SEARCH

SEARCH BY CITATION

Keywords:

  • brain-derived neurotrophic factor;
  • chronic mild stress;
  • depression;
  • neurogenesis;
  • sucrose preference

Abstract

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Exposure to chronic mild stress (CMS) is known to induce anhedonia in adult animals, and is associated with induction of depression in humans. However, the behavioral effects of CMS in young animals have not yet been characterized, and little is known about the long-term neurochemical effects of CMS in either young or adult animals. Here, we found that CMS induces anhedonia in adult but not in young animals, as measured by a set of behavioral paradigms. Furthermore, while CMS decreased neurogenesis and levels of brain-derived neurotrophic factor (BDNF) in the hippocampus of adult animals, it increased these parameters in young animals. We also found that CMS altered α-amino-3-hydroxy-5-methylisoxazole-4-propionate (AMPA) receptor GluR1 subunit levels in the hippocampus and the nucleus accumbens of adult, but not young animals. Finally, no significant differences were observed between the effects of CMS on circadian corticosterone levels in the different age groups. The substantially different neurochemical effects chronic stress exerts in young and adult animals may explain the behavioral resilience to such stress young animals possess.

Abbreviations used
AMPA

α-amino-3-hydroxy-5-methylisoxazole-4-propionate

BDNF

brain-derived neurotrophic factor

CMS

chronic mild stress

NAc

nucleus accumbens

PFC

prefrontal cortex

VTA

ventral tegmental area

Depression is characterized by an inability to experience pleasure (anhedonia) and by general loss of interest and motivation. In adolescents and children, disease symptoms mirror those of adults (Fleming and Offord 1990); however, there are many differences both in the neurobiological correlates and response to treatment. For example, depressed children and adolescents show no evidence of hypercortisolemia, a state frequently reported in adults (Kaufman and Ryan 1999) and they fail to respond to tricyclic antidepressants (Hazell et al. 1995). These differences may suggest that depression in the different age groups have somewhat different neuropathological bases.

The onset of depression results from an interaction between genetic predisposition and life stressors. Moreover, most episodes of major depression are preceded by stressful life events (Anisman and Matheson 2005). It is important to note that chronic stress (ongoing for weeks or months) is a stronger predictor of depressive symptoms than acute stressors (McGonagle and Kessler 1990) and multiple stressful events substantially increase the risk of a depressive onset (Kendler et al. 1998).

Several widely-used models for depression involve the induction of a stressful condition in animals and subsequent measurement of their coping behaviors. In the present study, we utilized the chronic mild stress (CMS) model of depression (Willner et al. 1987) that induces anhedonia by exposing rats to a chronic period of mild and unpredictable environmental stressors. This model has a high degree of predictive validity (behavioral changes are reversed with antidepressant drugs); face validity (chronic stress induces many of the behavioral alterations characterizing depressed patients), and construct validity (CMS decreases sensitivity in the brain reward system) (Borsini 1997). However, less is known about the long-term effects of chronic stress in juvenile animals.

Depression is associated with several changes in brain plasticity. Converging lines of evidence point to a critical role of brain-derived neurotrophic factor (BDNF) in depression. Animal and human studies have shown that BDNF levels are reduced during depression yet increased following treatments (Duman and Monteggia 2006). Control of BDNF expression is multifaceted. However, several studies have demonstrated that activation of glutamate α-amino-3-hydroxy-5-methylisoxazole-4-propionate (AMPA) receptors increase its expression both in vitro and in vivo (Hayashi et al. 1999). Moreover, accumulating evidence supports a role for glutamate and its receptors in depression, and specifically AMPA receptor levels in the rat hippocampus were found to be increased after chronic antidepressant treatments (Bajkowska et al. 1999).

Recent animal studies suggested that decreased hippocampal neurogenesis following stress, may contribute to hippocampal atrophy, observed in depression (Duman 2004). In contrast to the effect of stress, chronic antidepressant treatment and electroconvulsive shock therapy increase neurogenesis and BDNF levels in the hippocampus. Moreover, the increase in neurogenesis following antidepressant drugs was found to be necessary for their antidepressant action (Santarelli et al. 2003).

The purpose of this study was to compare between the behavioral effects of chronic mild stress in young and adult animals, and to evaluate the neurochemical alterations associated with neuroplasticity and depression in the different age groups. We found critical and surprising differences between the effects of chronic stress on adult and young animals.

Methods

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Animals

Ninety six young (30 day-old) male rats, 96 adult (60 day-old) male rats, and 10 adult (60 day-old) female rats, all of the Sprague-Dawley strain, were supplied by the Animal Breeding Center of The Weizmann Institute of Science and maintained under a 12 h light/12 h dark cycle (lights on at 8 am; lights off at 8 pm) with food and water ad libitum, except when the CMS procedure required deprivation or circadian changes.

Young and adult male rats were each divided into two groups. The experimental group was exposed to chronic mild stress, whereas the control group was given ordinary daily care. Behavioral tests were done individually for each rat, therefore all rats were housed singly. All male rats were tested as described below. The female rats were used only for the sexual behavior tests.

All animal experiments were conducted according to the rules of the Weizmann Institute’s Institutional Care and Use Committee, which are in complete accordance with NIH guidelines for the care and use of laboratory animals.

Chronic mild stress procedure

The CMS procedure was employed in our study for 4 weeks according to Willner and colleagues (Willner 1997), with slight modifications. Detailed description of the CMS procedure is provided in the supplementary information.

Sucrose preference

Rats had access to two drinking bottles positioned side-by-side at the rear of the cage for 9 days. The rats’ fluid consumption over 24 h was recorded by weighing the bottles every day between noon and 1 pm. Sucrose solution (0.2%, prepared in tap water) was placed in one bottle, and tap water in the other. After 4 days, tap water was filled in both bottles. The positions of the sucrose solution and water bottles were then switched and sucrose preference was tested for additional 4 days to control for the rats’ side preference. Sucrose preference was measured right after completion of the CMS procedure, and once again 6 weeks later.

Exploration and novelty-induced behavior

Rats were placed in a 40 × 40 cm exploration box (ActiMot System, TSE, Bad Homburg, Germany). The distance traveled, number of rearings and number of visits in the center of the arena were recorded automatically over 10 min. Detailed description of the exploration test is provided in the supplementary information.

Home-cage locomotion

Chronic monitoring of locomotion in the home cage was performed using a computerized Inframot system (TSE), based on infrared sensors located 3 cm above the home cage of each rat. Total overnight (between 8 pm and 8 am) activity was measured over six consecutive nights, starting 10 days after completion of the CMS procedure.

Spontaneous male sexual behavior test

The sexual behavior test was performed according to the method of Kalcheim et al. (1981) with some modifications (Kalcheim et al. 1981). Detailed description of the sexual behavior measurements is provided in the supplementary information.

Swim test

A modified forced swim test was conducted in a cylindrical tank 40 cm high and 18 cm in diameter, constructed in-house at the Weizmann Institute. The water temperature was kept at 26°C (2°C above room temperature) and the water level was such that the rat could not touch the bottom with its hind paws. Rats were given a single 10-min exposure to the swim tank, and their activities were videotaped. Any instances of diving behavior were noted. Video films of each forced swim test session were carefully analyzed by an observer blinded to the treatment group, using a computer-attached joystick (Gersner et al. 2005).

ELISA

The Elisa was done on a different group of rats that did not undergo behavioral tests, two weeks after completion of the CMS procedure. Their brains were extracted, immediately frozen in isoproponol and stored at −80°C. Bilateral tissue punches were obtained from coronal sections generated by a manual cut within the cryostat environment (at −20°C) and according to a rat brain atlas (Paxinous and Watson, 1998). The following coronal sections were used for each of the punches: dorsal hippocampus: −2 to −4.3 mm from bregma, ventral hippocampus: −4.8 to −6.3 mm, striatum: 2.7 to 0.7 mm, nucleus accumbens: 2.7 to 0.7 mm, and ventral tegmental area: −4.8 to −6.3 mm. Protein extraction and sandwich Elisa were performed as shown previously (Baker-Herman et al. 2004) and described in detail in the supplementary information.

Corticosterone measurement

Four additional groups were used to test the effect of CMS on plasma levels of corticosterone. Tail blood samples were collected by lancing the tail close to its tip and collecting a few drops of blood. Blood samples were collected during the CMS procedure (after 2 weeks) and 2 weeks after completion of the CMS procedure. At each of these time points samples were collected during the beginning of the light period (8 am) and during the beginning of the dark period (8 pm). Blood samples were centrifuged, plasma was collected and immediately stored at −80°C. Plasma corticosterone levels were measured using a commercially available enzyme immunoassay kit (Cayman Chemical, Ann Arbor, MI, USA).

Administration of BrdU and tissue preparation

BrdU administration, transcardial perfusion and preparation of brain slices were performed as described previously (Ziv et al. 2006; Levy et al. 2007). Detailed description of these procedures is provided in the supplementary information.

Immunohistochemistry, neurogenesis and quantification

Immunohistochemistry for analyzing GluR1, BDNF and neurogenesis was performed as described previously (Ziv et al. 2006; Levy et al. 2007) with slight modifications which are detailed in the supplementary information.

Data analysis

Data in text, figures and tables are expressed as mean ± SEM. Data were analyzed by two-way anovas (StatView 5.0, SAS, Cary, NC, USA), with the age (young vs. adult) and treatment (control vs. CMS) being between subject factors. Significant interaction effects were followed by two-tailed t-test analysis for independent samples of treatment groups within each age group. p-Value less than 0.05 was considered statistically significant.

Results

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Age-dependent effect of chronic mild stress on behavioral outcomes

Sucrose preference

The effect of chronic mild stress on sucrose preference was found to be age-dependent. While CMS in adult animals induced a significant reduction in sucrose preference, no such effect was observed in young animals exposed to the same CMS protocol (Fig. 1a). Two-way anova (age and treatment as the main factors) showed significant age X treatment interaction (F1,28 = 3.92, = 0.05) and further analysis revealed a significant difference between the adult CMS rats and their age-matched control group (t14 = 3.52, = 0.0034) but not between the young CMS rats and their age-matched controls. Even 6 weeks after completion of CMS (Fig. 1b), two-way anova again revealed a significant age X treatment interaction (F1,27 = 6.16, = 0.02). Only in adult CMS rats a reduced sucrose preference was apparent (t14 = 3.31, = 0.0056). The total fluid intake was similar for CMS treated and control animals.

image

Figure 1.  Effects of CMS in young and adult rats on sucrose preference. Data are presented as mean ± SEM of the percentage of sucrose (0.2%) intake over the course of 8 days, as calculated from total liquid consumption. Sucrose preference measurements were recorded (a) immediately after and (b) 6 weeks after the chronic mild stress procedure. **< 0.01 CMS vs. age-matched controls (n = 7–8 in each group).

Download figure to PowerPoint

Home-cage locomotion, exploration and activity in the forced swim test

Home-cage locomotion measured automatically over six nights was not affected by CMS (Fig. 2a). Two-way anova revealed no significant effect of age, treatment or interaction. Similarly, during the exploration test, in which locomotor activity was measured automatically in a novel environment, the total distance traveled (Fig. 2b) and the number of rearings (Fig. 2c) were not affected in either age group by the CMS procedure. On the other hand, the measurement of visits to the central portion of the exploration box (Fig. 2d) showed significant age X treatment interaction (F1,40 = 4.77, = 0.035) and further analysis revealed that only adult rats subjected to CMS had fewer center visits compared to age-matched controls (t26 = 3.27, = 0.003), a measure often interpreted as increased anxiety (Angrini et al. 1998). CMS in young rats did not decrease the number of visits to the center compared to their age matched controls, but both groups of younger animals tended to visit the central portion of the exploration boxes less often than the older groups.

image

Figure 2.  Effects of CMS in young and adult rats on locomotion, exploration, and performance in the forced swim test. Data are presented as means ± SEMs. (a) Average home-cage locomotion over six nights. (b) Total distance traveled in the exploration box during the 10-min test. (c) Number of rearings during the 10-min exploration test. (d) Number of times that the rats entered the center of the exploration box during the 10-min test. (e) Mobility score during the 10-min forced swim test. **< 0.01 CMS vs. age-matched controls (n = 7–8 in the young rats groups and n = 13–16 in the adult rats groups).

Download figure to PowerPoint

In the forced swim test, no differences were observed between the various groups (Fig. 2e). Exposure to CMS either in adult or young rats did not affect swimming behavior as revealed by two-way anova performed on either the swimming scores (Fig. 2e), or the immobility times (data not shown) during the single (modified) test.

Spontaneous male sexual behavior

The young groups had a lower number of mounts and intromissions than the adult groups (Fig. 3a and b) indicated by a significant main effect of age in number of mounts (F1,25 = 6.22, p = 0.02) and intromissions (F1,25 = 4.4, = 0.046) in two-way anova. However, no significant age X treatment interaction was found, indicating that CMS had no effect on the earlier stages of sexual activity in either age group. On the other hand, two-way anova showed significant age X treatment interaction in the number of ejaculations (F1,25 = 6.32, = 0.019; Fig. 3c). CMS only in adult rats induced a significant reduction in the number of ejaculations (t12 = 3.07, = 0.01, compared to the age-matched controls). No significant differences were observed in ejaculation latencies, hit rate, post-ejaculatory mount, post-ejaculatory interval and intracopulatory interval between young and adult CMS animals, and their controls (data not shown).

image

Figure 3.  Effects of CMS in young and adult rats on sexual behavior. Sexual behavior of each male rat was tested over 20 min after placing the female in the home-cage. Data are presented as means ± SEMs of the number of mounts (a), number of intromissions (b), and number of ejaculations (c) recorded in the adult or the young groups. **< 0.05 CMS vs. age-matched controls (n = 6–8 in each group).

Download figure to PowerPoint

Age-dependent effects of CMS on markers for neuroplasticity

Brain derived neurotrophic factor

Brain derived neurotrophic factor (BDNF) levels were measured by ELISA in punches from specific brain sites of rats killed 2 weeks after the completion of the CMS procedure and in age-matched controls. Separate groups of animals were used for these measurements in order to avoid possible effects of the behavioral tests on BDNF levels. Punches were taken from brain regions associated with reward function and the psychopathology of depression, including the dorsal and ventral hippocampus, striatum, nucleus accumbens and ventral tegmental area (Fig. 4).

image

Figure 4.  Effects of CMS on BDNF protein levels. BDNF protein levels were measured by ELISA in young and adults rats exposed to CMS or in age-matched controls. Data are presented as mean ± SEM levels measured in brain punches taken from the dorsal hippocampus (a), ventral hippocampus (b), striatum (c), nucleus accumbens (d) and ventral tegmental area (e). *< 0.05, ***< 0.001 CMS vs. age-matched controls. n = 7–8 in the young rats groups, n = 9–15 in the adult control group and 24–36 in the adult CMS group.

Download figure to PowerPoint

Chronic mild stress in adult, but not young, animals induced a reduction in BDNF levels in the dorsal hippocampus, as revealed by a significant age X treatment interaction (F1,54 = 7.74, = 0.007; Fig. 4a) followed by t-tests in each age group (t42 = 3.88, = 0.0004, in the adult group). On the other hand, in the ventral hippocampus we did not find significant main effects or interaction, but there was a significant increase in BDNF levels in young rats exposed to CMS when compared separately to their age-matched controls (t12 = 2.46, = 0.03; Fig. 4b). In the striatum there was a marginally significant interaction between age and treatment (F1,60 = 3.14, = 0.08; Fig. 4c) and a significant decrease in BDNF levels in adult (but not young) rats exposed to CMS when separately compared to their age-matched controls (t46 = 4.2, = 0.0001). In the nucleus accumbens (Fig. 4d) and the ventral tegmental area (Fig. 4e) there was no significant difference in BDNF levels between any of the groups.

To replicate these findings and identify differences in BDNF levels within specific sub-regions we utilized immunohistochemistry. Alterations in BDNF levels measured by immunohistochemistry in specific brain regions are presented in Table 1.

Table 1.   Effects of CMS in different age groups on BDNF expression
 PFCAnterior NAcDorsal DGVentral DG
  1. BDNF-specific immunostaining in the prefrontal cortex (PFC), anterior nucleus accumbens (NAc) and dorsal and ventral dentate gyrus (DG) are presented. All values are presented as percentages (± SEMs) of the average level of BDNF immunoreactivity measured in the adult control group (which was considered as the 100%).

  2. *Significant difference (p < 0.05) between CMS rats and the age-matched controls for the young and the adult groups, respectively (n = 5–8 in each group).

Young
 Control105 ± 4.991.7 ± 5.252.4 ± 6.5184.9 ± 27
 CMS111 ± 9.891.5 ± 7.794 ± 11.7*145.5 ± 7.3
Adult
 Control100 ± 5100 ± 5100 ± 14.3100 ± 8.1
 CMS110.4 ± 9.183.1 ± 5.3*62 ± 8.2*114.6 ± 17.4

In the prefrontal cortex (PFC), CMS did not alter BDNF levels significantly in either age group. In the anterior nucleus accumbens (NAc), there was no significant age X treatment interaction, but a separate analysis on each age group indicated that CMS in adult rats caused a significant reduction in BDNF levels in the anterior NAc (t12 = 2.32, = 0.039) while no such reduction was observed in young rats exposed to CMS (Table 1). In the dorsal dentate gyrus, a significant age X treatment interaction was observed (F1,27 = 13.39, = 0.0011; Table 1) and following analysis revealed that CMS in adult rats caused a significant decrease in BDNF levels (t14 = 2.3, = 0.037) while CMS in young rats caused a significant increase in BDNF levels (t13 = 2.99, = 0.01). In the ventral dentate gyrus, there was no significant main effect of treatment (CMS) or interaction between age and treatment, but a significant main effect of age was found (F1,20 = 14.2, = 0.0012) indicating higher BDNF levels in the young rats.

GluR1 subunit of the AMPA receptor

GluR1 levels were measured by immunohistochemistry in the prefrontal cortex, anterior and posterior nucleus accumbens, dorsal and ventral hippocampus and anterior and posterior ventral tegmental area (Table 2).

Table 2.   Effects of CMS on GluR1 expression
 PFCAnterior NAcPosterior NAcDorsal DGVentral DGAnterior VTAPosterior VTA
  1. GluR1-specific immunostaining in the prefrontal cortex (PFC), anterior nucleus accumbens (NAc), posterior NAc, dorsal dentate gyrus (DG), ventral DG, anterior ventral tegmental area (VTA) and posterior VTA are presented. All values are presented as percentages (± SEMs) of the average level of GluR1 immunoreactivity measured in the adult control group (which was considered as the 100%).

  2. *Significant difference (p < 0.05) between CMS rats and the age-matched controls for the young and the adult groups, respectively (n = 5–8 in each group).

Young
 Control58.5 ± 3.580.8 ± 9.5109 ± 16.560.9 ± 4.1127.4 ± 8.1114.8 ± 9.877.9 ± 9
 CMS42.9 ± 4.7*88.3 ± 17.290.4 ± 8.768.8 ± 9.9130.9 ± 10.795.9 ± 11.388 ± 11.3
Adult
 Control100 ± 8.3100 ± 8.4100 ± 6100 ± 11.4100 ± 10.4100 ± 17100 ± 22.9
 CMS79.5 ± 3.4*137.7 ± 15.1*108.2 ± 5.423.9 ± 10*94.2 ± 15.783.4 ± 13.683 ± 13

Two-way anova performed on the PFC data showed no significant age X treatment interaction, however it was evident that CMS significantly decreased GluR1 levels in both young CMS (t13 = 2.68, = 0.019) and adult CMS (t13 = 2.17, = 0.049) rats when compared to their age-matched controls. In the anterior nucleus accumbens, we found a significant main effect of age (F1,28 = 6.9, = 0.014) but no treatment or interaction effect. Interestingly, increase in GluR1 levels was noted only in adult CMS (t14 = 2.19, = 0.046) but not young CMS rats, when compared to the age-matched controls. In the posterior NAc, CMS did not induce any alterations in GluR1 levels in either age group (Table 2). In the dorsal dentate gyrus we found a significant age X treatment interaction (F1,23 = 19.61, = 0.0002) and, following analysis, revealed, in adult CMS rats only, a significant reduction in GluR1 levels when compared to their age-matched controls (t10 = 4.78, = 0.0008). In the ventral dentate gyrus, however, CMS did not induce any alterations in GluR1 levels in either age group. Finally, in the VTA (either the anterior or the posterior portions), there were no alterations in GluR1 levels in either age group (Table 2).

Neurogenesis in the hippocampus

Hippocampal neurogenesis was analyzed in a separate group of animals that were injected with BrdU two weeks after completion of the CMS procedure and killed 5 days after the BrdU injections. Two-way anova, on the number of proliferating cells (BrdU positive cells) counted in the dentate gyrus of the dorsal hippocampus, revealed a significant age X treatment interaction (F1,25 = 11.07, = 0.0027; Fig. 5a). CMS induced the opposite effect on proliferation in the two age groups. While in adult rats CMS induced a reduction in the number of BrdU positive cells (t12 = 2.97, = 0.012), in young rats CMS induced an increase in the number of BrdU positive cells (t13 = 2.34, = 0.036). Similarly, when the number of BrdU positive cells expressing the early differentiation marker doublecortin (DCX) was measured, a significant age X treatment interaction was found (F1,25 = 9.81, = 0.0044; Fig. 5b). CMS in adults induced a significant reduction in neuronal differentiation compared to their age-matched control (t12 = 4.46, = 0.0008), while CMS in young rats tended to increase neuronal differentiation (Fig. 5b). Furthermore, we found significant age X treatment interaction in the ratio of new neurons (BrdU+ DCX+ cells) to the total number of proliferating (BrdU+) cells (F1,25 = 10.4, = 0.0035). The percentage of neuronal differentiation was decreased in adult CMS rats compared to their age-matched control (mean ± SEM = 45.9 ± 10.7% vs. 80.4 ± 3%, respectively; = 0.002), while this was not altered in young CMS rats.

image

Figure 5.  Effects of CMS on neurogenesis in young and adult rats. Neurogenesis was analyzed by measurements of double immunostaining of BrdU and DCX in the dentate gyrus of the dorsal hippocampus. (a) Quantification of BrdU positive cells (b) Number of cells double-stained with BrdU and DCX. (c) Representative micrographs of the dentate gyri of young and adult rats that were double stained for BrdU (red). *< 0.05, ***< 0.001 compared to age-matched controls (n = 5–9 in each group).

Download figure to PowerPoint

Corticosterone

Corticosterone levels were measured in a separate group of rats. During the CMS procedure, corticosterone levels were higher in both CMS age groups compared to age-matched controls (Fig. 6a). Repeated measures two-way anova with treatment and age being between-subjects factors and time (am and pm) being a within-subjects factor revealed a significant main effect of treatment (F1,27 = 41.4, > 0.0001). Further analysis of each treatment showed significant main effect of time only in the control groups (F1,14 = 8.1, = 0.013) indicating a significant increased in corticosterone levels at 8 pm only in control and not CMS groups. There were no significant effects or interactions related to age.

image

Figure 6.  Effects of CMS on circadian corticosterone levels in young and adult rats. Morning and evening corticosterone levels were measured by ELISA in young and adult rats exposed to CMS or age-matched controls. Blood samples were taken at two different time points: (a) during the CMS procedure and (b) two weeks after completion of the CMS procedure. Data are presented as mean ± SEM. *< 0.05, **< 0.01, ***< 0.001 compared to age-matched controls (n = 7–8 in each group).

Download figure to PowerPoint

Two weeks after the end of the CMS procedure, a reduction in corticosterone levels were observed in both CMS groups relative to the age-matched controls (F1,28 = 9.7, = 0.0043). In addition, a significant main effect of time was observed (F1,28 = 9.7, < 0.0001) showing that corticosterone levels were higher at 8 pm (Fig. 6b) and time X treatment X age interaction (F1,28 = 4.2, = 0.046; Fig. 6b) was found.

Discussion

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Several studies have suggested that early life stress, in humans or animals, induces behavioral abnormalities in adulthood, including depression (Heim et al. 2004). However, the effects of chronic stress throughout the juvenile and adolescent period on motivational behaviors have not been analyzed in a controlled animal study. This study demonstrates substantial differences between the effects of chronic mild stress (CMS) on young and adult animals. These differences are expressed both in behavioral and site-specific neurochemical alterations associated with depression and imply that a special mechanism exists within young animals enabling them to better cope with chronic stress.

Reduced consumption of sucrose solutions by rats subjected to chronic mild stress has been used as a measure of anhedonia, a major symptom of depression (Willner 1997, 2005). As weight loss following CMS may influence sucrose intake and could be a confounding variable, we measured sucrose preference which is considered a better index for anhedonia (Matthews et al. 1995). It was found that adult rats responded with a decrease in sucrose preference that lasted even 6 weeks after completion of the CMS procedure. In contrast, young animals exposed to CMS, showed no change in preference for a sweetened solution, both immediately after, and up to at least 6 weeks following completion of the CMS procedure.

In this study, CMS did not affect locomotor activity in the home cage or in novel exploration boxes. However, we did find that adult (but not young) CMS animals visited the center of the exploration boxes less often than controls. This can be interpreted as increased anxiety (Angrini et al. 1998). In addition, we did not observe any changes in mobility in the forced swim test after CMS. Previous studies reported conflicting results on the effects of CMS in this test (Willner 2005). Differences in strain or gender may account for these discrepancies (Bielajew et al. 2003). Because anhedonia is a core symptom of depression, reaction to rewards was adopted as the major behavioral endpoint for the CMS procedure, however, it is likely that this model induces changes in reward sensitivity independently of changes in motivation or motor function (Willner 2005). This behavioral distinction is striking, particularly when comparing aspects of sexual behavior in rats subjected to CMS. Both young and adult CMS rats did not differ from age-matched controls in the earlier stages of mounting and intromission, indicating no effect of CMS on sexual drive. However, adult (but not young) CMS animals showed a significantly lower ejaculation frequency than controls. This decrease may indicate that adult CMS rats were unable to experience pleasure because decreased sexual behavior was reported to co-exist alongside hedonic deficits (D’Aquila et al. 1994; Gronli et al. 2005).

The age-dependant alterations in reward sensitivities following CMS were associated with several neurochemical changes. The decrease in BDNF expression in the dorsal hippocampus of adult CMS rats found in the present study matches the neurotrophic hypothesis of depression (Duman and Monteggia 2006) and the fact that CMS in young rats did not induce depressive behavior and reduction in hippocampal BDNF levels strengthen this theory. Indeed, hippocampal BDNF enhances neuroplasticity and learning mechanisms (Figurov et al. 1996), and infusion of BDNF into the hippocampus was reported to produce antidepressant-like effects (Shirayama et al. 2002). Therefore, it is possible that the increase in hippocampal BDNF expression in the young CMS group is part of a mechanism that enhances neuroplasticity and enables establishment of coping strategies with environmental changes in younger animals (including chronic unpredictable stressors).

Not all patients respond well to antidepressant treatment. In some cases this treatment actually increases suicidal tendencies, especially among younger patients (Tsai 2005). Moreover, it was suggested that mechanisms related to a BDNF-dependent pathway hold the key to poor response to antidepressants and treatment-emergent suicidal tendencies (Tsai 2005). The age-dependent effects of stress on behavior and BDNF levels, as revealed in the present study, support the notion that the neurochemical basis of depression may be different in some younger patients, and may account for the poor response to antidepressant treatments in these patients. However, it should be noted that depression is not a homogenous disorder in either adult or young subjects and that other neurochemical factors as well as other neurotrophic factors, may also mediate the effect of antidepressant treatments.

Apart from the hippocampus, increasing evidence implicates other brain regions in the pathophysiology of depression, making them targets for antidepressant drugs (Nestler and Carlezon 2006). We found decreased BDNF expression following CMS in the striatum. A decrease in striatal BDNF levels was already reported in animals subjected to maternal separation (Lippmann et al. 2007). The VTA-NAc pathway is strongly implicated in motivation, reward and the pathophysiology of depression (Nestler and Carlezon 2006). In this study we also found that CMS reduces BDNF levels in the anterior (but not the whole) NAc, only in the adult rats subjected to CMS. Our results are complementary to an earlier study showing that chronic antidepressant treatment increases BDNF levels in the VTA and NAc (Molteni et al. 2006), but not with recent studies demonstrating that blockade of BDNF activity in the VTA-NAc pathway actually exerts antidepressant activity and that microinjection of BDNF to the NAc increases immobility in the forced swim test (Eisch et al. 2003; Berton et al. 2006). These differences may result from the very different experimental procedures and species.

Several studies have demonstrated that AMPA receptor activation increases BDNF expression both in-vitro and in-vivo (Zafra et al. 1990; Hayashi et al. 1999; Lauterborn et al. 2000). Moreover, BDNF was also found to up-regulate GluR1 and GluR2/3 subunits of the AMPA receptor in vitro (Narisawa-Saito et al. 1999). In addition, accumulating data suggest that glutamate transmission plays an important role in depression (Paul and Skolnick 2003). For example, reduced expression of striatal GluR1 mRNA was found in bipolar depressive patients (Meador-Woodruff et al. 2001). In animal studies, chronic, but not acute, treatment with antidepressants increased hippocampal expression of GluR1 and GluR2/3 (Martinez-Turrillas et al. 2002). Moreover, AMPA receptor potentiators could be effectively employed as antidepressants (Li et al. 2001). In this study, we found that chronic mild stress induced a decrease in AMPA receptor subunit GluR1 levels in the PFC in both young and adult CMS rats, indicating that this change, in and of itself, cannot account for the anhedonia induced by CMS. In the anterior NAc and the dorsal dentate gyrus, on the other hand, CMS modified GluR1 levels (an increase and a decrease, respectively) only among the adult CMS group. Therefore, our observation of reduced GluR1 levels in the hippocampus in adult CMS rats is compatible with our findings on reduced BDNF levels, and with the hypothesis that GluR1 in the hippocampus can be a target for antidepressant treatment. However, the increase in GluR1 levels in the NAc, despite the decrease in BDNF in the same region, indicates a different interaction between GluR1 and BDNF in the NAc.

The age-dependant opposite effect of CMS on hippocampal neurogenesis revealed in this study is consistent with the alterations observed in hippocampal BDNF levels. As BDNF plays an important role in the regulation of neural progenitor cells (Lee et al. 2002), one can assume that the decrease in BDNF levels in adult animals exposed to CMS accounts for the decrease in hippocampal neurogenesis. In addition, alterations in AMPA receptor expression observed in the present study may also affect hippocampal neurogenesis, and the antidepressant-like activity of AMPA receptor potentiators has been attributed to the regulation of progenitor cell proliferation in the hippocampus (Li et al. 2001; Bai et al. 2003). It is unclear whether impaired neurogenesis plays a causal role in depressive pathology or is simply a product of stress. However, the ability of antidepressant treatments to increase neurogenesis and block the negative effects of stress (Duman et al. 2006; Duman and Monteggia 2006) provides strong evidence that adult neurogenesis plays a role in the treatment of depression and could possibly contribute to the illness itself. Moreover, it was reported that the increase in hippocampal neurogenesis is necessary for the behavioral effect of antidepressants (Santarelli et al. 2003). The hippocampus may contribute directly or indirectly to some of the symptoms of depression. Specifically, the hippocampus directly affects the activity and function of mesolimbic dopamine neurons that influence reward-related functions (Lisman and Grace 2005). Therefore, the effect of chronic mild stress on hippocampal plasticity can disrupt this pathway and thereby reduce reward sensitivities, as observed in the present study.

Depression is associated with altered hypothalamic-pituitary-adrenal axis physiology. More specifically, depression is often accompanied by hypersecretion of adrenal cortisol (Parker et al. 2003) and this may contribute to impaired hippocampal neurogenesis and induction of hippocampal atrophy, observed in depression (Duman 2004). We found in this study that CMS increased corticosterone levels in both young and adult rats and that the typical circadian changes in corticosterone levels were impaired. The impaired circadian corticosterone could result from disturbances of circadian rhythms induced by the CMS protocol and from the extremely heightened corticosterone levels during the morning (a ceiling effect) in these animals. Increased corticosterone levels and changes in circadian rhythms induced by CMS in adult animals have been previously reported (Bielajew et al. 2002; Ushijima et al. 2006). Here we observed similar alterations in both young and adult animals exposed to CMS. Interestingly, two weeks after the end of the CMS procedure, we observed a decrease in corticosterone levels in both CMS groups relative to age matched controls. Similarly, Pitman et al. observed a gradual decrease in basal corticosterone levels in rats, which were weekly exposed to a stressor (Pitman et al. 1990). Interestingly, hypocortisolism were also been found in posttraumatic stress disorder patients and in animal models for posttraumatic stress disorder (Rasmusson and Charney 1997; Pervanidou 2008). There are several possible mechanisms that may underlie the development of hypocortisolism: (i) reduced biosynthesis of corticosterone; (ii) increased feedback sensitivity; (iii) morphological changes such as atrophy or decreased hippocampal volume (Heim et al. 2000).

It is important to note that the behavioral and neural abnormalities were observed long after the CMS procedure, while increases in corticosterone levels were observed only during CMS. Therefore, corticosterone elevations per se do not mediate directly the behavioral abnormalities observed. Nevertheless, it is possible that these elevations induced long-lasting neural alterations (such as impaired neuroplasticity) which are more directly related to depressive-like behavior observed several weeks later.

While hypochorticosteronemia was observed in both age groups two weeks after completion of the CMS procedure, the pattern was different. Reduction in corticosterone levels were observed in 8 pm in the young CMS group while a reduction in corticosterone levels was noted in 8 am in the adult CMS group. This strikingly different pattern may be a possible explanation for the relative resilience of young rats to CMS, however the mechanism for this difference is yet to be studied. Another potential explanation for the different effect of CMS in young rats may come from the response to the corticosterone elevation. The high distribution of glucocorticoid receptors in the hippocampus contributes to its sensitivity to stress. Interestingly, the level of glucocorticoid receptors in dividing hippocampal cells increases substantially with age (Garcia et al. 2004). This can explain the relative resilience of the younger animals to chronic stress and suggests a mechanism by which CMS in the young animals did not induce reduction in hippocampal neurogenesis despite a similar effect of stress on corticosterone levels.

The age-dependent effects of chronic stress on depressive behavior and brain plasticity observed in the present study implies that chronic mild stress at a young age is less, or even not, harmful to brain plasticity and the associated risk of depressive induction. These results seem to contradict other studies demonstrating that exposure to stress early in life increases the risk for depression at least in some individuals (Heim and Nemeroff 2001; Kaufman and Charney 2001; Caspi et al. 2003; Teicher et al. 2006). However, great variability in methodologies related to stress severity and especially stress periods and lengths make it difficult to establish a clear conclusion on whether any stress in any young age increases the risk for depression (Hall 1998). The present study is the first one to measure depressive-like symptoms in a controlled animal model exposed to chronic mild stress starting at juvenility and continuing until adulthood. Indeed, exposure to acute or repeated stressors early in life can induce very long lasting behavioral impairments including depression, however, adaptation processes may evolve during chronic and prolonged exposure to mild stressors, as indicated by the present study. Such processes do not seem to develop in adult animals, despite a similar exposure to stress severity and duration and a similar effect of the stress on corticosterone levels.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

We would like to thank Dr. Alan Sholto for his editorial help. Dr. Zangen is an incumbent of the Joseph and Celia Reskin career development chair. Dr. Toth is supported by a Helena Rubinstein Postdoctoral Fellowship. This study was supported by the Israel Science Foundation and by the Rosenzweig-Coopersmith fund.

References

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Appendix  S1   Detailed description of methods.

Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

FilenameFormatSizeDescription
JNC_5642_sm_Appendix.doc45KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.