Transient Prepubertal Mifepristone Treatment Normalizes Deficits in Contextual Memory and Neuronal Activity of Adult Male Rats Exposed to Maternal Deprivation

Abstract

Early life adversity is a well-known risk factor for behavioral dysfunction later in life, including the formation of contextual memory; it is also (transiently) accompanied by hyperactivity of the stress system. We tested whether mifepristone (MIF) treatment, which among other things blocks glucocorticoid receptors (GRs), during the prepubertal period [postnatal days (PND)26-PND28] normalizes memory deficits in adult male rats exposed to 24-h maternal deprivation (MD) at PND3. MD reduced body weight gain and increased basal corticosterone (CORT) levels during the PND26, but not in adulthood. In adulthood, contextual memory formation of MD compared to noMD (i.e., control) male rats was significantly impaired. This impairment was fully prevented by MIF treatment at PND26-PND28, whereas MIF by itself did not affect behavior. A second behavioral test, a rodent version of the Iowa Gambling Task (rIGT), revealed that flexible spatial learning rather than reward-based aspects of performance was impaired by MD; the deficit was prevented by MIF. Neuronal activity as tested by c-Fos staining in the latter task revealed changes in the right hippocampal-dorsomedial striatal pathway, but not in prefrontal areas involved in reward learning. Follow-up electrophysiological recordings measuring spontaneous glutamate transmission showed reduced frequency of miniature postsynaptic excitatory currents in adult CA1 dorsal hippocampal and enhanced frequency in dorsomedial striatal neurons from MD versus noMD rats, which was not seen in MIF-treated rats. We conclude that transient prepubertal MIF treatment normalizes hippocampus-striatal-dependent contextual memory/spatial learning deficits in male rats exposed to early life adversity, possibly by normalizing glutamatergic transmission.

Keywords: Rodent Iowa Gambling Task; early life stress; glutamate transmission; mifepristone; reward-based decision making; spatial memory.

Figures

Figure 1.

Schematic overview of the experimental design in this study. A, Timeline for experimental series in which neuroendocrine data were collected (A1), either at PND26, PND29, or PND90. Animals were exposed to 24-h MD or control treatment at PND3. Part of the animals were treated twice daily with MIF or VEH administered through oral gavage directly into the stomach. At the indicated moment, we monitored body weight, adrenal weight (only at PND29), CORT levels, and collected hippocampal lobes for later Western blotting for MR and GR (typical examples shown in A2). B, Timeline for the object-in-(novel) context testing (B1). The details of the task are shown in B2, see Materials and Methods for details. C, Schematic representation of the rIGT maze model. D, Anatomic localization of brain region used for analysis of c-Fos expression. Bregma coordinates are indicated above each coronal section. DG, dentate gyrus; CA1, CA1 region of the hippocampus; CA3, CA3 region of the hippocampus; DSL, dorsolateral striatum; DMS, dorsomedial striatum; NaC, nucleus accumbens core; NaS, nucleus accubmens shell; Cg1, cingulate cortex; PrL, prelimbic cortex; IL, infralimbic cortex; INS, insular cortex; medOFC, medial orbital frontal cortex (OFC); latOFC, lateral OFC; venOFC, ventral OFC. E, Timeline for the electrophysiological recordings of CA1 pyramidal cells in hippocampal slices (E1). Typical example traces of mEPSCs are shown in E2, at a low time resolution (left) and a high time resolution (right). E3, Anatomic localization of brain region used for mEPSCs recording.

Figure 2.

Object-in-context memory formation. Discrimination index in the retention test on day 3 (n = 7–8 animals per group). The discrimination index is based on the duration of the object in the novel context during the entire 5-min period. One-sample t test revealed that noMD-VEH (control) group discriminated between the object in the matching and the object in the nonmatching context as scores were significantly above chance level (50%; #p < 0.001). Two-way ANOVA revealed a significant condition × treatment interaction effect (see main text). Post hoc analysis revealed significant differences between the MD-VEH and MD-MIF as well as noMD-VEH groups. All bars represent the mean + SEM. #p < 0.001 (paired t test against chance level); *p < 0.05; **p < 0.01 (post hoc comparisons after ANOVA).

Figure 3.

rIGT performance in male rats (PND90) exposed to MD or control treatment (noMD), receiving MIF or VEH between PND26 and PND28. All data were tested with a three-way ANOVA. For the parameters shown in A, B, we observed significant interaction effects of treatment and condition (see main text for statistical details). A, Mean (±SEM) fraction of empty arms choices across 12 trial blocks (empty arm visits as fraction of the total number visits per trial block; n = 10). The inset in the figure shows the scores of the last block of trials: 111–120. B, Mean (±SEM) fraction of disadvantageous arm choices across 12 trial blocks (disadvantageous arm visits as fraction of total number of visits to baited arms per trial block). The inset in the figure shows the scores of the last block of trials: 111–120.

Figure 4.

Changes in hippocampal and striatal mEPSCs frequency after MD are normalized by MIF treatment. A, The frequency of mEPSCs recorded in adult CA1 pyramidal neurons in vitro was found to be significantly reduced after 24-h MD experienced at PND3. This was normalized when animals were subjected to MIF treatment between PND26 and PND28, as opposed to VEH treatment. The data show the averaged mEPSC frequency ± SEM in a period of 5–10 min after establishing the whole cell recording configuration. Amplitude was not affected by either factor. Group sizes: n = 16 cells in noMD/VEH; n = 21 in noMD/MIF; n = 12 in MD/VEH; and n = 17 in MD/MIF. *p < 0.05, post hoc test. B, Comparable results were obtained when the cumulative frequency of mEPSC intervals was analyzed, using KS testing (see main text). C, Averaged data on mEPSCs frequency recorded in dorsomedial striatum neurons of adult male rats (PND90) exposed to MD or control treatment (noMD) at PND3, receiving MIF or VEH between PND26 and PND28. Two-way ANOVA revealed a significant interaction effect of treatment and condition (see main text and Table 4 for statistical details). Group sizes: n = 12 cells in noMD/VEH; n = 16 in noMD/MIF; n = 9 in MD/VEH; and n = 15 in MD/MIF. *p < 0.05, post hoc test. D, Amplitude of mEPSCs. No condition * treatment effect but a significant main effect of MD and a tendency toward a main effect of MIF treatment was found (see main text and Table 4).