Luteolin ameliorates chronic stress-induced depressive-like behaviors in mice by promoting the Arginase-1+ microglial phenotype via a PPARγ-dependent mechanism

Abstract

Accumulating evidence shows that neuroinflammation substantially contributes to the pathology of depression, a severe psychiatric disease with an increasing prevalence worldwide. Although modulating microglial phenotypes is recognized as a promising therapeutic strategy, effective treatments are still lacking. Previous studies have shown that luteolin (LUT) has anti-inflammatory effects and confers benefits on chronic stress-induced depression. In this study, we investigated the molecular mechanisms by which LUT regulates the functional phenotypes of microglia in mice with depressive-like behaviors. Mice were exposed to chronic restraint stress (CRS) for 7 weeks, and were administered LUT (10, 30, 40 mg· kg^-1 ·day^-1, i.g.) in the last 4 weeks. We showed that LUT administration significantly ameliorated depressive-like behaviors and decreased hippocampal inflammation. LUT administration induced pro-inflammatory microglia to undergo anti-inflammatory arginase (Arg)-1⁺ phenotypic polarization, which was associated with its antidepressant effects. Furthermore, we showed that LUT concentration-dependently increased the expression of PPARγ in LPS + ATP-treated microglia and the hippocampus of CRS-exposed mice, promoting the subsequent inhibition of the NLRP3 inflammasome. Molecular dynamics (MD) simulation and microscale thermophoresis (MST) analysis confirmed a direct interaction between LUT and peroxisome proliferator-activated receptor gamma (PPARγ). By using the PPARγ antagonist GW9662, we demonstrated that LUT-driven protection, both in vivo and in vitro, resulted from targeting PPARγ. First, LUT-induced Arg-1⁺ microglia were no longer detected when PPARγ was blocked. Next, LUT-mediated inhibition of the NLRP3 inflammasome and downregulation of pro-inflammatory cytokine production were reversed by the inhibition of PPARγ. Finally, the protective effects of LUT, which attenuated the microglial engulfment of synapses and prevented apparent synapse loss in the hippocampus of CRS-exposed mice, were eliminated by blocking PPARγ. In conclusion, this study showed that LUT ameliorates CRS-induced depressive-like behaviors by promoting the Arg-1⁺ microglial phenotype through a PPARγ-dependent mechanism, thereby alleviating microglial pro-inflammatory responses and reversing microglial phagocytosis-mediated synapse loss.

Keywords: PPARγ; depression; luteolin; microglia; neuroinflammation; phagocytosis.

Fig. 1. LUT alleviated depressive- and anxiety-like behaviors of mice induced by CRS exposure in mice.

a Molecular structure of luteolin (LUT). b Timeline of the animal experiments. CRS chronic restraint stress, Flx fluoxetine, OFT open field test, TST tail suspension test, FST forced swimming test, SPT sucrose preference test. c Sucrose preference rate of each group before treatment (3rd week) or after treatment (7th week) with either LUT or Flx. Values are shown as the mean ± SEM. ns not significant, *P < 0.05, **P < 0.01 and ***P < 0.001 by two-way ANOVA followed by Tukey’s post hoc test. d Effects of LUT or Flx administration on the immobility time of CRS-exposed mice in the FST, n = 10 per group. e Effects of LUT or Flx administration on the immobility time of CRS-exposed mice in the TST (n = 8–10 mice/group). f–h Effects of LUT or Flx administration on anxiety-like behaviors of CRS-exposed mice in the OFT. f The frequency and g cumulative duration of mice in the central area in the OFT (n = 10). h The total distance travelled of mice in the OFT (n = 10). The data are presented as the mean ± SEM. ns not significant, *P < 0.05, **P < 0.01 and ***P < 0.001 by one-way ANOVA followed by Tukey’s post hoc test.

Fig. 2. LUT promoted an Arg-1⁺ phenotype and reduced pro-inflammatory microglia in vivo.

a Representative images of Iba-1 immunofluorescence staining in hippocampus. b Quantification of the density of Iba-1⁺ microglia in the hippocampus of each group, n = 4 mice per group. c Proportion of Iba-1⁺ microglia in the hippocampus of each group, n = 4 mice per group. d Effects of LUT on the expression of Inos, Cd86, Arg-1 and Cd206 in the hippocampus of CRS-exposed mice, normalized to the control group (n = 5). e Representative images of immunofluorescence staining, and f percentages of iNOS⁺/Iba-1⁺ microglia in the hippocampus of each group, n = 4 mice per group; scale bar = 100 μm. g Representative images of immunofluorescence staining and h percentages of Arg-1⁺/Iba-1⁺ microglia in the hippocampus of each group, n = 4 mice per group; scale bar = 100 μm. The data are presented as the mean ± SEM. *P < 0.05, **P < 0.01 and ***P < 0.001 by one-way ANOVA followed by Tukey’s post hoc test. i, j Analyses of correlations between the proportions of iNOS⁺ microglia and Arg-1⁺ microglia with SPT outcomes (n = 4). k, l Analyses of the correlations between the proportions of iNOS⁺ microglia and Arg-1⁺ microglia with immobility time in the FST (n = 4). The correlation analysis was performed using Spearman’s correlation test, *P < 0.05, **P < 0.01 and ***P < 0.001.

Fig. 3. LUT promoted anti-inflammatory microglial polarization in vitro.

Primary hippocampal microglia were isolated from the mice and cultured with 10 μM or 20 μM LUT under LPS-induced conditions. a Images of immunofluorescence staining for Arg-1 in Iba-1⁺ microglia detected via fluorescence microscopy. Scale bar, 20 μm. b The mean fluorescence intensity of Arg-1 was quantified. c Effects of LUT on the expression of Inos and Cd86 (d), as well as Arg-1 and Cd206, in LPS-treated microglia. Values are shown as the mean ± SEM of at least three independent experiments. *P < 0.05, **P < 0.01 and ***P < 0.001 by one-way ANOVA followed by Tukey’s post hoc test.

Fig. 4. LUT inhibited microglial NLRP3 inflammasome activation.

a Representative images of NLRP3 immunofluorescence staining in the hippocampal microglia of each group. b Mean fluorescence intensity ratio of NLRP3 in the hippocampal microglia of each group (n = 4). c–f Hippocampal expression of the NLRP3, cleaved caspase-1, and cleaved IL-1β proteins was measured in the control, LUT, CRS and CRS + LUT groups (n = 5). g, h Detection of NLRP3 inflammasome activation induced by LPS and ATP in primary microglia. Microglial levels of the NLRP3, cleaved caspase-1 and cleaved IL-1β proteins were detected. (n = 4). i, j Primary microglia were exposed to LPS + ATP and LUT, and the proportion of cells was assessed with Pycard specks and quantified via fluorescence microscopy. Scale bar = 10 μm. Data are expressed as the mean ± SEM of at least three independent experiments, ns not significant, *P < 0.05, **P < 0.01, and ***P < 0.001 by one-way ANOVA followed by Tukey’s post hoc test.

Fig. 5. LUT regulated PPARγ expression in LPS + ATP-induced microglia and the hippocampus of CRS-exposed mice.

a Primary microglia were exposed to LPS + ATP and LUT, and PPARγ expression was assessed via fluorescence microscopy. Scale bar, 10 μm. b, c Microglial expression of the PPARγ protein in the control, LPS + ATP, LPS + ATP + LUT (10 μM) and LPS + ATP + LUT (20 μM) groups was measured by Western blotting (n = 4). d Representative images of immunofluorescence staining showing PPARγ expression in the hippocampus of the control, LUT, CRS, and CRS + LUT groups. Scale bar, 200 μm. e Mean fluorescence intensity of PPARγ in the hippocampus of each group (n = 4). f, g PPARγ protein expression was measured by Western blotting (n = 5). Values are shown as the mean ± SEM. ns not significant, *P < 0.05, **P < 0.01 and ***P < 0.001 by one-way ANOVA followed by Tukey’s post hoc test.

Fig. 6. LUT interacts with PPARγ.

a The binding mode of LUT with the PPARγ protein. b MD simulates of RMSD trace values for the PPARγ-luteolin complex or free PPARγ over a given course of the simulation period (100 ns). c MD simulations of RMSF trace values of the PPARγ-luteolin complex or free PPARγ over a given course of the simulation period (100 ns). d PPARγ-luteolin complex or free PPARγ Rg traces. e MD simulation of the SASA of the PPARγ-luteolin complex or free PPARγ. f MD simulation of the number of hydrogen bonds in the PPARγ-luteolin complex. g The MST traces of LUT and PPARγ. h Dose-response curve of the ligand fitted by MST analysis software of PPARγ with LUT.

Fig. 7. PPARγ blockade reversed the LUT-elicited anti-inflammatory microglial polarization in vitro.

a Schematic diagram of the experimental strategy. b Effects of LUT and GW9662 on the expression of Inos, Cd86, Arg-1 and Cd206 in LPS-exposed microglia. Gapdh was used as the reference, and the data were normalized to the control group. c Representative images of immunofluorescence staining for Arg-1 in primary microglia from the control, LPS, LPS + LUT (20 μM), and LPS + LUT (20 μM) + GW9662 groups. Scale bar, 10 μm. d Mean fluorescence intensity of Arg-1 in microglia from each group (n = 4). e Microglial expression of the NLRP3, caspase-1, cleaved caspase-1, IL-1β, and cleaved IL-1β proteins was assessed by Western blotting (n = 4). Data are expressed as the mean ± SEM of at least three independent experiments, ns not significant, *P < 0.05, **P < 0.01, and ***P < 0.001 by one-way ANOVA followed by Tukey’s post hoc test.

Fig. 8. PPARγ blockade reversed the LUT-elicited anti-inflammatory microglial polarization in vivo.

a Schematic depiction of the experimental timeline used in the animal study. b Representative images of immunofluorescence staining and c percentages of Arg-1⁺/Iba-1⁺ microglia in the hippocampi of the control, CRS, CRS + LUT and GW groups; n = 3; scale bar = 50 μm. d The expression of the Inos, Cd86, Arg-1 and Cd206 mRNAs in the hippocampus of the control, CRS, CRS + LUT and GW groups was normalized to the levels in the control group (n = 4). e, f The hippocampal levels of proteins in the NLRP3 inflammasome (NLRP3, CAS-1, c-CAS-1, IL-1β and c-IL-1β) were detected via Western blotting and normalized to the β-actin level (n = 4). The data are presented as the mean ± SEM, *P < 0.05, **P < 0.01, and ***P < 0.001 by one-way ANOVA followed by Tukey’s post hoc test.

Fig. 9. PPARγ blockade eliminated the effects of LUT on alleviating hippocampal microglial synaptic phagocytosis and depressive behaviors in mice exposed to CRS.

a Phagocytosed microspheres were detected to assess the phagocytic activity of Iba-1⁺ primary microglia via fluorescence microscopy. Scale bar, 10 μm. b The densities of the beads in the control, LPS, LPS + LUT, and LPS + GW9662 + LUT groups were analyzed and quantified. c Representative images showing immunofluorescence staining for PSD95 in Iba-1⁺ microglia in the hippocampus, as assayed via fluorescence microscopy. d The density of PSD95 in Iba-1⁺ cells from the control, CRS, CRS + LUT, and GW groups was quantified and analyzed (n = 3). e, f The PSD95, TrkB, and BDNF protein expression was determined by Western blotting, with β-actin used as the reference (n = 4). Effects of LUT on CRS-induced depressive behaviors in mice via the activation of PPARγ, including the sucrose preference rate (g, n = 10), immobility time in the tail suspension test (h, n = 8–10 mice per group), immobility time in the forced swim test (i, n = 10), and total exploration distance in the open field test (j, n = 10). The data are presented as the mean ± SEM, ns means not significant, *P < 0.05, **P < 0.01, ***P < 0.001 by one-way ANOVA followed by Tukey’s post hoc test.

Fig. 10. Mechanisms of luteolinʼs antidepressant effects via PPARγ activation.

Luteolin alleviates chronic stress-induced depressive-like behaviors in mice by promoting the polarization of microglia toward an Arginase-1+ anti-inflammatory phenotype. Luteolin inhibits microglial NLRP3 activation and hippocampal neuroinflammation in mice with depressive-like behaviors. Additionally, luteolin administration reduces the excessive phagocytosis of synapses by overactivated microglia, promoting neuroprotection. These effects are mediated through a PPARγ-dependent mechanism.