Effects of repeated adolescent stress and serotonin transporter gene partial knockout in mice on behaviors and brain structures relevant to major depression

Abstract

In humans, exposure to stress during development is associated with structural and functional alterations of the prefrontal cortex (PFC), amygdala (AMY), and hippocampus (HC) and their circuits of connectivity, and with an increased risk for developing major depressive disorder particularly in carriers of the short (s) variant of the serotonin transporter (5-HTT) gene-linked polymorphic region (5-HTTLPR). Although changes in these regions are found in carriers of the s allele and/or in depressed patients, evidence for a specific genotype × developmental stress effect on brain structure and function is limited. Here, we investigated the effect of repeated stress exposure during adolescence in mice with partial knockout of the 5-HTT gene (HET) vs. wildtype (WT) on early-adulthood behavioral measures and brain structure [using magnetic resonance imaging (MRI)] relevant to human major depression. Behaviorally, adolescent stress (AS) increased anxiety and decreased activity and did so to a similar degree in HET and WT. In a probabilistic reversal learning task, HET-AS mice achieved fewer reversals than did HET-No-AS mice. 5-HTT genotype and AS were without effect on corticosterone stress response. In terms of structural brain differences, AS reduced the volume of two long-range white matter tracts, the optic tract (OT) and the cerebral peduncle (CP), in WT mice specifically. In a region-of-interest analysis, AS was associated with increased HC volume and HET genotype with a decreased frontal lobe volume. In conclusion, we found that 5-HTT and AS genotype exerted long-term effects on behavior and development of brain regions relevant to human depression.

Keywords: adolescent stress; development; magnetic resonance imaging; major depression; mouse model; serotonin transporter gene.

Figure 1

The time line of the study. Animals were weaned at postnatal week 4 and assigned to experiment 1 or 2. All mice were exposed to the adolescence stress procedure and context re-exposure. Mice in experiment 1 were subsequently tested during late adolescence in the novelty-induced hypophagia (NIH) and elevated plus maze (EPM) tests. In adulthood, a serial blood sampling (SBS) procedure was conducted, afterwards animals were trained and tested in the probabilistic reversal learning (PRL) task. Mice in experiment 2 were tested during late adolescence in the open field (OF), and perfused in adulthood to collect the brains for structural magnetic resonance imaging.

Figure 2

Results of the adolescent stress procedure on freezing behavior. Mean percent time spent freezing during the adolescent stress sessions (S1–S9) for mice in experiment 1 (A) and 2 (C). Mean percent time spent freezing during the context re-exposure over 5 intervals (I) of 4-min each for mice in experiment 1 (B) and 2 (D). Results are reported as mean ± s.e.m.

Figure 3

Results of the novelty-induced hypophagia test (experiment 1). Lines indicate the cumulative feeding over a 30 min session (1800 s) and dots the time point at which the feeding (event) occurred. No effect of genotype was found.

Figure 4

Results of the elevated plus maze test (experiment 1). Mean percent time spent in the open arm (A) and mean percent open arm entries (B). Percent open arm entries was lower in HET-AS mice relative to HET-No-AS mice and WT-AS mice (^*p < 0.03). Results are reported as mean ± s.e.m.

Figure 5

Total distance moved during the open field test (experiment 2) over six intervals (I) of 5 min each. Results are reported as mean ± s.e.m.

Figure 6

Corticosterone titers during the serial blood sampling procedure (experiment 1). Results are reported as mean ± s.e.m.

Figure 7

Analysis of the normalized volumes of 62 structures showed that a significant effect of AS in two white matter regions: the cerebral peduncle (10% FDR) and the optic tract (5% FDR). WT-AS mice showed a smaller cerebral peduncle (A) and optic tract (B) volumes relative to WT-No-AS mice and to HET-AS mice (^*p < 0.02). Results are reported as mean ± s.e.m.

Figure 8

The whole brain voxel-wise analysis showed a significant effect of genotype (10% FDR) in two clusters of right temporal-parietal cortex (A and C). HET mice showed tissue volume reduction relative to WT mice (B and D, ^*p < 0.00001). Results are reported as mean ± s.e.m.

Figure 9

The region-of-interest analysis of the normalized frontal lobe volume demonstrated a smaller frontal lobe in HET relative to WT animals (^*p < 0.008). Results are reported as mean ± s.e.m.