Ferulic acid suppresses expression of tryptophan metabolic key enzyme indoleamine 2, 3-dioxygenase via NFκB and p38 MAPK in lipopolysaccharide-stimulated microglial cells

Abstract

Ferulic acid (FA) is a phenol compound found in plants that has anti-inflammatory properties. Indoleamine 2, 3-dioxygenase (IDO) is a tryptophan catabolic enzyme induced in immune cells, including glial cells, during inflammation. Enhanced IDO expression leads to reduced tryptophan levels and increased levels of toxic metabolites, including quinolinic acid. Therefore, inhibition of IDO expression may be effective in suppressing progression of neurodegenerative diseases. In this study, we examined the effect of FA in microglial cells on IDO expression levels and related inflammatory signal molecules. FA suppressed LPS-induced IDO mRNA expression and also suppressed nuclear translocation of NF-κB and phosphorylation of p38 MAPK. However, FA did not affect the production of LPS-induced inflammatory mediators and phosphorylation of JNK. Our results indicate that FA suppresses LPS-induced IDO mRNA expression, which may be mediated by inhibition of the NF-κB and p38 MAPK pathways in microglial cells.

Graphical abstract

Ferulic acid suppresses lipopolysaccharide-induced ｍRNA expression of indoleamine 2,3-deoxygenase via inhibition of NFκB and p38 MAPK pathways in microglial cells.

Key words:

• ferulic acid
• indoleamine 2, 3-dioxygenase, tryptophan
• inflammation
• microglia

Ferulic acid (FA) is a phenol compound present in the cell wall of plants and also found in fibrous foods of plant origin, such as rice and wheat bran.¹⁾ In addition to consuming unrefined foods, supplementation of FA has become possible in recent years. FA has antioxidative^2–4) and anti-inflammatory⁵⁾ properties and is reported to exert suppressive effects against obesity,⁶⁾ hypertension,^7,8) and hyperlipidemia.⁹⁾

Microglial cells are resident immune cells that play a role in monitoring and maintaining neuronal cells in the central nervous system (CNS).^10,11) Conversely, once the inflammatory cytokines, tumor necrosis factor-alpha (TNF-α) and interleukin-6 (IL-6), and nitric oxide (NO) are secreted from microglial cells, they accelerate CNS inflammation.^12,13) Signaling involving nuclear factor-kappa B (NF-κB)^14,15) or mitogen-activated protein kinases, including extracellular signal-regulated kinase (Erk),¹⁶⁾ c-jun N-terminal kinase (JNK),¹⁷⁾ and p38 MAPK,^16,18) plays an important role in inducing microglial inflammation. Microglial cells are reported to accumulate around affected brain regions in Alzheimer’s, Parkinson’s, and Huntington’s diseases, suggesting they may be involved in inflammation of neurodegenerative diseases.^19,20)

Previously, quinolinic acid (QUIN), a tryptophan metabolite, was observed to increase in brains of patients with neurodegenerative diseases.²¹⁾ QUIN is a ligand of N-methyl-D-aspartate (NMDA) receptors, which are involved in the induction of neuroexcitotoxicity.²²⁾ Indoleamine 2, 3-dioxygenase (IDO) is a rate limiting enzyme in the CNS tryptophan-metabolizing pathway, present upstream of this pathway producing QUIN. Moreover, IDO expression is induced in microglial cells during various inflammatory situations,^21,23–26) leading to increased QUIN levels and suggesting that inhibition of IDO expression may be effective to suppress neurodegenerative diseases.

FA can pass through the blood–brain barrier that is critical for controlling permeability of materials from the blood to brain.²⁷⁾ Indeed, transvenous FA administration suppresses neuronal cell apoptosis in rats with cerebral ischemic injury.²⁸⁾ Additionally, in vivo studies indicate that FA, or its ethyl ester body, have suppressive effects against amyloid β-induced neurotoxicity via anti-inflammatory properties.^29,30) However, little is known about the relationship between microglial IDO expression and the anti-inflammatory effect of FA in the CNS.

Therefore, in the present study, we investigated the effect of FA on IDO expression in microglial cells involved in neurodegenerative disease pathogenesis.

Material and methods

Materials

Lipopolysaccharide (LPS) (E. coli O-111) was purchased from Sigma-Aldrich (St Louis, MO, USA), and FA from Tokyo Chemical Industry (Tokyo, Japan). LPS and FA were dissolved in phosphate-buffered saline (PBS) and dimethylsulfoxide, respectively.

Cell culture

The MG6 cell line, derived from mouse microglial cells,^31,32) was obtained from the RIKEN Cell Bank. MG6 cells were cultured in 100 mm dishes in Dulbecco’s modified Eagle’s medium (Nissui Pharmaceutical, Tokyo, Japan) with 10% fetal bovine serum (Biological Industries, Kibbutz Beit Haemek, Israel). Cells were seeded on 12-well plates (0.5 × 10⁶ cells/well) or 60 mm dishes (4.0 × 10⁶ cells/dish) for total RNA or protein extraction, respectively. MG6 cells were treated with LPS (300 ng/mL) and/or FA (10, 50, 250 μM) for 24 h for the detection of TNF-α, IL-6, and NO secretion and mRNA expression levels.

Protein collection was performed after 5 (for p38), 15 (for JNK and Erk), or 30 (for IκBα and NF-κB) min, when activation of these proteins by LPS were clearly observed, according to the results of preliminary experiments (data not shown).

Cell viability assay

Cell viability was determined using CellTiter 96® AQueous One Solution Cell Proliferation Assay (Promega), according to the manufacturer’s protocol. Briefly, cells were seeded on 96-well plates and treated with or without FA (10, 50, 250 μM). After 24 h FA treatment, 20 μL of Cell Titer regent was added to each well, incubated for 15 min at 37 °C, and the absorbance at 450 nm measured.

TNF-α and IL-6 detection

After incubating MG6 cells for 24 h, TNF-α and IL-6 concentrations in the cell medium were measured using ELISA MAX™ (BioLegend, San Diego, CA, USA), according to the manufacturer’s protocol.

Nitric oxide (NO) assay

After incubating MG6 cells for 24 h, NO concentration in the medium was measured by the Griess reaction method.³³⁾ Briefly, 100 μL per well of culture medium was applied to 96-well plates, and 100 μL of an equal volume mixture of Griess reagent (0.2% N-1-naphthyl ethylenediamine (Wako pure chemical Co., Osaka, Japan) and 2% sulfanilamide (Sigma-Aldrich) in 10% H₃PO₄ (Wako)) added to each well. After incubating for 10 min at room temperature, absorbance at 550 nm was measured using an Infinite® F200 Pro (Tecan, Männedorf, Switzerland).

Total RNA extraction and cDNA synthesis

After cell treatments, 500 μL per well of RNAiso plus regent (Takara, Shiga, Japan) was added to cells seeded on 12-well plates. Total RNA was extracted from cells according to the manufacturer’s protocol. Extracted total RNA was reverse-transcribed to cDNA using the PrimeScript® RT Reagent Kit (Takara), according to the manufacturer’s protocol.

Quantitative PCR analysis

Reverse-transcribed cDNA was amplified using SYBR® Premix Ex Taq™ (Takara) and specific primers of the following sequences: IDO 5′-primer (forward), 5′-GGGATGACGATGTTCGAAAG-3′; and 3′-primer (reverse), 5′-GTGGACCCAGACACGTTTTT-3′; 36B4 5′-primer (forward), 5′-TGTGTGTCTGCAGATCGGGTAC-3′; and 3′-primer (reverse), 5′-CTTTGGCGGGATTAGTCGAAG-3′. Reactions and fluorescence detection were performed using ABI PRISM 7000 Sequence Detection System (Life Technologies, Carlsbad, CA, USA). 36B4 mRNA expression levels were used as the internal standard for determination of IDO mRNA expression levels.

Protein extraction

After treatment, cells were washed with PBS and collected in protein extraction buffer (20 mM Tris-HCl, 150 mM NaCl, and 1% Triton) containing protease inhibiter (Calbiochem, La Jolla, CA, USA) and phosphatase inhibitor (Wako) cocktails. Solutes were fractionated using an ultrasonic disintegrator, and lysates centrifuged at 13,000 × g, 4 °C for 5 min. Centrifuged supernatants were collected as total protein solutions. Cell nuclear proteins were extracted by the following method. First, cells were collected in PBS and centrifuged at 1200 × g, 4 °C for 5 min. Collection buffer (25 mM Tris-HCl, 250 mM sucrose, and 1 mM EDTA) containing protease inhibitor cocktail was added to the precipitate. Cell membranes were fractured using a glass-Teflon homogenizer, and homogenates centrifuged at 3000 × g, 4 °C for 5 min. Precipitates were dissolved in RIPA buffer (150 mM NaCl, 50 mM Tris-HCl, 1% Triton, 0.5% sodium deoxycholate, and 0.1% sodium dodecyl sulfate), and the resulting solution was collected as nucleus protein solution. Total protein solution was used to detect IκBα, p38, Erk, and JNK protein expressions, and nucleus protein solution for NF-κB protein expression.

Western blot analysis

Protein solutions were separated by 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride membranes (Millipore, Billerica, MA, USA). Membranes were incubated in 5% skimmed milk or nonfat dried milk solution for 1 h and then washed three times in PBS or TBS containing 0.05% Tween 20 (PBST and TBST). Next, membranes were treated overnight with the following antibodies: IκBα (1:200) (sc-371), Lamin B (1:1000) (sc-6217) (Santa Cruz Biotechnology, Santa Cruz, CA, USA), NF-κB (1:1000) (#4764), p-p38 (1:100) (#9215), p38 (1:1000) (#9212S), p-Erk (1:750) (#9101S), Erk (1:1000) (#4695), p-JNK (1:300) (#9251), JNK (1:1000) (#9252), or β-actin (1:1000) (#4967) (all from Cell Signaling Technology, Beverly, MA, USA). After washing unreacted primary antibodies with PBST or TBST, membranes were treated with secondary antibody, specifically, anti-rabbit IgG horseradish peroxidase (HRP) conjugate (1:1000–5000) (W4018) (Promega, Madison, WI, USA). Membranes were treated with Immobilon Western Chemiluminescent HRP Substrate (Millipore), and fluorescence detected by Fuji LAS-1000 (Fujifilm, Tokyo, Japan). Luminescence intensity was quantified using Image J software (National Institutes of Health, Bethesda, MD, USA).

Statistical analysis

All values were shown as mean ± standard error (SE). Results were analyzed by Tukey’s test after one-way ANOVA or Shirley–Williams test for the detection of significant differences using Excel-Toukei ver. 6.0 (ESUMI, Co., Ltd., Tokyo, Japan). A value of *p < 0.05 was considered statistically significant.

Results

Inflammatory marker production

First, we measured the cytotoxicity of FA. 250 μM FA without LPS treatment suppressed cell viability (supplemental Fig. 1). However, >90% survival rate were observed in cells treated with 10–250 μM of FA in the presence of LPS (Supplemental Fig. 2). FA did not affect cell viability in the presence of LPS.

Production of proinflammatory cytokines, such as TNF-α and IL-6, precede IDO expression during CNS inflammation.^34,35) Therefore, we measured levels of these cytokines as markers of inflammation. In addition, we measured NO, which is also used as a marker of inflammation and considered a putative suppressor of IDO expression.³⁶⁾

TNF-α (Fig. 1(A)) and IL-6 (Fig. 1(B)) productions were increased in LPS-stimulated MG6 cells, and FA did not affect these production levels. NO production was also induced by LPS treatment and was not affected by FA (Fig. 1(C)).

Fig. 1. TNF-α, IL-6, and NO production.

Notes: Production of inflammatory mediators in MG6 cells treated with LPS and/or FA for 24 h. (A), TNF-α; (B), IL-6; and (C), NO. Values indicate mean ± SE.

IDO mRNA expression

Our preliminary experiments indicated that IDO mRNA expression maximally increases after 24 h of LPS treatment (data not shown). Thus, we examined the effect of FA on IDO mRNA expression in microglial cells treated with LPS for 24 h. FA suppressed LPS-induced IDO mRNA expression in a dose-dependent manner (Fig. 2).

Fig. 2. IDO mRNA expression.

Notes: IDO mRNA expression in MG6 cells treated with LPS and/or FA for 24 h. Values indicate mean ± SE. *p < 0.05, **p < 0.01 vs. culture treated with LPS alone (Tukey’s test).

Expression of NF-κB and MAPK pathway proteins

NF-κB and MAPK are crucial pathways for inducing expression of inflammatory mediators.^37,38) We found LPS-induced degradation of NF-κB inhibitor alpha (IκBα), nuclear localization of NF-κB (Fig. 3(A)), phosphorylation of p38 MAPK (Fig. 3(B)) were suppressed by FA in a dose-dependent manner. Erk Phosphorylation was slightly attenuated by 250 μM of FA treatment (Fig. 3(B)). LPS-induced phosphorylation of JNK (Fig. 3(B)) was not affected by FA.

Fig. 3. Expression of proteins in NF-κB and MAPK pathways.

Notes: Protein expression in MG6 cells treated with LPS and/or FA. (A) IκBα in whole cell and NF-κB in nuclear extract. (B) p38, Erk, and JNK phosphorylation in whole cell extract.

Discussion

In response to inflammatory stimuli, microglial cells produce many inflammatory substances including TNF-α, IL-6, and NO.^{12,13,39–41)} In addition to the direct effect of TNF-α, IL-6, NO, and other inflammatory elements, enhanced IDO expression in microglial cells or astrocytes is an important mechanistic factor in CNS inflammation.^40–42) Inflammatory cytokines produced in glial cells in the early period of inflammation promote further inflammation through cross talk between glial and neuronal cells.^36,37) Induced IDO expression in activated microglial cells and astrocytes,⁴³⁾ enhances tryptophan degradation and QUIN production and causes neuronal excitotoxicity.⁴⁴⁾ Increased QUIN induces neuronal death by NMDA glutamate receptor activation, leading to CNS failure.^44,45) Moreover, amyloid β, which accumulates in Alzheimer’s disease lesions, induces IDO expression and QUIN production, suggesting IDO involvement in neurodegenerative diseases.²¹⁾ Thus, regulation of IDO expression through diet may be useful for the prevention of neurodegenerative diseases. Yamamoto et al. reported that some kind of phytochemicals suppress IDO expression.⁴⁶⁾ However, there is no information whether FA can suppress IDO expression. In this study, we investigated the effect of FA on IDO expression in microglial cells.

We treated microglial cells with LPS to mimic the inflammatory state in neurodegenerative patients. LPS is an endotoxin derived from Gram-negative bacteria and used experimentally to induce inflammation.⁴⁷⁾

In the present study, we found LPS-induced IDO mRNA expression was significantly suppressed by FA treatment in microglial cells.

Fujigaki et al. reported that NF-κB and p38 MAPK are involved in gene expression of IDO, and the predicted NF-κB response element is located in the 5’ upstream region of the IDO gene.⁴⁸⁾ JNK is also involved in the regulation of IDO expression.⁴⁹⁾ Therefore, to examine the suppressive mechanism of IDO expression by FA, we investigated expression of NF-κB and MAPK-signaling pathway proteins.

In the unstimulated state, IκBα binds to transcriptional factor NF-κB and inhibits nuclear localization of it. However, phosphorylation of IκBα causes its degradation and allows NF-κB to translocate to the nucleus.⁵⁰⁾ Erk, JNK, and p38 are MAPK family members that are phosphorylated and undergo nuclear translocation in response to inflammatory stimuli.⁵¹⁾

In this study, LPS-induced degradation of IκBα in cytoplasm and nuclear translocation of NF-κB were both suppressed by FA treatment in a dose-dependent manner. Furthermore, LPS-induced phosphorylation of p38 was attenuated by FA treatment. The effect of FA on LPS-induced Erk phosphorylation was not remarkable, and FA did not affect JNK phosphorylation. These results suggest that inhibition of p38 phosphorylation and nuclear translocation of NF-κB would be involved in the inhibitory mechanism of IDO expression by FA.

In this experiment, production of the inflammatory factors TNF-α, IL-6, and NO, were not affected by FA treatment under this experimental condition, which may be caused by maintained phosphorylation level of JNK and Erk.

FA affects IDO expression in microglial cells; hence, it is feasible that FA may have a suppressive effect on CNS inflammation in vivo by the same mechanism. IDO is closely involved in neurodegenerative disease pathogenesis.^52,53) Thus, suppression of IDO expression by oral intake of FA supplements or food containing large amounts of FA including unrefined grains may be helpful for avoiding CNS inflammation.

Author contributions

Y.E. designed the study. M.K. and H.K. conducted the experiments. M.K., S.H. and Y.E. drafted the manuscript. All authors read and approved the final manuscript.

Disclosure statement

No potential conflict of interest was reported by the authors.

Funding

This work was supported by grants-in-aid to Y.E. from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

Supplemental material

Supplemental material for this article can be accessed at http://dx.doi.org/10.1080/09168451.2016.1274636.