Antidepressant responsiveness in adulthood is permanently impaired after neonatal destruction of the neurogenic pool

Abstract

The dynamic turnover of hippocampal neurons is implicated in the regulation of cognitive and affective behavior. Extending our previous demonstration that administration of dexamethasone (ND) to neonatal rats depletes the resident population of neural precursor cells (NPC) and restrains the size of the neurogenic regions, we now show that the adverse effects of ND persist into adulthood. Specifically, ND impairs repletion of the neurogenic pool and neurogenesis; ND also compromises cognitive performance, the ability to actively adapt to an acute stressor and, the efficacy of glucocorticoid (GC) negative feedback. Interestingly, although ND depletes the neurogenic pool, it does not permanently abolish the proliferative machinery of the residual NPC population; however, ND increases the susceptibility of hippocampal granule neurons to apoptosis. Although the antidepressant fluoxetine (FLX) reverses the latter phenomenon, it does not replenish the NPC pool. Treatment of ND-treated adult rats with FLX also improves GC negative feedback, albeit without rescuing the deleterious effects of ND on behavior. In summary, ND leads to protracted disruption of mental functions, some of which are resistant to antidepressant interventions. We conclude that manipulation of the NPC pool during early life may jeopardize the therapeutic potential of antidepressants in adulthood.

Figure 1

Spatial and novel object-recognition memory are impaired after neonatal dexamethasone (ND) exposure and fluoxetine (FLX) cannot rescue these cognitive deficits. The schematic (a) summarizes the paradigms used to assess two object-recognition memory. Discrimination between the objects was determined using a discrimination ratio, calculated as the difference in time spent exploring the novel object vs the familiar object divided by the total time spent exploring both objects. Initially, the distance traveled and speed of movement of all experimental groups was assessed in the open field arena without objects; the results are shown in b. Results of the 2-object relocated object preference test are shown in c. This test revealed that ND-treated rats are impaired in their ability to discriminate between a relocated object and an object in its original location, and that this impairment of spatial recognition is not restored after treatment with FLX. Examination of novel object preference (d) revealed that ND are retarded in their ability to discriminate between a novel and familiar object, and that this impairment is cannot be rescued with FLX. Data are mean±s.e.m. (n=13, 13, 13 and 12 for CON, ND, FLX and ND–FLX groups, respectively). *P<0.05 vs control.

Figure 2

Persistent and fluoxetine (FLX)-irreversible depletion of the neurogenic pool in the subgranular zone (SGZ) of the hippocampus. (a and b) Representative confocal images of the SGZ in the adult control (CON, a) and ND-treated (ND, b) rat; sections were double-stained for Sox2 (neural precursor cells (NPC)) and Ki67 (mitotic cells) as well as Hoechst 33342 (cell nuclei); Sox2- and Ki67-immunopositive are labeled with arrows. The dotted lines indicate arbitrary demarcation of the SGZ (defined as three-cell layer zone at the granule cell layer (GCL)-hilus border; the insets in a and b are enlarged in A1 and B1, respectively. Scale bars in a, b and A1, B1 are 50 μm and 10 μm, respectively. (c–e) The effects of ND, FLX and ND–FLX treatment on the number of Ki67-, Sox2/Ki67- and Sox2-positive cells in the SGZ at the time of killing. There was an ~50% reduction in the number of proliferating cells (c)/proliferating NPC (d) in the SGZ of ND animals, an effect that was reversed in a subgroup of ND-treated animals that were treated with FLX during adulthood. Note the ~22% reduction in the number of NPC (e) in the SGZ of adult animals that had been exposed to ND, and the significantly lower number of NPC in the ND–FLX subgroup, as compared with the corresponding FLX group. (f) The effects of ND, FLX and ND–FLX treatments on the density of NPCs (Sox2+) in the SGZ. It is important to note that, FLX failed to restore SGZ Sox2+ cell density to control levels when the influence of volumetric effects of FLX (cf. Supplementary Figure S2A) is considered. (g) ND reduces the number of migrating neural progenitors (judged by number of Sox2+ cells in GCL), an effect that can be rescued by FLX during adulthood. Numerical data are mean±s.e.m. (n=8 for all groups). *indicates P<0.05 as compared with CON (c–g) or pairs of treatment groups (e and f). ^#indicates P<0.05 vs corresponding ND-treated groups.

Figure 3

Fluoxetine (FLX) attenuates neonatal dexamethasone (ND)-induced susceptibility of hippocampal cells to apoptosis. (a) The number of activated caspase 3-positive cells in the subgranular zone (SGZ) did not differ between control and ND-treated rats; stereological estimates were made at the end of the experiment (cf. Figure 2). (b, upper) Western blots showing the influence of ND and/or FLX on the regulation of Bcl-2 family members in the hippocampus of adult rats; the lower panel presents the semi-quantitative results of the immunoblotting analysis (n=4 replicates). Note the increased expression ratios of Bax (pro-apoptotic molecule) to Bcl-2 or Bcl-xL (anti-apoptotic molecules) in the hippocampi of ND-treated rats and the restoration of these ratios to those found in controls by FLX. Numerical data represent mean±s.d. *P<0.05 vs control; ^#P<0.05 vs ND.

Figure 4

Fluoxetine (FLX) cannot reverse neonatal dexamethasone (ND)-induced increases in stress-coping (depressive-like) behaviors and anxiety. (a) Adult rats that had been exposed to ND treatment showed signs of anxiety (less time and less entries into the open arm of an elevated plus maze) as compared with controls (CON) animals, a feature that was not altered by FLX treatment. Note, that FLX itself increased anxiety-like behavior in CON rats. (b) ND rats exhibited significantly less rears and spent less time in the central area of an open field arena, indicating their increased state of anxiety. The effects of ND were not counteracted by FLX and rats treated with FLX alone spent less the time in the central area of the arena. (c) ND was associated with increased depressive-like behavior, as assessed by the number of floating episodes and time spent floating in the forced-swim test paradigm; FLX did not reverse these effects of ND. All data shown represent mean±s.e.m. (n=27, 13, 26, 13 for CON, ND, FLX and ND–FLX groups, respectively). *P<0.05 vs control. P<0.05 vs ND.

Figure 5

Basal and stress-provoked corticosterone (CORT) secretion. (a) Comparison of basal levels of serum CORT and serum CORT concentrations following exposure to an acute stressor after 30 and 120 min. The rate of return to pre-stress levels was used an indicator of glucocorticoid (GC) negative feedback efficacy which is often reduced in depressed patients. (b) GC negative feedback efficacy was also evaluated with the dexamethasone suppression test (DST). Animals were held on a 12 l:12D schedule and DEX (10 μg kg⁻¹) was injected at ZT6; tail-blood samples were harvested for estimation of CORT at ZT12 when the peak of CORT secretion normally occurs. The latter CORT measurements were compared with levels found in blood samples obtained at ZT12 in each of the treatment groups, and expressed as a percentage. Note that ND animals showed the least suppression of CORT after DEX administration, indicating impaired GC negative feedback. All numerical values are mean±s.e.m. (n=27, 13, 26 and 13 for CON, ND, FLX and ND–FLX groups, respectively). *P<0.05 vs the corresponding control value, ^#P<0.05 vs ND-treated counterparts. DEX, dexamethasone; FLX, fluoxetine; ND, neonatal DEX.