases in pre-mRNA editing without compensatory changes in Gαq binding. This resembles the response to chronic stress[128] and may suggest adverse effects of fluoxetine on the serotonin system in normal adolescent animals. This study provides a strong indication that effects of SSRIs on the developing brain may differ depending on prior history and depressive symptomatology.

Dopamine System The mesocorticolimbic dopamine system, originating from the VTA and connecting with the PFC, amygdala, hippocampus and nucleus accumbens, is thought to play a role in the regulation of motivation, hedonic state, reward, social behavior, cognition and emotional control.[112] Given that anhedonia and loss of motivation are two of the core symptoms of depression, it is unsurprising that dopamine has been implicated in depression and the antidepressant response.[131,132]

With the exception of high-dose sertraline, SSRIs have low affinity for components of the dopamine system, yet they are capable of influencing dopaminergic function after both chronic and acute administration.[133] Temporary attenuation of mesolimbic dopaminergic activity by acute SSRI treatment, mediated by serotonergic activation of 5-HT2C receptors, is thought to contribute to the early anxiogenic effects and delay in efficacy of the SSRIs.[134] Conversely, chronic SSRI treatment in adults has been associated with increased firing of mesocorticolimbic dopaminergic neurons[135] and increases in synaptic dopamine,[136] suggesting that disinhibition of the dopamine system may be important for the therapeutic effects of the SSRIs.[137]

Only one study has investigated whether SSRIs have similar effects on the dopamine system in adolescents. Karanges et al. conducted a direct adolescent versus adult comparison of the behavioral and neural effects of chronic paroxetine in rats, reporting developmental differences in the effects of the drug on dopamine metabolites and turnover in the striatum (see Figure 2), and dopamine transporter binding density in the nucleus accumbens.[86] Specifically, paroxetine increased measures of dopamine turnover, homovanillic acid (a dopamine metabolite) and dopamine transporter in adult rats, with no such effects in adolescents. As reviewed earlier, the developing brain may not respond to chronic SSRI treatment with desensitization of the 5-HT receptor subtypes involved in the moderation of dopamine release, thus preventing the dopaminergic upregulation commonly seen in adults. These findings potentially explain some of the adverse behavioral effects and lack of therapeutic efficacy reported in adolescents.

Norepinephrine System Dysfunction of the noradrenergic system has been implicated in depression and anxiety disorders,[138] particularly with regard to symptoms associated with arousal, energy and vigilance.[139] Antidepressant drugs such as the TCAs and SNRIs have direct effects on noradrenergic function, while SSRIs appear to affect this system primarily through serotonergic mechanisms.[139] Specifically, chronic treatment with SSRIs in adults has been associated with reductions in extracellular norepinephrine in the amygdala and LC[140] and reductions in spontaneous and sensory-evoked firing of LC neurons.[141] Interestingly, inhibition of LC neuronal activity has been reliably associated with other antidepressant therapies including TCAs, MAOIs and electroconvulsive shock.[135,141] This has been proposed as a mechanism by which antidepressants facilitate dopamine release from the VTA,[142] contributing to the relief of depression-related symptoms such as anhedonia.

However, a recent study by West et al. (see, [87] and its addendum[143]) demonstrates that short-term treatment with some antidepressants may actually produce opposite effects in adolescent rats. In contrast to the decrease in LC neuronal activity found in adults, short-term administration (over 2–4 days) of paroxetine or venlafaxine increased LC activity in adolescents, with reductions emerging after 8 or more days of treatment. Compellingly, the directional changes in LC activity reflected depressive-like behaviors in the forced swim test, suggesting that hyperactivity of LC neurons may contribute to depressogenic effects of antidepressants in some adolescents. Indeed, increases in LC neuronal activity have been previously observed in conjunction with depression-like behaviors in animal models of chronic stress.[142]

However, certain components of the norepinephrine system are not commonly modulated by SSRI treatment. With the exception of paroxetine, which is known to block norepinephrine reuptake at high doses,[144] SSRIs do not appear to modulate norepinephrine transporter binding or affinity[105] or total tissue norepinephrine[86] in adult or adolescent rodents.

Effects on Neurogenesis & Synaptic Plasticity

The neurotrophic hypothesis of depression and antidepressant action, reviewed extensively elsewhere,[145–149] proposes that reductions in hippocampal neurogenesis and/or neurotrophic factors play a role in the etiology of depression, and that antidepressants act to normalize these deficits. Supporting evidence for the role of neurogenesis in depression includes the hippocampal atrophy and reduced concentrations of neurotrophic factors such as brain-derived neurotrophic factor (BDNF) in individuals with depression,[146] and decreases in neurogenesis and BDNF expression in animals exposed to chronic stress.[149] Conversely, stimulation of neurogenesis is a key feature of many antidepressant therapies, including the SSRIs and other antidepressant drugs,[150] exercise[151] and electroconvulsive shock,[152] and suppression of these neurotrophic actions can prevent the relief of certain depression- or anxiety-like symptoms by such treatments.[150,153,154] Antidepressant treatment in adults has also been associated with upregulation of BDNF and other neurotrophic proteins,[147,155,156] downregulation of proapoptotic proteins[155] and stimulation of dendritic arborization and synaptic plasticity.[154,157] Together, these findings suggest that the actions of antidepressants on neurogenesis may be important for their therapeutic effects.[148,154]

Although it is not the purpose of this article to critique this hypothesis (see [158] for a recent critique), it is worth noting that not all studies support a causal relationship between stimulation of neurogenesis and antidepressant action. The behavioral effects of antidepressants appear to be neurogenesis independent in certain strains of rodents such as the BALB/cJ mouse,[88] and stimulation of hippocampal neurogenesis appears to be required for relief of anxiety-like but not depression-like symptoms in some animal models.[154] These findings and others have led to the proposal that it is the stimulation of neuronal plasticity and associated processes rather than neurogenesis per se that underlies the behavioral response to antidepressants.[149,154]

Regardless of whether the effect of antidepressants on neurotrophic processes are an epiphenomenon, the ability of antidepressants to affect synaptic plasticity has important implications for the treatment of adolescents, given the elevation of baseline synaptic plasticity and neurogenesis and the malleability of limbic–cortical links during adolescence.[28] The adolescent response to antidepressants has therefore been investigated more heavily with relation to neuroplasticity than any other aspect of the neural response. However, as with the adult literature, the adolescent literature is complicated by differences in prior stressor exposure, strain, drug, dose and variation in the dependent variables investigated.

Several studies have found no effect of antidepressant treatment on measures of hippocampal cell proliferation, differentiation and/or survival in standard-reared adolescent rodents. For example, following treatment with fluoxetine for 25 days, Cowen et al. report no differences in dentate gyrus volume, cell proliferation or cell survival in adolescent rats.[28] Similar findings have been reported in mice treated with fluoxetine during the juvenile and early adolescent periods.[90,159] In addition, in contrast to commonly observed effects in adults, adolescent antidepressant treatment does not appear to stimulate expression of neurotrophic factors, and may even disrupt associated signaling pathways.[104,160] However, it should be noted that these studies either lack an adult comparison group[90,104,159] or show none of the commonly observed neurogenic effects of antidepressant treatment in adults,[28,160] limiting robust conclusions on developmental differences.

However, two studies have employed direct adult versus adolescent comparisons investigating the effects of chronic fluoxetine on hippocampal neurogenesis. While both studies report differential age effects, they are seemingly contradictory in direction and mediation by sex. Following administration of fluoxetine (5 mg/kg) to rats for 14–18 days, Hodes et al. report increased DNA synthesis and cell proliferation in adult male rats, with no such effects in sex-matched adolescents.[161] The pattern differed in females: aside from a trend toward decreased cell survival in fluoxetine-treated adolescents, fluoxetine did not stimulate hippocampal neurogenesis in either age group. In direct contrast, Navailles et al. show no effects of chronic fluoxetine (16 or 25 mg/kg) on neurogenesis in adult mice of either sex, while observing increases in some, but not all, measures of neurogenesis in standard-reared adolescents.[98] The reason for these contradictory findings is unclear, but may be related to dose or species differences. Indeed, higher doses, such as those used by Navailles et al.,[98] may be required to stimulate neurogenesis in adolescent rodents, which are known to metabolize drugs more rapidly than their adult counterparts.[12] Furthermore, granule cell proliferation and maturation follow different time courses in rats and mice, and are of greater functional importance in rats, with suggestions that the rat hippocampus may better model that of the human.[162]

Rodent models of chronic stress may provide a more etiologically valid environment in which to examine the effects of antidepressants on neurogenesis. The studies relevant to this article employing such models have demonstrated modulation of adolescent responses to chronic fluoxetine and tianeptine by early life or adolescent stress paradigms.[98,159,163] In the study conducted by Navailles et al. reviewed earlier, IMS abolished the neurogenic responses to fluoxetine observed in standard facility reared adolescents.[98] Similarly, exposure to adolescent social stress removed the stimulatory effects of tianeptine on synaptophysin, a marker of synaptic plasticity.[163] By contrast, fluoxetine increased hippocampal cell proliferation and survival in adolescent mice exposed to social isolation rearing (SIR), with no such effects in standard facility-reared mice.[159] These conflicting findings may again reflect procedural differences: SIR and IMS may have different neural effects,[164] influencing antidepressant action and neurogenesis. Furthermore, the inhibitory effects on hippocampal neurogenesis were only reported following the SIR manipulation in these studies.

Several studies have pointed to a relationship between adult hippocampal neurogenesis, hypothalamic–pituitary–adrenal axis activity and glucocorticoids. Glucocorticoids inhibit hippocampal neurogenesis[165] and the release of neurotrophic factors such as BDNF via activation of the glucocorticoid receptor (GR).[166] Furthermore, antidepressants stimulate neurogenesis by GR-dependent mechanisms[167] and have been associated with reductions in cortisol and adrenocorticotropic hormone concentrations in treatment responders, but not in treatment nonresponders.[168] As such, changes in corticosterone concentrations with treatment may provide an indication of antidepressant efficacy. Studies investigating adolescent responses to SSRIs have uniformly reported no effects on corticosterone or adrenocorticotropic hormone concentrations during or shortly after treatment.[124,161] However, these studies were conducted in 'normal' animals, who are less likely to show alterations in corticosterone with treatment.[169]

As previously inferred, the effects of antidepressants are not restricted to neurogenesis or even to the hippocampal region, but extend to related processes such as neuroplasticity, synaptic remodeling and synaptogenesis (Box 2 & Table 3). Unsurprisingly, there are indications that antidepressant treatment during adolescence may cause lasting perturbations in normal developmental processes, altering dendritic spine development and synaptic outgrowth. For example, chronic treatment of rats with fluoxetine from P21 until P42 prevented the normal age-related increase in dendritic spine density in the CA1 region of the hippocampus.[30] This contrasts with reports that fluoxetine inhibits stress-induced atrophy of dendritic spines in adults,[154] suggesting that hippocampal plasticity may be differentially affected by fluoxetine in adolescents. However, these effects may be specific to dendritic arborization, given indications that hippocampal N-acetyl aspartate, a marker of neuronal density and function, appears to increase in response to SSRI treatment in both adult humans[170] and adolescent rodents.[171]

Chronic SSRI treatment during adolescence also seems to moderate synaptic plasticity in the amygdala. The polysialylated form of the neural cell adhesion molecule (PSA-NCAM) is a promoter of neurite and synaptic outgrowth and plays a key role in neuronal development.[172] Generally, upregulation of PSA-NCAM expression is indicative of increased synaptic remodeling, while reductions may indicate regressive structural changes.[173] In adult rodents, fluoxetine increases PSA-NCAM expression in the medial PFC and parts of the hippocampus, but decreases expression in the amygdala.[173] In a recent study, Homberg et al. measured lasting changes (14–17 days post-treatment) in PSA-NCAM in adult and adolescent rats treated chronically with fluoxetine.[125] In line with previous findings, fluoxetine tended to reduce PSA-NCAM concentrations in the amygdala of adult rats. By contrast, however, there was a trend towards increased PSA-NCAM expression in the amygdala of adolescent rats, suggesting increased amygdala plasticity in this age group in response to fluoxetine treatment. It is possible that these neuroplastic effects may underlie the increases in behavioral despair observed selectively in adolescent rats in this study.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: