Sex-specific transcriptional signatures in the medial prefrontal cortex underlying sexually dimorphic behavioural responses to stress in rats

Abstract

Background: Converging evidence suggests that stress alters behavioural responses in a sex-specific manner; however, the underlying molecular mechanisms of stress remain largely unknown.

Methods: We adapted unpredictable maternal separation (UMS) and adult restraint stress (RS) paradigms to mimic stress in rats in early life or adulthood, respectively. The sexual dimorphism of the prefrontal cortex was noted, and we performed RNA sequencing (RNA-Seq) to identify specific genes or pathways responsible for sexually dimorphic responses to stress. We then performed quantitative reverse transcription polymerase chain reaction (qRT-PCR) to verify the results of RNA-Seq.

Results: Female rats exposed to either UMS or RS showed no negative effects on anxiety-like behaviours, whereas the emotional functions of the PFC were impaired markedly in stressed male rats. Leveraging differentially expressed genes (DEG) analyses, we identified sex-specific transcriptional profiles associated with stress. There were many overlapping DEGs between UMS and RS transcriptional data sets, where 1406 DEGs were associated with both biological sex and stress, while only 117 DEGs were related to stress. Notably, Uba52 and Rpl34-ps1 were the first-ranked hub gene in 1406 and 117 DEGs respectively, and Uba52 was higher than Rp134-ps1, suggesting that stress may have led to a more pronounced effect on the set of 1406 DEGs. Pathway analysis revealed that 1406 DEGs were primarily enriched in ribosomal pathway. These results were confirmed by qRT-PCR.

Limitations: Sex-specific transcriptional profiles associated with stress were identified in this study, but more in-depth experiments, such as single-cell sequencing and manipulation of male and female gene networks in vivo, are needed to verify our findings.

Conclusion: Our findings show sex-specific behavioural responses to stress and highlight sexual dimorphism at the transcriptional level, shedding light on developing sex-specific therapeutic strategies for stress-related psychiatric disorders.

Figure 1

Stress has different effects on anxiety- and depression-like behavioural phenotypes in female and male rats. (A) Schematic timeline and experimental paradigm. (B–E) Comparisons of behavioural effects of unpredictable maternal separation (UMS) in female and male rats, which were affected differently. (B) Distance moved in the open field test (OFT). There was a main effect of sex on distance moved (F_1,43 = 59.86, p < 0.0001). Among control rats, males showed a decreased distance moved relative to females (post hoc test, p < 0.001). (C) Duration in central zone in the OFT (2-way analysis of variance [ANOVA], interaction: F_1,43 = 7.202, p = 0.0103; effect of UMS: F_1,43 = 3.694, p = 0.0612; effect of sex: F_1,43 = 0.004930, p = 0.9443). Among male rats, time in central zone was significantly decreased (post hoc test, p = 0.0232) after UMS, and no detectable behavioural effect of UMS was observed in females in the OFT (post hoc test, p = 0.9320). From left to right in the figure panel: n = 15, n = 13, n = 10 and n = 9 rats). (D) Open arm ratio in the elevated plus maze (EPM) test (2-way ANOVA, interaction: F_1,33 = 0.008304, p = 0.92793; effect of UMS: F_1,33 = 21.46, p < 0.0001; effect of sex: F_1,33 = 19.85, p < 0.0001). Although time spent in open arms was significantly decreased both in females (post hoc test, p = 0.0164) and males (post hoc test, p = 0.0096), the decrease was more pronounced in males. Additionally, among control rats, males showed a decreased open arm ratio relative to females (post hoc test, p = 0.0184). From left to right in the figure panel: n = 9, n = 9, n = 10 and n = 9 rats. (E) Sucrose consumption in the sucrose preference test (SPT) (2-way ANOVA, interaction: F_1,41 = 1.458, p = 0.2342; effect of UMS: F_1,41 = 11.39, p = 0.0016; effect of sex: F_1,41 = 3.571, p = 0.0659). Among male rats, preference for sucrose was significantly decreased (post hoc test, p = 0.0192). From left to right in the figure panel: n = 12, n = 13, n = 11 and n = 9 rats. (F–I) Comparisons of behavioural effects of adult restraint stress (RS) in female and male rats. (F) distance moved in the OFT. There was a main effect of sex on distance moved (F_1,35 = 26.65, p < 0.0001). Among control rats, males showed a decreased distance moved relative to females (post hoc test, p = 0.036). (G) Duration in central zone in the OFT (2-way ANOVA, interaction: F_1,35 = 8.419, p = 0.0064; effect of RS: F_1,35 = 5.529, p = 0.0245; effect of sex: F_1,35 = 15.66, p = 0.0004). In males, RS decreased the duration in the central zone in the OFT (post hoc test, p = 0.0033). Notably, females did not show an elevated anxiety phenotype after exposure to RS (post hoc test, p = 0.9805). From left to right in the figure panel: n = 8, n = 10, n = 9 and n = 10 rats. (H) Open arm ratio in the EPM (2-way ANOVA, interaction: F_1,33 = 4.579, p = 0.0399; effect of RS: F_1,33 = 4.617, p = 0.0391; effect of sex: F_1,33 = 38.59, p < 0.0001). Male rats showed a decreased open arm ratio in the EPM (post hoc test, p = 0.0210) after RS. Notably, females did not show an elevated anxiety phenotype after exposure to RS (post hoc test, p > 0.9999). Additionally, among control rats, males showed a decreased open arm ratio relative to females (post hoc test, p = 0.0305). From left to right in the figure panel: n = 9, n = 9, n = 10 and n = 10 rats. (I) Sucrose consumption in the SPT (2-way ANOVA, interaction: F_1,34 = 2.118, p = 0.1548; effect of RS: F_1,34 = 0.00, p = 0.9972; effect of sex: F_1,34 = 2.048, p = 0.1615). There was no difference in sucrose preference in male (post hoc test, p = 0.7159) and female (post hoc test, p = 0.7501) rats. From left to right in the figure panel: n = 9, n = 9, n = 10 and n = 10 rats. Ctrl = control; mPFC = medial prefrontal cortex; PND = postnatal day. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 2

Differential gene expression analysis between female and male rats in the unpredictable maternal separation (UMS) analyses. (A) Heat map analysis. (B) Volcano plot of expression levels of the medial prefrontal cortex region in UMS rats; 759 genes were found to be downregulated in male rats compared with female rats. Female rats had 1380 genes decline compared with male rats (p ≤ 0.05, false discovery rate–corrected, fold change ≥ 2). (C) Network enrichment analysis for differentially expressed genes using DAVID (2021 update). The colour of the nodes correspond to adjusted p values. Two pathways (nodes) are connected if they share 30% (default, adjustable) or more genes. Green and red represent downregulated and upregulated pathways, respectively. Darker nodes are more significantly enriched gene sets, bigger nodes represent larger gene sets, and thicker edges represent more overlapped genes. F = female; logFC = log fold change; M = male; MHC = major histocompatibility complex.

Figure 3

Differential gene expression analysis between female and male rats in the adult restraint stress (RS) analyses. (A) Heat map analysis. (B) Volcano plot of expression levels of the medial prefrontal cortex region in RS rats; 417 genes were found to be downregulated in male rats compared with female rats. Female rats had 1278 genes decline compared with male rats (p ≤ 0.05, false discovery rate–corrected, fold change ≥ 2). (C) Network enrichment analysis for differentially expressed genes using DAVID (2021 update). Connected gene sets share more genes. The colour of the nodes correspond to adjusted p values. Two pathways (nodes) are connected if they share 30% (default, adjustable) or more genes. Green and red represent downregulated and upregulated pathways, respectively. Darker nodes are more significantly enriched gene sets, bigger nodes represent larger gene sets, and thicker edges represent more overlapped genes. F = female; logFC = log fold change; M = male; MHC = major histocompatibility complex.

Figure 4

Network analyses identify sex-specific transcriptional signatures associated with stress. (A) Differentially expressed genes (p < 0.05) and the top associated gene ontology (GO) terms among each indicated comparison. (B) Network of 117 genes associated with stress. Green and red represent downregulated and upregulated genes, respectively. Bigger nodes represent larger gene sets. (C) Network of ranked hub gene. CTRL = control; F = female; M = male; RS = adult restraint stress; TAP = transporter associated with antigen processing; UMS = unpredictable maternal separation.

Figure 5

Network analyses identify sex-specific transcriptional signatures associated with stress. (A) Fold change of hub genes (male v. female, genes associated with both sex and stress) expression in control rats and stressed rats. p < 0.05 for all listed genes. (B) Fold change of hub genes (male v. female, genes associated with stress) expression in adult restraint stress (RS) rats and unpredictable maternal separation (UMS) rats. p < 0.05 for all listed genes. (C) The expression of Uba52 in control rats, RS rats and UMS rats in RNA sequencing. (D) Associations of the behavioural test and ribosome pathways in rats. The Spearman correlation coefficient was calculated. Positive relationships are indicated in blue, and negative relationships are indicated in red. F = female; M = male.

Figure 6

Validation of the selected genes associated with both sex and stress in RNA sequencing data by quantitative reverse transcription polymerase chain reaction (qRT-PCR). (A–G) The level of Rps6, Uba52, Fau, Pten, Rpl19, Rsl1d1l1 and Rpl8 mRNA expression in the medial prefrontal cortex. All data from the qRT-PCR were analyzed using 2-way analysis of variance, with Bonferroni post hoc multiple comparisons correction. CTRL = control; F = female; M = male; RS = adult restraint stress; UMS = unpredictable maternal separation.