β-sitosterol reduces anxiety and synergizes with established anxiolytic drugs in mice

Abstract

Anxiety and stress-related conditions represent a significant health burden in modern society. Unfortunately, most anxiolytic drugs are prone to side effects, limiting their long-term usage. Here, we employ a bioinformatics screen to identify drugs for repurposing as anxiolytics. Comparison of drug-induced gene-expression profiles with the hippocampal transcriptome of an importin α5 mutant mouse model with reduced anxiety identifies the hypocholesterolemic agent β-sitosterol as a promising candidate. β-sitosterol activity is validated by both intraperitoneal and oral application in mice, revealing it as the only clear anxiolytic from five closely related phytosterols. β-sitosterol injection reduces the effects of restraint stress, contextual fear memory, and c-Fos activation in the prefrontal cortex and dentate gyrus. Moreover, synergistic anxiolysis is observed when combining sub-efficacious doses of β-sitosterol with the SSRI fluoxetine. These preclinical findings support further development of β-sitosterol, either as a standalone anxiolytic or in combination with low-dose SSRIs.

Keywords: CNS drugs; SSRIs; anxiolytics; drug repositioning; fluoxetine; metabolomics; phytosterol; transcriptomics; β-Sitosterol.

Graphical abstract

Figure 1

In silico screening for anxiolytic compounds mimicking the importin α5 mutant phenotype (A) The CMap approach was used to compare DEGs from hippocampal RNA-seq analysis of importin α5 knockouts to DEGs from cell lines treated with drugs and drug candidates. (B) Compounds functionally related to the query state (CMap score >0.75). The candidate compounds were prioritized based on lack of a prior indication for anxiety, ability to cross the blood-brain barrier, compatibility with oral application, and low likelihood of side effects (LSE). (C) The top five drugs meeting these criteria were tested for anxiolytic properties in mice using the open-field test 1 h after i.p. injection, as compared to vehicle. The time spent, the number of rearings, the number of visits, and the distance traveled in the open-field center are represented normalized to vehicle treatment values for each group. Data are shown for β-sitosterol (100 mg/kg), n = 5, vehicle, n = 8; alvespimycin (75 mg/kg), n = 10, vehicle, n = 10; oxamniquine (15 mg/kg), n = 5, vehicle, n = 5; fluspirilene (10 mg/kg), n = 8, vehicle, n = 7 and primaquine (60 mg/kg) n = 5, vehicle, n = 5. The bottom panel depicts group heatmap representations of mouse activity over 10 min of open-field exploration. See Table S2 for raw data analysis.∗p < 0.05, ∗∗p < 0.01; two-tailed t test, mean ± SEM (D and E) β-sitosterol was further studied for its ability to reduce anxiety when given by oral gavage, in both dose-response and time window experiments. The compound showed significant anxiolytic properties in the open-field test at a dosage of 100 mg/kg and 1 h after its administration (D: n = 8 mice per dosage, 1-way ANOVA followed by Dunnett’s multiple-comparisons test with the vehicle group as control; E: β-sitosterol, n = 10, vehicle, n = 9 mice at 1 h and n = 5 mice in both groups at 6 h, 2-way ANOVA followed by Sidak’s multiple-comparisons test); ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001; mean ± SEM. (F and G) β-sitosterol effects were studied in the elevated plus maze. The compound showed significant anxiolytic properties at a dosage of 100 mg/kg (i.p) 1 h after its administration (F) and no effects when tested 6 h after injection (G). (F and G: β-sitosterol, n = 10, vehicle, n = 10 mice in each time points, two-tailed t test; ∗p < 0.05; mean ± SEM). See also Figure S1.

Figure 2

Comparison of the anxiolytic activities of plant sterols (A) Chemical structures of β-sitosterol, stigmasterol, campesterol, brassicasterol, and fucosterol. All phytosterols were diluted in corn oil (vehicle) and administered intraperitoneally at a final dosage of 100 mg/kg, 1 h before evaluation of anxiolytic effects in the open-field test. The bottom panel depicts group heatmap representations of mouse activity over 10 min of open-field exploration. (B and C) β-sitosterol was the only compound tested that caused a significant increase in the distance traveled (B) and the number of visits in the open-field center (C) compared to vehicle-treated mice. Vehicle, n = 32; β-sitosterol, n = 10; stigmasterol, n = 5; campesterol, n = 10; brassicasterol, n = 5; fucosterol, n = 5. The effect of the different sterols on the OF center exploration was analyzed by a 1-way ANOVA followed by Dunnett’s multiple-comparisons test with the vehicle group as control. ∗∗∗p < 0.0001, mean ± SEM. See also Figure S1.

Figure 3

Effects of β-sitosterol treatment on gene expression in mouse hippocampus (A) Heatmap representation of standardized log₂-normalized counts of DEGs after β-sitosterol treatment (n = 4 mice/group, 2 groups: 1 h, 6 h after 100 mg/kg β-sitosterol i.p. injection). K-means clustering analysis revealed a specific pattern of gene downregulation in cluster 4 at 1 h after treatment. (B) Heatmap based on gene sets of cluster 4, Log₂ ratio of treated (drug) versus untreated (vehicle) is represented. A clear pattern of downregulation in the sitosterol 1 h condition not visible after 6 h or under stigmasterol treatment (see Figure S2). (C) Timeline and schematic representation of the contextual fear conditioning experiment. Gray bars indicate 1 h intervals. On the second day (acquisition phase) and third day (context test), animals were injected with β-sitosterol (100 mg/kg, i.p.) or its vehicle solution (corn oil) 1 h before the experiment. After completion of the context test on day 3, brains were harvested for immunohistochemistry. (D) Total freezing duration in the context text. n = 8 mice per group. ∗∗p < 0.01; two-tailed t test. Mapping of c-Fos activated neurons in vivo. (E) Representative images of the prefrontal cortex (PFC) and the dorsal dentate gyrus (DG). Scale bar, 200 μm. (F and G) Quantifications of the number of c-Fos-positive neurons in the PFC (F) and the DG (G) after contextual fear conditioning. n = 3 mice per group. ∗p < 0.05; ∗∗p < 0.01; two-tailed t test, mean ± SEM. See also Figure S2.

Figure 4

Fluoxetine synergizes with β-sitosterol for anxiolysis in mice The effect of fluoxetine/β-sitosterol co-treatments was studied by injecting mice with 5 mg/kg fluoxetine (daily i.p. injection for 3 weeks) and then the indicated doses of β-sitosterol (1 h before the test). Control littermates received both vehicles (saline and corn-oil) only. (A) Group heatmap representations of mouse activity over 10 min of open-field exploration. (B and C) Distance traveled (B) and time spent (C) in the open-field center (in cm) monitored over 10 min in the open-field test for vehicle-injected animals (vehicle, n = 9) and mice receiving one of the 6 combinations as indicated (n ≥ 9 per combination). The effects on OF center exploration in (B) and (C) were analyzed by a two-sided Kruskal-Wallis test with vehicle as control. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. Error bars represent the median with a 95% confidence interval (CI). See also Figure S3. (D and E) Mice were subjected to 1.5 h restraint stress for 3 weeks and treated with vehicle (RSV), fluoxetine 20 mg/kg (RSF), β-sitosterol 100 mg/kg (RSB), or the combination of fluoxetine 5 mg/kg + β-sitosterol 20 mg/kg (RSC) and compared to naive mice in the novelty-suppressed feeding test (NSFT). Data are expressed as the cumulative percentage of animals that have eaten (latency to eat, in seconds), over the test session (naive, n = 10; RSF, n = 10; RSB, n = 9; RSC, n = 10). Data were analyzed using the Mantel-Cox log-rank test with the naive group set as the control. ∗p < 0.05. (F) Timeline and schematic representation of the different drug treatments preceding the LC-MS/MS experiment. (G) Heatmap representation of the metabolite levels (expressed as a log2 fold change between vehicle and treated mice) in each of the four conditions: β-sitosterol 100 mg/kg (1), fluoxetine 5 mg/kg (2), fluoxetine 20 mg/kg (3), and the combination fluoxetine 5 mg/kg + β-sitosterol 20 mg/kg (4). (H and I) Volcano plot representations of the neurochemical alterations induced by the 100 mg/kg β-sitosterol treatment (i.p.) in the PFC (H) and the HPC (I). (J)–(M) Volcano plot representations of the neurochemical alterations induced by a treatment combining fluoxetine (5 mg/kg, daily i.p. injection for 3 weeks) and β-sitosterol (20 mg/kg, 1 h before the test, i.p.) in the PFC (J), the CPu (K), the HPC (L), and the SN (M). For results from (H)–(M), changes in compound concentration when the log₂ fold change (vehicle versus β-sitosterol) >0.58 were considered significant at p < 0.05 (represented on the plot as –log₁₀(p value) >1.3). See also Figure S4 and Table S4.