Lack of interferon regulatory factor 3 leads to anxiety/depression-like behaviors through disrupting the balance of neuronal excitation and inhibition in mice

Abstract

Disrupting the balance of neuronal excitation and inhibition (E/I) is an important pathogenic mechanism of anxiety and depression. Interferon regulatory factor 3 (IRF3) plays a key role in the innate immune response, and activation of IRF3 triggers the expression of type I interferons and downstream interferon-stimulated genes, which are associated with anxiety and depression. However, whether IRF3 participates in the pathogenesis of anxiety/depression by regulating E/I balance remains poorly understood. Here, we reported that global knockout (KO) of IRF3 (IRF3^-/-) significantly increased anxiety/depression-like behaviors, but did not affect normal spatial learning and memory. Compared with wild type (WT) control mice, the E/I balance was disrupted, as reflected by enhanced glutamatergic transmission and decreased GABAergic transmission in the neurons of hippocampal CA1 and medial prefrontal cortex (mPFC) in IRF3-KO mice. Importantly, genetic rescue of IRF3 expression by adeno-associated virus (AAV) was sufficient to alleviate anxiety/depression-like behaviors and restore the neuronal E/I balance in IRF3-KO mice. Taken together, our results indicate that IRF3 is critical in maintaining neuronal E/I balance, thereby playing an essential role in ensuring emotional stability.

Keywords: Anxiety; Depression; Hippocampus; IRF3; Medial prefrontal cortex; Synaptic transmission.

Figure 1

IRF3-KO mice have no effect on spatial learning and memory. (A, B)IRF3 KO mice were verified by PCR (A) and Western blotting (B). (C) The escape latency to the hidden platform during the Morris water maze training (n = 10–16 per group). Repeated measures ANOVA: 1.5 m, F(1,19) = 76.335, P = 0.955; 3 m, F(1,30) = 76.335, P = 0.426; 6 m, F(1,20) = 256.528, P = 0.368. (D) The time spent in the target quadrant during the Morris water maze test. Unpaired Student's t-test: 1.5 m, t = 2.284, P = 0.321; 3 m, t = 1.291, P = 0.170; 6 m, t = 0.308, P = 0.760. (E) The number of entries to the platform zone during the Morris water maze test. Unpaired Student's t-test: 1.5 m, t = 0.617, P = 0.916; 3 m, t = 5.366, P = 0.131; 6 m, t = 0.099, P = 0.212. Data are expressed as mean ± SEM. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001; ns, no significant difference.

Figure 2

IRF3-KO mice exhibit age-dependent anxiety/depression-like behaviors. (A) Representative heat maps in the open field test. (B) The time spent in the center zone during the open field test (n = 10–13 per group). Unpaired Student's t-test: 1.5 m, t = 3.537, P = 0.159; 3 m, t = 0.498, P = 0.001; 6 m, t = 3.756, P < 0.001. (C) The number of entries in the center zone during the open field test. Unpaired Student's t-test: 1.5 m, t = 4.468, P = 0.416; 3 m, t = 0.001, P < 0.001; 6 m, t = 4.797, P < 0.001. (D) Representative heat maps in the elevated plus maze test. (E) The time spent in the open arm during the elevated plus maze test (n = 10–19 per group). Unpaired Student's t-test: 1.5 m, t = 0.584, P = 0.921; 3 m, t = 1.281, P < 0.001; 6 m, t = 30.364, P = 0.009. (F) The number of entries to the open arm during the elevated plus maze test. Unpaired Student's t-test: 1.5 m, t = 0.919, P = 0.184; 3 m, t = 0.172, P < 0.001; 6 m, t = 0.602, P = 0.004. (G) The total immobility time in tail suspension test (TST) (n = 8–17 per group). Unpaired Student's t-test: 1.5 m, t = 1.601, P = 0.535; 3 m, t = 5.315, P = 0.003; 6 m, t = 5.876, P = 0.002. (H) The latency to immobility in TST. Unpaired Student's t-test: 1.5 m, t = 6.421, P = 0.512; 3 m, t = 6.714, P = 0.241; 6 m, t = 2.690, P = 0.008. (I) The total immobility time in forced swimming test (FST) (n = 8–17 per group). Unpaired Student's t-test: 1.5 m, t = 3.454, P = 0.429; 3 m, t = 4.967, P = 0.007; 6 m, t = 2.017, P < 0.001. (J) The latency to immobility in FST. Unpaired Student's t-test: 1.5 m, t = 1.148, P = 0.074; 3 m, t = 4.296, P = 0.267; 6 m, t = 12.528, P = 0.004. Data are expressed as mean ± SEM. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001; ns, no significant difference.

Figure 3

Restoring IRF3 expression alleviates anxiety/depression-like behaviors in 3-month-old IRF3-KO mice. (A) Detection of AAV infection in neurons (green) in the mPFC and hippocampus using immunofluorescence assay. (B) The expression of endogenous IRF3 (∼45KD) and IRF3-EGFP fusion protein (∼75KD) in the mPFC and hippocampus following i.c.v. microinjection of AAV_IRF3 and its control AAV_EGFP assessed by Western blotting. (C) Representative heat maps in the open field test. (D) The time spent in the center zone during the open field test (n = 11–17 per group). One-way ANOVA: F(2,37) = 12.085, P < 0.001. (E) The number of entries in the center zone during the open field test. One-way ANOVA: F(2,37) = 19.412, P < 0.001. (F) Representative heat maps in the elevated plus maze test. (G) The time spent in the open arm during the elevated plus maze test (n = 11–17 per group). One-way ANOVA: F(2,37) = 15.433, P < 0.001. (H) The number of entries to the open arm during the elevated plus maze test. One-way ANOVA: F(2,37) = 26.417, P < 0.001. (I) The total immobility time in tail suspension test (TST) (n = 11–17 per group). One-way ANOVA: F(2,37) = 10.293, P < 0.001. (J) The latency to immobility in TST. One-way ANOVA: F(2,37) = 7.275, P = 0.002. (K) The total immobility time in forced swimming test (FST) (n = 11–17 per group). One-way ANOVA: F(2,37) = 8.658, P < 0.001. (L) The latency to immobility in FST. One-way ANOVA: F(2,37) = 9.350, P < 0.001. Data are expressed as mean ± SEM. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001; ns, no significant difference.

Figure 4

IRF3-KO results in neuronal hyperexcitability in the mPFC. (A) Representative AP firing in the mPFC neurons. (B) Bar graphs of the resting membrane potential (RMP) in mPFC neurons (n = 15–18 cells from 3 mice per group). One-way ANOVA: F(2,46) = 19.412, P = 0.498. (C) Bar graphs of the APs threshold in mPFC neurons. One-way ANOVA: F(2,46) = 4.708, P = 0.023. (D) Bar graphs of the number of APs evoked by current injections from 0 to 180 pA. Repeated measures ANOVA: F(2,46) = 186.142, P = 0.020. (E) Representative mEPSC traces in the mPFC neurons. (F) Bar graphs of mEPSCs amplitude (n = 19–23 cells from 3 mice per group). One-way ANOVA: F(2,61) = 9.997, P < 0.001. (G) Bar graphs of mEPSCs frequency. One-way ANOVA: F(2,61) = 23.620, P < 0.001. (H) Representative mIPSCs traces in the mPFC neurons. (I) Bar graphs of mIPSCs amplitude (n = 17–19 cells from 3 to 4 mice per group). One-way ANOVA: F(2,52) = 41.788, P < 0.001. (J) Bar graphs of mIPSCs frequency. One-way ANOVA: F(2,52) = 11.195, P < 0.001. Data are expressed as mean ± SEM, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001; ns, no significant difference.

Figure 5

IRF3-KO results in neuronal hyperexcitability in the CA1 area of the hippocampus. (A) Representative AP firing in the hippocampal CA1 neurons. (B) Bar graphs of the RMP in the hippocampal CA1 neurons (n = 17–22 cells from 3 mice per group). One-way ANOVA: F(2,56) = 0.017, P = 0.983. (C) Bar graphs of the APs threshold. One-way ANOVA: F(2,56) = 0.173, P = 0.842. (D) Bar graphs of the number of APs evoked by current injections from 0 to 180 pA. Repeated measures ANOVA: F(2,56) = 254.068, P = 0.980. (E) Representative mEPSCs traces of the hippocampal CA1 neurons. (F) Bar graphs of mEPSCs amplitude (n = 11–15 cells from 3 to 4 mice per group). One-way ANOVA: F(2,37) = 9.976, P < 0.001. (G) Bar graphs of mEPSCs frequency. One-way ANOVA: F(2,37) = 2.372, P = 0.107. (H) Representative mIPSCs traces of the hippocampal CA1 neurons. (I) mIPSCs amplitude of CA1 neurons (n = 18–21 cells from 3 mice per group). One-way ANOVA: F(2,56) = 9.772, P < 0.001. (J) Bar graphs of mIPSCs frequency. One-way ANOVA: F(2,56) = 7.089, P = 0.002. Data are expressed as mean ± SEM, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001; ns, no significant difference.