In general, dietary antigens are tolerated by the gut associated immune system. Impairment of this so-called oral tolerance is a serious health risk. We have previously shown that activation of the ligand-dependent transcription factor aryl hydrocarbon receptor (AhR) by the environmental pollutant 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) affects both oral tolerance and food allergy. In this study, we determine whether a common plant-derived, dietary AhR-ligand modulates oral tolerance as well. We therefore fed mice with indole-3-carbinole (I3C), an AhR ligand that is abundant in cruciferous plants. We show that several I3C metabolites were detectable in the serum after feeding, including the high-affinity ligand 3,3´-diindolylmethane (DIM). I3C feeding robustly induced the AhR-target gene CYP4501A1 in the intestine; I3C feeding also induced the aldh1 gene, whose product catalyzes the formation of retinoic acid (RA), an inducer of regulatory T cells. We then measured parameters indicating oral tolerance and severity of peanut-induced food allergy. In contrast to the tolerance-breaking effect of TCDD, feeding mice with chow containing 2 g/kg I3C lowered the serum anti-ovalbumin IgG1 response in an experimental oral tolerance protocol. Moreover, I3C feeding attenuated symptoms of peanut allergy. In conclusion, the dietary compound I3C can positively influence a vital immune function of the gut.
Citation: Hammerschmidt-Kamper C, Biljes D, Merches K, Steiner I, Daldrup T, Bol-Schoenmakers M, et al. (2017) Indole-3-carbinol, a plant nutrient and AhR-Ligand precursor, supports oral tolerance against OVA and improves peanut allergy symptoms in mice. PLoS ONE 12(6): e0180321. https://doi.org/10.1371/journal.pone.0180321
Editor: Hossam M. Ashour, Wayne State University, UNITED STATES
Received: February 8, 2017; Accepted: June 14, 2017; Published: June 30, 2017
Copyright: © 2017 Hammerschmidt-Kamper et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper.
Funding: This work was supported by the Deutsche Forschungsgemeinschaft, ES103/6-1. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Abbreviations: AhR, aryl hydrocarbon receptor; DC, dendritic cell; DIM, 3, 3'-diindolylmethane; FICZ, 6-formylindolo[3,2-b]carbazol; I3C, indolo-3-carbinol; ICZ, indolo[3,2-b]carbazol; OT, oral tolerance; OVA, ovalbumin; PE, peanut extract; RA, retinoic acid; SC, Standard chow; TCDD, 2,3,7,8-tetrachlorodibenzo-p-dioxin
Oral tolerance (OT) is a vital and unique feature mediated by the mucosal immune system. OT denotes that the immune system remains unresponsive to food antigens encountered in the gut. OT was first described many decades ago, and models of inducing OT in the mouse have shown that the antigen dose is important for the mechanism underlying OT. Thus, it was proposed that high doses of antigen in the diet result in anergic T cells, while low doses result in an induction of regulatory T cells (Treg) by dendritic cells (DC) of the small-intestine-draining mesenteric lymph nodes [1–3]. This rigid difference is challenged, however, by the finding that high numbers of Treg develop in high-dose tolerizing protocols , and the difficulties of providing direct proof of T cell deletion and anergy.
OT against food antigens is both local and systemic, and must be distinguished from local tolerance against gut residing commensal bacteria. Food allergies and celiac disease are thought to result from impaired oral tolerance induction. On the flip side, OT is a promising therapeutic approach against autoimmune reactions. These approaches aim at establishing OT against the disease-causing autoantigens by feeding the respective protein or peptide [5–7].
Retinoic acid (RA) produced by DC enhances the conversion of T cells into Tregs . Furthermore, RA produced by DC imprints the a4ß7 integrin gut-homing surface markers on T cells . RA is generated by retinal dehydrogenases, which are highly expressed in DC. Interestingly, the aryl hydrocarbon receptor (AhR), a ligand-activated transcription factor, can enhance RA-mediated differentiation in myeloblastic leukemia cells , and the physiological AhR ligand 6-formylindolo[3,2-b]carbazol (FICZ) enhances differentiation in these leukemia cells . AhR-activation by the AhR ligand drug VAF347 shifts the monocyte-mediated differentiation of T cells to a T helper (Th)-22 phenotype, which may protect from autoimmunity [12, 13]. The AhR is an old evolutionary transcription factor. It is activated upon binding of a diverse range of small molecular weight chemicals, among them certain xenobiotics (such as planar polyhalogenated hydrocarbons), dietary compounds (flavonoids, glucosinolates) or endogenously generated tryptophan derivatives . AhR signaling is critical for many immune functions. The AhR is highly expressed in gut epithelium and gut immune cells. The AhR is also involved in gut epithelial barrier integrity, differentiation of innate lymphoid cells, and the Th17/ Treg balance [15–17]. Of note, AhR-ligands are abundant in the diet, e.g., plant components such as flavonoids or glucosinolates, or in animal tissue in which xenobiotic ligands such as dioxins accumulate.
We had previously shown that exposure to 2,3,87,8-tetrachlorodibenzo-p-dioxin (TCDD) destabilized oral tolerance in mice. Thus, TCDD treated mice produced antibodies against the model antigen ovalbumin (OVA) after several boosting immunizations, despite previous tolerization . TCDD is an environmental pollutant, and policies seek to reduce pollution and thereby exposure—including by dietary uptake—as much as possible. Certain plant-derived AhR-ligands are part of a healthy diet, however, and some are even sold as dietary supplements in an uncontrolled market. We therefore sought to investigate how oral dosing of one such AhR-ligand, indole-3-carbinol (I3C), affects oral tolerance, and whether this correlated with an increase in food allergies. We here report the surprising finding that the naturally occurring ligand I3C was able to boost oral tolerance against OVA and improved the anaphylactic score in a model of peanut allergy.
Material and methods
Ultra-high performance liquid chromatography coupled with photodiode array detection (UPLC-PDA)
I3C or its metabolites 3,3´-diindolomethane (DIM) and indolo[3,2-b]carbazol (ICZ) were quantified in blood serum samples up to 7 days by feeding with 2 g/kg I3C containing chow. Briefly, 50 μl serum was mixed with 500 μl tert-butylmethylether and 20 μl ISTD (4-methoxyindole; 0,5 ng/μl in MeOH) for organic extraction, followed by drying under a stream of nitrogen N2. The extracts were then split and dissolved in either 70% methanol (for I3C) or acetonitrile (for DIM/ICZ). Blank human serum samples were spiked with defined concentrations of all three analytes to obtain calibrators at levels of 25, 50, 100, 250, 500 and 1000 ng/ml serum. Analyses were performed on a Waters Acquity® ultra high performance LC-PDA system using MassLynx® software, and an Acquity UPLC® HSS C18 1.8 μm; 2.1 x150 mm column. The detector type was an UPLC LG 500 nm. Sera from two mice were analyzed and quantified by linear regression of the peak-area ratios (analyte/internal standard).
Mice and dietary intervention
We used C57BL/6 mice or conditional AhRΔCD11c and AhRΔvillin. We used littermates as WT controls. Mice were bred in our specific-pathogen free animal facility; mice were under a 12 hour/12 hour light–dark cycle, and had access to food and water ad libitum. I3C (2 g/kg) enriched normal chow (NC) was purchased from Sniff Special Diets (Soest, Germany). The 2 g/kg dose was chosen because it had been successfully used in other long-term dietary intervention studies of gut immune parameters dependent on AhR [16, 19]. The daily food uptake/per mouse was equal for both chows (data not shown), and the calculated daily I3C dose per mouse was approximately 5–6 mg.
To determine gene induction in the small intestine, a bolus of 200 μg or 20 mg I3C/mouse or 10 μg/kg TCDD, dissolved in DMSO and diluted in olive oil (10% DMSO and 90% Olive oil) was fed by gavage.
Peanut extract (PE) (30 mg/ml) was prepared from peanuts from the Golden Peanut Plant (provided by Intersnack Nederland BV, The Netherlands) as described previously . Peanut extracts were checked for protein content by BCA analysis (Pierce, Rockford, IL). Cholera toxin (CT) was purchased from List Biological Laboratories, Inc. (Campbell, CA).
Mice were sacrificed by cervical dislocation for organ resections. All experiments were performed with the permission of the relevant governmental bodies in accordance with relevant German animal welfare laws (permission by LANUV, Az.:84–02.04.2012 A166).
Oral tolerance induction
For tolerization, mice were gavaged with 20 mg of OVA (grade V;#A5503, Sigma-Aldrich, Munich, Germany) in PBS/20 g body weight on day 0 and day 2 of the experiment. Mice were then immunized i.p. day 9 and boosted on days 20, 27, 34, and 42 with 10 μgOVA/20 g body weight, dissolved in 50 μl of PBS and 50 μl of incomplete Freund´s adjuvant (IFA). Blood samples were taken on the days of immunization.
OVA-specific IgG1 antibodies in mouse serum were determined by ELISA. 96-well plates were coated with 100 μg/ml overnight at 4°C. Serum samples were titrated onto the plates and detected with goat anti-mouse IgG1 antiserum (Southern Biotech, Birmingham, USA) coupled to biotin. The ELISA was developed with avidin-HRP/TMB and values measured at 450 nm after the addition of 0.75 μl of H2SO4. All values are expressed related to a standard from pooled sera of mice immunized and boosted with OVA/IFA.
Flow cytometry of lamina propria immune cells
To isolate immune cells from the Lamina propria, we adapted protocols available in the literature, leaving out a gradient step to avoid the possible selective loss of viable cells . Briefly, the small intestine was resected, washed and cut into small pieces. The pieces were placed for 20 minutes in a solution containing 0.2 mg/ml collagenase and DNAse (Roche, Mannheim, Germany). The solution was vigorously vortexed and then filtered through a 100 μm strainer. Cells were washed in medium and counted in trypan blue. Fluorescence was measured on a FACSCanto-II™ flow cytometer (Becton-Dickinson, Heidelberg, Germany). At least 30,000 cells were collected in list mode. Debris, dead cells, and doublets were gated out by staining with Fixable Viability Dye eFluor 506 (eBioscience, Frankfurt/Main, Germany staining. Lymphocytes/leucocytes were gated out according to scatter characteristics. Finally, the frequency of DC was determined in the remaining population by MHCII (M5/114.15.2 BD-Biosciences, Heidelberg, Germany), CD11c (N418), and CD103 (2E7, eBioscience, Frankfurt, Germany)
Allergic reaction to peanut extracts (PE) was performed as described [22, 23]. Briefly, mice were sensitized to PE by intragastric exposure to PE (6 mg PE, 200 μl/mouse) with CT (15 μg/mouse) on three consecutive days (day 0, 1, and 2) followed by weekly dosing (days 7, 14, 21, and 28). One day before termination, mice were challenged i.p. with PE (10 μg PE/g body weight). Control mice received PBS only during sensitization and PE for challenge. Mice were fed chow containing I3C 2 g/kg throughout the experiment. Mice were scored for allergic symptoms on a 1–4 scale (1 = scratching at the snout, 2 = decreased activity, some swelling around nose, 3 = longer than 1 minute of no motions, 4 = no reaction upon touching of fibrillae, 5 = tremor and death—animals were killed before reaching this stage). The mice’s rectal temperatures were measured at the time points indicated. Blood samples for measuring PE-specific IgG and IgE were taken three days after challenge.
RT-PCR of RNA from small intestine
Small intestine pieces were stored at -80°C. Tissue was homogenized in RNAmagic (BioBudget, Krefeld, Germany) with one 3 mm stainless steel or tungsten carbide bead (Qiagen, Hilden, Germany) at 30 Hz for 1 min using the TissueLyserII (Qiagen, Hilden, Germany). RNA was extracted from the homogenate by chloroform/isopropanol, after which the pellet was washed in ethanol and finally resuspended in water. RNA was reverse transcribed with MMLV reverse transcriptase. Quantitative PCR reactions were performed with the Rotor-Gene SYBR Green PCR Kit (Qiagen, Hilden, Germany) in a Rotor-GeneQ thermo cycler (Qiagen, Hilden, Germany). The following primers were used: Cyp1a1-For 5’-TCCTTGCATGTCCATGTTTC-3’, Cyp1a1-Rev 5’-TGCATAAGCAAAATACAGTCCA-3’, RPS6-For 5’-TACTGTGCCTCGTCGGTTG-3’, RPS6-Rev 5’-TGAATCTTGGGTGCTTTGGT-3’, AhR-For 5’-AGACCGGCTGAACACAGAGT-3’, AhR-Rev 5’-GTCAGCAGGGGTGGACTTTA-3’, Aldh1a1-For 5’-TGTGGGAATACCGTGGTTGTC-3’, Aldh1a1-Rev 5’-GTGAAGAGCCGTGAGAGGAG-3’ (Vicente-Suarez et al., 2015), Aldh1a3-For 5’-GAGCAGCAATTTCCTCCCATC-3’, Aldh1a3 5’-GAGCCGGTGAAGGCTATCT-3’ (Vicente-Suarez et al., 2015), GAPDH-For 5’-CGTCCCGTAGACAAAATGGT-3’, GAPDH-Rev 5’-TTGATGGCAACAATCTCCAC-3’. Expression levels were calibrated to the expression of rps6  in the same sample using the 2ΔΔCT method .
Oral exposure to I3C induces AhR-target gene CYP1A1 in the small intestine
We first assessed whether feeding mice with I3C induces the AhR-target gene cyp1a1 in the small intestine. We gavaged mice with 200 μg or 20 mg I3C per mouse; we then sacrificed the mice 4 or 24 hours later. Control mice were fed solvent alone (low control) or 10 μg/kg body weight TCDD (high control). As shown in Fig 1, the lower dose of 200 μg of significantly induced cyp1A1 gene expression in the duodenum after 4 hours. In the more distal small intestine sections, induction was visible, but did not reach significance. In contrast, the high 20 mg I3C dose induced strong cyp1a1 expression along the entire small intestine (in duodenum, jejunum and ileum) at both 4 hours and 24 hours. Except for the ileum, the increase was significant. The I3C effect was long-lasting and still detectable 24 hours after gavage. The cyp1a1 induction, however, was an order of magnitude lower than the induction observed by 10 μg of TCDD. Oral tolerance protocols take several weeks. I3C is metabolized and thus a single bolus may not be enough for continuous AhR-activation. We therefore tested whether the enrichment of mouse chow with 2 g/kg (resulting in an approximate dose of 5–6 mg/mouse/day) induces AhR-target gene cyp1a1. Mice were fed for 49 days with this chow, and as shown in Fig 1E, cyp1a1 induction was high along the entire small intestine, significantly in the duodenum and jejunum. Thus, continuous feeding of an AhR ligand does not lead to exhaustion, but to persistent AhR-activation in the gut. We note, however, that the single bolus of 20 mg I3C/mouse and the long-term dietary intervention led to approximately the same AhR induction over the control, indicating a plateau level of AhR-activation (compare Fig 1D with 1E). The expression of the AhR gene itself was not affected by I3C feeding (data not shown). We used the dose of 2 g/kg chow in further dietary intervention studies.
A-D: Mice were fed by gavage with TCDD, I3C or DMSO (as solvent for TCDD) at the indicated doses. Shown is cyp1a1 expression relative to the housekeeping gene rps6 in the three sections of the small intestine 4 hours after gavage (A, B), or 24 hours after gavage (C, D). Note the scale differences for TCDD and I3C. N = 3; Mean ± SEM. Two-way ANOVA followed by Sidak’s multiple comparison test. *≤0.05, ** ≤0.01, *** ≤0.001, **** ≤0.0001. E: Mice were fed for 49 days with chow containing 2 g/kg I3C. On day 49, RNA was isolated from small intestine and cyp1a1 expression was analyzed by quantitative PCR. The results of two pooled independent experiments are shown. N = 11–15. Mean ± SEM. Two-way ANOVA followed by Sidak’s multiple comparison test. *≤0.05, ** ≤0.01, *** ≤0.001, **** ≤0.0001.
Metabolites of dietary I3C are detectable the serum
We took serum samples 15 minutes, 2 days and 7 days after giving I3C by gavage (15 min) or via supplemented chow; we analyzed the samples by UPLC coupled to UV-detection. I3C is known to convert in vitro within minutes in an acid solution into a complex mixture, with approximately 2–6% DIM, and <0.01% ICZ . Similarly, the group of Bjeldanes reported that DIM and ICZ were present in gut tissue after I3C gavage (yields of 0.1% and <0.01%, respectively). Confirming and extending their results, we analyzed the serum and found that the I3C metabolite DIM was easily detectable, but ICZ was below our limit of detection (Fig 2). In addition to DIM, eight possible metabolites of I3C were detected; however, we did not characterize them further. The concentrations of I3C, DIM and metabolites detected at the end of the feeding periods are shown in Table 1.
Mice were fed for up to 7 days with chow containing 2 g/kg I3C. For the 15 minute value, mice received 5 mg/I3C per mouse diluted in 10%DMSO/olive oil by gavage. Serum samples were prepared and analyzed for the presence of I3C, DIM, and ICZ. Sera were spiked with 4-methoxyindole as the internal standard to calibrate for possible differences in extraction efficiency. Note that I3C is detectable only after the bolus feeding, but not when I3C was given in the chow, and is thus eaten in smaller amounts over the entire day. Y-axis: arbitrary units.
Dietary intervention with I3C licenses oral tolerance against OVA
Several naturally occurring ligands of AhR, including indoles, reportedly ameliorated autoimmune diseases such as EAE or allergic asthma [27, 28]. We developed an OVA feeding protocol in such a way that the OVA dose was no longer high enough to induce oral tolerance in mice fed normal chow. Mice on normal or I3C enriched chow (starting already at day minus 7) were fed 20 mg OVA/20 g body weight on days 0 and 2, followed by immunization and 3 boosts. Fig 3 shows that in this situation, feeding of I3C led to five times lower titers of anti-OVA IgG1 antibodies compared to normal chow-fed control mice, i.e., I3C enrichment of the chow supported the formation of oral tolerance.
Mice were fed a placebo (black bars) or I3C containing diet (white bars) and were either not tolerized (PBS = control) or tolerized with OVA (dotted bars). Mice were immunized and then boosted 2 (A), 3 (B), and 4 weeks (C) later. Antibody titers were determined at the day of each boost and one week after the third boost (D). Antibody titers are depicted as % of placebo/PBS-treated animals that were set to 100%. Data are pooled from three independent experiments. N = 12–13 mice/group. Mean ± SEM. One-way ANOVA followed by Tukey’s multiple comparison test. *≤0.05, ** ≤0.01, *** ≤0.001, **** ≤0.0001.
Characterization of gut immune cells upon I3C feeding
The lamina propria is the mucosal effector site where many DC reside and contribute to gut homeostasis and a non-inflammatory state . DCs are MHC-II+CD11c+CD103+, but several subsets—possibly differing in functions—exist and can be identified by further surface molecules. We analyzed lamina propria dendritic cells by flow cytometry in I3C fed mice versus control mice. Mice were fed for 28 days with 2 g/kg I3C supplemented chow and during the following 21-day tolerizing protocol (see Material and methods). Interestingly, the percentage of CD11c+ increased significantly in the lamina propria in these mice fed I3C. Within this subset, the number of CD103+MHC-II+, i.e., those DC with tolerogenic properties, increased compared to control DC (Fig 4A and 4B). At the same time, feeding of I3C induced retinaldehyde dehydrogenase aldh1a1 in the small intestine Fig 4C. In contrast, aldh1a3 was not induced by I3C feeding (Fig 4D). The enzyme Aldh1a1 converts vitamin A to retinoic acid, which in turn, induces TGFβ production by DC, and thus is critical for the generation of regulatory T cells in the gut and to gut-homing of Treg . AhR activity can drive Aldh1 , although the evidence is indirect so far. Aldh1 enzymes are expressed in both intestinal epithelial cells and in DC [32–34]. We therefore tested “I3C tolerogenicity” in conditional AhR-deficient mice, where AhR is lost either in intestinal epithelial cells (AhRΔVillin) or in CD11c (AhRΔCD11c). The results show that in both mouse lines, I3C supported tolerance, indicating that the presence of AhR in either cell type alone is not a sine qua non for the positive I3C effect (Fig 5). AHR-null mice were not tested upon I3C feeding as their digestion and detoxification would not be comparable due to their patent ductus venosus [35, 36]
A-B: aldh1 expression after continuous feeding of I3C. Mice were fed for 49 days with normal or I3C enriched chow and expression of retinaldehyde dehydrogenase aldh1a1 (A) and aldh1a3 (B), analyzed by quantitative PCR. Fold change compared to housekeeping gene GAPDH is shown. Shown is the mean ± SEM of pooled data from two independent experiments. C-D: mice were fed continuously for 28 days normal chow (“placebo”) or I3C-enriched chow before tolerance induction. Samples were taken on day 21, i.e., 49 days after start of dietary intervention. For gating strategy see Material and methods C: Percentage of CD11c+ (MHCII positive or negative) in living lamina propria cells of the small intestine (left), and the CD103+ subpopulation in CD11c+MHCII (right) DC in the lamina propria. D. typical histograms N = 9. Mean ± SEM. Unpaired Student’s t-test. *≤0.05, ** ≤0.01, *** ≤0.001.
OVA specific IgG1 antibody titers of AhRΔCD11c (A) and AhRΔvillin (B) mice that lack the AhR in dendritic cells or intestinal epithelial cells, respectively. AhRΔCD11c and AhRΔvillin mice (= cre+; white bars) and WT littermates (= cre-; black bars) were fed an I3C containing diet and either not tolerized (PBS = control) or tolerized with OVA. Antibody titers were determined at the end of the experiment. Antibody titers are depicted as % of I3C/cre-/PBS-treated animals that were used as controls. Mean ± SEM. One-way ANOVA followed by Tukey’s multiple comparison test. *≤0,05, ** ≤0,01, *** ≤0,001, **** ≤0,0001.
Peanut allergy after I3C feeding
TCDD exposure suppresses the allergic reaction in a model of peanut allergy in mice . We therefore tested this model in I3C fed mice. Mice were first sensitized and then challenged i.p. with peanut extract; we then measured anti-peanut IgG1 and IgE antibodies in the serum. In contrast to the TCDD exposure , feeding I3C did not suppress peanut allergy (Fig 6). Peanut extract induced a strong antibody response for both isotypes, but feeding of I3C had no effect on the response strength (Fig 6A and 6B). The strength of an allergic response can be assessed by the typical sudden drop in body temperature upon antigen challenge, and by scoring the bodily reaction (from 1–5, see Material and methods). The mean body temperature of the mice at the onset of the challenge was 38.23 ±0.26°C. The drop in body temperature is shown in Fig 6C. Indeed, only 2/5 mice fed with I3C but 4/5 mice on normal chow had a drop in body temperature of more than 2% (i.e., ≤37.4°C) upon allergen challenge. Additionally, when scoring the overall body reaction, a significant improvement by I3C was observed. While the allergic reaction in untreated mice resulted on average in a score of 3, i.e., swelling around the nose and >1 min motionlessness, the I3C-fed mice scored 1, i.e., they displayed only some scratching (Fig 6D).
Mice were fed with normal (NC) I3C-enriched chow (I3C), sensitized to peanut extract, and then challenged by an i.p. injection with peanut extract (PE). A: IgG1 B: IgE antibody titers. C: rectal temperatures were taken after peanut injection. Mean rectal body temperature for placebo control was 38.3°C. The number of mice with a drop in temperature of >2% (considered a symptom of allergy), i.e., with a temperature below 37.4°C, were counted at 10, 30, 50, 70, 90, and 110 minutes. D: Anaphylactic scores. 0 = no symptoms; (N = 4 for control groups, N = 5 for PE groups. Mean± SEM is shown; A, B, C. statistical significance calculated with One-way ANOVA followed by Tukey’s multiple comparisons test: *≤0,05, **≤0,01, ***≤0,001,. NC = normal chow; I3C = 2 g/kg I3C supplemented chow graph.
Thus, our data in the two models of gut-mediated immune responses indicate that I3C can improve both oral tolerance and ameliorate the symptoms of a food allergy. Together, this suggests that dietary AhR-ligands are part of gut immune homeostasis.
The health benefits of plants of the Brassica family are well known. The focus in the respective research is mainly on the anti-tumorigenic properties, however, where studies demonstrated beneficial effects of broccoli on some (but not all) cancers . The phytochemicals mediating these effects are often glucosinolates, indoles, and their various derivatives, in particular their acid-derivatives. Indoles can bind to the AhR, leading to gene expression, e.g., of phase I and phase II enzymes , and eventually cellular and systemic responses. In the rat, increased CYP1A1 enzyme activity has been shown in the liver after feeding of 57 mg I3C/kg for 7 days . We extended this study to show that oral exposure leads to induction of the cyp1a1 along the murine small intestine, i.e., our site of immunological interest. Moreover, we found that induction did not subside upon long-term exposure of a comparatively moderate dose, indicating constantly high AhR-activity by the diet. I3C is converted to several metabolites in the acid environment of the stomach, most abundantly into DIM, as well as into other products [26, 41]. Both I3C and DIM are AhR ligands, with binding affinities of approximately 2x10-10 (ICZ) and 9x10-8 (DIM) [26, 41], compared to 10–12 M for TCDD. Of note, we detected I3C derivatives in serum, i.e., they can transfer systemically. This extends older studies, which analyzed I3C derivatives DIM and ICZ locally in the gut and feces within a day of I3C feeding .
In contrast to cancer, very few studies examined immunological consequences of dietary AhR-ligands. AhR is important for the immune system and critically contributes to the balance of regulatory T cells and Th17 pro-inflammatory T cells [17, 42, 43]. Indeed, it was shown that removal of AhR ligands from the diet mimics the adverse effects of AhR-deficiency in the gut [15, 16], such as failure to expand innate lymphoid cells after birth. Re-addition was able to rescue the effect in wild-type, but not in AhR-deficient mice, demonstrating its AhR-dependency . We here investigated the effects of I3C on the vital immune function of oral tolerance, and found that dietary exposure of I3C licensed oral tolerance under conditions that would not normally induce this immunosuppressive state. Thus, I3C exposure had the opposite outcome of TCDD exposure, which broke tolerance. Rather, the I3C effect we observed is reminiscent of an experiment by the group of Howard Weiner, who applied 2-(1'H-indole-3'-carbonyl)-thiazole-4-carboxylic acid methyl ester (ITE) and the autogenic peptide MOG35-55 embedded in nanoparticles to mice. This protocol of oral tolerization treatment improved the outcome of experimental autoimmune encephalitis (EAE), by inducing Treg . Supporting tolerization or induction of tolerance is of great interest as a therapeutic approach [2, 45]. Beyond oral tolerance, it has also been tested in other tolerizing systems, e.g., successfully using anterior-associated immune deviation to improve collagen-induced arthritis or EAE [46, 47]. Amelioration of EAE by injections of AhR ligands have been studied for several years [48, 49]; the discussed mechanism includes both the induction of Treg via changes in DC tolerogenicity, or the induction of T cell cycle arrest by certain micro RNAs . As a caveat, the group of Nagarkatti showed that dietary I3C and DIM suppressed a DTH-response against a model protein, while the high-affinity ligand FICZ did not .
Why did I3C behave so differently from TCDD in our experimental OT model and why do differences exist between AhR-ligands in the other studies? For OT, several possibilities exist, namely (i) the antigen dose, (ii) exposure timing, or (iii) nature of the AhR-binding peculiars. To start with the latter, the differing effects of ligands on AhR-activation outcome are well-known but little understood, and still highly controversial . Whether different ligands bind to AhR at slightly differing sites and with differing outcomes on its ligand-induced conformation changes, DNA-binding capacity or resistance to down-stream degradation is not known, and must remain speculative as only the PAS-A domain of AhR has been crystallized so far [51–53]. Another obvious explanation is the persistence of TCDD in the body, while other ligands are often quickly degraded. By feeding I3C continuously (which continuously triggered high cyp1a1 expression), we tried to equalize this parameter. Indeed, our data show that continuous feeding of I3C results in a plateau level of DIM in the serum. Thus, although DIM has an approximately three orders of magnitude lower binding affinity than TCDD, the dose is effective. Recently, we have found that the skin barrier can also be improved by I3C feeding; thus, I3C derivatives also can act far beyond the gut itself (Haas et al, in press).
Oral tolerance is foremost an observable phenomenon: food contains numerous proteins (>100 g protein per day in an average human diet) that can serve as antigens and are immunogenic if encountered under immunizing conditions; yet, we normally do not react to food antigens [54, 55]. Models of experimental oral tolerance (EOT) in animals suggested that the dose and feeding regimes can influence which mechanism is triggered, i.e., either the induction of regulatory T cells or the induction of T cell anergy [1, 2, 56]. Possibly, both mechanisms occur side-by-side in a real-life situation, and possibly there is also a “no observed effect level” dose, i.e., where food antigens might be too rare to necessitate an active oral tolerance response. In our hands, comparatively small changes in the antigen exposure scheme resulted in a situation in which tolerance was suboptimal but could be unmasked by I3C. Both the humoral and cellular immune responses are triggered in EOT. It involves “stepwise and anatomically defined processes of induction, maintenance and migration of Tregs” . Central to the process are CD103+DC of the gut lamina propria, which take up antigen, transport it to MLN and induce a gut homing phenotype in T cells. The cytokine milieu can influence DC tolerogenicity depending on the local availability and context. For instance, GM-CSF can have both pro- and anti-inflammatory properties . Lack of GM-CSF, which in the gut is mainly produced by RORγt(+) innate lymphoid cells (which, in turn, may need AHR for their development ), can lead to low numbers of Treg cells and impaired oral tolerance . CD103+ DC derived RA supports the conversion of naïve T cells into regulatory T cells and helps to maintain their phenotype [29, 60, 61]. In a very recent study, the group of Joshua Mezrich demonstrated that mice fed an I3C-enriched diet had more Tregs, ILC3s and γδ T cells in cecal cells of the gut, associated with a better protection from Clostridium difficile infection . We note, however, that in an in-vitro study where CD103+DC were differentiated from bone marrow cells, the addition of DIM suppressed their ability to induce Treg . This result provides additional evidence of the importance of tissue spatial context and dose. While we failed to see a significant increase in Treg in the intestinal tissue in our study (data not shown), we identified increased frequencies of CD11c+ and CD103+ cells by I3C in our OT model. Moreover, the key enzyme for RA production was significantly higher in all sections of the small intestine after continuous exposure to I3C. Together, our findings offer an explanation of why oral tolerance was induced upon long-term I3C feeding, and can be tested further. Of note, epithelial cells are also an important source of RA and promote CD103+ DC differentiation . Our results with I3C-fed conditional AhR-deficient mice (DC or IEC specific) support the argument that AhR-deficiency in only one of these cell types is not sufficient to loose I3C-enhanced tolerance.
Finally, we tested the effects of dietary I3C on peanut allergy. Food allergies can be understood as the breakdown of oral tolerance, resulting in systemic anaphylaxis and usually a Type I allergic response. Food allergies result in symptoms ranging from mild itching, to abdominal pain, bronchospasms and even anaphylactic shock. While difficult to estimate, it is thought that food allergies affect approximately 2–5% of the European population . We found that feeding of I3C ameliorated the allergic symptoms in our mice, but not the sensitization (i.e., antibody formation). Thus, there seems to be no (protective) effect of I3C regarding allergy induction and B cell activity, which is in line with results regarding peanut allergy in AhR-KO mice . Conceivably, challenge responses are more easily influenced by interventions compared to the antibody levels, i.e., sensitization itself. This has also been observed in other types of intervention , although we do not yet have a mechanistic explanation for this phenomenon. Possibly, I3C or its metabolites directly affect mast-cells, or interfere with mast-cell/Treg interaction.
While we only tested one dose regime, it is possible that more I3C could have even higher positive effects. Thus, more research is needed regarding other doses and timing. Importantly, while indoles have positive health effects and have even been clinically tested in certain settings as safe , they can turn cancerogenic at too high doses .
In conclusion, we showed that dietary I3C or its metabolites such as DIM activate AhR in the gut and contribute to the generation of anti-inflammatory milieu in the gut. This supports vital immune functions of oral tolerance and may ameliorate or prevent food allergies. We thus provide a new angle on the positive health effects of this important plant compound.
We thank Babette Martiensen and Swantje Steinwachs for skilled technical support.
We gratefully acknowledge funding by the Deutsche Forschungsgemeinschaft (DFG grant ES103/6-1).
- Conceptualization: CE CH RP.
- Formal analysis: CH MB IS TD.
- Funding acquisition: CE.
- Investigation: CH IS DB KM.
- Methodology: CE RP TD.
- Project administration: CE CH MB.
- Resources: MB RP TD.
- Supervision: CE RP.
- Visualization: CH IS.
- Writing – original draft: CE CH.
- Writing – review & editing: CE CH IS MB RP TD KM.
- 1. Friedman A and Weiner HL. Induction of anergy or active suppression following oral tolerance is determined by antigen dosage. Proc Natl Acad Sci U S A. 1994; 91: 6688–6692. pmid:8022835
- 2. Weiner HL, da Cunha AP, Quintana F, Wu H. Oral tolerance. Immunol Rev. 2011; 241: 241–259. pmid:21488901
- 3. Friedman A, al-Sabbagh A, Santos LM, Fishman-Lobell J, Polanski M, Das MP, et al. Oral tolerance: a biologically relevant pathway to generate peripheral tolerance against external and self antigens. Chem Immunol. 1994; 58: 259–290. pmid:8011155
- 4. Siewert C, Lauer U, Cording S, Bopp T, Schmitt E, Hamann A, et al. Experience-driven development: effector/memory-like alphaE+Foxp3+ regulatory T cells originate from both naive T cells and naturally occurring naive-like regulatory T cells. J Immunol. 2008; 180: 146–155. pmid:18097014
- 5. Faria AM and Weiner HL. Oral tolerance: therapeutic implications for autoimmune diseases. Clin Dev Immunol. 2006; 13: 143–157. pmid:17162357
- 6. Huber A, Diedrichs-Mohring M, Wildner G. Spontaneously relapsing-remitting experimental autoimmune uveitis in rats allows successful therapeutic oral tolerance induction in ongoing disease. Mol Immunol. 2015; 63: 215–226. pmid:25085538
- 7. Thurau SR and Wildner G. Oral tolerance for treating uveitis—new hope for an old immunological mechanism. Prog Retin Eye Res. 2002; 21: 577–589. pmid:12433378
- 8. Coombes JL, Siddiqui KR, rancibia-Carcamo CV, Hall J, Sun CM, Belkaid Y, et al. A functionally specialized population of mucosal CD103+ DCs induces Foxp3+ regulatory T cells via a TGF-beta and retinoic acid-dependent mechanism. J Exp Med. 2007; 204: 1757–1764. pmid:17620361
- 9. Iwata M, Hirakiyama A, Eshima Y, Kagechika H, Kato C, Song SY. Retinoic acid imprints gut-homing specificity on T cells. Immunity. 2004; 21: 527–538. pmid:15485630
- 10. Bunaciu RP and Yen A. Activation of the aryl hydrocarbon receptor AhR Promotes retinoic acid-induced differentiation of myeloblastic leukemia cells by restricting expression of the stem cell transcription factor Oct4. Cancer Res. 2011; 71: 2371–2380. pmid:21262915
- 11. Bunaciu RP and Yen A. 6-Formylindolo (3,2-b)carbazole (FICZ) enhances retinoic acid (RA)-induced differentiation of HL-60 myeloblastic leukemia cells. Mol Cancer. 2013; 12: 39-. pmid:23656719
- 12. Baba N, Rubio M, Kenins L, Regairaz C, Woisetschläger M, Carballido JM, et al. The aryl hydrocarbon receptor (AhR) ligand VAF347 selectively acts on monocytes and naive CD4(+) Th cells to promote the development of IL-22-secreting Th cells. Hum Immunol. 2012; 73: 795–800. pmid:22609446
- 13. Pot C. Aryl hydrocarbon receptor controls regulatory CD4+ T cell function. Swiss Med Wkly. 2012; 142: w13592-. pmid:22653785
- 14. Denison MS, Soshilov AA, He G, DeGroot DE, Zhao B. Exactly the same but different: promiscuity and diversity in the molecular mechanisms of action of the aryl hydrocarbon (dioxin) receptor. Toxicol Sci. 2011; 124: 1–22. pmid:21908767
- 15. Li Y, Innocentin S, Withers DR, Roberts NA, Gallagher AR, Grigorieva EF, et al. Exogenous Stimuli Maintain Intraepithelial Lymphocytes via Aryl Hydrocarbon Receptor Activation. Cell. 2011; pmid:21999944
- 16. Kiss EA, Vonarbourg C, Kopfmann S, Hobeika E, Finke D, Esser C, et al. Natural aryl hydrocarbon receptor ligands control organogenesis of intestinal lymphoid follicles. Science. 2011; 334: 1561–1565. pmid:22033518
- 17. Esser C and Rannug A. The aryl hydrocarbon receptor (AhR) in barrier organ physiology, immunology, and toxicology. Pharmacol Rev. 2015; 67: 259–279. pmid:25657351
- 18. Chmill S, Kadow S, Winter M, Weighardt H, Esser C. 2,3,7,8-Tetrachlorodibenzo-p-dioxin impairs stable establishment of oral tolerance in mice. Toxicol Sci. 2010; 118: 98–107. pmid:20729464
- 19. Ito S, Chen C, Satoh J, Yim S, Gonzalez FJ. Dietary phytochemicals regulate whole-body CYP1A1 expression through an arylhydrocarbon receptor nuclear translocator-dependent system in gut. J Clin Invest. 2007; 117: 1940–1950. pmid:17607366
- 20. van Wijk F., Hoeks S, Nierkens S, Koppelman SJ, van KP, Boon L, et al. CTLA-4 signaling regulates the intensity of hypersensitivity responses to food antigens, but is not decisive in the induction of sensitization. J Immunol. 2005; 174: 174–179. pmid:15611239
- 21. Goodyear AW, Kumar A, Dow S, Ryan EP. Optimization of murine small intestine leukocyte isolation for global immune phenotype analysis. J Immunol Methods. 2014; 405: 97–108. pmid:24508527
- 22. Marcondes RM, Hassing I, Bol-Schoenmakers M, Bleumink R, Boon L, van BJ, et al. CD4(+) CD25(+) T regulatory cells do not transfer oral tolerance to peanut allergens in a mouse model of peanut allergy. Clin Exp Allergy. 2011; 41: 1324–1333. pmid:21338425
- 23. Bol-Schoenmakers M, Marcondes RM, Bleumink R, Boon L, Man S, Hassing I, et al. Regulation by intestinal gammadelta T cells during establishment of food allergic sensitization in mice. Allergy. 2011; 66: 331–340. pmid:20880210
- 24. Frericks M and Esser C. A toolbox of novel murine house-keeping genes identified by meta-analysis of large scale gene expression profiles. Biochim Biophys Acta. 2008; 1799: 830–837.
- 25. Livak KJ and Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods. 2001; 25: 402–408. pmid:11846609
- 26. Bjeldanes LF, Kim JY, Grose KR, Bartholomew JC, Bradfield CA. Aromatic hydrocarbon responsiveness-receptor agonists generated from indole-3-carbinol in vitro and in vivo: comparisons with 2,3,7,8-tetrachlorodibenzo-p-dioxin. Proc Natl Acad Sci U S A. 1991; 88: 9543–9547. pmid:1658785
- 27. Jeong KT, Hwang SJ, Oh GS, Park JH. FICZ, a tryptophan photoproduct, suppresses pulmonary eosinophilia and Th2-type cytokine production in a mouse model of ovalbumin-induced allergic asthma. Int Immunopharmacol. 2012; 13: 377–385. pmid:22561122
- 28. Singh NP, Singh UP, Rouse M, Zhang J, Chatterjee S, Nagarkatti PS, et al. Dietary Indoles Suppress Delayed-Type Hypersensitivity by Inducing a Switch from Proinflammatory Th17 Cells to Anti-Inflammatory Regulatory T Cells through Regulation of MicroRNA. J Immunol. 2016; 196: 1108–1122. pmid:26712945
- 29. Jaensson E, Uronen-Hansson H, Pabst O, Eksteen B, Tian J, Coombes JL, et al. Small intestinal CD103+ dendritic cells display unique functional properties that are conserved between mice and humans. J Exp Med. 2008; 205: 2139–2149. pmid:18710932
- 30. von Boehmer H. Oral tolerance: is it all retinoic acid? J Exp Med. 2007; 204: 1737–1739. pmid:17620364
- 31. Stanford EA, Wang Z, Novikov O, Mulas F, Landesman-Bollag E, Monti S, et al. The role of the aryl hydrocarbon receptor in the development of cells with the molecular and functional characteristics of cancer stem-like cells. BMC Biol. 2016; 14: 20-. pmid:26984638
- 32. Bhat PV. Retinal dehydrogenase gene expression in stomach and small intestine of rats during postnatal development and in vitamin A deficiency. FEBS Lett. 1998; 426: 260–262. pmid:9599020
- 33. Coombes JL and Powrie F. Dendritic cells in intestinal immune regulation. Nat Rev Immunol. 2008; 8: 435–446. pmid:18500229
- 34. Lampen A, Meyer S, Arnhold T, Nau H. Metabolism of vitamin A and its active metabolite all-trans-retinoic acid in small intestinal enterocytes. J Pharmacol Exp Ther. 2000; 295: 979–985. pmid:11082432
- 35. Julliard W, De Wolfe TJ, Fechner JH, Safdar N, Agni R, Mezrich JD. Amelioration of Clostridium difficile Infection in Mice by Dietary Supplementation With Indole-3-carbinol. Ann Surg. 2017; 265: 1183–1191. pmid:27280500
- 36. Schmidt JV, Su GH, Reddy JK, Simon MC, Bradfield CA. Characterization of a murine Ahr null allele: involvement of the Ah receptor in hepatic growth and development. Proc Natl Acad Sci U S A. 1996; 93: 6731–6736. pmid:8692887
- 37. Schulz VJ, Smit JJ, Willemsen KJ, Fiechter D, Hassing I, Bleumink R, et al. Activation of the aryl hydrocarbon receptor suppresses sensitization in a mouse peanut allergy model. Toxicol Sci. 2011; 123: 491–500. pmid:21804081
- 38. Latte KP, Appel KE, Lampen A. Health benefits and possible risks of broccoli—an overview. Food Chem Toxicol. 2011; 49: 3287–3309. pmid:21906651
- 39. Verhoeven DT, Verhagen H, Goldbohm RA, van den Brandt PA, van PG. A review of mechanisms underlying anticarcinogenicity by brassica vegetables. Chem Biol Interact. 1997; 103: 79–129. pmid:9055870
- 40. Staack R, Kingston S, Wallig MA, Jeffery EH. A comparison of the individual and collective effects of four glucosinolate breakdown products from brussels sprouts on induction of detoxification enzymes. Toxicol Appl Pharmacol. 1998; 149: 17–23. pmid:9512722
- 41. Nguyen LP and Bradfield CA. The search for endogenous activators of the aryl hydrocarbon receptor. Chem Res Toxicol. 2008; 21: 102–116. pmid:18076143
- 42. Stockinger B, Di MP, Gialitakis M, Duarte JH. The aryl hydrocarbon receptor: multitasking in the immune system. Annu Rev Immunol. 2014; 32: 403–432. pmid:24655296
- 43. Hanieh H. Toward understanding the role of aryl hydrocarbon receptor in the immune system: current progress and future trends. Biomed Res Int. 2014; 2014: 520763-. pmid:24527450
- 44. Yeste A, Nadeau M, Burns EJ, Weiner HL, Quintana FJ. Nanoparticle-mediated codelivery of myelin antigen and a tolerogenic small molecule suppresses experimental autoimmune encephalomyelitis. Proc Natl Acad Sci U S A. 2012; 109: 11270–11275. pmid:22745170
- 45. Park KS, Park MJ, Cho ML, Kwok SK, Ju JH, Ko HJ, et al. Type II collagen oral tolerance; mechanism and role in collagen-induced arthritis and rheumatoid arthritis. Mod Rheumatol. 2009; 19: 581–589. pmid:19697097
- 46. Farooq SM, Elkhatib WF, Ashour HM. The in vivo and in vitro induction of anterior chamber associated immune deviation to myelin antigens in C57BL/6 mice. Brain Behav Immun. 2014; 42: 118–122. pmid:24953428
- 47. Farooq SM, Kumar A, Ashour HM. Eye-mediated immune tolerance to Type II collagen in arthritis-prone strains of mice. J Cell Mol Med. 2014; 18: 2512–2518. pmid:25211510
- 48. Quintana FJ, Basso AS, Iglesias AH, Korn T, Farez MF, Bettelli E, et al. Control of T(reg) and T(H)17 cell differentiation by the aryl hydrocarbon receptor. Nature. 2008; 453: 65–71. pmid:18362915
- 49. Rouse M, Rao R, Nagarkatti M, Nagarkatti PS. 3,3'-diindolylmethane ameliorates experimental autoimmune encephalomyelitis by promoting cell cycle arrest and apoptosis in activated T cells through microRNA signaling pathways. J Pharmacol Exp Ther. 2014; 350: 341–352. pmid:24898268
- 50. Duarte JH, Di MP, Hirota K, Ahlfors H, Stockinger B. Differential influences of the aryl hydrocarbon receptor on Th17 mediated responses in vitro and in vivo. PLoS One. 2013; 8: e79819-. pmid:24244565
- 51. Soshilov AA and Denison MS. Ligand promiscuity of aryl hydrocarbon receptor agonists and antagonists revealed by site-directed mutagenesis. Mol Cell Biol. 2014; 34: 1707–1719. pmid:24591650
- 52. Wu D, Potluri N, Kim Y, Rastinejad F. Structure and dimerization properties of the aryl hydrocarbon receptor PAS-A domain. Mol Cell Biol. 2013; 33: 4346–4356. pmid:24001774
- 53. Fraccalvieri D, Soshilov AA, Karchner SI, Franks DG, Pandini A, Bonati L, et al. Comparative analysis of homology models of the AH receptor ligand binding domain: verification of structure-function predictions by site-directed mutagenesis of a nonfunctional receptor. Biochemistry. 2013; 52: 714–725. pmid:23286227
- 54. Husby S, Mestecky J, Moldoveanu Z, Holland S, Elson CO. Oral tolerance in humans. T cell but not B cell tolerance after antigen feeding. J Immunol. 1994; 152: 4663–4670. pmid:8157979
- 55. Kapp K, Maul J, Hostmann A, Mundt P, Preiss JC, Wenzel A, et al. Modulation of systemic antigen-specific immune responses by oral antigen in humans. Eur J Immunol. 2010; 40: 3128–3137. pmid:20957752
- 56. Faria AM and Weiner HL. Oral tolerance. Immunol Rev. 2005; 206: 232–259. pmid:16048553
- 57. Pabst O and Mowat AM. Oral tolerance to food protein. Mucosal Immunol. 2012; 5: 232–239. pmid:22318493
- 58. Bhattacharya P, Budnick I, Singh M, Thiruppathi M, Alharshawi K, Elshabrawy H, et al. Dual Role of GM-CSF as a Pro-Inflammatory and a Regulatory Cytokine: Implications for Immune Therapy. J Interferon Cytokine Res. 2015; 35: 585–599. pmid:25803788
- 59. Mortha A, Chudnovskiy A, Hashimoto D, Bogunovic M, Spencer SP, Belkaid Y, et al. Microbiota-dependent crosstalk between macrophages and ILC3 promotes intestinal homeostasis. Science. 2014; 343: 1249288-. pmid:24625929
- 60. Sun CM, Hall JA, Blank RB, Bouladoux N, Oukka M, Mora JR, et al. Small intestine lamina propria dendritic cells promote de novo generation of Foxp3 T reg cells via retinoic acid. J Exp Med. 2007; 204: 1775–1785. pmid:17620362
- 61. Agace WW and Persson EK. How vitamin A metabolizing dendritic cells are generated in the gut mucosa. Trends Immunol. 2012; 33: 42–48. pmid:22079120
- 62. Park JH, Choi AJ, Kim SJ, Jeong SY. 3,3'-Diindolylmethane Inhibits Flt3L/GM-CSF-induced-bone Marrow-derived CD103(+) Dendritic Cell Differentiation Regulating Phosphorylation of STAT3 and STAT5. Immune Netw. 2015; 15: 278–290. pmid:26770182
- 63. Iliev ID, Spadoni I, Mileti E, Matteoli G, Sonzogni A, Sampietro GM, et al. Human intestinal epithelial cells promote the differentiation of tolerogenic dendritic cells. Gut. 2009; 58: 1481–1489. pmid:19570762
- 64. Nwaru BI, Hickstein L, Panesar SS, Roberts G, Muraro A, Sheikh A. Prevalence of common food allergies in Europe: a systematic review and meta-analysis. Allergy. 2014; 69: 992–1007. pmid:24816523
- 65. van den Elsen LW, Bol-Schoenmakers M, van Esch BC, Hofman GA, van de Heijning BJ, Pieters RH, et al. DHA-rich tuna oil effectively suppresses allergic symptoms in mice allergic to whey or peanut. J Nutr. 2014; 144: 1970–1976. pmid:25342698
- 66. Minich DM and Bland JS. A review of the clinical efficacy and safety of cruciferous vegetable phytochemicals. Nutr Rev. 2007; 65: 259–267. pmid:17605302