Neurosteroid allopregnanolone (3α,5α-THP) inhibits inflammatory signals induced by activated MyD88-dependent toll-like receptors

Abstract

We have shown that endogenous neurosteroids, including pregnenolone and 3α,5α-THP inhibit toll-like receptor 4 (TLR4) signal activation in mouse macrophages and the brain of alcohol-preferring (P) rat, which exhibits innate TLR4 signal activation. The current studies were designed to examine whether other activated TLR signals are similarly inhibited by 3α,5α-THP. We report that 3α,5α-THP inhibits selective agonist-mediated activation of TLR2 and TLR7, but not TLR3 signaling in the RAW246.7 macrophage cell line. The TLR4 and TLR7 signals are innately activated in the amygdala and NAc from P rat brains and inhibited by 3α,5α-THP. The TLR2 and TLR3 signals are not activated in P rat brain and they are not affected by 3α,5α-THP. Co-immunoprecipitation studies indicate that 3α,5α-THP inhibits the binding of MyD88 with TLR4 or TLR7 in P rat brain, but the levels of TLR4 co-precipitating with TRIF are not altered by 3α,5α-THP treatment. Collectively, the data indicate that 3α,5α-THP inhibits MyD88- but not TRIF-dependent TLR signal activation and the production of pro-inflammatory mediators through its ability to block TLR-MyD88 binding. These results have applicability to many conditions involving pro-inflammatory TLR activation of cytokines, chemokines, and interferons and support the use of 3α,5α-THP as a therapeutic for inflammatory disease.

Fig. 1. 3α,5α-THP inhibits the activation of MyD88-dependent TLR pathways in RAW264.7 cells.

RAW264.7 cells (n = 5–10/grp) were treated with Pam3Cys (10 µg/ml; 30 min) (A), imiquimod (IMQ; 3 µg/ml; 24 h) (B), or Poly(I:C) (25 µg/ml; 24 h) (C) with or without 3α,5α-THP (1 µM). Cells were harvested at 24 h after treatment initiation and examined for the expression of MyD88-dependent (A, B) and TRIF-dependent (C) signal activation. A Pam3Cys caused a significant increase in the levels of TRAF6 (n = 8/grp), pERK1/2 (n = 10/grp), pCREB (n = 10/grp), pATF2 (n = 10/grp), and TNF-α (n = 5/grp) relative to vehicle control (CTL), and these increases were completely blocked by 3α,5α-THP (One-way ANOVA, Tukey’s post hoc test: *p < 0.05; **p < 0.01). B IMQ caused a significant increase in the levels of pIRF7 (n = 9/grp) and TNF-α (n = 5/grp) relative to CTL, and these increases were completely blocked by 3α,5α-THP (One-way ANOVA, Tukey’s post hoc test: *p < 0.05, **p < 0.01, ***p < 0.001). C Poly(I:C) significantly increased the levels of pIRF3 (n = 5/grp), IP-10 (n = 6/grp), and TNF-α (n = 5/grp) relative to CTL (One-way ANOVA, Tukey’s post hoc test: *p < 0.05). The increases of pIRF3, IP-10, and TNF-α were not inhibited by 3α,5α-THP (One-way ANOVA, Tukey’s post hoc test: p > 0.05). D 3α,5α-THP does not target the non-activated TLR signals. RAW264.7 cells untreated with TLR agonist but exposed to vehicle or treated with 3α,5α-THP (1 µM) were harvested after 24 h. The levels of TRAF6 (n = 6/grp), pERK1/2 (n = 6/grp), pCREB (n = 6/grp), pATF2 (n = 6/grp), TNF-α (n = 6/grp), pIRF7 (n = 6/grp), IP-10 (n = 6/grp), and pIRF3 (n = 6/grp) were similar in the 3α,5α-THP-treated and untreated cells (t-test, p > 0.05). The data indicate that 3α,5α-THP specifically targets only the activated TLR signal.

Fig. 2. 3α,5α-THP inhibits the innately activated MyD88-dependent TLR4 and TLR7 signals in the nucleus accumbens (NAc) and amygdala of P rats.

A, B Male and female alcohol-preferring (P) rats (n = 8/grp) were treated (IP; 30 min) with 3α,5α-THP (15 mg/kg) or vehicle (45% w/v 2-hydroxypropyl-β-cyclodextrin) control and the amygdala and NAc were examined for MCP-1 and TLR4 expression. A There is no a significant sex difference for MCP-1 expression in the amygdala (Two-way ANOVA: F(1,28) = 0.02030, p = 0.8877, n = 8/grp). However, in the NAc there is a significant sex difference in the MCP-1 level (Two-way ANOVA: F(1,28) = 72.27, p < 0.0001, n = 8/grp). Specifically, baseline MCP-1 levels are significantly higher in the NAc from males than females (Two-way ANOVA, Tukey’s post hoc test: ****p < 0.0001, n = 8/grp). 3α,5α-THP administration significantly reduces MCP-1 expression in the male and female amygdala (n = 8/grp) and NAc (n = 8/grp) (Two-way ANOVA, Tukey’s post hoc test: *p < 0.05, **p < 0.01). However, the total values of MCP-1 in the presence of 3α,5α-THP are also higher in males than females (Two-way ANOVA, Tukey’s post hoc test: ****p < 0.0001, n = 8/grp) in P rat NAc. B The levels of TLR4 are similar in the males and females NAc tissues (Two-way ANOVA: F(1,28) = 0.06052, p = 0.8075, n = 8/grp) and they are not altered by 3α,5α-THP treatment (Two-way ANOVA: F(1,28) = 0.4218, p = 0.5213, n = 8/grp). C The TLR7 signal is innately activated in female P rats. NAc tissues from male and female alcohol-preferring (P) (n = 6/grp) and non-preferring (NP) control (n = 6/grp) rats were assayed for the expression of TLR7 and pIRF7, which is indicative of TLR7 signal activation. The levels of TLR7 and pIRF7 are significantly higher in female P than NP rats (t-test: *p < 0.05, n = 8/grp) consistent with innate TLR7 activation in female P rats. The levels of TLR7 and pIRF7 are similar in male P and NP rats (SI, Fig. S1). D, E Female P rats (n = 8/grp) were treated (IP; 30 min) with 3α,5α-THP or vehicle control and the NAc and amygdala were examined for expression of TLR7 and pIRF7 (indicative of TLR7 activation). 3α,5α-THP administration significantly reduced pIRF7 expression in the NAc and amygdala from female P rats (t-test: *p < 0.05, n = 8/grp) (D), but it had no effect on TLR7 expression in female P rat NAc (t-test, p > 0.05, n = 8/grp) (E).

Fig. 3. 3α,5α-THP inhibits MyD88, but not TRIF binding to TLRs.

Protein extracts obtained from the NAc of male and female P rats (n = 3–4/grp) after administration of 3α,5α-THP (15 mg/kg) or vehicle (45% w/v 2-hydroxypropyl-β-cyclodextrin) were immunoprecipitated (IP) with antibody to MyD88 (A, B) or TRIF (C). Proteins immunoprecipitated with MyD88 antibody were analyzed by immunoblotting (IB) with TLR4 and MyD88 antibodies (A) and TLR7 and MyD88 antibodies (B). Proteins immunoprecipitated with TRIF antibody were analyzed by immunoblotting with TLR4 and TRIF antibodies (C). Normal IgG was used as immunoprecipitation control. MyD88 co-precipitated with TLR4 (A) and TLR7 (B), and TRIF co-precipitated with TLR4 (C) but MyD88 and TRIF did not co-precipitate with normal IgG (A–C). A The levels of TLR4 co-precipitating with MyD88 were significantly reduced by 3α,5α-THP both in males (n = 3/grp) and females (n = 4/grp) (t-test, *p < 0.05). B The levels of TLR7 co-precipitating with MyD88 were also significantly reduced by 3α,5α-THP in females (t-test, *p < 0.05, n = 3/grp). C The levels of TLR4 co-precipitating with TRIF were not altered by 3α,5α-THP treatment both in males (n = 3/grp) and females (n = 3/grp) (t-test, p > 0.05).

Fig. 4. Schematic of TLR signal inhibition by 3α,5α-THP.

TLR4, TLR2, and TLR7 signal activation initiates with TLR binding of MyD88, and this binding is inhibited by 3α,5α-THP. This results in the inhibition of pathway activation, including TRAF6 and activated (phosphorylated) downstream members (pERK1/2, pCREB, pATF2, and pIRF7). The activated transcription factors translocate to the nucleus and initiate the production of various inflammatory mediators (IP-10, MCP-1, and TNF-α). TLR4 also binds TRIF, which is the only adaptor protein bound by TLR3, leading to the phosphorylation (activation) of IRF3. Significantly, 3α,5α-THP does not inhibit TRIF binding and its resulting signal activation.