Nuclear factor-E2-related factor 2 (Nrf2) is a transcription factor that promotes antioxidant responses. Nrf2 binds to the E3 ubiquitin ligase adaptor protein Keap1 and is subsequently ubiquitinated and degraded by the proteasome. Sulforaphane (SFN), a compound found in broccoli sprouts, has been shown to activate Nrf2 by inactivating Keap1. We previously demonstrated that Nrf2 stability was also regulated by WDR23 independently of Keap1. In this study, SFN treatment significantly increased Nrf2 protein levels in A549 cells harboring a loss-of-function Keap1 mutation, supporting a Keap1-independent component of Nrf2 regulation. Based on these findings, we focused on the PI3K/AKT/FOXO pathway as a candidate mechanism underlying SFN-induced Nrf2 upregulation. Co-treatment with SFN and either a PI3K or an AKT inhibitor significantly suppressed SFN-induced Nrf2 accumulation. Moreover, immunoblotting suggested that SFN altered FOXO3A phosphorylation in an AKT inhibitor-sensitive manner. SFN also decreased the protein levels of DDB1 and CUL4, core components of the WDR23 E3 ubiquitin ligase complex. Together, these findings are consistent with the possible involvement of AKT/FOXO/DDB1/WDR23-related signaling in SFN-induced Nrf2 accumulation. Because SFN has been reported to exert anti-inflammatory effects, we examined its effects on LPS-induced inflammatory gene expression. SFN reduced LPS-induced interleukin (IL)-1β gene expression in Hep3B and A549 cells. In addition, Nrf2 overexpression was associated with reduced IL-1β gene expression. Taken together, these findings indicate that SFN-induced Nrf2 upregulation is associated with suppression of IL-1β gene expression.
Nuclear factor-E2-related factor 2 (Nrf2) is a key transcription factor that regulates cellular detoxification and antioxidant responses by inducing the transcription of antioxidant factors and drug-metabolizing enzymes. Under basal conditions, Nrf2 binds to the E3 ubiquitin ligase adaptor protein Keap1 and is subsequently ubiquitinated and degraded by the proteasome. When cells are exposed to radicals such as reactive oxygen species, Keap1 undergoes oxidative modifications that change its conformation and impair its binding to Nrf2 (Zhang and and Hannink, 2003; Wakabayashi et al., 2004; Cleasby et al., 2014). As a result, Nrf2 is stabilized and then translocates into the nucleus, where it promotes the transcription of antioxidant factors such as HO-1 and NQO1 (Motohashi et al., 2004; Cullinan et al., 2004; Katsuoka et al., 2005). Nrf2 is also thought to regulate cellular defense mechanisms against xenobiotic toxicity through the induction of drug-metabolizing enzymes via the Keap1 pathway, particularly cytochrome P450s (CYPs) (Siswanto et al., 2022a).
Nrf2 is regulated not only by the Keap1 pathway but also by pathways such as the PI3K/AKT/GSK3β pathway, in which AKT is a serine-threonine kinase that regulates various cellular functions, including metabolism, cell growth, survival, and protein synthesis (Irie et al., 2005; Riggio et al., 2017). GSK3β, a downstream kinase of AKT, phosphorylates Nrf2 and promotes the nuclear export of Nrf2. Phosphorylated Nrf2 that is exported to the cytosol is then degraded by the proteasomes. When AKT is activated, GSK3β is phosphorylated by AKT and becomes inactive, allowing Nrf2 to remain in the nucleus and escape degradation (Nguyen et al., 2003; Brugge et al., 2007; Carnero et al., 2008; Liu et al., 2009). An alternative Keap1-independent pathway involving WDR23 has also been reported to regulate the stability and activity of Nrf2. Lo et al. demonstrated that WDR23 regulates Nrf2 proteostasis through a WDR23–DDB1–CUL4 axis independently of the canonical Keap1–CUL3 system (Lo et al., 2017). Consistently, our previous study showed that WDR23 regulates the expression of Nrf2-driven drug-metabolizing enzymes in mammalian cells (Siswanto et al., 2020). WDR23 functions as a substrate receptor in a CUL4–DDB1 ubiquitin ligase complex (Choe et al., 2009; Lo et al., 2017), and WDR23 isoforms have been shown to differentially regulate WDR23 substrates (Spatola et al., 2019).
Sulforaphane (SFN) is a naturally occurring isothiocyanate found in broccoli sprouts and other plants. It is produced by the hydrolysis of glucoraphanin via the enzyme myrosinase. SFN possesses both antioxidant and anti-inflammatory properties and has therefore attracted considerable attention for its potential therapeutic applications (Al-Sawaf et al., 2015). It has been suggested that SFN stabilizes Nrf2 by reacting with specific reactive cysteine residues in Keap1, thereby inducing a conformational change in Keap1 (Shapiro et al., 2001; Kubo et al., 2017). However, our previous study demonstrated that chlorogenic acid activated the Nrf2 homolog SKN-1 and extended lifespan in Caenorhabditis elegans, which lacks a Keap1 homolog (Siswanto et al., 2022b).
In the present study, we investigated potential Keap1-independent mechanisms underlying SFN-induced Nrf2 upregulation and examined the effects of SFN on inflammatory gene expression using A549 cells harboring a loss-of-function Keap1 mutation and Hep3B cells expressing wild-type Keap1.
The human hepatoma cell line Hep3B was obtained from the Cell Resource Center for Biomedical Research at the Institute of Development, Aging, and Cancer of Tohoku University (Sendai, Japan). Hep3B cells were cultured in high glucose Dulbecco’s modified Eagle’s medium (DMEM) (Wako Pure Chemicals, Osaka, Japan) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Sigma-Aldrich, St. Louis, MO, USA), and 1% penicillin-streptomycin (Sigma-Aldrich). The human lung cancer cell line A549 (Resource No. RCB0098) was purchased from the RIKEN BioResource Center (Ibaraki, Japan) and cultured under the same conditions.
Treatment of cells with reagentsSFN (Cayman Chemicals, Ann Arbor, MI) was dissolved in 99.5% ethanol and added to Hep3B cells at a final concentration of 10 µM for 24 hr and to A549 cells at a final concentration of 30 µM for 8 hr. Unless otherwise noted, the experiments were conducted under these conditions for SFN. The AKT inhibitor XI (Cayman Chemicals) was dissolved in dimethyl sulfoxide (DMSO) and added to cells at a final concentration of 1 µM for 25 hr in Hep3B cells and for 9 hr in A549 cells. The inhibitor was added 1 hr prior to SFN treatment. Unless otherwise noted, the experiments involving inhibitor treatment were conducted using this method.
Overexpression ExperimentsHep3B and A549 cells (5 x 105 cells) were seeded in 6-cm dishes (Thermo Fisher Scientific, MA) to reach 70–80% confluence after 2 days and were transfected with 12 µg (Hep3B cells) or 6 µg (A549 cells) of purified plasmid using the calcium phosphate method. After transfection, the cells were cultured for 16 hr, and the medium was replaced. The cells were then cultured for an additional 24 hr before treatment with the reagents. For Nrf2 and FOXO3 overexpression, human Nrf2 cDNA and human FOXO3 cDNA were ligated into pcDNA3.1 and FLAG-pcDNA3.1 as described previously (Siswanto et al., 2022b).
Knockdown ExperimentsHuman AKT1-targeting shRNA was expressed using the pBAsi-hU6 Neo Vector (Takara Bio, Shiga, Japan), with sh-GFP used as the control. The target sequences are listed in Table 1. The plasmids were transfected using the calcium phosphate method. After 16 hr, the medium was replaced, and the cells were cultured for an additional 24 hr before SFN treatment. Hep3B cells were then treated with 10 µM SFN for 24 hr, whereas A549 cells were treated with 30 µM SFN for 8 hr.
Reverse transcription and semi-quantitative PCRTotal RNA was extracted from cells using ISOGEN (Nippon Gene, Tokyo) according to the manufacturer’s instructions, and cDNA was synthesized using Revert Aid™ M-MuLV Reverse Transcriptase (New England Biolabs, Ipswich, MA). The reverse transcriptase reaction was performed at 25°C for 15 min, 42°C for 60 min, and 70°C for 10 min. cDNA was then amplified with Go Taq polymerase (Promega, Madison, WI) and each of the primer sets listed in Table 2.
Quantification of PCR productsThe polymerase chain reaction (PCR) products were electrophoresed on a 2% agarose gel (Nippon Gene) containing TAE buffer (40 mM Tris, 40 mM acetic acid, and 1 mM EDTA) plus ethidium bromide (Wako Pure Chemicals, Osaka, Japan). Bands were visualized with a video camera (model AE-6905H Image Save HR, Atto Co., Tokyo) and quantified using ImageJ software.
ImmunoblottingImmunoblotting was performed with anti-Cullin4A (Cul4) and anti-DDB1 antibodies (Santa Cruz Biotechnology, Dallas, TX), anti-FOXO3A (Abcam, Carlsbad, CA), and anti-phosphorylated FOXO3A (p-FOXO3A) antibodies (Cell Signaling Technology, Danvers, MA). Anti-Nrf2 and anti-β-actin antibodies were prepared in our laboratory (Siswanto et al., 2022b). Cells were collected in PBS and whole-cell lysates were prepared using an ultrasonic homogenizer (BRANSON). The protein concentration of the lysates was measured using the Lowry method. The cell lysates mixed with SDS sample buffer were heated at 95°C for 5 min, followed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE). After electrophoresis, the proteins were transferred to a nitrocellulose membrane (BIO-RAD) using a semi-dry blotting system (ATTO) and transfer buffer (100 mM Tris, pH 7.5, 190 mM glycine, 5% methanol). After blotting, the membrane was blocked for 1 hr with 5% skim milk in TPBS (PBS + 0.05% Tween 20). The membrane was incubated with the primary antibody overnight at 4°C. The following day, the membrane was washed with TPBS, and the membrane was incubated with the secondary antibody for 45 min at room temperature. The membrane was then washed with PBS. The membrane was incubated with a chromogenic reagent prepared by dissolving 4-chloro-1-naphthol in methanol and adding hydrogen peroxide. The stained bands were imaged using an imager, and the band intensities were quantified using ImageJ.
For the detection of FOXO and p-FOXO, we performed blotting using a PVDF membrane. For this, the transfer buffer consisted of 0.48 M Tris, 0.39 M glycine, and methanol, and we used the same semi-dry blotting apparatus as described above. After transfer, we blocked the membrane with Blocking One-P for 1 hr at room temperature. The membrane was incubated with the primary antibody overnight at 4°C. The next day, the membrane was washed with TBS-T (0.4 M Tris, 3 M NaCl, 0.02% Tween-20). The membrane was incubated with the secondary antibody at room temperature for 45 min. After washing the membrane with TBS-T, bands were detected using chemiluminescence and visualized using an imager.
Immunofluorescence StainingFlag-FOXO3-overexpressing Hep3B and A549 cells were seeded in 3.5-cm glass-bottom dishes. For immunofluorescence analysis, both cell lines were treated with 10 µM SFN for 24 hr. AKT inhibitor XI alone was applied at 1 µM for 24 hr. For co-treatment, cells were pretreated with 1 µM AKT inhibitor XI for 1 hr before SFN addition and were then treated with 10 µM SFN for 24 hr. Cells were stained with an anti-DYKDDDDK tag antibody and DAPI, followed by Alexa Fluor 488-conjugated goat anti-mouse IgG, and observed using a confocal laser microscope (Leica SP8, Leica Microsystems).
Statistical analysisA t-test was used for the comparisons between two groups. A one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test was performed for multiple comparisons.
To confirm whether Nrf2 was appropriately induced in the mammalian cell lines used in this study, we examined the Nrf2 expression levels in the Hep3B and A549 cells after treating the cells with a Keap1 inhibitor (tertiary butylhydroquinone [tBHQ]) and SFN (Fig. 1). Hep3B cells express wild-type Keap1, whereas A549 cells harbor a Keap1 loss-of-function mutation, resulting in impaired Keap1-mediated repression of Nrf2 (Singh et al., 2006; Ohta et al., 2008; Hast et al., 2014). In the Hep3B cells, the expression of Nrf2 was significantly increased by the addition of tBHQ at a final concentration of 60 µM (Fig. 1A) and by SFN at final concentrations of 10 and 25 µM (Fig. 1B), whereas the addition of tBHQ to A549 cells at a final concentration of 60 µM did not significantly affect Nrf2 protein levels, consistent with impaired Keap1 function (Fig. 1C).

The induction of Nrf2 protein expression in Hep3B and A549 cells by tertiary butylhydroquinone (tBHQ) and sulforaphane (SFN). Hep3B cells were treated with 60 µM tBHQ for 4 hr (A) and 10 or 25 µM SFN for 24 hr (B), and A549 cells were treated with 60 µM tBHQ for 4 hr (C) and 30 or 40 µM SFN for 8 hr (D). Cell lysates were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Western blotting. Columns indicate the mean ± SD of Nrf2 protein levels (normalized by β-actin) from three independent dishes, with the control set at 1.0. *p<0.05, **p<0.01 (t-test).
Since an increase in Nrf2 was not observed in A549 cells after treatment with SFN at 10 µM for 24 hr (data not shown), we examined different treatment conditions and found that treatment of A549 cells with 30 µM SFN for 8 hr significantly increased Nrf2 protein levels despite the presence of a Keap1 loss-of-function mutation (Fig. 1D). These results suggest that a Keap1-independent mechanism may contribute to SFN-induced Nrf2 upregulation in A549 cells.
The inhibition of SFN-induced Nrf2 expression by treatment with inhibitors of PI3K and AKTBecause we previously reported a Keap1-independent Nrf2 regulatory pathway involving the AKT/FOXO/DDB1 axis (Siswanto et al., 2022b), we first examined whether PI3K/AKT signaling contributes to SFN-induced Nrf2 upregulation. To explore the potential signaling pathways involved in Keap1-independent Nrf2 regulation, Nrf2 expression levels were examined in Hep3B and A549 cells treated with SFN in the presence or absence of 5 µM PI3K inhibitor LY294002 (Fig. 2) or 1 µM AKT inhibitor XI (Fig. 3). Both PI3K inhibition and AKT inhibition significantly reduced SFN-induced Nrf2 protein levels in Hep3B and A549 cells (Figs. 2 and 3). These results suggest that PI3K and AKT signaling contribute to SFN-induced Nrf2 upregulation.

Nrf2 expression in Hep3B and A549 cells treated with the PI3K inhibitor LY294002 in the presence of SFN. Hep3B cells were treated with 5 µM LY294002 and 10 µM SFN for 24 hr (A), and A549 cells were treated with 5 µM LY294002 and 30 µM SFN for 8 hr (B). The inhibitor was added 1 hr before the addition of SFN. Cell lysates were used for the immunoblotting of Nrf2 and β-actin. Columns indicate the mean ± SD of Nrf2 protein levels (normalized by β-actin) from three independent dishes, with the control set at 1.0. *p<0.05, **p<0.01 (Tukey’s test).

Nrf2 expression in Hep3B and A549 cells treated with the AKT inhibitor XI in the presence of SFN. Hep3B cells were treated with 1 µM AKT inhibitor XI and 10 µM SFN for 24 hr (A), and A549 cells were treated with 1 µM AKT inhibitor XI and 30 µM SFN for 8 hr (B). The inhibitor was added 1 hr before the addition of SFN. Cell lysates were used for the immunoblotting of Nrf2 and β-actin. Columns indicate the mean ± SD of Nrf2 protein levels (normalized by β-actin) from three independent dishes, with the control set at 1.0. *p<0.05, **p<0.01 (Tukey’s test).
To further examine the involvement of AKT suggested by the inhibitor experiments, we evaluated SFN-induced Nrf2 expression after transfection with an shRNA designed to target AKT1 in Hep3B and A549 cells. Although the shRNA was designed against AKT1, the expression levels of both AKT1 and AKT2 were decreased (Fig. 4). This reduction in AKT2 may have resulted from an off-target effect due to sequence similarity between the shRNA target sequence and AKT2 mRNA, or alternatively, from a secondary effect of AKT1 knockdown on AKT2 expression. In sh-GFP-transfected control cells, SFN treatment increased Nrf2 expression in both Hep3B and A549 cells. Consistent with the AKT inhibition experiment shown in Fig. 3, AKT1-targeting shRNA significantly attenuated the SFN-induced increase in Nrf2 expression in both cell lines (Fig. 4). In A549 cells transfected with AKT1-targeting shRNA, Nrf2 expression did not differ significantly between the vehicle- and SFN-treated groups.

Effect of AKT1 knockdown on SFN-induced Nrf2 expression in Hep3B and A549 cells. Hep3B (A) and A549 (B) cells were transfected with sh-GFP or AKT1-targeting shRNA and treated with 10 µM SFN for 24 hr (A) and 30 µM SFN for 8 hr (B). The levels of mRNA of AKT1 and AKT2 were assessed by semi-quantitative RT-PCR. Nrf2 and β-actin proteins were detected by immunoblotting. Data are shown as the mean ± SD of three independent dishes, with Nrf2 levels normalized to β-actin and the corresponding control set to 1.0. **p<0.01 for the indicated comparisons; N.S., not significant (one-way ANOVA followed by Tukey’s post hoc test).
Because we previously reported a Keap1-independent Nrf2 regulatory pathway involving the AKT/FOXO/DDB1 axis (Siswanto et al., 2022b), we examined total FOXO3A and phosphorylated FOXO3A (p-FOXO3A) in whole-cell lysates under the present experimental conditions (Fig. 5). In A549 cells, SFN treatment was accompanied by an apparent increase in the p-FOXO3A signal without a clear change in total FOXO3A, whereas co-treatment with AKT inhibitor XI appeared to attenuate this change. In Hep3B cells, SFN treatment tended to decrease FOXO3A and p-FOXO3A signals, whereas co-treatment with AKT inhibitor XI appeared to restore these signals. We confirmed that SFN treatment decreased nuclear FOXO3 levels, which were attenuated by treatment with an AKT inhibitor in both Hep3B and A549 cells, (Fig. S2).

The phosphorylation of FOXO3A in Hep3B and A549 cells treated with SFN and/or the AKT inhibitor XI. A549 cells were treated with 1 µM AKT inhibitor XI and 30 µM SFN for 8 hr, and Hep3B cells were treated with 1 µM AKT inhibitor XI and 10 µM SFN for 24 hr. The inhibitor was added 1 hr before the addition of SFN. FOXO3A and phosphorylated FOXO3A (p-FOXO3A) were separated by SDS-PAGE and detected by immunoblotting using chemiluminescence. The arrow indicates phosphorylated FOXO3A. Representative immunoblots from two independent experiments are shown. No densitometric quantification or statistical analysis was performed for this figure.
In mammalian cells, FOXO has been shown to regulate the transcription of DDB1, an adaptor component of the CRL4 ubiquitin ligase complex associated with WDR23-mediated substrate recognition (Siswanto et al., 2022b). We therefore examined the protein expression levels of DDB1 and CUL4, which are components of the CUL4–DDB1–RBX1 E3 ubiquitin ligase complex (Fig. 6). SFN treatment significantly reduced the expression of both DDB1 and CUL4 in Hep3B cells (Figs. 6A and 6B) and A549 cells (Figs. 6C and 6D).

DDB1 and CUL4 protein levels in Hep3B and A549 cells treated with SFN. Hep3B cells were treated with 10 µM SFN for 24 hr, and DDB1 and CUL4 protein levels were determined by immunoblotting (A, B, respectively). A549 cells were treated with 30 µM SFN for 8 hr, and DDB1 and CUL4 protein levels were determined by immunoblotting (C, D, respectively). Data are presented as the mean ± SD of three independent dishes, with protein levels normalized to β-actin and the control set at 1.0. *p<0.05, **p<0.01 (t-test).
Previous studies have reported anti-inflammatory effects associated with Nrf2 activation by phytochemicals, including SFN (Dokumacioglu et al., 2017; Ma et al., 2023; Liu et al., 2023). However, the extent to which Nrf2 contributes to the anti-inflammatory effects of SFN remains unclear. We therefore examined the effect of SFN on LPS-induced expression of IL-1β, an NF-κB-responsive inflammatory gene (Fig. 7). SFN significantly induced HO-1 gene expression, a representative Nrf2 target gene, in both Hep3B and A549 cells, whereas LPS alone did not (Figs. 7A and 7B). LPS increased IL-1β gene expression in both Hep3B and A549 cells, and SFN reduced this LPS-induced IL-1β gene expression (Figs. 7C and D). In addition, Nrf2 overexpression in Hep3B cells was associated with a decrease in IL-1β gene expression and an increase in HO-1 gene expression (Suppl. Fig. S1).

Effect of SFN on LPS-induced IL-1β gene expression in Hep3B and A549 cells. Hep3B cells were treated with 200 ng/mL LPS and 10 µM SFN for 24 hr, followed by PCR to detect HO-1 mRNA (A) and IL-1β mRNA (C). A549 cells were treated with 200 ng/mL LPS and 30 µM SFN for 8 hr, followed by PCR to detect HO-1 mRNA (B) and IL-1β mRNA (D). LPS was added 1 hr before the addition of SFN. Data are presented as the mean ± SD of three independent dishes. Expression levels were normalized to β-actin, with the control value set to 1.0. *p<0.05, **p<0.01 (one-way ANOVA followed by Tukey’s post hoc test).
SFN is electrophilic and reacts with thiols on the cysteine residues of Keap1, changing its conformation and stabilizing Nrf2 (Zhang and Hannink, 2003; Wakabayashi et al., 2004; Cleasby et al., 2014). We demonstrated that SFN treatment significantly increased Nrf2 protein levels in A549 cells, in which Keap1 is mutated at the Nrf2-binding site (Siswanto et al., 2022a). Since these findings suggest that SFN regulates Nrf2 protein levels through pathways other than the Keap1 pathway, the primary objective of the present study was to investigate novel mechanisms for the induction of Nrf2 by SFN. We focused on PI3K/AKT signaling because co-treatment with either PI3K inhibitor, LY294002 or AKT inhibitor XI significantly attenuated the SFN-induced increase in Nrf2 protein levels in both Hep3B and A549 cells. These findings suggest that PI3K/AKT signaling contributes to SFN-induced Nrf2 upregulation. Consistent with the inhibitor experiments, transfection with AKT1-targeting shRNA attenuated the SFN-induced increase in Nrf2 expression in both cell lines (Fig. 4), that is, further supporting a contribution of AKT signaling to SFN-induced Nrf2 upregulation. Because the AKT1-targeting shRNA also reduced AKT2 expression, the knockdown experiment does not distinguish the contributions of individual AKT isoforms. Furthermore, these findings do not distinguish the FOXO/WDR23-related pathway from other AKT-dependent mechanisms, including the GSK3β pathway.
In the present study, SFN decreased both DDB1 and CUL4 protein levels in Hep3B and A549 cells. This finding is consistent with the possibility that SFN affects the WDR23-associated CRL4 machinery, because WDR23 has been reported to regulate Nrf2 through a DDB1-CUL4 axis independently of Keap1 (Lo et al., 2017), and CRL4 complexes are composed of a CUL4 scaffold, the RING protein RBX1, DDB1 as an adaptor, and DCAF substrate-recognition proteins (Jackson and Xiong, 2009; Bosu and Kipreos, 2008). However, the mechanism underlying the decrease in CUL4 protein should be interpreted cautiously. In our previous study, chlorogenic acid activated Nrf2/SKN-1 through the AKT-FOXO3/DAF16a-DDB1 pathway (Siswanto et al., 2022b). Importantly, that study showed that DDB1 mRNA, but not CUL4 mRNA, was decreased after chlorogenic acid treatment, suggesting that CUL4 is unlikely to be regulated in the same manner as DDB1 through the FOXO-dependent transcriptional pathway (Siswanto et al., 2022b). In addition, CUL4A or CUL4B transcription has been reported to be regulated by several transcriptional mechanisms in specific cellular contexts, including CREB, the CARM1-p300-c-Myc-Max transcriptional complex, HIF-1α, and the ERα–GATA3 axis (Ashok et al., 2019; Ashok et al., 2020; Lu et al., 2020; Huang et al., 2021). Therefore, the decrease in CUL4 protein observed in the present study should not be interpreted as direct FOXO-dependent transcriptional repression. Since CRL activity is regulated not only by the abundance of cullin proteins but also by complex-level and post-translational mechanisms, including cullin neddylation/deneddylation and COP9 signalosome-mediated regulation (Bosu and Kipreos, 2008), the decrease in CUL4 protein may represent a secondary, post-transcriptional, or post-translational change associated with altered CRL4 regulation. Such regulation may be biologically meaningful because CRL4 complexes control the stability of multiple substrates involved in cell-cycle progression, DNA damage responses, chromatin regulation, and stress-related processes (Jackson and Xiong, 2009). Further studies examining CUL4A mRNA expression, CUL4 protein turnover, neddylated CUL4 levels, and WDR23-DDB1-CUL4 complex formation will be required to clarify how SFN reduces CUL4 protein expression.
The second objective of the present study was to examine the effect of SFN on inflammatory gene expression. SFN reduced LPS-induced IL-1β mRNA expression in both Hep3B and A549 cells. In addition, Nrf2 overexpression in Hep3B cells was associated with lower IL-1β mRNA expression (Suppl. Fig. S1), suggesting that Nrf2 upregulation may contribute, at least in part, to the suppression of IL-1β expression by SFN. Previous studies have shown that Nrf2 can suppress the transcription of proinflammatory cytokines, including IL-1β (Kobayashi et al., 2016), which is consistent with our findings. However, the present data do not establish a causal link between the AKT–FOXO–WDR23-related changes and the suppression of IL-1β expression. Further studies examining whether disruption of this pathway attenuates the inhibitory effect of SFN on IL-1β expression will be required to clarify its contribution to the anti-inflammatory effects of SFN.
Although Nrf2 activation is generally cytoprotective under oxidative or inflammatory stress, excessive or sustained Nrf2 activation may not always be beneficial. Previous studies have shown that persistent Nrf2 activation can induce reductive stress and contribute to pathological remodeling in the heart, and that constitutive Nrf2 activation in cancer cells can support metabolic adaptation and therapy resistance (Rajasekaran et al., 2011; Kannan et al., 2013; Wu et al., 2019; Weiss-Sadan et al., 2023). Therefore, the biological effects of SFN should be interpreted in a context- and dose-dependent manner. In the present study, SFN treatment was associated with Nrf2 induction and reduced LPS-induced IL-1β expression; however, we did not examine markers of reductive stress or the long-term effects of Nrf2 activation. Further studies will be required to determine the effects of prolonged or excessive Nrf2 activation under SFN treatment.
The present findings support a contribution of PI3K/AKT signaling to SFN-induced Nrf2 upregulation and are consistent with the possible involvement of FOXO/WDR23-related regulation. SFN-induced Nrf2 upregulation was associated with reduced LPS-induced IL-1β expression; however, the present data do not establish that the AKT–FOXO–WDR23 pathway causally mediates the anti-inflammatory effect of SFN.
The authors thank Dr. Masako Tanaka for insightful suggestions and discussions.
FundingThis study was supported by a grant from the Japan Society for the Promotion of Science (no. 17K08581 to S.I.) and a Grant-in-Aid from Kwansei Gakuin University (to S.I.).
Conflict of interestThe authors declare that there is no conflict of interest.
Data availabilityThe datasets generated and/or analyzed in this study are available from the corresponding author upon reasonable request.
Author contributionsS.I. designed and coordinated the study. M.H., H.D. and A.T. performed experiments and data analyses. A.O. and Y.K. provided resources. M.H., A.T. and S.I. wrote the manuscript. All authors contributed to the review and editing of the final manuscript.
Ethical approval and consent to participateNot applicable.
Patient consent for publicationNot applicable.