Elevated paternal glucocorticoid exposure alters the small noncoding RNA profile in sperm and modifies anxiety and depressive phenotypes in the offspring

Abstract

Recent studies have suggested that physiological and behavioral traits may be transgenerationally inherited through the paternal lineage, possibly via non-genomic signals derived from the sperm. To investigate how paternal stress might influence offspring behavioral phenotypes, a model of hypothalamic-pituitary-adrenal (HPA) axis dysregulation was used. Male breeders were administered water supplemented with corticosterone (CORT) for 4 weeks before mating with untreated female mice. Female, but not male, F1 offspring of CORT-treated fathers displayed altered fear extinction at 2 weeks of age. Only male F1 offspring exhibited altered patterns of ultrasonic vocalization at postnatal day 3 and, as adults, showed decreased time in open on the elevated-plus maze and time in light on the light-dark apparatus, suggesting a hyperanxiety-like behavioral phenotype due to paternal CORT treatment. Interestingly, expression of the paternally imprinted gene Igf2 was increased in the hippocampus of F1 male offspring but downregulated in female offspring. Male and female F2 offspring displayed increased time spent in the open arm of the elevated-plus maze, suggesting lower levels of anxiety compared with control animals. Only male F2 offspring showed increased immobility time on the forced-swim test and increased latency to feed on the novelty-supressed feeding test, suggesting a depression-like phenotype in these animals. Collectively, these data provide evidence that paternal CORT treatment alters anxiety and depression-related behaviors across multiple generations. Analysis of the small RNA profile in sperm from CORT-treated males revealed marked effects on the expression of small noncoding RNAs. Sperm from CORT-treated males contained elevated levels of three microRNAs, miR-98, miR-144 and miR-190b, which are predicted to interact with multiple growth factors, including Igf2 and Bdnf. Sustained elevation of glucocorticoids is therefore involved in the transmission of paternal stress-induced traits across generations in a process involving small noncoding RNA signals transmitted by the male germline.

Figure 1

Effect of corticosterone (CORT) on F0 behavior and physiology. CORT administration did not affect F0 behavior on the elevated-plus maze (n=23–24) (a), the light–dark apparatus (n=12; b) or the forced-swim test (n=24; c). CORT administration significantly increased the level of serum CORT 3 h following the beginning of the active phase (n=4–5; c) and resulted in a significant decrease in the post-stress response of serum CORT (n=6; d). CORT caused a significant decrease in the adrenal weights of the F0 animals (n=8–10; e) and, although there were no changes to the hippocampal expression of mineralocorticoid receptor (MR), there was a significant reduction in glucocorticoid receptor (GR) mRNA (f). *P<0.05, **P<0.01, ***P<0.001, values represent means±s.e.m.

Figure 2

Paternal CORT treatment changes early-life fear conditioning, with no changes to behavior in adulthood in female offspring. Paternal CORT has no effect on the number of vocalizations made by female animals (a). At 2 weeks of age, the mean levels of freezing to the conditioned stimulus (CS) during the conditioning trial, and extinction trial day 1 and day 2, respectively n=22–25 (b) per group. At 8 weeks of age, there were no differences in the mean levels of freezing to the CS during the conditioning trial and extinction trial days 1 and 2, respectively (c) n=8–12 per group. As adult offspring, paternal CORT treatment had no effect on the time spent in the open of the elevated-plus maze (d; n=23 per group), or the time spent on the light side of the light–dark apparatus (n=12–14 per group; e). Paternal CORT had no effect on the total time immobile in the forced-swim test (n=23–14 per group; f). At 2 weeks of age, there were no changes in the expression of Igf2 mRNA in the hippocampus; however, there was a significant decrease in the level of hippocampal Bdnf exon IV mRNA (g; n=4 per group). At 8 weeks of age, Igf2 mRNA was significantly decreased in the hippocampus of F1 females born to CORT fathers (n=6–7 per group), whereas there was a significant increase in the level of Bdnf exon IV (n=8 per group; h). *P<0.05, **P<0.001, values represent means±s.e.m. CORT, corticosterone.

Figure 3

Paternal CORT treatment is associated with altered anxiety-related behaviors of male offspring. Male offspring of CORT-treated mice made fewer calls during the last minute of separation (a; n=19–21 per group). Both 2-week (b; n=24 per group) and 8-week-old (c; n=15 per group) male F1 offspring display no differences in fear conditioning and extinction trials. Male offspring of CORT-treated mice spent less time in the open arms of the elevated plus-maze (EPM) (d; n=19 per group) and less time in the light half of the light–dark box (e; n=12 per group). No difference was observed on the forced swim test (FST) (f; n=23–24 per group). There was a significant increase in the level of Igf2 mRNA expression in the hippocampus; however, there were no changes in the expression of Bdnf exon IV expression in the hippocampus (g) n=8. *P<0.05, **P<0.001, values represent means±s.e.m. CORT, corticosterone.

Figure 4

Transgenerational effects of paternal CORT treatment on F2 anxiety and depression-related behaviors. F2 female mice born to F1 CORT male mice exhibited greater time in the open arms of the elevated-plus maze (n=7–13 per group; a), but no difference was observed in the light–dark box (n=9–14 per group; b). There was no significant difference in depression-related behaviors as assessed with the forced-swim (n=10–11 per group; c) and novelty-suppressed feeding (n=7–13 per group; d) tests. There were no significant changes to hippocampal mRNA expression of Igf2 or Bdnf exon 4 (n=8 per group; e). F2 male mice born to F1 CORT male mice also exhibited more time in the open arms of the elevated-plus maze (EPM) (n=13–14; f) but with no difference observed in the light–dark box (n=12–14 per group) (g). However, F2 CORT male mice exhibited pro-depressive behaviors in the forced-swim (n=12–17 per group; h) and novelty-suppressed feeding (n=6–10 per group; i) tests. There was a significant increase in the level of Igf2 mRNA expression in the hippocampus of F2 males; however, there were no changes in the expression of Bdnf exon IV expression in the hippocampus (n=7–8 per group; j). Data represented as mean±s.e.m. Unpaired t-test: *P<0.05, **P<0.01. CORT, corticosterone.

Figure 5

CORT treatment alters microRNAs (miRNAs) in sperm. Distribution of various types of small RNAs in the sperm of control and CORT-treated mice (a). Expression value of genes with altered expression compared with the average of the control animals (b). Genes that have predicted binding sites for miRNAs found to have altered expression in the sperm of CORT-treated animals (c). Fold change (relative to SNORD95) of top miRNA candidates from sequenced sperm (d). *P<0.05, **P<0.01, values represent means±s.e.m. CORT, corticosterone.