Many previous studies have shown that MAPKs play a role in the induction of Nrf2. summarizes key findings of all previous studies that support the involvement of MAPKs in Nrf2 activation. Manipulation of the catalytic activity of MAPKs, either through pharmaceutical intervention (kinase inhibitors) or genetic engineering (overexpression of wild-type or dominant-negative kinases, siRNA-mediated knockdown, or gene knockout), coupled with a readout of Nrf2 activity (ARE-dependent luciferase reporter gene, or Nrf2 nuclear accumulation), was used as the major approach to address the potential role of MAPKs on Nrf2 (). All three well-characterized categories of MAPKs (ERKs, JNKs, and p38) have been implicated in modulating Nrf2, with some contradictory results among different groups. Since MAPKs regulate bewilderingly diverse cellular programs, it is not clear whether the observed effects of MAPKs on Nrf2 are highly specific or just bystander effects. It remains elusive whether MAPKs regulate Nrf2 through direct phosphorylation of Nrf2 or through indirect mechanisms. Furthermore, it is not known whether or not MAPKs phosphorylate Nrf2 in vivo at all. Another key question that remains open is whether MAPKs affect Nrf2 protein stability that is primarily controlled by Keap1. Therefore, the exact role of MAPKs in Nrf2 activation, as well as the underlying molecular mechanism, remains poorly understood.
In this study, we report the first evidence that Nrf2 is phosphorylated by MAPKs in vivo. We have identified the major phosphorylation sites and used site-mutagenesis to decipher the function of direct phosphorylation of Nrf2 by MAPKs. As demonstrated by data obtained from ARE-dependent luciferase reporter gene assay and qRT-PCR analysis, a loss of Nrf2 phosphorylation caused only moderate changes in the transcriptional activity of Nrf2 ( and ). The Keap1-mediated control of Nrf2 protein levels was not affected by Nrf2 phosphorylation in vivo under both basal and induced conditions (), nor was the interaction between Nrf2 and Keap1 (). However, the nuclear accumulation of Nrf2 was slightly enhanced by its phosphorylation (). We concluded that direct phosphorylation of Nrf2 by MAPKs has a limited contribution in regulating the Nrf2-dependent antioxidant responses.
We are aware of the fact that our mass spectrometry analysis may not have identified all phosphorylated residues on Nrf2. There might be signal-induced Nrf2 phosphorylation on specific sites that were missed in the current identification procedure whereby overexpressed Nrf2 protein itself was purified from HEK293T cells under basal conditions followed by LC-MS/MS. However, the fact that the Nrf2-5A mutant has significantly decreased phosphorylation levels compared to wild-type Nrf2 in the presence of overexpressed JNK2 suggests that these five residues are the major targets of phosphorylation in vivo (). Neither inducible ARE-dependent transcription nor inducible Nrf2 protein levels was dramatically altered by combined mutations on the identified phosphorylation sites in the presence of several Nrf2 inducers such as tBHQ, sulforaphane, PGJ2, arsenite or cadmium, suggesting that phosphorylation at these sites has limited contribution in regulating inducible Nrf2 signaling (). Additionally, in an experiment without using mutations, overexpression of MAPKs failed to cause any observable effect on the interaction between wild-type Nrf2 and Keap1 (, lane 1–5), although Nrf2 phosphorylation was significantly enhanced under the same conditions (). Collectively, these data suggest that direct phosphorylation of Nrf2 plays a limited role in regulating Nrf2 activity.
Consistent with our data, Zipper and Mulcahy showed that mutational disruption of six MAPK consensus sites in Nrf2 (including S215, S408, S577 that were also characterized in this study) failed to significantly change the GCLM ARE-dependent reporter gene activities in the absence or presence of pyrrolidine dithiocarbamate (PDTC), an Nrf2 inducer 
. Here we provide additional evidence that mutational disruption of these sites does not dramatically alter the endogenous mRNA levels of NQO1, HO-1 or GCLM (), excluding possibilities that phosphorylation might interfere with molecular events that are specific to native chromosomal architectures. In addition, we showed that neither Nrf2 protein levels nor ARE-dependent luciferase activities were dramatically altered by the combined mutation in cells either left untreated or treated with divergent Nrf2 inducers, such as tBHQ, sulforaphane, PGJ2, arsnite, oridonin, cadmium and H2
(), making it a much less attractive hypothesis that phosphorylation might be involved in the action of some particular Nrf2 inducers, if not all. Furthermore, we performed a co-immunoprecipitation assay and directly demonstrated that the physical interaction between Nrf2 and Keap1 was not altered in vivo
in the presence of overexpressed MAPKs or with Nrf2 phosphorylation-deficient mutants (). Another previous report from Kong and colleagues also showed that site-directed mutagenesis of Nrf2 at several MAPK consensus sites did not affect the transactivation activity of Nrf2, although these sites do not overlap with the sites characterized in our present study 
. Taken together, MAPK-mediated regulation of the Nrf2 signaling pathway is not likely to be through direct phosphorylation of Nrf2.
One possible indirect mechanism is translational control on Nrf2 protein synthesis by MAPKs. MAPKs have been shown to modulate mTOR signaling pathways in controlling eukaryotic initiation factor 4 (eIF4) complex assembly, a critical step in Cap-dependent translational initiation, through inhibiting tuberous sclerosis complex 2 (TSC2) and activating 90-kD ribosomal protein S6 kinases (RSKs) 
. MAPKs also promote phosphorylation of eIF4E, eIF4B-BP1, and translation elongation factor 2 kinase (eEF2 kinase) through MAPK-interacting kinase 1 and 2 (Mnk1/2), mitogen- and stress-activated protein kinase 1 (Msk1), and MAPK-activated kinase 2/3 (MK2/3), respectively, which is thought to enhance translation of some mRNAs 
. On the other hand, Nrf2 has been shown to be regulated at the translational level under several conditions 
. Thus, MAPKs-mediated indirect control of Nrf2 protein synthesis, along with the moderate role of direct phosphorylation on Nrf2, may represent the underlying mechanism by which MAPKs regulates the Nrf2 signaling pathway.