Naringenin promotes recovery from colonic damage through suppression of epithelial tumor necrosis factor–α production and induction of M2-type macrophages in colitic mice

N U TR IT ION RE S EAR CH 6 4 ( 2 01 9 ) 8 2 –9 2

Available online at www.sciencedirect.com

ScienceDirect www.nrjournal.com

Naringenin promotes recovery from colonic damage through suppression of epithelial tumor necrosis factor–α production and induction of M2-type macrophages in colitic mice Yuta Chaen, Yoshinari Yamamoto, Takuya Suzuki⁎ Department of Biofunctional Science and Technology, Graduate School of Biosphere Science, Hiroshima University, 1-4-4 Kagamiyama, Higashi-Hiroshima 739-8528, Japan

ARTI CLE I NFO

A BS TRACT

Article history:

Our previous study demonstrated that supplemental naringenin reduced the development

Received 20 August 2018

of colitis induced by dextran sodium sulfate (DSS) in mice, however, the effect of naringenin

Revised 7 January 2019

on the recovery from colonic damage was totally unknown. The primary purpose was to

Accepted 15 January 2019

investigate if naringenin promoted recovery from colonic damage in DSS-administered mice and colonic tissues. When mice were fed diets lacking or containing naringenin (0.3%,

Keywords:

w/w) for 11 days after colitis induction through DSS administration, the supplemental

Naringenin

naringenin was found to promote a reversal of body weight loss and suppress tumor

Colitis

necrosis factor (TNF)-α mRNA expression in the DSS-administered mice. Moreover, protein

Tumor necrosis factor–α

expression of two tight junction proteins, claudin-3 and junctional adhesion molecule-A,

Macrophages

was higher in DSS-administered mice that were fed naringenin than in the mice that did

Interleukin-10

not receive naringenin. To examine the early mechanisms underlying the naringeninmediated reduction of colonic damage, the inflamed colonic tissues of DSS-administered mice were incubated with or without naringenin for 24 hours; in tissues incubated with naringenin, TNF-α production was lower and interleukin (IL)-10 and CD206 mRNA expression was higher than in tissues incubated without naringenin, but naringenin did not affect the expression of the tight junction proteins. Flow cytometry results further demonstrated that naringenin reduced TNF-α–positive epithelial cells, but not macrophages, and promoted the polarization of M2-type macrophages in the colonic tissues. Thus, supplemental naringenin promoted recovery from colonic damage in mice with colitis, and suppression of epithelial TNF-α production and induction of M2-type macrophages might represent the early mechanisms underlying this naringenin effect. © 2019 Elsevier Inc. All rights reserved.

Abbreviations: CXCL-2, chemokine C-X-C motif ligand-2; DSS, dextran sodium sulfate; HBSS, Hanks' balanced salt solution; IBD, inflammatory bowel disease; IL, interleukin; JAM, junctional adhesion molecule; LPL, lamina propria lymphocyte; PI3K, phosphatidylinositol 3-kinase; TJ, tight junction; TNF, tumor necrosis factor; ZO, zonula occludens. ⁎ Corresponding author at: 1-4-4, Kagamiyama, Higashi-Hiroshima 739-8528, Japan. Tel.: +81 82 424 7984; fax: +81 82 424 7916. E-mail address: [email protected] (T. Suzuki). https://doi.org/10.1016/j.nutres.2019.01.004 0271-5317/© 2019 Elsevier Inc. All rights reserved.

N U TR IT ION RE S EA RCH 6 4 ( 2 01 9 ) 8 2 –9 2

1.

Introduction

Inflammatory bowel disease (IBD) is a chronic inflammatory condition that includes ulcerative colitis and Crohn's disease. The IBD clinical symptoms are mainly diarrhea, rectal bleeding, fever, and abdominal pain and cramping, and the patients experience periods of relapse and remission [1]. A notable fraction of patients respond positively to pharmacotherapy, but the medications can occasionally produce severe adverse side effects, such as drug hypersensitivity, nephrotoxicity, hepatitis, fever, and pancreatitis [2]. Moreover, although biological drugs such as antibodies against tumor necrosis factor (TNF)-α have been successfully developed and commonly used in IBD therapy, long-term and extensive usage of these drugs could cause considerable economic burden to patients. Accordingly, novel therapeutic approaches involving dietary and nutritional intervention have been investigated in IBD management. IBD etiology remains poorly elucidated, but the inflamed tissues of IBD patients are known to be characterized by an impaired epithelial barrier and an uncontrolled and robust production of proinflammatory cytokines [1]. The impaired tight junction (TJ) barrier allows luminal bacterial products to permeate into the intestinal mucosa and induce inflammatory reactions in immune and nonimmune cells [3], and the immunologically activated cells produce various proinflammatory cytokines, such as TNF-α, interleukin (IL)-1β, and IL-6, and develop chronic inflammation. Among the cytokines, TNF-α, which is produced by activated macrophages, dendritic cells, and epithelial cells, plays a central role in IBD pathogenesis [4,5]. TNF-α has been shown to induce apoptosis and barrier disruption of epithelial cells, necrosis of Paneth cells, and inflammatory cytokine production by macrophages in the inflamed tissues in IBDs [4]. Conversely, an anti-inflammatory cytokine, IL-10, which is highly produced by regulatory T cells and type-2 macrophages, plays an essential role in the resolution of inflammation [6]. IL-10 has been shown to inhibit the activated immune cells and repair the epithelial barrier [4,7]. Recently, natural products and functional dietary components have been attracting considerable research attention in IBD management. Supplementation with probiotics, dietary fibers, and polyphenols has been shown to help manage the active disease or maintain the remission in patients with IBDs and in rodent models of IBDs [8-13]. Escherichia coli Nissle 1917, Lactobacillus rhamnosus GG and multispecies mixture VSL#3 reportedly prolong the remission or prevent the relapse in patients with IBDs [8-10]. The VSL#3 also reduces the endoscopic and histological scores. Psyllium fiber (Plantago ovata, ispaghula husk) is found to be superior to placebo in relieving intestinal symptoms in patients with ulcerative colitis [11]. Germinated barley foodstuff, which mainly consists of dietary fiber and glutamine-rich protein, has been shown to reduce clinical activity and prolong the remission in ulcerative colitis [12]. Compared to probiotics and dietary fibers, there is less clinical evidence regarding the use of polyphenols for IBD therapy, but curcumin, an active ingredient found in turmeric, significantly reduces clinical and endoscopic activity indices in patients with ulcerative colitis, leading to the reduction of relapse rates [13]. However, the mechanisms underlying these beneficial effects have not been clearly delineated in most cases.

83

We previously reported how colitis was affected by supplementation with naringenin (4′,5,7-trihydroxyflavanone), a major polyphenol isolated from citrus fruits: naringenin supplementation before colitis induction by dextran sodium sulfate (DSS) administration reduced the disease development in mice, which indicated the prophylactic effect of naringenin [14]. Our results suggested that one potential mechanism underlying the naringenin-mediated reduction of colitis is the promotion of intestinal barrier integrity through the regulation of epithelial TJ structure [14,15], although the effect of naringenin on recovery from colonic inflammation or damage remains to be investigated. Dietary components that can promote colonic repair as well as prevent colonic inflammation could serve as effective agents in IBD management. In this study, mice were fed naringenin-supplemented diets after colitis induction by DSS administration, and then the effect of naringenin on the recovery from and resolution of colitis was investigated. Moreover, the early mechanisms involved in the naringenin-mediated effect were investigated by examining organ cultures prepared using the inflamed colon tissues from the DSS-administered mice. We hypothesized that naringenin promoted the recovery from colonic damages through restoration of TJ barrier and/or alteration of immune reaction.

2.

Methods and materials

2.1.

Reagents

DSS (molecular weight: 36000–50 000) was purchased from MP Biomedicals, Inc (Santa Ana, CA, USA). The following antibodies were from commercial sources: rabbit anti-zonula occludens (ZO)-1, ZO-2, occludin, claudin-3, claudin-4, claudin-7, and junctional adhesion molecule (JAM)-A, mouse FITC-conjugated anti-TNF-α, and goat Alexa Fluor 488-conjugated anti-rabbit IgG, Thermo Fisher Scientific (Waltham, MA, USA); rabbit anti-pAkt, pERK1/2, and pNF-κB p65, Cell Signaling Technology (Danvers, MA, USA); horseradish peroxidase-conjugated anti-rabbit IgG, SeraCare (Milford, MA, USA); mouse APC-conjugated anti-pancytokeratin, Sigma (St Louis, MO, USA); and mouse APCconjugated anti-F4/80, FITC-conjugated anti-CD11c, and FITCconjugated anti-CD206, BioLegend (San Diego, CA, USA). Cell culture reagents and supplies were purchased from Thermo Fisher Scientific. All other chemicals were obtained from Wako Pure Chemical Industries (Osaka, Japan).

2.2.

Animals

All study protocols were preapproved by the Animal Use Committee of Hiroshima University, and all mice were maintained in accordance with the Hiroshima University guidelines for the care and use of laboratory animals (authorization no. C15–10-3). Male Balb/c mice aged 7 weeks old were obtained from Charles River Japan (Yokohama, Japan). Throughout the study, the mice were housed under the following conditions: controlled temperature of 22 ± 2°C, relative humidity of 40%–60%, and light exposure from 0800 to 2000 hours. The mice were allowed to acclimatize to the laboratory environment, with free access to an AIN-93G-

84

N U TR IT ION RE S EAR CH 6 4 ( 2 01 9 ) 8 2 –9 2

formula control diet and distilled water, for 1 week prior to the start of the experiments.

2.3.

Test of recovery from DSS-induced colitis

To assess the effect of naringenin on recovery from colitis, DSS-administered mice were fed diets without or with naringenin (Table 1): Mice were administered 2% (w/v) DSS through drinking water for 6 days and divided into two groups, DSS and DSS + naringenin, and the induction period was followed by a 11-day recovery period during which distilled water was provided to the mice. The DSS group was fed the control diet throughout the experiment, whereas the DSS + naringenin group was fed the control diet during the colitis-induction period (6 days) and then a diet containing 0.3% naringenin (w/w) during the recovery period (11 days). Our previous study showed that the 0.3% naringenincontaining diet reduced the development of colitis induced by DSS in mice [14]. Control mice received the control diet plus distilled water throughout the experiment (17 days). Body weights of the mice were measured daily. To assess the severity of colitis, stool scores were determined based on a standard scoring system as described previously [16]. In summary, the stool score was defined as the mean values of the diarrheal stool score (0, normal stool; 1, mildly soft stool; 2, very soft stool; 3, watery stool; and 4, more watery stool) and the bloody stool score (0, normal color stool; 1, brown color stool; 2, reddish color stool; 3, bloody stool; and 4, more bloody stool). At the end of the experiment, the mice were euthanized by exsanguination under isoflurane anesthesia, and then the colon was dissected and used in immunoblotting and qRT-PCR analyses, as described in 2.5 and 2.6.

2.4.

Organ culture of DSS-induced colitic tissues

To examine the early mechanisms involved in naringeninmediated reduction of colitis, colonic tissues from DSSadministered mice were cultured. As described in 2.3, colitis

was induced in mice through DSS administration for 6 days, and then the colon was removed, cut into ~5-mm-long segments, washed with PBS, and cultured in Dulbecco's modified Eagle medium supplemented with 10% fetal bovine serum in the absence and presence of naringenin (100 and 200 μmol/L) for 24 hours. Colonic segments were incubated with inhibitors of phosphatidylinositol 3-kinase (PI3K; LY294002) and MEK (U0126) for 1 hour before naringenin administration. The cultured colonic segments were used in immunoblotting, qRT-PCR, and flow cytometry assays as described in 2.5, 2.6, and 2.9. TNF-α production in culture supernatants was determined by performing ELISA (Thermo Fisher Scientific).

2.5.

Mouse colon tissue (25 mg) was homogenized in 0.5 mL of lysis buffer containing 1% (w/v) SDS, 1% (v/v) Triton X-100, and 1% (w/v) sodium deoxycholate in 30 mmol/L Tris with protease and phosphatase inhibitors (pH 7.4) by using a Polytron®-type homogenizer (KINEMATICA AG, Lucerne, Switzerland). Immunoblotting for ZO-1, ZO-2, JAM-A, occludin, claudin-3, claudin-7, pAkt, pERK1/2, and pNF-κB p65 was performed as previously described [17-20].

2.6.

Ingredient

Control diet e g/kg diet

a

Casein α-Corn starch b Sucrose Soy bean oil Choline bitartrate L-cystine Mineral mixture c Vitamin mixture c Cellulose d a

200 529.5 100 70 2.5 3 35 10 50

Casein (ALACID; New Zealand Daily Board). α-Corn starch (Amylalpha CL; Chuo-Shokuryou). c Mineral and vitamin mixtures were prepared according to the AIN-93G formulation. d Powdered cellulose (Just fiber; International Fiber Corporation). e Naringenin-containing diet was prepared by adding naringenin (0.3%, w/w) to control diet. b

qRT-PCR analysis

Total RNA from colonic tissues was isolated using NucleoSpin RNA II (Macherey-Nagel GmbH, Düren, Germany) and reversetranscribed using a ReverTra Ace qPCR RT kit (Toyobo, Osaka, Japan), according to the manufacturer's instructions. The mRNA expression level of TNF-α, IL-1β, IL-6, chemokine C-X-C motif ligand-2 (CXCL-2), IL-17A, interferon (IFN)-γ, F4/80, CD206, occludin, claudin-3, claudin-4, and JAM-A was determined using qRT-PCR, as previously described [14,17-19,21]. The primer sequences used for PCR are shown in Table 2. Data were analyzed using the comparative threshold cycle (ΔΔCT) method and were normalized to the expression level of glyceraldehyde 3-phosphate dehydrogenase, the internal control.

2.7. Table 1 – Composition of test diets

Immunoblotting analysis

Immunofluorescence analysis

Mouse colon tissue was embedded in optimal cutting temperature compound (Sakura Finetek Japan, Tokyo, Japan) and frozen sections (8 μm) were prepared on glass slides. The cryosections were fixed in 4% paraformaldehyde, permeabilized with 0.2% Triton X-100, blocked in 4% skim milk, and incubated for 16 hours at 4°C with APC-conjugated anti-F4/80 plus either anti-pAkt or anti-pERK, and then incubated for 1 hour with the secondary antibody, goat Alexa Fluor 488-conjugated anti-rabbit IgG. The immunofluorescence signal was visualized using an LCM700 confocal laser scanning microscope (Carl Zeiss, Oberkochen, Germany).

2.8. Isolation of colonic epithelial cells and lamina propria lymphocytes (LPLs) Epithelial cells and LPLs of the colon were isolated for flow cytometry analysis, as described previously [22,23]. Briefly, mouse colonic tissue was agitated in Hanks' balanced salt solution (HBSS) containing 10 mmol/L dithiothreitol for

85

N U TR IT ION RE S EA RCH 6 4 ( 2 01 9 ) 8 2 –9 2

Table 2 – Primer sequences for qRT-PCR Target gene

Forward

Reverse

References

Mouse Mouse Mouse Mouse Mouse Mouse Mouse Mouse Mouse Mouse Mouse Mouse Mouse Mouse Mouse

5′-TCGTAGCAAACCACCAAGTG-3′ 5′-TATCTGGAGGAACTGGCAAA-3′ 5′-TTCCTCCTTGCCTCTGATGG-3′ 5′-CTGATGCTGGTGACAACCAC-3′ 5′-AGTGAACTGCGCTGTCAATG-3′ 5′-CCTTTCAATTGCTCTCATCC-3′ 5′-TGGATTCAGAGGCAGATTCA-3′ 5′-ATACCCTCCAGCACATCCAG-3′ 5′-ATTGAGGGTGGGTGTCAGGA-3′ 5′-TTTGAGTGGAGTGATGGAACC-3′ 5′-CACACTTGCTTGGGACAGAG-3′ 5′-AACTGCGTACAAGACGAGACG-3′ 5′-CTGGAGTGGATGTCCTCATGT-3′ 5′-TAACTACGCTGCGCTTCAGA-3′ 5′-TCAAGAAGGTGGTGAAGCAG-3′

5′-CTTTGAGATCCATGCCGTTG-3′ 5′-TGACGCTTATGTTGTTGCTG-3′ 5′-ATGTGCTGGTGCTTCATTCA-3′ 5′-TCCACGATTTCCCAGAGAAC-3′ 5′-ACTTTTTGACCGCCCTTGAG-3′ 5′-ATCTCCCTGGTTTCTCTTCC-3′ 5′-CAGTTTGGGACCCCTTTACA-3′ 5′-AGTTTGCCATCCGGTTACAG-3′ 5′-GTTGGGCATTGGGTTCTTGT-3′ 5′-ACAGCATGGCTTTGTGATACC-3′ 5′-TAGCCATAGCCTCCATAGCC-3′ 5′-ATCCCTGATGATGGTGTTGG-3′ 5′-GAGTAGCGCTGGAGTAACGTG-3′ 5′-AACCCCTTTTCCAACCAATC-3′ 5′-AAGGTGGAAGAGTGGGAGTTG-3′

19 14 19 14 14 19 14 This 21 This 20 20 This This 14

Tnfa Ifng Il1b Il6 Cxcl2 IL-10 Il17a F4/80 Foxp3 CD206 Occludin Claudin-3 Claudin-4 JAM-A Gapdh

10 minutes and then in HBSS containing 30 mmol/L EDTA for 10 minutes. The tissue samples were vigorously shaken for 3 minutes, which released epithelial cells, and the supernatants were centrifuged at 5000 × g for 2 minutes to sediment the epithelial cells. The harvested epithelial cells were treated with 2 mg/mL dispase (Thermo Fischer Scientific) and 50 mg/mL DNase I (Roche Applied Science, Upper Bavaria, Germany) at 37°C for 1 hour with vortexing every 5 minutes. The resultant single epithelial cells were purified through centrifugation performed using a discontinuous 20% and 40% Percoll gradient (GE Healthcare, Chicago, IL, USA). The colonic tissues that remained after the epithelial cells were released were collected using 100-μm-pore-size nylon membranes, minced, and incubated with Advanced-RPMI 1640 medium containing 2.1 mg/mL collagenase D at 37°C for 20 minutes. The undigested tissue debris was removed using 30-μm-pore-size nylon membranes, and LPLs were purified on a discontinuous 40% and 80% Percoll gradient (GE Healthcare). The obtained epithelial cells and LPLs were used in flow cytometry analysis, as described next.

2.9.

Flow cytometry

To identify the cells that are responsible for altering TNF-α levels in colonic tissues, the isolated epithelial cells and LPLs were fixed and permeabilized with Cytofix/Cytoperm solution (BD Biosciences, Franklin Lakes, NJ, USA) and then incubated with FITC-conjugated anti–TNF-α plus APC-conjugated antipan-cytokeratin or anti-F4/80 antibody at 4°C for 30 minutes. To examine macrophage polarization, LPLs were incubated with FITC-conjugated anti-CD11c or anti-CD206 antibody at 4°C for 30 minutes, permeabilized with Cytofix/Cytoperm solution, and incubated for 30 minutes with the APC-conjugated anti-F4/80 antibody. Species- and isotype-matched antibodies of unrelated specificity were used as controls. Fluorescence intensity was measured using a Guava EasyCyte flow cytometry system (MERCK, Kenilworth, NY, USA).

2.10.

Statistical analyses

All data are presented as means ± standard error of the mean (SEM). Statistical analyses were performed using one-way ANOVA followed by the Tukey–Kramer post hoc test by using

study study

study study

Statcel 3 program (OMS Publishing, Saitama, Japan). The sample size was calculated using the POWER procedure for 1-way ANOVA, considering P < .05 with a power of 0.80 (SAS Institute, Cary, NC, USA) and using the results in our previous studies. Differences were considered significant at P < .05.

3.

Results

3.1.

Effect of naringenin on recovery from DSS-induced colitis

The body weights of mice in the different treatment groups were measured daily (Fig. 1A): DSS administration induced body weight loss in mice, and the body weight change in the DSS group was lower than that in the control group between Days 6 and 16. In the DSS + naringenin group, the body weight change was lower than that in the control group between Days 6 and 11. Although the change in body weight showed no statistically significant difference between the DSS and DSS + naringenin groups throughout the experiment, the mean values in the DSS + naringenin group at and after Day 12 were higher than those in the DSS group. The body weight changes in the DSS + naringenin group between Days 12 and 17 did not significantly differ from those in the control group. The food intake in the DSS + naringenin group at and after Day 12 were apparently and slightly higher than that in the DSS group (data not shown), however, any statistical comparisons were not performed due to the multiple housing of mice (3 or 4 mice/cage). The stool scores in the DSS and DSS + naringenin groups at and after Day 5 were higher than that in the control group (Fig. 1B). The mean values of the stool score in the DSS + naringenin group at and after Day 12 were lower than those in the DSS group, but there was no significant difference between these 2 groups through the experiment. We next examined how naringenin affected the expression of TJ- and inflammation-associated molecules. TJs localize at the intercellular junctions of epithelial cells and play a critical role in maintaining intestinal barrier integrity [7]. Immunoblotting revealed that the TJ proteins JAM-A and claudin-3 were expressed at higher levels in the colon of mice in the DSS + naringenin group than the DSS group (Fig. 1C). Furthermore, RT-PCR results showed that TNF-α and F4/80 mRNA expression in the DSS group,

86

N U TR IT ION RE S EAR CH 6 4 ( 2 01 9 ) 8 2 –9 2

A

B

15

4

0

* *

-5

* *

* * * * *

-10

* Control * DSS * * * * DSS+Naringenin * *

-20 0

Protein density (AU)

Stool score

5

-15

C

3 2.5

β-actin #

#

1 0.5 0

*

* *

* * * 2

*

*

*

* 1 *

* *

*

* * *

Control DSS * DSS+Naringenin * *

0 0

D Control DSS DSS+Naringenin

* * *

*

3 6 9 12 15 18 Time after start of DSS administration (d)

2 1.5

* *

3

mRNA expression (AU)

Body weight change (% of the initial values)

10

40

30

3 6 9 12 15 18 Time after start of DSS administration (d)

Control DSS DSS+Naringenin

20

10

0

* * *

Fig. 1 – Effects of naringenin on body weight change, protein expression of TJ-associated molecules, and mRNA expression of inflammation-associated molecules in DSS-administered mice. Mice were administered 2% (w/v) DSS through drinking water for 6 days, and this was followed by a 11-day recovery period during which distilled water was provided. The mice were fed diets with or without 0.3% (w/w) naringenin during the recovery period. Body weight change (A) and stool score (B) were evaluated daily. Protein expression of TJ-associated molecules (C) and mRNA expression of inflammation-associated molecules (D) were analyzed using immunoblotting and qPCR analyses, respectively. Each immunoblot was representative of 7 mice. Protein and mRNA expressions were normalized to the values from the control group. Values are expressed as means ± SEM, n = 7. *P < .05 vs Control group; #P < .05 vs DSS group (Tukey–Kramer post hoc test).

but not in the DSS + naringenin group, was higher than that in the control group (Fig. 1D). The mRNA levels of IL-1β, IL-6, and IL-17A in the DSS group were also higher than those in the control group, although these increases were not statistically significant. The mean values of TNF-α and IL-6 mRNA levels in the DSS + naringenin groups tended to be lower than those in the DSS group.

3.2. Effect of naringenin on inflammation- and TJ-associated molecules in cultured colonic tissues To examine the early mechanisms underlying the naringeninmediated reduction of colonic damage, the colonic tissues of DSS-administered mice were cultured for 24 hours with or

without naringenin and then used in mRNA expression analyses (Fig. 2A). TNF-α mRNA expression in the DSS group was higher than that in the control group, but the expression levels in the DSS + 100 μmol/L naringenin and DSS+ 200 μmol/L naringenin groups did not differ from that in the control group; however, the reduced TNF-α mRNA level induced by naringenin treatments was not significantly different from the expression level in the DSS group. The mRNA levels of IL-6, IL-10, and F4/80 in the DSS group tended to be higher than those in the control group, and naringenin treatment further increased the mRNA levels of IL-10 and CD206 in colonic tissues, with these levels in the DSS+ 200 μmol/L naringenin group being higher than those in the DSS group. Lastly, the mRNA expression of the TJ proteins occludin, claudin-3, claudin-4, and JAM-A in the DSS group was

87

N U TR IT ION RE S EA RCH 6 4 ( 2 01 9 ) 8 2 –9 2

A

Control

DSS

DSS+100 μM Naringenin

DSS+200 μM Naringenin

8 mRNA expression (AU)

7 6

*#

5 *

4 3

*

*# *

2 1

* **

0

***

***

** *

B TNF-α production (pg/mL)

100 *

80 60

* #

40 #

20 0 Control

0

100

200

Fig. 2 – Effects of naringenin on mRNA expression of inflammation- and TJ-associated molecules and TNF-α protein production in cultured colonic tissues of DSS-administered mice. Two days after DSS administration for 6 days, colonic tissues of mice were dissected and cultured for 24 hours with or without naringenin (100 and 200 μmol/L). (A) The mRNA expression of inflammation- and TJ-associated molecules was analyzed using qPCR and were normalized to the values from the control group. (B) TNF-α protein concentration in culture media was measured using ELISA. Values shown are means ± SEM, n = 7. *P < .05 vs Control group; #P < .05 vs DSS group (Tukey–Kramer post hoc test).

lower than that in the control group, and naringenin treatment did not influence this DSS-induced reduction in expression in the 24-h culture. The TNF-α mRNA expression result was corroborated by ELISA results showing that DSS administration increased TNF-α production by the cultured colonic tissue (Fig. 2B). Naringenin treatment suppressed this DSS-induced TNF-α production, and the TNF-α levels in the DSS+ 100 and DSS+ 200 μmol/L naringenin groups were lower than that in the DSS group.

3.3. Effect of naringenin on inflammatory signaling in cultured colonic tissues TNF-α is recognized to induce distinct inflammatory signaling pathways such as the NF-κB, MEK, and PI3K pathways, which are associated with chronic inflammation [4]. Immunoblotting results demonstrated that the phosphorylation (i.e., the activation) of Akt, ERK1/2, and NF-κB p65 was higher in DSSadministered mice than in control mice, although the increase

in pERK1/2 level was not statistically significant (Fig. 3). Naringenin treatment suppressed the phosphorylation of these signaling molecules, and their phosphorylation levels were significantly lower in the DSS+ 200 μmol/L naringenin group than in the DSS group.

3.4. Effect of naringenin on TNF-α production in epithelial cells and macrophages of cultured colonic tissues Because the reduction in TNF-α production appeared to be associated with the naringenin-mediated suppression of colonic damage, the TNF-α-positive cell populations among the epithelial cells and macrophages of colonic tissues were quantified using flow cytometry. Here, pan-cytokeratin and F4/80 were used as markers of epithelial cells and macrophages, respectively. Whereas DSS administration increased both pan-cytokeratin+ TNF-α+ cells and F4/80+ TNF-α+ cells in the colonic tissues, naringenin treatment decreased the population of pan-cytokeratin+ TNF-α+ cells but not the F4/80+ TNF-α+ macrophages (Fig. 4).

88

N U TR IT ION RE S EAR CH 6 4 ( 2 01 9 ) 8 2 –9 2

β-actin 6

Control DSS DSS+100 μM Naringenin DSS+200 μM Naringenin

Protein density (AU)

5

*

4 #

3

*

2 #

#

1 0 pAkt

pERK1/2

pp65

Fig. 3 – Effects of naringenin on phosphorylation of Akt, ERK1/2, and NF-κB p65 in cultured colonic tissues of DSS-administered mice. Two days after DSS administration for 6 days, colonic tissues of mice were dissected and cultured for 24 hours with or without naringenin (100 and 200 μmol/L). Phosphorylation of Akt, ERK1/2, and NF-κB p65 was analyzed through immunoblotting and densitometry. Each immunoblot was representative of 7 mice. Protein density was normalized to the values from the control group. Values are means ± SEM, n = 7. *P < .05 vs Control group; #P < .05 vs DSS group (Tukey–Kramer post hoc test).

3.5. Effect of naringenin on macrophage polarization in cultured colonic tissues Macrophages are commonly classified into cells of two distinct phenotypes, the M1 and M2 types, and the M2-type macrophages are characterized by the expression of CD206 and the anti-inflammatory molecule IL-10 [6]. To examine how naringenin affects macrophage polarization, flow cytometry was used to detect the F4/80+ CD11c+ and F4/80+ CD206+ cells, which indicate the M1- and M2-type macrophages, respectively. The populations of both F4/80+ CD11c+ and F4/80+ CD206+

10

14

* #

12 * #

8 6 4 2

*

10 8 6 4 2

DSS+ Naringenin

0

0

DSS

DSS+ Naringenin

DSS

Control

DSS+ Naringenin

DSS

Control

Fig. 4 – Effects of naringenin on TNF-α-positive epithelial cells and macrophages in cultured colonic tissues of DSSadministered mice. Two days after DSS administration for 6 days, colonic tissues of mice were dissected and cultured for 24 hours with or without 200 μmol/L naringenin. Isolated epithelial cells (A) and LPLs (B) were stained for pan-cytokeratin plus TNF-α and F4/80 plus TNF-α, respectively, and analyzed using flow cytometry. Values are means ± SEM, n = 7. *P < .05 vs Control group; #P < .05 vs DSS group (Tukey–Kramer post hoc test).

*

Control

0.5

B

12

F4/80+ CD206+ (%)

1

0

0

A

DSS+ Naringenin

10

*

DSS

20

*

1.5

F4/80+ CD11c+ (%)

#

Macrophage polarization is regulated by distinct cellular signaling pathways, such as the PI3K and MEK pathways [6,24,25]. To test the involvement of these pathways in the

2

* 30

3.6. Cellular signaling pathways associated with naringenin-mediated M2-type macrophage polarization in cultured colonic tissues

Control

B

40

F4/80+ TNF-α+ (%)

Pan-cytokeratin+ TNF-α+ (%)

A

macrophages in the DSS group were higher than those in the control group, and naringenin treatment decreased the F4/80+ CD11c+ cells and increased the F4/80+ CD206+ cells (Fig. 5). These results suggest that naringenin induced the polarization of M2-type macrophages.

Fig. 5 – Effects of naringenin on M1-M2 polarization of macrophages in cultured colonic tissues of DSS-administered mice. Two days after DSS administration for 6 days, colonic tissues of mice were dissected and cultured for 24 hours with or without 200 μmol/L naringenin. Isolated LPLs were stained for F4/80 and CD11c (A) or CD206 (B) and analyzed using flow cytometry. Values are means ± SEM, n = 7. *P < .05 vs Control group; #P < .05 vs DSS group.

89

N U TR IT ION RE S EA RCH 6 4 ( 2 01 9 ) 8 2 –9 2

B a

20

ab

a

15 b

a b

10

c

5

b

b

0

10 8

1.5 ab b

1

a b b

0.5

b 0

b

6

b

4 2

ab

bc

c

0

2 a

a

12

IL-10 mRNA expression (AU)

25

CD206 mRNA expression (AU)

CD206 mRNA expression (AU)

IL-10 mRNA expression (AU)

A

2 a 1.5

ab

ab 1

b

b

0.5 0

Naringenin

0

0

200

200

200

Naringenin

0

0

200

200

200

LY294002

0

0

0

12.5

25

U0126

0

0

0

5

10

Control

DSS mice

Control

DSS mice

Fig. 6 – Effects of signaling inhibitors on naringenin-mediated IL-10 and CD206 expression in cultured colonic tissues of DSS-administered mice. Two days after DSS administration for 6 days, colonic tissues of mice were dissected and cultured for 24 hours with or without 200 μmol/L naringenin in the presence and absence of LY294002 (A) or U0126 (B), and then qPCR was used to analyze the mRNA expression of IL-10 and CD206. The mRNA expressions were normalized to the values from the control group. Values are means ± SEM, n = 7. Means marked without a common letter are significantly different, with P < .05 (Tukey–Kramer post hoc test).

naringenin effects observed here, we pretreated colonic tissues with LY294002 (a PI3K inhibitor) and U0126 (a MEK inhibitor) before naringenin treatment. IL-10 and CD206 expression was lower in colonic tissues treated with naringenin in the presence of 25 μmol/L LY294002 and 5 or 10 μmol/L U0126 than in the absence of these inhibitors, although the suppression of IL-10 expression by 10 μmol/L U0126 was not statistically significant (Fig. 6A and B). Immunofluorescence analysis further revealed that the phosphorylation levels of Akt and ERK1/2 in the F4/80-positive cells of the naringenin-treated colonic tissues were higher than those in the corresponding cells of tissues that were not treated with naringenin (Fig. 7). Consistent with the observation in the flow cytometry, the numbers of F4/80-positive cells in 2 DSS groups were apparently higher than that in the control tissues.

4.

Discussion

Our previous study demonstrated that naringenin supplementation before DSS administration reduced the development of colitis in mice [14], which suggested that the use of supplemental naringenin could serve as a preventive method against IBDs. However, in IBD management, it is also critical to alleviate

the inflammatory response in the active disease stage and direct the disease condition into remission, because the patients experience periods of relapse and clinical remission for an extended period [1]. The results of this study conducted using colitic mice and the inflamed colonic tissues obtained from the mice showed that naringenin promoted recovery from the colonic damage and inflammation. Moreover, the early molecular mechanisms involved in this naringenin-mediated effect appeared to be the suppression of TNF-α expression in the epithelial cells and promotion of M2-type polarization in the macrophages of the colonic tissues. Although IBD pathogenesis remains poorly elucidated, accumulating evidence suggests that impairment of the intestinal barrier and uncontrolled and robust expression of inflammatory cytokines in the disease are factors that potentially underlie the development of intestinal inflammation [1,3,4]. Because these cellular processes interact and influence each other, identifying the early mechanisms responsible for recovery from colitis can be challenging. Here, naringenin treatment of cultured colonic tissues for 24 hours did not restore the expression of any TJ proteins tested, although feeding of naringenin-supplemented diets for 11 days up-regulated claudin-3 and JAM-A in DSSadministered mice. By contrast, reduction of TNF-α expression and increase of IL-10 and CD206 expression were observed

90

N U TR IT ION RE S EAR CH 6 4 ( 2 01 9 ) 8 2 –9 2

DSS

DSS+Naringenin

B

pERK1/2

F4/80

Merge

Control

Merge

DSS

F4/80

DSS+Naringenin

pAkt

Control

A

Fig. 7 – Effects of naringenin on phosphorylation of Akt and ERK1/2 in macrophages of cultured colonic tissues of DSS-administered mice. Two days after DSS administration for 6 days, colonic tissues of mice were dissected and cultured for 24 hours with or without 200 μmol/L naringenin. Cryosections of the colonic tissues were stained for F4/80 and pAkt (A) or pERK (B) and analyzed using confocal microscopy. Green signal: immunofluorescence of pAkt or pERK1/2; red signal: immunofluorescence of F4/80. Images shown are representative of the labeling in 7 mice used for each treatment.

following 24-h naringenin treatment in the colonic tissues. These results suggest that in the effect produced by naringenin, the down-regulation of TNF-α and the induction of M2-type macrophages represent mechanistic steps that precede the regulation of the TJ barrier. The decreased TNF-α and increased IL-10 possibly result in the restoration of TJ barrier, because the TNF-α and IL-10 negatively and positively regulate the TJ barrier, respectively. Alternatively, the effect of naringenin on TJ regulation in the inflamed tissues may occur in the later phase. However, the current findings or observations are inadequate for further specific speculation, and additional investigations must be conducted to elucidate the precise mechanisms underlying the naringenin effects. TNF-α is produced by various immune and nonimmune cells in the inflamed intestines, including macrophages and epithelial cells [4,5]. Intriguingly, naringenin reduced the population of pan-cytokeratin+ TNFα+ cells, but not F4/80+ TNF-α+ cells, in the cultured colonic tissues of DSS-administered mice; this finding indicates diminished TNF-α production in the epithelial cells of the tissue. TNF-α has been shown to exert pleiotropic effects in the inflamed mucosa in IBD, such as induction of apoptosis and barrier loss of epithelial cells, activation of macrophages, and necrosis of Paneth cells [4,7]. Therefore, decreased production of epithelial TNF-α is likely to at least partly contribute to the naringenin-mediated recovery from colonic damage. PinhoRibeiro et al. demonstrated that naringenin suppressed lipopolysaccharide-induced TNF-α production in RAW 264.7

macrophages by inhibiting NF-κB activation [26]. Although our study has not precisely elucidated how naringenin suppresses epithelial TNF-α expression, the reduced activation of NF-κB in colonic tissues might represent a step upstream of TNF-α down-regulation. As an essential component of innate immunity, macrophages perform multiple functions in both tissue damage and repair [6]. Although the naringenin treatment used here did not affect F4/80 expression in cultured colonic tissues, naringenin influenced the balance of M1-M2 polarization and induced an increase in M2-type macrophages. M1 macrophages are stimulated by microbial products or proinflammatory cytokines and are characterized by high production of inflammatory mediators, and these cells are therefore implicated in initiating and sustaining inflammation [6,27]. Conversely, M2 macrophages are associated with the resolution of chronic inflammation through the production of anti-inflammatory cytokines such as IL-10 [6,27]. Our results suggest that the up-regulation of IL-10 production by M2-type macrophages is another potential mechanism responsible for the naringenin-mediated colonic repair. Accumulating evidence indicates that IL-10 plays an essential role in intestinal homeostasis, protection, and healing. In mouse models, genetic IL-10 deficiency causes the spontaneous development of colitis [28], and inducible overexpression of IL-10 decreases the colitic phenotype elicited by DSS administration [29]. Accordingly, two human studies have demonstrated that mutations in IL-10 or its receptor are

N U TR IT ION RE S EA RCH 6 4 ( 2 01 9 ) 8 2 –9 2

associated with the severity of IBDs [30,31]. We further found that naringenin-mediated expression of IL-10 and CD206 was sensitive to pharmacological inhibition of PI3K and MEK. Previously, the induction of M2-type macrophages by luteolin and medroxyprogesterone was shown to involve the PI3K and MEK pathways, respectively [24,25]. Here, the results of immunofluorescence labeling confirmed that the phosphorylation of PI3K and ERK1/2 was increased in the F4/80+ macrophages of colonic tissues treated with naringenin. In agreement with our results, naringenin has been found to activate these signaling pathways in pancreatic β-cells or myocardial cells, although the precise underlying mechanisms are unknown [32,33]. Further studies are required to unveil the regulation of macrophage polarization by naringenin. IBDs are lifelong and relapsing disorders for which therapy to induce remission is followed by therapy to maintain remission [1]. Although medication represents the first-line therapy in IBD management, both relapse and remission are influenced by diets and nutrition [2]. Therefore, in addition to medical therapy, nutritional and dietary interventions are considered to be critical in IBD management. In several clinical trials, dietary supplements such as probiotics, prebiotics, and vitamin D were investigated in the management of IBDs [2,34,35]; however, the results were inconsistent, possibly due to heterogeneity of patients or disease etiology. Our findings potentially suggest another option in effective dietary intervention for IBD management. However, the precise mechanisms underlying the regulation of intestinal inflammation by naringenin should be elucidated before using naringenin for human consumption in future studies. Naringenin was used here at ~200 μmol/L because the colonic naringenin concentrations in mice fed a 0.3% (w/w) naringenin diet were 100–200 μmol/L in our preliminary study. However, the naringenin intake and concentration used in this study may not be achieved by regular meals in humans. The daily intake of naringenin from regular meals providing 2800 kcal has been estimated at 56 mg/d in a clinical study [36]. The 0.3% (w/w) naringenin diet is approximately equivalent to 2.1 g/2800 kcal of diet. Therefore, a naringenin supplement might be required to effectively provide the effects observed in this study. In this regard, the high intake of naringenin in humans seems to be still safe. The animal study using diabetic rats demonstrated that no mortality was found in the naringenin intake of 100 mg/kg body weight, which was equivalent to 6 g/60 kg body weight [37]. Moreover, the additional mechanisms underlying the naringenin effects must be elucidated in further investigations. This study did not consider the interaction between the epithelial cells and macrophages in the analysis of the naringenin effect. The suppression of epithelial TNF-α expression might influence the polarization of M2 macrophages or the production of IL-10, or vice versa. Not only epithelial cells and macrophages but also other cells may contribute to the naringenin-mediated recovery from colonic damage. For example, the regulatory T cells, which produce the IL-10, may in part have a role in the naringeninmediated IL-10 expression. Wang et al. demonstrated that naringenin promoted the differentiation of regulatory T cells in murine splenocytes [38]. In conclusion, naringenin exhibits considerable potential to promote recovery from DSS-induced colitis, and suppression of

91

epithelial TNF-α production and induction of M2-type macrophages potentially represent early mechanisms underlying the naringenin effect.

Acknowledgment This research was partially supported by a Grant-in-Aid for Scientific Research(C) (JSPS Kakenhi 16K07737) and Mishima Kaiun Memorial Foundation. The authors declare that there are no conflicts of interest. Y. C. and T. S. designed the research; Y. C. conducted the study and performed statistical analyses; Y. C., Y. Y., and T. S. analyzed the data; T.S. wrote the manuscript and had primary responsibility for the final content. We would like to thank Editage (www.editage.jp) for English language editing.

REFERENCES

[1] Ananthakrishnan AN. Epidemiology and risk factors for IBD. Nat Rev Gastroenterol Hepatol 2015;12:205–17. [2] Haskey N, Gibson DL. An examination of diet for the maintenance of remission in inflammatory bowel disease. Nutrients 2017;9:259. [3] Turner JR. Intestinal mucosal barrier function in health and disease. Nat Rev Immunol 2009;9:799–809. [4] Neurath MF. Cytokines in inflammatory bowel disease. Nat Rev Immunol 2014;14:329–42. [5] Stadnyk AW. Intestinal epithelial cells as a source of inflammatory cytokines and chemokines. Can J Gastroenterol 2002;16:241–6. [6] Wang N, Liang H, Zen K. Molecular mechanisms that influence the macrophage m1-m2 polarization balance. Front Immunol 2014;5:614. [7] Suzuki T. Regulation of intestinal epithelial permeability by tight junctions. Cell Mol Life Sci 2013;70:631–59. [8] Zocco MA, dal Verme LZ, Cremonini F, Piscaglia AC, Nista EC, Candelli M, et al. Efficacy of lactobacillus GG in maintaining remission of ulcerative colitis. Aliment Pharmacol Ther 2006; 23:1567–74. [9] Losurdo G, Iannone A, Contaldo A, Ierardi E, Di Leo A, Principi M. Escherichia coli Nissle 1917 in ulcerative colitis treatment: systematic review and meta-analysis. J Gastrointestin Liver Dis 2015;24:499–505. [10] Miele E, Pascarella F, Giannetti E, Quaglietta L, Baldassano RN, Staiano A. Effect of a probiotic preparation (VSL#3) on induction and maintenance of remission in children with ulcerative colitis. Am J Gastroenterol 2009;104:437–43. [11] Hallert C, Kaldma M, Petersson BG. Ispaghula husk may relieve gastrointestinal symptoms in ulcerative colitis in remission. Scand J Gastroenterol 1991;26:747–50. [12] Hanai H, Kanauchi O, Mitsuyama K, Andoh A, Takeuchi K, Takayuki I, et al. Germinated barley foodstuff prolongs remission in patients with ulcerative colitis. Int J Mol Med 2004;13:643–7. [13] Hanai H, Iida T, Takeuchi K, Watanabe F, Maruyama Y, Andoh A, et al. Curcumin maintenance therapy for ulcerative colitis: randomized, multicenter, double-blind, placebo-controlled trial. Clin Gastroenterol Hepatol 2006;4:1502–6. [14] Azuma T, Shigeshiro M, Kodama M, Tanabe S, Suzuki T. Supplemental naringenin prevents intestinal barrier defects and inflammation in colitic mice. J Nutr 2013;143:827–34. [15] Noda S, Tanabe S, Suzuki T. Naringenin enhances intestinal barrier function through the expression and cytoskeletal association of tight junction proteins in Caco-2 cells. Mol Nutr Food Res 2013;57:2019–28.

92

N U TR IT ION RE S EAR CH 6 4 ( 2 01 9 ) 8 2 –9 2

[16] Kawabata A, Van Hung T, Nagata Y, Fukuda N, Suzuki T. Citrus kawachiensis Peel powder reduces intestinal barrier defects and inflammation in Colitic mice. J Agric Food Chem 2018;66:10991–9. [17] Hung TV, Suzuki T. Dietary fermentable Fiber reduces intestinal barrier defects and inflammation in Colitic mice. J Nutr 2016;146:1970–9. [18] Hung TV, Suzuki T. Short-chain fatty acids suppress inflammatory reactions in Caco-2 cells and mouse colons. J Agric Food Chem 2018;66:108–17. [19] Ogata M, Ogita T, Tari H, Arakawa T, Suzuki T. Supplemental psyllium fibre regulates the intestinal barrier and inflammation in normal and colitic mice. Br J Nutr 2017;118:661–72. [20] Miyoshi Y, Tanabe S, Suzuki T. Cellular zinc is required for intestinal epithelial barrier maintenance via the regulation of claudin-3 and occludin expression. Am J Physiol Gastrointest Liver Physiol 2016;311:G105–16. [21] Kaikiri H, Miyamoto J, Kawakami T, Park SB, Kitamura N, Kishino S, et al. Supplemental feeding of a gut microbial metabolite of linoleic acid, 10-hydroxy-cis-12-octadecenoic acid, alleviates spontaneous atopic dermatitis and modulates intestinal microbiota in NC/nga mice. Int J Food Sci Nutr 2017;88:941–51. [22] Miyauchi E, Ogita T, Miyamoto J, Kawamoto S, Morita H, Ohno H, et al. Bifidobacterium longum alleviates dextran sulfate sodium-induced colitis by suppressing IL-17A response: involvement of intestinal epithelial costimulatory molecules. PLoS One 2013;8:e79735. [23] Miyamoto J, Mizukure T, Park SB, Kishino S, Kimura I, Hirano K, et al. A gut microbial metabolite of linoleic acid, 10hydroxy-cis-12-octadecenoic acid, ameliorates intestinal epithelial barrier impairment partially via GPR40-MEK-ERK pathway. J Biol Chem 2015;290:2902–18. [24] Jiang Q, Pan D, Yang Y, Hu Y, Fang L, Shang P, et al. Luteolin regulates macrophage polarization via the PI3K/Akt pathway to inhibit the apoptosis stimulated by angiotensin. Curr Pharm Biotechnol 2018;19:428–37. [25] Tsai YC, Tseng JT, Wang CY, Su MT, Huang JY, Kuo PL. Medroxyprogesterone acetate drives M2 macrophage differentiation toward a phenotype of decidual macrophage. Mol Cell Endocrinol 2017;452:74–83. [26] Pinho-Ribeiro FA, Zarpelon AC, Mizokami SS, Borghi SM, Bordignon J, Silva RL, et al. The citrus flavonone naringenin reduces lipopolysaccharide-induced inflammatory pain and leukocyte recruitment by inhibiting NF-kappaB activation. J Nutr Biochem 2016;33:8–14.

[27] Shapouri-Moghaddam A, Mohammadian S, Vazini H, Taghadosi M, Esmaeili SA, Mardani F, et al. Macrophage plasticity, polarization, and function in health and disease. J Cell Physiol 2018;233:6425–40. [28] Kuhn R, Lohler J, Rennick D, Rajewsky K, Muller W. Interleukin-10-deficient mice develop chronic enterocolitis. Cell 1993;75:263–74. [29] Cardoso A, Gil Castro A, Martins AC, Carriche GM, Murigneux V, Castro I, et al. The dynamics of Interleukin-10-afforded protection during dextran sulfate sodium-induced colitis. Front Immunol 2018;9:400. [30] Glocker EO, Kotlarz D, Boztug K, Gertz EM, Schaffer AA, Noyan F, et al. Inflammatory bowel disease and mutations affecting the interleukin-10 receptor. N Engl J Med 2009;361: 2033–45. [31] Shouval DS, Biswas A, Goettel JA, McCann K, Conaway E, Redhu NS, et al. Interleukin-10 receptor signaling in innate immune cells regulates mucosal immune tolerance and anti-inflammatory macrophage function. Immunity 2014;40: 706–19. [32] Lin CY, Ni CC, Yin MC, Lii CK. Flavonoids protect pancreatic beta-cells from cytokines mediated apoptosis through the activation of PI3-kinase pathway. Cytokine 2012;59:65–71. [33] Han X, Pan J, Ren D, Cheng Y, Fan P, Lou H. Naringenin-7-Oglucoside protects against doxorubicin-induced toxicity in H9c2 cardiomyocytes by induction of endogenous antioxidant enzymes. Food Chem Toxicol 2008;46:3140–6. [34] Salaritabar A, Darvishi B, Hadjiakhoondi F, Manayi A, Sureda A, Nabavi SF, et al. Therapeutic potential of flavonoids in inflammatory bowel disease: a comprehensive review. World J Gastroenterol 2017;23:5097–114. [35] Orel R, Kamhi Trop T. Intestinal microbiota, probiotics and prebiotics in inflammatory bowel disease. World J Gastroenterol 2014;20:11505–24. [36] Ranka S, Gee JM, Biro L, Brett G, Saha S, Kroon P, et al. Development of a food frequency questionnaire for the assessment of quercetin and naringenin intake. Eur J Clin Nutr 2008;62:1131–8. [37] Mathi S, Murugaiyah K. Determination of LD50 of naringenin for its effects on diabetic nephropathy in rats-a pilot study. J Chem Pharm Res 2015;7:550–3. [38] Wang HK, Yeh CH, Iwamoto T, Satsu H, Shimizu M, Totsuka M. Dietary flavonoid naringenin induces regulatory T cells via an aryl hydrocarbon receptor mediated pathway. J Agric Food Chem 2012;60:2171–8.