Kaempferol Exerts Anti-Inflammatory Effects by Accelerating Treg Development via Aryl Hydrocarbon Receptor-Mediated and PU.1/IRF4-Dependent Transactivation of the Aldh1a2/Raldh2 Gene in Dendritic Cells

Abstract

Retinaldehyde dehydrogenase 2 (Raldh2) encoded by the Aldh1a2 gene is expressed in intestinal dendritic cells (DCs) and converts retinal toward retinoic acid (RA), which functions as a ligand of the nuclear receptor RAR. DC-derived RA accelerates the development of Tregs by promoting the RAR-dependent transactivation of the Foxp3 gene, which encodes a master transcription factor of Tregs [1, 2]. In the present study, we screened food ingredients with the expectation of finding dietary compounds that exert beneficial effects on intestinal immune tolerance and identified kaempferol, a flavonoid, as the compound that most effectively increased Aldh1a2 mRNA levels in DCs (Table S1, Figure S1A–C). The kaempferol treatment enhanced the enzyme activity of Raldh2 in bone marrow-derived DCs (BMDCs) (Figure 1A) and in migratory DCs (migDCs) isolated from mesenteric lymph nodes (MLNs) (Figure 1B). The development of Foxp3⁺ cells from OT-II-derived naïve CD4⁺ T cells was enhanced by a co-culture with ovalbumin (OVA) peptide-pulsed BMDCs in the presence of kaempferol (Figure 1C) and was accompanied by the suppression of T-cell proliferation (Figure S1D). The frequency of Foxp3⁺ cells was also increased by a co-culture with kaempferol-pretreated BMDCs (Figure S1E), suggesting that the kaempferol treatment conferred enhanced Treg-inducing activity on DCs.

To investigate the molecular mechanisms underlying the effects of kaempferol on Raldh2, we analyzed the aryl hydrocarbon receptor (AhR), a reported target of kaempferol [3], and found that kaempferol increased the levels of both the mRNA (Figure 1D right) and proteins (Figure 1E) of AhR. Knockdown (KD) experiments using siRNA revealed that under AhR-KD conditions, Aldh1a2 mRNA levels (Figure 1D left), and the number of Raldh2-expressing cells (Figure S2A) in BMDCs increased and became unresponsive to kaempferol. The mRNA level and enzymatic activity of Raldh2 were reduced by the treatment with TCDD, an agonist of AhR known as a persistent environmental contaminant (Figure 1F), and were significantly enhanced by the treatment with CH-223191, an antagonist of AhR (Figure 1G). Although Raldh2 activity was not affected by the supplementation with kynurenine, an endogenous agonist of AhR (Figure S2B), KD of Ido1 and Ido2, the rate-limiting enzymes of kynurenine synthesis (indoleamine 2,3-dioxygenases) in DCs, increased Aldh1a2 mRNA levels in DCs (Figure S2C), and this was attenuated by the addition of kynurenine (Figure 1H). These results suggest that kaempferol up-regulated Aldh1a2 gene expression by antagonizing AhR, which repressed Aldh1a2 gene expression with constitutive activation by endogenously synthesized kynurenine in DCs.

We examined the effects of kaempferol on the expression and function of the transcription factors PU.1 (encoded by the Spi1 gene) and IRF4, which transactivate the Aldh1a2 gene by binding to the enhancer element in approximately −2 kb of the Aldh1a2 gene in DCs [4]. The kaempferol treatment increased Irf4 transcripts (Figure 2A) and the protein levels of PU.1 and IRF4 (Figure 2B, Figure S3A), and also accelerated the recruitment of PU.1 to the Aldh1a2 gene (Figure 2C). Furthermore, the kaempferol-induced increase in Aldh1a2 mRNA levels was abolished in DCs following the KD of Spi1 or Irf4 (Figure 2D). In addition, kaempferol-induced increases in PU.1 protein levels were eliminated by Ahr KD, whereas not only the kaempferol treatment but also Ahr KD raised IRF4 protein levels (Figure 2E). Ahr KD increased the mRNA levels of Spi1 and Irf4, and the increased mRNA levels of Irf4 but not Spi1 in Ahr KD DCs were further elevated by kaempferol treatment (Figure S3B). Based on these observations, we indicated that the release of PU.1 and IRF4 from transcriptional repression by AhR is responsible for the kaempferol-induced transactivation of the Aldh1a2 gene in DCs, and suggest that AhR may be involved in post-transcriptional regulation of PU.1, while the Irf4 gene is also upregulated by kaempferol through another pathway independent of AhR.

We then examined the effects of AhR antagonization on Raldh2 activation in DCs in vivo. The frequency of DCs exhibiting Raldh2 activity in the MLNs of mice was increased by the intraperitoneal (i.p.) administration of an AhR antagonist (Figure 2F). We also observed an increase in Tregs in the Peyer's patches of C57BL/6 mice i.p. administered kaempferol-treated BMDCs (Figure 2G). We utilized an OVA-induced food allergy model of Balb/c mice to examine the effects of kaempferol in vivo (Figure S4A) and confirmed that the rapid decrease in body temperature and allergic diarrhea observed just after the OVA challenge were significantly suppressed in mice administered kaempferol (Figure 2H). The administration of kaempferol during the sensitization phase (from Day −7 to Day 28) was intended to limit the effect of kaempferol on challenge stage, given the accumulating evidence, including ours, indicating that activation of mast cells (MCs) is suppressed by kaempferol [5-7]. Although the potential impact of kaempferol on MC function cannot be ruled out, at least our preliminary experiments suggest that kaempferol–AhR axis may not be involved in suppression of MCs, because the inhibitory effect of kaempferol on MC activation was not affected by Ahr KD and AhR mRNA level was quite low in MCs (data not shown). Further analysis is needed to clarify the anti-allergic effects of kaempferol on food allergy. In addition, as limitations of this study, we state that these results should be replicated in human cells.

The essential role of Raldh2 in neonatal skin homeostasis was demonstrated in a recent study, which showed that the DC-specific deficiency of Raldh2 limited commensal specific Treg generation in mice [8]. Therefore, the upregulation of gene expression and/or the function of Raldh2 in DCs contribute to anti-inflammatory immunoresponses systemically.

M.T. performed experiments, analyzed data, and prepared figures. K.N. analyzed data, prepared figures, and wrote the manuscript. Y.W. performed experiments, analyzed data, and prepared figures. M.Y., K.I., T.H., N.M., M.K., W.Z., N.I. and T.Y. performed experiments and analyzed data. C.N. supervised, designed research, and wrote the manuscript.

The authors declare no conflicts of interest.