To confirm the consistency of the Nrf2 induction potency of the plant-derived compounds evaluated using the two assays (ARE luciferase reporter and the conventional NQO1 assay), the CD values obtained for the 12 plant-derived compounds using these assays were compared. Most compounds showed an overall consistency, although several compounds did not. Therefore, although the NQO1 assay can be used as an evaluation method to estimate the Nrf2 induction potency of a compound, a reporter assay may offer greater precision.
A relatively high consistency was observed for isothiocyanates, such as sulforaphane, erucin, and benzyl isothiocyanate. In contrast, no consistency was observed for several of the compounds evaluated, including resveratrol and kaempferol. The reporter assay directly assesses the potency of a compound in promoting the nuclear accumulation of Nrf2, whereas the NQO1 assay indirectly assesses the Nrf2 induction potency by measuring the enzymatic activity of NQO1, which is induced by Nrf2 in response to oxidative stress or exposure to electrophilic compounds. NQO1 is also known to be induced by other transcription factors, such as the aryl hydrocarbon receptor (AhR) [25]. Among the 12 compounds, resveratrol and genistein have previously been reported to induce AhR at > 10 µM, while curcumin, quercetin, and myricetin have also been reported to slightly induce AhR at > 100 µM [26, 27]. Since the concentrations at which Nrf2 induction was observed for curcumin, quercetin, and myricetin (CD values shown in Table 1) were below the concentration at which AhR is induced (> 100 µM), it can be assumed that the NQO1 or Nrf2 induction potency evaluated in this study was not affected by AhR. The results of the NQO1 assay for resveratrol and genistein were assumed to be affected by both Nrf2 and AhR, and the CD values of the NQO1 assay were expected to be lower than those of the reporter assays. Contrary to this expectation, the CD values of the NQO1 assay were higher than those of the reporter assays. In particular, resveratrol showed the largest difference in CD values between the reporter and NQO1 assays. Previous reports showed that the NQO1 induction potency of resveratrol was CD > 50 µM (induction potency expressed as fold change converted to CD value) [21] and CD > 100 µM (induction potency expressed as nmol/min/mg protein converted to CD value) [19], suggesting that the NQO1 induction potency of resveratrol is not particularly high. Although the precise reason for this is unknown, it may be due to the influence of the 20S proteasome, which contributes to NQO1 degradation. NQO1 is degraded by the 20S proteasome at low concentrations of flavin adenine dinucleotide (FAD) [28], and the 20S proteasome can be activated by low molecular weight compounds [29]. Although the effects of resveratrol on the FAD concentration or the 20S proteasome remain unclear, our results suggest that one reason for the low NQO1 induction potency of resveratrol is the fact that resveratrol induces NQO1 via Nrf2 and AhR, while simultaneously decreasing FAD concentrations. This may in turn promote the degradation of NQO1 by the 20S proteasome. However, further studies are needed to confirm this hypothesis. In conclusion, the inconsistency in Nrf2 induction potency evaluated by the reporter and NQO1 assays for some plant-derived compounds, including resveratrol, may be due to multiple factors regulating NQO1 expression and activity other than Nrf2, and each compound has a different degree of effect on these factors. Further research is needed because the factors that regulate NQO1 expression and activity, and the effects of each plant-derived compound on each of these factors are not yet known.
Many of the compounds in the upper ranks were electrophilic, including isothiocyanates (sulforaphane and erucin), triterpenoids (withaferin A), diterpenoids (andrographolide and tanshinone IIA), and curcuminoids (curcumin). Isothiocyanates and α, β-unsaturated carbonyl groups can promote the nuclear accumulation of Nrf2 by Michael addition reactions to the thiol group of Keap1, forming covalent bonds with the cysteine residue of Keap1 [20, 30–32]. Many of the lower-ranked compounds were glycosides, such as naringin, ginsenoside Rd, and mangiferin. The NQO1 induction potency of flavonoids is known to be weakened when in glycoside form [33], which supports the low Nrf2 induction potency of various glycosides observed in the present study. Among the flavonoids, luteolin and apigenin tended to show a relatively high Nrf2 induction potency, whereas kaempferol, morin, myricetin, and epigallocathechin-3-gallate showed a relatively low potency. Thus, it is possible that the 3-C hydroxyl modification of the flavone backbone, a structure common to flavonols and flavanols, may be an obstacle to their interaction with Keap1, which may reduce their Nrf2 induction potency. Several induction and activation mechanisms of Nrf2 by flavonoids, other than its interaction with Keap1, have been reported, such as Nrf2 phosphorylation by kinases, including ERK1/2, Akt, and p38MAP kinase [34]. Therefore, each chemical structure of the compounds may lead to differences in the primarily activated pathway, depending on various factors (i.e., electron affinity, hydrophobicity, and molecular size). However, further studies are needed to elucidate the details of this hypothesis.
Limitations
Since only a limited number of studies were compared in this report, the consistency of Nrf2 induction potencies of plant-derived compounds evaluated by the reporter and NQO1 assays has not been fully investigated. In addition, the NQO1 induction potency has been reported to vary between cell types, which this study did not consider [11]. Despite these limitations, to the best of our knowledge, this study is the first to examine the consistency of Nrf2 induction potency evaluated by reporter and NQO1 assays for multiple compounds. Future comparisons of CD values by each assay across cell types may be used to demonstrate consistency between the assays, as well as reveal those factors that influence Nrf2 induction potency.