2024 Volume 47 Issue 11 Pages 1937-1945
Nurr1 (NR4A2) is a member of nuclear receptor superfamily that regulates gene transcription in midbrain dopaminergic neurons and also inhibits nuclear factor-κB-mediated inflammatory responses in brain microglial cells. To date, various compounds have been reported to stimulate transcriptional activity of Nurr1 on neuronal genes, but their anti-inflammatory actions are poorly characterized. The present study examined the effects of three kinds of Nurr1 ligands, amodiaquine, 1,1-bis(3′-indolyl)-1-(p-chlorophenyl)-methane (C-DIM12) and 5-chloronaphthalen-1-amine (5-CNA), on inflammatory responses of microglial BV-2 cells. Lipopolysaccharide (LPS)-induced upregulation of interleukin-1β mRNA and tumor necrosis factor α mRNA was inhibited by all three compounds, whereas upregulation of interleukin-6 mRNA and inducible nitric oxide synthase (iNOS) mRNA was significantly inhibited only by 5-CNA. On the other hand, LPS-induced nuclear translocation of p65 subunit of nuclear factor-κB was prevented only by amodiaquine. C-DIM12 increased nuclear localization of Nurr1 and transiently upregulated Nurr1 protein expression, whereas amodiaquine and 5-CNA had no effect on these parameters. Notably, inhibitory effect of 5-CNA on iNOS mRNA upregulation was reversed by co-application of amodiaquine. Conversely, inhibitory effect of amodiaquine on p65 nuclear translocation was cancelled by 5-CNA. These results reveal distinct characteristics of anti-inflammatory actions of Nurr1 ligands.
Nurr1 (NR4A2) is a member of nuclear receptor superfamily originally identified as a brain-specific transcription factor.1) It acts as monomer, homodimer, or heterodimer with retinoid X receptors (NR2Bs) to recognize specific promoter sequences of DNA and regulates expression of downstream genes.2) One of the most well-known functions of Nurr1 is its role in generation and maintenance of dopaminergic neurons in the midbrain.3) On the other hand, Nurr1 has also been shown to play an anti-inflammatory role in the central nervous system, via functions distinct from its role as a classical nuclear receptor. That is, Nurr1 binds directly to p65 subunit of nuclear factor-κB (NF-κB) in the nucleus and thereby inhibits NF-κB-mediated expression of pro-inflammatory genes in microglia and astrocytes.4)
Although Nurr1 was initially considered as a ligand-independent nuclear receptor that exhibits constitutive transcriptional activity,5) diverse sets of natural and synthetic compounds are now recognized to act as Nurr1 ligands.6,7) Notable examples are 4-amino-7-chloroquinolines such as amodiaquine and chloroquine, which have been identified as Nurr1 ligands by high-throughput assay based on stimulatory activity on tyrosine hydroxylase gene promotor.8) Many other kinds of Nurr1 ligands have also been identified primarily based on their stimulatory activities on Nurr1-mediated gene transcription as well as on their abilities to bind to Nurr1.9,10) In addition, several studies have attempted to compare the efficacy and potency of diverse Nurr1 ligands in terms of their stimulatory activities on gene transcription.11,12) By contrast, NF-κB-related anti-inflammatory actions of various Nurr1 ligands are left uncharacterized for the most part.
In the present study, we examined anti-inflammatory actions of three kinds of Nurr1 ligands (Fig. 1), namely, amodiaquine, 1,1-bis(3′-indolyl)-1-(p-chlorophenyl)methane (C-DIM12) and 5-chloronaphthalen-1-amine (5-CNA). Amodiaquine has been shown to suppress lipopolysaccharide (LPS)-induced upregulation of mRNAs encoding interleukin (IL)-1β, IL-6, tumor necrosis factor (TNF) α and inducible nitric oxide synthase (iNOS) in microglial BV-2 cells and/or rat primary microglial cells,8) but the detailed mechanisms of these actions are not reported. C-DIM12 is one of methylene-substituted diindolylmethanes that was shown to activate Nurr1 in bladder cancer cells,13) and inhibitory actions of this compound on NF-κB-mediated inflammatory responses in BV-2 cells have been reported by De Miranda et al.14) 5-CNA has been identified as a potent Nurr1 ligand through a systematic screening of compounds possessing the core structure related to amodiaquine and chloroquine,15) and to date, no reports are available for its biological actions other than Nurr1 transcriptional activities. We show here that these three compounds produce anti-inflammatory effects in BV-2 cells, but the underlying mechanisms of their actions may differ from each other.
Amodiaquine dihydrochloride dihydrate was obtained from Cypex Ltd. (Dundee, U.K.). C-DIM12 was synthesized as described previously.16) 5-CNA was purchased from Sigma-Aldrich Japan (Tokyo, Japan), and (E)-3-(4-methylphenyl)sulfonylprop-2-enenitrile (BAY 11-7082) was from Tokyo Chemical Industry (Tokyo, Japan). These compounds were dissolved in dimethyl sulfoxide and added to the culture medium at 1000-fold dilution.
Cell Culture and TreatmentBV-2 cells were maintained in Dulbecco’s modified Eagle medium (DMEM) containing 10% fetal bovine serum (FBS), 100 U/mL penicillin, 100 µg/mL streptomycin, and 2 mM glutamine at 37 °C in a humidified incubator under 5% CO2. For every set of experiments, cells were seeded into 35 mm dish at a density of 4.0 × 105 cells per dish. After 24 h, treatment with drugs (Nurr1 ligands or BAY 11-7082) was initiated with DMEM containing 1% FBS, 100 U/mL penicillin, 100 µg/mL streptomycin, and 2 mM glutamine. One hour after initiation of drug treatment, LPS (from Escherichia coli 0111:B4; Sigma-Aldrich Japan) was added to the medium at 100 ng/mL.
RT-PCRQuantification of mRNAs by RT-PCR was performed essentially according to the methods described previously.17) At 24 h after initiation of LPS treatment, total RNA was extracted with RNAiso Plus (TaKaRa Bio Inc., Shiga, Japan). RT-PCR was performed with Prime Script RT Master Mix (TaKaRa Bio Inc.). The thermal cycling program consisted of 95 °C for 30 s for polymerase activation, and then 40 cycles at 95 °C for 15 s, 53 °C for 45 s and 72 °C for 30 s. Reactions were quantified by the comparative threshold cycle method. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) mRNA was used as the internal control. Reproducibility of the results was confirmed by at least two independent sets of experiments. Primer sequences were as follows: IL-1β forward, 5′-TGAAGGGCTGCTTCCAAACC-3′; IL-1β reverse, 5′-TGTCCATTGAGGTGGAGAG-3′; IL-6 forward, 5′-TCCAGTTGCCTTCTTGGGAC-3′; IL-6 reverse, 5′-GTGTAATTAAGCCTCCGACTTG-3′; TNFα forward, 5′-TTCTGTCTACTGAACTTCGGGGTGATCGGT-3′; TNFα reverse, 5′-GTATGAGATAGCAAATCGGCTGACGGTGTG-3′; iNOS forward, 5′-TGCTTTTGTCCGGAGTCTCAGT-3′; iNOS reverse, 5′-CGGACCATCTCCTGCATTTCT-3′; GAPDH forward, 5′-ACCATCTTCCAGGAGCGAGA-3′; GAPDH reverse, 5′-CAGTCTTCTGGGTGGCAGTG-3′.
ImmunofluorescenceProcedures for immunofluorescence basically followed those of our previous study.18) At 30 min after initiation of LPS treatment, cells were fixed with 4% paraformaldehyde in phosphate-buffered saline (PBS). Cells were then incubated in PBS containing 0.3% Triton X-100 and 3% donkey serum for 1 h, before overnight incubation with primary antibodies. Rabbit anti-NF-κB (1 : 800; cat# 8242, Cell Signaling Technology Japan, K.K., Tokyo, Japan) and rabbit anti-Nurr1/NR4A2 (1 : 500, #10975-2-AP, Proteintech Japan, Tokyo, Japan) were used as primary antibodies, and Alexa Fluor 488 goat anti-rabbit immunoglobulin G (IgG) (H + L) (1 : 500; Invitrogen, Life Technologies Japan, Tokyo, Japan) was used as the secondary antibody. Hoechst 33342 (Nacalai Tesque, Kyoto, Japan) was used for nuclear counterstaining. Fluorescence signals were observed with the use of THUNDER microscope (Leica Microsystems Co., Tokyo, Japan). The intensity of green fluorescence inside and outside the nuclear area was quantified to evaluate the levels of nuclear NF-κB p65 and Nurr1 with the use of LAS AF software (Leica Microsystems Co.).
Western BlottingAfter indicated periods of treatment with Nurr1 ligands, cells were collected in lysis buffer (150 mM NaCl, 50 mM Tris–HCl, 5 mM ethylenediaminetetraacetic acid (EDTA), 1% NP-40, 0.1% sodium orthovanadate, 50 mM NaF, 0.1% sodium dodecyl sulfate (SDS) and 0.5% sodium deoxycholate containing 1% protease inhibitor cocktail). With sample buffer containing 0.5 M Tris–HCl (pH 6.8), 10% SDS, 25% 2-mercaptoethanol, 10% glycerol and 1% bromophenol blue, each sample of lysates and supernatant was heated at 100 °C for 10 min. SDS-polyacrylamide gel electrophoresis was performed on 5.4% stacking gel with 10% separating gel. After electrophoresis, proteins were transferred onto polyvinylidene difluoride membrane. The membrane was washed with Tris-buffered saline/Tween 20 and blocked for 1 h with Blocking One (Nacalai Tesque). The membrane was incubated with rabbit anti-Nurr1/NR4A2 (1 : 500, Proteintech) and mouse anti-β-actin antibody (1 : 1000, #A5441, Sigma-Aldrich Japan) overnight. After incubation with horseradish peroxidase-conjugated secondary antibodies for 1 h, bands were detected with Clarity™ Western ECL Substrate (ThermoFisher Scientific K.K.. Tokyo, Japan) on a luminoimaging analyzer (FUSION SOLO; Vilber-Lourmat, France) and analyzed by EvolutionCapt Edge software (Vilber-Loumat).
Statistical AnalysisAll numerical data sets are presented as mean ± standard error of the mean in addition to individual data points. GraphPad Prism software (Graph Pad, San Diego, CA, U.S.A.) was used for statistical analysis. Data were analyzed by one-way ANOVA followed by Tukey’s multiple comparisons test or analyzed by non-parametrical Kruskal–Wallis test followed by Dunn’s multiple comparisons test. Two-tailed probability values less than 5% were considered significant.
To compare anti-inflammatory effects of Nurr1 ligands, we examined their counteracting effects on LPS-induced upregulation of mRNAs encoding IL-1β, IL-6, TNFα and iNOS. Nurr1 ligands amodiaquine, C-DIM12 and 5-CNA were applied to BV-2 cells, and 1 h later, LPS (100 ng/mL) was added to the culture in the presence of Nurr1 ligands. We applied Nurr1 ligands at 3, 10, and 30 µM, but amodiaquine and C-DIM12 were cytotoxic at 30 µM, and the effect of these two compounds were assessed at 3 and 10 µM only. Under the present experimental conditions, amodiaquine suppressed IL-1β mRNA upregulation at 3 and 10 µM, and suppressed TNFα mRNA upregulation at 10 µM, whereas it had no significant effect on upregulation of IL-6 mRNA and iNOS mRNA (Figs. 2A–D). The anti-inflammatory effect of C-DIM12 was weak relative to that of amodiaquine. That is, C-DIM12 significantly suppressed TNFα mRNA upregulation and tended to suppress IL-1β mRNA upregulation at 10 µM, whereas it had no significant effect on iNOS mRNA upregulation, and rather, augmented IL-6 mRNA upregulation at 10 µM (Figs. 2E–H). On the other hand, 5-CNA was effective in suppressing upregulation of all four pro-inflammatory factors: the suppressing effect of 5-CNA on upregulation of IL-1β mRNA was significant at 30 µM, and the compound significantly suppressed upregulation of IL-6, TNFα and iNOS mRNAs at 3–30 µM (Figs. 2I–L). We also examined the effect of BAY 11-7082 on LPS-induced upregulation of pro-inflammatory genes. BAY 11-7082 is a specific and irreversible inhibitor of IκBα phosphorylation, thereby inhibiting NF-κB-mediated inflammatory gene activation.19) When applied from 1 h before LPS treatment, BAY 11-7082 (3 µM) virtually abolished upregulation of IL-1β and IL-6 mRNAs and significantly attenuated upregulation of iNOS mRNA, whereas it had no effect on TNFα mRNA upregulation (Figs. 2M–P).
Nurr1 ligands at indicated concentrations (in µM) and BAY 11-7082 (BAY; 3 µM) were applied to BV-2 cells, and 1 h later, LPS (100 ng/mL) was added. Total RNA was obtained at 24 h after initiation of LPS treatment, and RT-PCR was run to quantify mRNA levels of IL-1β (A, E, I, M), IL-6 (B, F, J, N), TNFα (C, G, K, O) and iNOS (D, H, L, P). Effects of amodiaquine (AQ; A–D), C-DIM12 (E–H), 5-CNA (I–L) and BAY (M–P) are shown. n = 3 culture dishes for each treatment group. * p < 0.05, ** p < 0.01, *** p < 0.001 vs. control; # p < 0.05, ## p < 0.01, ### p < 0.001 vs. LPS alone.
Nuclear translocation of NF-κB in response to toll-like receptor-mediated signaling cascade is considered as one of the critical steps leading to upregulation of pro-inflammatory gene expression. Therefore, we addressed the effect of Nurr1 ligands on subcellular localization of NF-κB p65 subunit (Fig. 3). Treatment of BV-2 cells with 100 ng/mL LPS for 30 min prominently increased the level of p65 within the nucleus. As expected, 60-min pretreatment with BAY 11-7082 (3 µM) abolished LPS-induced nuclear translocation of p65. Amodiaquine (10 µM) also prevented LPS-induced nuclear localization of p65, whereas C-DIM12 (10 µM) and 5-CNA (30 µM) showed no significant effect. These results suggest a difference among Nurr1 ligands in terms of regulation of NF-κB trafficking.
Amodiaquine (AQ; 10 µM), C-DIM12 (10 µM), 5-CNA (30 µM) or BAY 11-7082 (BAY; 3 µM) was applied to BV-2 cells, and 60 min later, LPS (100 ng/mL) was added. After 30-min treatment with LPS, cells were fixed and immunostained with anti-NF-κB p65 antibody and fluorescent secondary antibody. (A) Representative photographs showing immunofluorescence of NF-κB p65 subunit (green) with Hoechst 33342 nuclear counterstaining (blue). Scale bars = 20 µm. (B) Quantitative results on the intensities of nuclear p65-immunopositive signals. Values are presented as the ratio of nuclear p65 over total p65 in individual cells. n = 24 cells in three culture dishes for each treatment group. *** p < 0.001 vs. control; ### p < 0.001 vs. LPS alone.
We also examined the effect of three ligands on subcellular localization of Nurr1, as Nurr1 is considered to act in the nucleus both as a classical nuclear receptor2) and as an NF-κB regulator.4) In the absence of LPS treatment, 90-min treatment of BV-2 cells with C-DIM12 (10 µM) significantly increased the level of Nurr1 localized within the nucleus, whereas amodiaquine (10 µM) and 5-CNA (30 µM) showed no significant effect. Similar results were obtained when LPS (100 ng/mL) was added to the culture for 30 min after 60-min pretreatment with Nurr1 ligands, where C-DIM12, but not amodiaquine and 5-CNA, significantly increased nuclear localization of Nurr1 (Figs. 4A, B). Under the present experimental conditions, LPS (100 ng/mL) alone showed modest and insignificant effect on subcellular localization of Nurr1, although a previous study employing a higher concentration (1 µg/mL) of LPS reported an increase in nuclear Nurr1 by LPS in BV-2 cells.20)
Amodiaquine (AQ; 10 µM), C-DIM12 (10 µM) or 5-CNA (30 µM) was applied to BV-2 cells for 90 min, then cells were fixed and immunostained with anti-Nurr1 antibody and fluorescent secondary antibody. In several groups, LPS (100 ng/mL) was added at 60 min after initiation of treatment with Nurr1 ligands. (A) Representative photographs showing immunofluorescence of Nurr1 (green) with Hoechst 33342 nuclear counterstaining (blue). Scale bars = 20 µm. (B) Quantitative results on the intensities of nuclear Nurr1-immunopositive signals. Values are presented as the ratio of nuclear Nurr1 over total Nurr1 in individual cells. n = 24–26 cells in three culture dishes for each treatment group. *** p < 0.001 vs. control; ## p < 0.01 vs. LPS alone.
We next examined the expression levels of total Nurr1 protein after treatment with Nurr1 ligands. After the change of medium from 10% FBS-DMEM to 1% FBS-DMEM, control cultures maintained a stable Nurr1 protein level for up to 24 h (Fig. 5A, Supplementary Fig. 1). Application of C-DIM12 (10 µM) significantly increased the level of total Nurr1 protein at 1.5 h, which gradually returned to the baseline level by 24 h (Fig. 5C). In contrast, amodiaquine (10 µM) and 5-CNA (30 µM) did not increase Nurr1 protein (Figs. 5B, D).
Amodiaquine (10 µM), C-DIM12 (10 µM) or 5-CNA (30 µM) was applied to BV-2 cells for indicated periods, and then Western blotting was performed to quantify protein levels of Nurr1 and β-actin. Shown are representative blots and quantitative results of Nurr1 protein levels after indicated periods of treatment with vehicle (A), 10 µM amodiaquine (B), 10 µM C-DIM12 (C) or 30 µM 5-CNA (D). n = 3 culture dishes for each treatment group. ** p < 0.01 vs. 0 h.
Amodiaquine and 5-CNA share structural characteristics as the latter was identified by a structure–activity relationship study on amodiaquine analogs.15) Therefore, these two compounds may share the binding pocket of Nurr1 and compete each other concerning their binding to the receptor. To address this hypothesis, we examined potential counteracting effect of amodiaquine against 5-CNA, and vice versa. As shown above (Fig. 2), upregulation of IL-1β and TNFα mRNAs was suppressed by both amodiaquine and 5-CNA, whereas upregulation of IL-6 and iNOS mRNAs was suppressed only by 5-CNA. When we applied amodiaquine (10 µM) and 5-CNA (30 µM) simultaneously, the suppressive effect of 5-CNA on iNOS mRNA upregulation was substantially reduced by amodiaquine (Fig. 6D), although the effect of 5-CNA on IL-6 mRNA upregulation was unaffected by amodiaquine (Fig. 6B). Additive effect of amodiaquine and 5-CNA was not evident for the expression levels of IL-1β and TNFα mRNAs (Figs. 6A, C).
(A–D) 5-CNA (30 µM) with or without amodiaquine (AQ; 10 µM) was applied to BV-2 cells, and 1 h later, LPS (100 ng/mL) was added. Expression levels of mRNAs encoding IL-1β (A), IL-6 (B), TNFα (C) and iNOS (D) were quantified at 24 h after initiation of LPS treatment. n = 3 culture dishes for each treatment group. ** p < 0.01, *** p < 0001 vs. control; # p < 0.05, ### p < 0001 vs. LPS alone; $$ p < 0.01. (E) Representative photographs showing immunofluorescence of NF-κB p65 subunit (green) with Hoechst 33342 nuclear counterstaining (blue). AQ (10 µM) with or without 5-CNA (30 µM) was applied to BV-2 cells, and 60 min later, LPS (100 ng/mL) was added. After 30-min treatment with LPS, cells were fixed and immunostained with anti-NF-κB p65 antibody and fluorescent secondary antibody. Scale bars = 20 µm. (F) Quantitative results on the intensities of nuclear p65-immunopositive signals. The effect of BAY 11-7082 (BAY; 3 µM) was also examined as a reference. n = 27 cells in three culture dishes for each treatment group. *** p < 0001 vs. control; ## p < 0.01, ### p < 0001 vs. LPS alone; $$$ p < 0.001.
On the other hand, when applied alone, amodiaquine but not 5-CNA prevented LPS-induced nuclear translocation of NF-κB p65 subunit (Fig. 3). Interestingly, the inhibitory effect of 10 µM amodiaquine on nuclear translocation of NF-κB was cancelled by simultaneous treatment with 30 µM 5-CNA (Figs. 6E, F).
Nurr1 is considered as a potential molecular target for therapies of several neurological disorders such as Parkinson disease and Alzheimer disease,2,7) as this nuclear receptor not only regulates the functions of dopaminergic neurons via its transcriptional activity but also inhibits inflammatory responses of glial cells possibly via suppression of NF-κB signaling. Efforts are being made by several studies to explore and develop novel Nurr1 ligands.7,21,22) However, the screening of compounds in those studies is based on the binding affinity to Nurr1 and/or stimulatory actions on Nurr1-mediated gene transcription. For example, 5-CNA was found as a Nurr1 agonist with higher affinity and potency than chloroquine in a Gal4-Nurr1 hybrid reporter gene assay.15) In the same study, this compound was also shown to stimulate tyrosine hydroxylase mRNA expression more potently than chloroquine in T98G cells,15) but its anti-inflammatory actions have not been examined. Accordingly, the present study aimed to reveal the characteristics of Nurr1 ligands as anti-inflammatory agents.
In the first series of experiments, 5-CNA was found to be a less cytotoxic and more effective anti-inflammatory compound than the other two Nurr1 ligands. In addition, we found that the anti-inflammatory profiles of amodiaquine, C-DIM12 and 5-CNA were different from each other and that these profiles also show differences from those of BAY 11-7082, an inhibitor of NF-κB-mediated signaling.19) Namely, all four compounds including BAY 11-7082 suppressed LPS-induced upregulation of IL-1β mRNA, but upregulation of IL-6 and iNOS mRNAs was significantly suppressed only by 5-CNA and BAY 11-7082. On the other hand, upregulation of TNFα mRNA was suppressed by three Nurr1 ligands amodiaquine, C-DIM12 and 5-CNA, but not by BAY 11-7082. The latter results suggest that, at least under the present experimental conditions, LPS induces TNFα mRNA upregulation via NF-κB-independent manner and therefore that the suppressive effects of Nurr1 ligands on TNFα mRNA expression are independent from NF-κB regulation. In this context, a study using primary microglial cells has shown that Nurr1 might directly bind to TNFα gene promoter to repress downstream gene transcription.23) To the best of our knowledge, no evidence is currently available for direct binding of Nurr1 to the promoters of IL-1β, IL-6 and iNOS genes.
Negative regulation of NF-κB by Nurr1 has been first demonstrated by Saijo et al.4) Their study examined the function of Nurr1 without ligands and showed that Nurr1 binds to NF-κB p65 in the nucleus and recruits CoREST repressor complex to transrepress NF-κB-mediated gene transcription (Fig. 7A). According to this scheme, ligand-bound Nurr1 might exhibit augmented binding affinities to NF-κB p65 and/or CoREST repressor complex in the nucleus, leading to potent suppression of inflammatory gene upregulation. However, the findings in the present study indicate that Nurr1 ligands such as amodiaquine and C-DIM12 may provide anti-inflammatory actions via mechanisms other than augmented functions of Nurr1 in the nucleus. Firstly, amodiaquine inhibited nuclear translocation of NF-κB without affecting the nuclear level of Nurr1, suggesting that amodiaquine-bound Nurr1 preferentially interacts with cytosolic NF-κB, thereby prevents translocation of NF-κB from the cytosol into the nucleus (Fig. 7B). It should also be noted that amodiaquine suppressed upregulation of IL-1β mRNA but not of IL-6 and iNOS mRNAs, although upregulation of all three mRNAs was sensitive to BAY 11-7082 and therefore should be dependent on NF-κB. In this context, a previous study using milder conditions than those of the present study reported that upregulation of IL-1β, IL-6 and iNOS mRNAs induced by 4-h treatment of BV-2 cells with 10 ng/mL LPS was suppressed by 10–15 µM amodiaquine.8) Therefore, the anti-inflammatory effect of amodiaquine afforded by cytosolic Nurr1 may be too weak and/or transient to sufficiently inhibit pro-inflammatory gene upregulation in the present study.
(A) Nurr1 in its free (ligand-unbound) form can recruit CoREST repressor complex in the nucleus of microglial cells to transrepress NF-κB-mediated pro-inflammatory gene expression.4) (B) Amodiaquine-bound Nurr1 may interact with NF-κB in the cytosol to prevent nuclear translocation of NF-κB, thereby diminishing pro-inflammatory gene expression. (C) C-DIM12 may not bind to Nurr1 directly12) but may increase the expression of Nurr1 protein via unidentified mechanisms. Consequently, increased Nurr1 in the nucleus leads to enhancement of CoREST-mediated transrepression. (D) 5-CNA neither prevents NF-κB translocation nor increases Nurr1 protein, but potently suppresses upregulation of NF-κB-dependent pro-inflammatory genes. A plausible explanation of this action is that 5-CNA binding to Nurr1 may augment transrepression activity of Nurr1, for example, by increasing the affinity of Nurr1 to NF-κB and/or CoREST repressor complex.
Secondly, C-DIM12 did not inhibit nuclear translocation of NF-κB, whereas it transiently increased total level of Nurr1 protein and increased nuclear level of Nurr1. Interestingly, a previous study demonstrating the binding of amodiaquine to Nurr1 ligand-binding domain demonstrated that C-DIM12 did not directly bind to Nurr1,12) suggesting potential indirect actions of this compound on Nurr1-mediated biological effects. Accordingly, C-DIM12 may increase the expression level of Nurr1 protein via unidentified mechanisms, thereby increases ligand-unbound Nurr1 in the nucleus, contributing to the enhanced transrepression of anti-inflammatory genes (Fig. 7C). It should be noted again that the anti-inflammatory effect of C-DIM12 was evident for IL-1β but not IL-6 and iNOS, which might be dependent on the experimental conditions employed in the present study. Indeed, a previous study showed that upregulation of IL-1β, IL-6 and iNOS mRNAs induced by 24-h treatment of BV-2 cells with 1 µg/mL LPS was prevented by 10 µM C-DIM12.14) The same study also demonstrated enhanced nuclear localization of Nurr1 in C-DIM12-treated primary microglial cells,14) which is consistent with our observations in BV-2 cells.
In contrast to amodiaquine and C-DIM12, 5-CNA had no effect on nuclear localization of either NF-κB p65 or Nurr1, yet this compound effectively suppressed LPS-induced upregulation of IL-1β, IL-6 and iNOS mRNAs. Interestingly, amodiaquine and 5-CNA acted as an antagonist of each other with respect to their effects on iNOS mRNA expression and NF-κB nuclear translocation. Based on the mechanisms of the action of ligand-unbounded Nurr1 as proposed by Saijo et al.,4) the results obtained in the present study suggest that 5-CNA is a potent agonist of Nurr1 as a negative NF-κB regulator. That is, 5-CNA-bound Nurr1 in the nucleus may show enhanced transrepression activity on NF-κB-regulated genes as compared to ligand-unbound Nurr1 (Fig. 7D). Meanwhile, 5-CNA-bound Nurr1 in the cytosol may not interact with cytosolic NF-κB, unlike in the case of amodiaquine-bound Nurr1. The reason why 5-CNA was more dominant than amodiaquine in the cytosol (that means, 5-CNA completely prevented the effect of amodiaquine on NF-κB translocation) while amodiaquine can compete with 5-CNA in the nucleus (that means, amodiaquine diminished the suppressive effect of 5-CNA on iNOS mRNA upregulation) is unclear. A fact deserving consideration is post-translational modification of Nurr1 in the nucleus. Namely, Saijo et al. reported that covalent modification with small ubiquitin-like modifier (SUMO) peptides, so-called SUMOylation, was required for Nurr1-mediated transrepression.4) They also showed that phosphorylation of Nurr1 by Nemo-like kinase contributed to Nurr1-CoREST interaction. Therefore, nuclear Nurr1 involved in transrepression of NF-κB-regulated genes may have received these modifications, and consequently, nuclear Nurr1 may prefer a different conformation from that of unmodified cytosolic Nurr1. Accordingly, different conformations of Nurr1 may result in different binding affinities of respective ligands to Nurr1 in respective subcellular compartments. In any case, the present study did not directly address the effect of 5-CNA and other ligands on molecular interactions of Nurr1 with its binding partners in the cytosol and the nucleus, and elucidation of the precise mechanisms of actions of these compounds requires further investigations. In addition, we cannot exclude the possibility that off-target actions unrelated to Nurr1 are involved in the anti-inflammatory actions of these compounds.
In the present study, we demonstrated for the first time that a Nurr1 ligand 5-CNA provided anti-inflammatory effect in microglial cells, via different mechanisms from those of other well-known Nurr1 ligands such as amodiaquine and C-DIM12. It should be noted that, although BV-2 cells have been widely utilized for in vitro examinations of inflammatory responses in the brain,24) the validity of these cells as a model of microglia was sometimes called into question because they exhibit several different profiles of cytokine expression from those of primary microglia.25) Therefore, validation of the anti-inflammatory effect of 5-CNA under in vivo conditions may be required in future investigations. Indeed, Nurr1 ligands are considered as promising drug candidates for various neurological disorders.26) We previously reported that both amodiaquine and C-DIM12 were effective in alleviating neurological symptoms and neuropathological injuries in a mouse model of intracerebral hemorrhage.16,27) Potential effect of 5-CNA as a therapeutic agent for neurological disorders is an interesting issue to be explored, and our investigations on the effect of 5-CNA in vivo are underway.
This work was supported by The Smoking Research Foundation; JSPS KAKENHI, MEXT, Japan Grants 20H04126, 22K19756, 23H03332, and 23K28022.
The authors declare no conflict of interest.
This article contains supplementary materials.
