SEARCH

SEARCH BY CITATION

Keywords:

  • antioxidant response element;
  • Bach1;
  • small maf proteins;
  • CREB-binding protein;
  • glutamate cysteine ligase catalytic subunit

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. Author contributions
  9. References
  10. Supporting Information

NFE2-related factor 2 (Nrf2) transcriptionally governs the cellular response to harmful electrophiles, xenobiotics, and reactive oxygen species. Its nuclear levels decline with age (Suh et al., 2004a), which in part explains the age-related loss of phase II detoxification. However, little work has yet characterized how age affects Nrf2 DNA binding or the role that alterations to the Nrf2 transcriptional apparatus plays in modulating Nrf2-mediated gene expression. In this study, we used immunoprecipitation assays to show that Nrf2 bound to the active antioxidant response element (ARE) of the catalytic subunit of glutamate cysteine ligase (GCLC) is significantly lower in hepatic chromatin from aged vs. young rats. Moreover, the activity at this ARE locus is diminished during aging because of the presence of Bach1 and the absence of CREB-binding protein (CBP), a transcriptional repressor and co-activator, respectively. Further analysis reveals that Nrf2 occupies an alternate ARE site located −2.2 kb downstream from the normally active ARE binding site in livers of old rats, indicating an age-specific adaptation to maintain gene expression. Our results, thus, show that the conversion of Nrf2 binding from an active ARE to an alternative ARE element is not adequate to maintain basal expression of hepatic Gclc in old rats, which provides a potential mechanism for the age-related loss of glutathione synthetic and other phase II enzymes.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. Author contributions
  9. References
  10. Supporting Information

A hallmark of the aging process is a decreased response to both exogenous and endogenous stresses (Finkel & Holbrook, 2000). In the liver, this inadequate stress response is reflected by a lower detoxification capacity (Hagen et al., 2002), which is exemplified by diminished levels of certain low molecular weight antioxidants like glutathione (GSH) and related phase II detoxification and antioxidant enzymes (Suh et al., 2004a,b). In this context, we and others previously showed that GSH levels decline in the liver of aging Fischer 344 rats (Hall et al., 2001; Suh et al., 2004a,b), in part from a loss in GSH synthesis.

The de novo synthesis of GSH from its constituent amino acids involves two ATP-requiring enzymatic steps: the formation of γ-glutamylcysteine, followed by its conjugation to glycine (Huang et al., 1993a,b). Glutamate cysteine ligase (GCL) catalyzes the first and rate-limiting step of de novo synthesis, making it a major determinant of overall GSH synthetic capacity (Huang et al., 1993a,b; Lu et al., 1999). Data from many laboratories conclusively show that hepatic activity, protein levels, and gene expression of GCL are significantly lower in aging rat liver (Stio et al., 1994; Suh et al., 2004a). The GCL protein is a heterodimer that can be dissociated under nondenaturing conditions into a catalytic (GCLC, 73 kDa) and a modulatory (GCLM, 29 kDa) subunit, which are encoded by separate genes (Lu, 2009). Although the heavy subunit contains the active site, its association with GCLM, which is present in far less amounts than GCLC, modulates the overall activity of the enzyme (Lee et al., 2006; Lu, 2009). Although GCLC is elaborately regulated at kinetic, post-translational and transcriptional levels (Toroser et al., 2006), its transcriptional regulation produces a more persistent effect, and thus is more important for the maintenance of GSH homeostasis in response to oxidative stress (Zipper & Mulcahy, 2000; Song et al., 2005; Yang et al., 2005; Shenvi et al., 2009).

Previously, we showed that the age-related loss of rat hepatic GCL is linked to lower nuclear steady-state levels of the transcription factor, NFE2-related factor 2 (Nrf2) (Suh et al., 2004a,b). Coincident with lower nuclear Nrf2 is diminished binding of the transcription factor to its cognate DNA sequence, the antioxidant response element (ARE), in the 5′-flanking region of Gclc. Although, lower steady-state nuclear Nrf2 levels seem likely to be causally linked to the age-associated decline in GSH, the exact consequences of Nrf2 loss on the Gclc transcriptome still remains to be elucidated.

In this study, we used Gclc as a representative Nrf2-/ARE-mediated gene to study the transcriptional mechanism of age-related deficiency in GSH synthesis. We previously showed that the 5′-flanking region of the rat Gclc gene has three ARE elements, but only one of these sequences displays Nrf2-binding and transcriptional activity (Shenvi et al., 2009). Therefore, we hypothesized that in aged rats, there is lower Nrf2 binding to the Gclc, which leads to a transcriptional remodeling and contributes to a decline in gene expression.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. Author contributions
  9. References
  10. Supporting Information

Nuclear steady-state levels of Nrf2 decline with age

Nuclear and cytosolic Nrf2 levels were measured to establish the extent of age-related changes in this key transcription factor. Compared with hepatocytes from young rats, those from old animals exhibited a significant 50 ± 17% decline (P ≤ 0.05) in steady-state nuclear Nrf2 levels (Fig. 1). Although results show that there is an age-associated change in cytosolic Nrf2 compared with the nuclear fraction, the change was not found to be significant by a Student’s t-test. These results show that the age-related loss of Nrf2 is primarily restricted to the nuclear compartment.

image

Figure 1.  Nuclear levels of Nrf2 decline with age. (A) The nuclear fraction of hepatocytes was isolated and subjected to Western blot analysis for Nrf2 levels. (B) Results show that nuclear Nrf2 levels decline significantly by 50% (*P ≤ 0.05) with age. Blots are representative of N = 3 (young) and N = 4 (old) animals.

Download figure to PowerPoint

Nrf2 enrichment at ARE4 is lower in hepatocytes from old rats

Figure 2(A) shows a schematic representation of the putative ARE elements present in the 5′-untranslated region of Gclc. Only one (‘ARE2’) of the four ARE sequences that is located 2.1 kb upstream from the transcriptional start site does not contain all the elements that are characteristics of typically active AREs. To determine the consequences of lower nuclear Nrf2 levels on binding of this transcription factor to individual ARE promoters, chromatin immunoprecipitation (ChIP) assays were performed using an Nrf2 antibody. While Nrf2 does not bind to the two sequences located 1.0 kb and 3.1 kb upstream of the transcriptional start site (designated: ‘ARE1’ and ‘ARE3’, respectively) in hepatocytes isolated from both young and old rats (Shenvi et al., 2009), this transcription factor associates with the fourth ARE (‘ARE4’) 3.9 kb upstream of the transcriptional start site. Nrf2 binding to ARE4 decreases significantly (59 ± 17%, P ≤ 0.05) in old rat hepatocytes (Fig. 2B,C and Table 1) relative to young controls. Thus, the loss of Nrf2–ARE4 binding in old rat hepatocytes is consistent with lower nuclear Nrf2 levels.

image

Figure 2.  Aging induces promoter shifting of Nrf2 from ARE4 to an alternate antioxidant response element (ARE) site (ARE2). (A) Schematic representation of 5′-flanking region of Gclc showing locations of the four ARE promoters. (B) Amount of Nrf2 binding to Gclc ARE4 declines with age, while Nrf2 binds to the ARE2 promoter only in hepatocytes from old rats. ChIP assays were performed on young and old rat hepatocytes using an antibody to Nrf2 and amplifying the region spanning the ARE4 and ARE2 using specific primers. The sheared input chromatin is used as a positive control. (C) Quantification of immunoprecipitated chromatin by qRT-PCR shows that Nrf2 binding to ARE4 is 59 ± 17% (P ≤ 0.05) lower in liver of old rats as compared to young. Nrf2 is also more enriched at the ARE2 promoter (50 ± 3; P < 0.05) compared with the ARE4 promoter with age. No Nrf2 was detected at the ARE2 promoter in young rat hepatocytes, showing that the promoter switching was age specific. Results are representative of ChIPs performed on hepatocytes from three young and three old animals.

Download figure to PowerPoint

Table 1.   Age-related alterations in transcription factor binding to Gclc ARE4 and ARE2
Pull downARE 4ARE 2
Fold change (from young control)Fold change (from young control)
  1. Figures marked with a star and bold values are calculated as statistically different by student’s t-test (P < 0.05).

  2. ARE, antioxidant response element; CBP, CREB-binding protein.

Nrf20.41*49.83*
mafK/G0.860.92
c-jun0.04*0.38*
c-fos0.12*0.5*
CBP0.02*1.12
Bach149.48*1.61
IgG1.361.47
Fra-10.641.02

We further asked the question whether the ARE2 element displays any Nrf2-binding activity. ChIP analysis revealed no detectable binding of Nrf2 to the ARE2 promoter in hepatocytes from young rats (Fig. 2B,C and Table 1), indicating that ARE4 likely mediates the increase in Gclc expression by Nrf2. In contrast, Nrf2 was reproducibly associated with the ARE2 sequence in chromatin isolated from old rat hepatocytes (Fig. 2B,C and Table 1). Taken together, these results demonstrate that Nrf2–ARE4 complex formation is attenuated in the aged rat liver, but there is an age-specific recruitment of Nrf2 to an alternate ARE promoter in Gclc.

Aging induces transcriptional remodeling of the Gclc/ARE4 complex

While the age-related loss of nuclear Nrf2 levels may be responsible for Nrf2 binding to the ARE4 site, it cannot explain why Nrf2 is partially re-directed in aged tissue to the ARE2 element, which is neither transcriptionally active nor contains the core Nrf2 binding sequence. Thus, the simple decline in steady-state nuclear Nrf2 levels may not completely address the age-associated loss in Gclc transcriptional activity. We hypothesized that, along with lower Nrf2 associated with ARE4, a repressive-type of transcriptional protein complex develops at that site, while Nrf2 binding at the ARE2 locus may be a compensation for this loss. Initially, this hypothesis was explored by using ChIP assays to identify Nrf2 partner proteins that bind to both the ARE4 and ARE2 sites in chromatin from young and old rat hepatocytes. Table 1 shows that in the young rat liver, along with Nrf2, small maf proteins were present at ARE4. However, because no antibodies specific to mafG, mafK, or mafF are commercially available, it was not possible to identify the precise small maf protein partnering with Nrf2. In addition, c-Jun and c-Fos, which are known to bind and recognize the ARE, were also detected at this locus. Finally, the histone acetyltransferase, CREB-binding protein (CBP), is also present on ARE4 in hepatocytes from young rats (Table 1 and Fig. 3). While these ChIP assays cannot absolutely quantify a given amount of a transcription factor because of differences in antibody-binding affinities, relative quantification by qPCR analysis shows that the ARE4 in cells from young rats is transcriptionally active because only permissive binding partners with Nrf2 are present at significant levels above a nonspecific IgG protein (see Table 1). Thus, transcription factors that would permit Gclc expression reside at the ARE4 locus in hepatocytes from young rats.

image

Figure 3.  Aging induces loss of CREB-binding protein (CBP) binding to the active Gclc ARE4 transcriptome. ChIP assays performed on hepatocytes from young rats show the presence of the histone acetyltransferase, CBP at the ARE4 promoter with Nrf2. With age, CBP binding to the ARE4 drops below that of the nonspecific IgG control (see Table 1), indicating the absence of this key co-activator at the ARE4 locus. Results are representative of ChIPs performed on hepatocytes from three young and three old rats.

Download figure to PowerPoint

Using a similar strategy to map age-related changes to the ARE4 locus, we found that the ARE4 transcriptome of Gclc in hepatocytes from old rats is significantly altered relative to young rats (Table 1). First, chromatin from aged cells shows the presence of Bach1, a known transcriptional repressor of ARE-mediated gene expression (Dhakshinamoorthy et al., 2005; Reichard et al., 2007; Kaspar & Jaiswal, 2010). Bach1 has a DNA-binding motif and associates with the ARE promoter when Nrf2 is limiting, but has no trans-activating domain. Thus, the presence of Bach1 at ARE4 suggests a repressive transcriptional motif developed on the Gclc promoter with age.

Also, in sharp contrast to young hepatocytes, ChIP analysis for the presence of CBP at the ARE4 locus revealed a complete lack of amplification for this key transcriptional co-activator at ARE4. These results suggest no co-activation of Nrf2 at this site (Table 1 and Fig. 3). Taken together, we interpret the observed age-associated changes to the Nrf2–ARE transcriptome to mean that a repressive transcriptional environment occurs at the Gclc-ARE4 site. These results correlate with our data showing an age-related loss in Nrf2 binding to the ARE and also reflect diminished Gclc expression shown previously (Suh et al., 2004a).

Nrf2 occupies ARE2 in old rats along with AP1 transcription factors

Further ChIP experiments were undertaken to understand whether the age-specific formation of the Nrf2–ARE2 complex was accompanied by a permissive transcription factor profile. Table 1 shows that, of the transcription factors examined in young rats, the ARE2 motif binds only c-Jun and c-Fos, not Nrf2. As c-jun and c-Fos constitute the typical AP1 transcription heterodimer, these results may indicate that gene expression through the ARE2 element could be possible and partially contribute to the expression of Gclc, but in an Nrf2 independent manner. It should also be noted that elements of a core AP1 sequence are contained in the ARE2 promoter. Consistent with the observations made in hepatocytes from young rats, c-jun and c-Fos also associated with ARE2 in old rat hepatocytes; however, AP1 binding was also accompanied by Nrf2 interacting with the ARE2 element (Table 1). No evidence of Bach1 binding to this element was observed. Thus, only co-activating transcription factors and no repressors were observed at ARE2 site with age.

‘Nrf2 promoter switching’ in old rats does not compensate for the ‘age-related’ decline in ARE-mediated transcriptional activity

To examine the consequences of the age-specific changes in Nrf2 binding to both ARE4 and the ARE2 elements, we transiently transfected primary hepatocytes from young and old rats with minimal promoter–luciferase reporter constructs, each containing three copies of either ARE4 (pGL4.23Gclc[3X]ARE4.luc2) or the ARE2 (pGL4.23Gclc[3X]ARE2.luc2) locus. To verify that no reporter activity occurred through the ARE1 and ARE3 sites even in the absence of Nrf2 binding, transfection experiments were carried out with the ARE1 and ARE3 luciferase reporters as well. Additionally, the pHRL-CMV Renilla luciferase construct was co-transfected as a control for transfection efficiency. Data from these experiments show that the ARE1 and ARE3 promoters are not transcriptionally active (Shenvi et al., 2009), suggesting that these are pseudo-ARE sites and not involved in gene expression. However, when the ARE4-luc construct was transfected into hepatocytes, robust activity was evident. With age, there was a significant 47 ± 14% decline in luciferase activity induced by the ARE4 promoter, indicating that lower Nrf2 binding at this motif correlates well with a decline in transcriptional activity (Fig. 4). Interestingly and in concert with the ARE4 site, luciferase activity induced by the ARE2 promoter also exhibited an age-related decline to the same extent as seen in the ARE4 promoter (Fig. 4). Given the age-specific occupation of the ARE2 promoter by Nrf2, these results were somewhat surprising. Based on the aforementioned results, we hypothesized that even though Nrf2 was found on the ARE2 site in aging rat liver, the amount of Nrf2 actually bound to ARE2 was inadequate to compensate for lower age-related expression through the normal ARE4 element. To test this hypothesis, hepatocytes from old rats were co-transfected with an Nrf2 expression plasmid along with the ARE4 and ARE2 luciferase reporter constructs to substantially increase nuclear Nrf2 levels. Figure 4 shows the results of Nrf2 over-expression on ARE4 and ARE2 luciferase activities in old animals compared with baseline activities in the young. An examination of both the ARE4 as well as ARE2-driven luciferase activities reveals that Nrf2 over-expression compensates for the overall loss of ARE-governed transcription in aging primarily through the ARE2 element, if not exclusively. However, over-expression of Nrf2 in hepatocytes from young rats only increases ARE4-driven, but not ARE2-controlled, luciferase activity. These results confirm that Nrf2 levels are limiting to transcriptional activity of the ARE2 promoter in the old rat liver. Additionally, these data also provide credibility to the hypothesis that loss in activity of the ARE2 site is governed by a two-pronged mechanism: lower Nrf2 binding as well as the formation of a repressive transcriptional motif.

image

Figure 4.  Nrf2 promoter switching in the old rat liver does not compensate for the age-related decline in antioxidant response element (ARE) transcriptional activity. Primary hepatocytes from young and old rats were transfected with a luciferase reporter construct containing three copies of either ARE4 or ARE2 and luciferase activities determined after 24 h. A Renilla luciferase construct was co-transfected to normalize for transfection. Results show that ARE4-driven luciferase activity drops 47 ± 14% (P ≤ 0.05) with age. Nrf2 over-expression resulted in a statistical significant increase in luciferase activity driven by ARE4 (P ≤ 0.05). In contrast, ARE4 co-transfected with Nrf2 in hepatocytes isolated from old rats was not found to be statistically different. Conversely, Nrf2 over-expression did not resulted in a statistical significant increase in luciferase activity driven by ARE2 in young, but ARE2 co-transfected with Nrf2 in hepatocytes isolated from old rats was found to be statistically different. Luciferase reporter activities are calculated as Firefly/Renilla and are normalized to the activities of the empty vector in young and old rat hepatocytes, respectively. Results are depicted as mean ± SEM and are representative of reporter assays from four young and four old rats.

Download figure to PowerPoint

The lesion in Nrf2–ARE-mediated transcription contributes to age-related decline in glutamate cysteine ligase activity and glutathione levels

We verified the consequences of an age-related decline in the Nrf2–ARE regulon on hepatic Gclc mRNA expression, protein levels, GCL activity, and GSH levels. Gclc mRNA levels measured using qPCR revealed a significant 27 ± 10% (P ≤ 0.05) decline in message levels in hepatocytes isolated from old rats compared with young (Fig. 5A). A Western blot using total cell extracts from old rat hepatocytes shows that GCLC levels decline 42 ± 8% (P ≤ 0.05) relative to that seen in young rat hepatocytes (Fig. 5B). Furthermore, an examination of GCL activity showed a significant 60 ± 1.5% (P ≤ 0.05) age-dependent loss in GSH synthetic capacity (Fig. 5C). Lastly, we ascertained that old rats exhibited 40% lower total GSH levels, but this difference was not statistically significant (Fig. 5D). These results suggest that aberrant Nrf2–ARE-mediated transcriptional regulation in old rats contributes to the attenuated hepatic GSH content with age.

image

Figure 5.  Age-related decline in hepatic glutathione (GSH) synthesis. (A) Gclc mRNA levels are lower in liver tissues from old rats. Quantitative PCR analysis revealed a significant 27 ± 10% (P < 0.05) decline in message levels. (B) Glutamate cysteine ligase catalytic subunit (GCLC) protein levels decline with age in hepatocytes. Western analysis of GCLC shows a 42 ± 8% loss in old rat hepatocytes compared with young. (C) Glutamate cysteine ligase activity declines approximately 60% with age in rat hepatocytes (*P ≤ 0.05). (D) Measurement of GSH levels in young and old rat hepatocytes show a 40% loss with age. Results are expressed as mean ± SEM and are representative of four young and four old rats.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. Author contributions
  9. References
  10. Supporting Information

The findings in this article mark a novel observation regarding age-related changes in the transcriptional machinery of an important stress response gene, Gclc. To our knowledge, this is the first comprehensive report of the protein composition of the ARE transcriptome and how the aging process modifies its composition. The Nrf2–ARE system is the master regulator of over 200 genes related to detoxication, antioxidant defense, and anti-inflammatory processes. Other than the loss of nuclear Nrf2, our results demonstrate that the make-up of the permissive transcriptional machinery of the ARE is fundamentally altered with age. As older individuals are at increased risk for cancer, artherosclerosis, and other diseases associated with aging, these observations are significant as it points to a rationale why stressors that may be effectively mitigated in younger individuals enhance risk for these pathologies in the elderly (Ungvari et al., 2011a,b).

The current data using a primary hepatocyte culture model are consistent with our previous studies using isolated rat liver and suggests that the culture system recapitulates the aging phenotype with respect to age-related decline in Nrf2 nuclear tenure (Suh et al., 2004a) (Shenvi et al., 2008). It should be emphasized that diminished basal nuclear Nrf2 values were not accompanied by equivalent losses in cytosolic Nrf2 levels, which suggests that Nrf2 expression and translation, or its typical Keap1/cul3 ubiquitin ligase-dependent turnover is not affected with age. In this regard, we found no evidence for an age-associated loss of Nrf2 mRNA levels (data not shown), which is also in keeping with the consensus view that Nrf2 is constantly transcribed at high rates and its expression does not change in response to antioxidants or electrophiles (Jaiswal, 2004). Thus, the mechanisms for the age-dependent and nuclear-specific loss of Nrf2 are presently unknown, but points to a fundamental perturbation in Nrf2 homeostatic mechanism(s) that have the potential to be remediated. For example, caloric restriction as well as lipoic acid treatment in aging increases Nrf2 nuclear translocation, which indicates that a greater nuclear pool of Nrf2 is available under those regimens to initiate gene transcription (Hyun et al., 2006; Pearson et al., 2008) (Ungvari et al., 2011a,b). As Nrf2 has constitutively active nuclear localization signals, but elaborately regulated nuclear export mechanisms, our data may point to enhanced nuclear efflux of Nrf2 with age (Li et al., 2006; Theodore et al., 2008). While the mechanisms associated with the age-associated disruption in Nrf2 nuclear tenure has not been fully identified, nevertheless, such a loss may profoundly influence basal profiles of phase II detoxification enzymes as well as overall xenobiotic stress response.

Nrf2 is obligated to bind to ARE sequences as a heterodimer with other bZIP proteins to elicit gene expression (Venugopal & Jaiswal, 1996; Dhakshinamoorthy & Jaiswal, 2000; Li et al., 2008; Levy et al., 2009). Recent data indicate that Nrf2-binding proteins serve an overall modulatory role to fine-tune the degree of gene transcription. When Nrf2 concentrations are low, such as our data suggest happens with age, a repressive-type transcriptome may develop at the ARE and modulate gene expression. For instance, small Maf homodimers may form on ARE sites and actually limit expression of detoxification genes (Dhakshinamoorthy & Jaiswal, 2000). In keeping with this concept that age modifies an active Nrf2 transcriptional complex, it is interesting to note we observed that Bach1 resides on the ARE4 locus of the Gclc promoter in hepatocytes isolated from old rats. Bach1 is a cellular heme sensor that represses the NFE2 promoter, a responsive element that displays sequence homology to maf recognition elements (Igarashi et al., 1998). Bach1 binds to ARE sequences when either Nrf2 or small maf proteins are limiting and results in silencing of both HO-1 and NQO1 expression (Kaspar & Jaiswal, 2010). Furthermore, Yoshida and coworkers showed that Bach1 represses transcription by complexing with small maf homodimers, thereby initiating DNA loop formation (Igarashi et al., 1998). Thus, Bach1 acts as a potent architectural chromatin-remodeling agent, which may ultimately result in the observed decline in Gclc expression. We, therefore, interpret our results to mean that lower Nrf2 levels in aging rat hepatocytes limit Nrf2 availability for ARE4 and instead permit Bach1 association. Further experiments will be needed to assess the nature of this hypothesized Bach1-mediated chromatin modification, which is outside the scope of the present article.

Along with the presence of Bach1, the age-related loss of CBP in the Nrf2 transcriptional complex in hepatocytes from old rats would also theoretically inhibit Gclc transcription. CBP is a histone acetyltransferase involved in chromatin opening (Zhu & Fahl, 2001; Shen et al., 2004), and its levels decline with age in the liver, kidney, and cerebral cortex of rats (Bandyopadhyay et al., 2002; Chung et al., 2002). It is normally recruited to ARE sequences via its association with the Neh4 and Neh5 domain of Nrf2 (Katoh et al., 2001; Shen et al., 2004) and directly modulates Nrf2 association with ARE sequences through acetylation of Nrf2. Interestingly and in relation to the present study, CBP inhibitors also lower nuclear Nrf2 localization similar to that seen in aging (Katoh et al., 2001; Sun et al., 2009). Thus, the age-related disappearance of CBP from the ARE4 site may link lower nuclear levels of Nrf2 with the repressive transcriptional phenotype that develops at that locus with age. Additionally, loss of CBP may lead to the lack of chromatin opening at the ARE4 site, prompting Nrf2 binding to an alternate locus (i.e. ARE2) as seen in this study. We are currently assessing the role of CBP as a central mediator for the decline in Nrf2 stress response in aging.

There are multiple AREs in the Gclc promoter, and, although we and others have looked at the functional significance of each of these, it was not known until this report whether aging affects Nrf2–ARE binding universally or specifically affects only certain ARE sequences (Wasserman & Fahl, 1997; Erickson et al., 2002; Shenvi et al., 2009) (Malhotra et al., 2010). Aside from age-related alterations to the Nrf2/ARE4 transcriptome, we observed that Nrf2 also binds to an alternative ARE (ARE2) in old rats; no Nrf2 binding was detected at this site in young rat hepatocytes. This is an intriguing result with large ramifications for transcriptional adaptation in aged animals. ARE2 contains a consensus sequence in the anti-sense strand and a partial ARE in the sense. In contrast, as we previously reported (Shenvi et al., 2009), ARE4 contains a consensus ARE in both the sense and anti-sense strand. Further research is necessary to determine whether the observed age-related ARE2-mediated Nrf2-induced expression is exclusively sequence dependent. While the consequences of activating this unique ARE2 sequence are not yet fully known, it is interesting that only co-activating bZip partner proteins and no repressors (e.g. Bach1) were found to bind at ARE2 in cells from aged animals. Moreover, our data from the Nrf2 over-expression experiments suggest that the ARE2 site can increase expression of Gclc. Previously, we showed that the nuclear Nrf2 as well as GCLC lost during aging was replenished by feeding old rats with (R)-α-lipoic acid, a redox-active dithiol compound known to induce a phase II response. This finding is very important, especially because it provides credence to our hypothesis that the age-associated transcriptional lesion in ARE4 can be remediated only through Nrf2–ARE2 binding. Thus, promoter switching by Nrf2 during aging may potentially be an attempt at compensating for the loss in expression of Gclc. Taken together, these results suggest that ARE2 is comparatively more accessible to Nrf2 in the aging liver and can serve as a potential means to compensate for the decline in GCLC expression through ARE4 (Fig. 6).

image

Figure 6.  Schematic representation of age-dependent alterations in the Gclc-antioxidant response element (ARE) transcriptome. Nrf2 binds to the active ARE4 promoter along with the histone acetyl transferase CREB-binding protein (CBP) and partner proteins small maf, c-jun, and c-Fos in the young rat liver. The transcriptional activity of ARE4 significantly declines with age by a two-pronged alteration: loss in Nrf2 binding accompanied by loss of CBP and binding of the negative regulator Bach1. Age-related decline in Nrf2–ARE4 binding is attempted to be compensated by Nrf2–ARE2 binding in the old liver.

Download figure to PowerPoint

Partial promoter switching to maintain gene activity with age has literature precedent. For example, Iakova et al. recently reported that aging induces growth arrest of proliferative pathways in the liver by a C/EBP alpha promoter-switching mechanism (Iakova et al., 2003). Similarly, there are numerous reports of promoter silencing by methylation-dependent mechanisms during aging (Burzynski, 2005; Strunnikova et al., 2005; Gupta et al., 2006). Of relevance to the results obtained in this study, Barzilai and coworkers (Thompson et al., 2010) recently showed that the aging liver, in particular, is more susceptible to hypermethylation at promoters of genes involved in metabolic regulation and stress response. These findings are important because ARE promoters, especially ARE4, are extremely GC-rich (89%) and are plausibly susceptible to hypermethylation.

In summary, this report identifies a novel Nrf2-mediated promoter switching mechanism that, nevertheless, cannot compensate for the age-related decline in hepatic GSH synthetic capacity. Moreover, we have preliminary evidence that the expression of other ARE-containing genes (e.g. GST2A, HO-1, and NQO1) also decline with age (Shenvi et al., in preparation), which indicates that the information contained in this report may not only pertain to Gclc. Experiments are ongoing as to whether alternative ARE site(s) and/or promoter switching is also responsible for their diminished expression in aged tissue. Thus, there is the potential that a common mechanism has been identified that leads to compromised phase II stress response with age.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. Author contributions
  9. References
  10. Supporting Information

Reagents

Restriction enzymes and T4 DNA ligase for subcloning were from New England Bio Laboratories (Boston, MA, USA). The Dual Luciferase Reporter Assay system and reporter plasmids, pGL4 minimal promoter vector, and phRL-CMV vector were from Promega (Madison, WI, USA). The expression vector for Nrf2 (pcDNA3.1-Nrf2) was a kind gift provided by Dr. Anil Jaiswal at University of Maryland School of Medicine, Baltimore, Maryland. Custom oligonucleotides used in PCR cloning, subcloning, and DNA sequencing were purchased from Invitrogen (Carlsbad, CA, USA). Sequence service was provided by Center for Gene Research and Biotechnology, Oregon State University. Rabbit anti-Nrf2 (H-300), anti-lamin B1, small maf antibodies, and Nrf2 siRNA as well as scrambled oligonucleotide sequences were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Other antibodies were purchased from the following suppliers: c-Jun and Bach1 (Abcam, Cambridge, MA, USA), c-Fos (eBioscience, San Diego, CA, USA), CBP (Upstate Biotechnology, New York, USA), and IgG (Calbiochem, San Diego, CA). Protein A-Sepharose was purchased from Sigma (St. Louis, MO, USA). All chemicals used were at least analytical grade.

Animals

Rats (Fischer 344, virgin male, outbred albino), both young (2–5 months: = 12) and old (24–28 months, n = 25; National Institute of Aging animal colonies), were acclimatized in the Oregon State University animal facilities for at least 1 week before experimentation. Animals were maintained on a standard chow diet, and food and water were given ad libitum. All animal procedures were performed in accordance with the Oregon State University guidelines for animal experimentation.

Cell isolation and culture

Hepatocytes were plated on collagen-coated culture dishes in William’s Medium E supplemented with 5% FBS, 1 mm dexamethasone, 100 ng mL−1 insulin, 2 mm l-glutamine, 100 U mL−1 penicillin, and 100 mg mL−1 streptomycin for 4 h in 5% CO2 at 37 °C to allow attachment to the dishes. The medium was then replaced with fresh supplemented William’s Medium E, and the cells were cultured for an additional 48 h before chromatin immunoprecipitation or transfection with the appropriate ARE-luciferase constructs.

Preparation of nuclear extracts and Nrf2 analysis

Nuclear and cytosolic proteins isolated from rat hepatocytes were resolved by SDS-PAGE and Nrf2 levels were determined by Western blotting using Nrf2 antibody (1:1000) as described before. Anti-lamin B1 and β-actin were used as loading controls for nuclear and cytosolic proteins, respectively.

Chromatin immunoprecipitation (ChIP) assay

Chromatin immunoprecipitation analysis was conducted using control rabbit IgG, anti-Nrf2, anti-small maf, anti-c-Jun, anti-c-Fos, anti-CBP, and anti-Bach1 antibodies. PCR primers are described in Table S1 (Supporting information). Eighteen million hepatocytes were used for each chromatin immunoprecipitation experiment.

Construction of luciferase reporter vectors

The ARE- and ARE-like-luciferase reporter plasmids were generated using the pGL4 minimal promoter vector (Promega) containing a minimal TATA promoter upstream of the firefly luciferase gene. The sequences of the inserts used in the different plasmids are summarized in Table S2 (Supporting information). Single-stranded oligonucleotides were first annealed to form double-stranded oligonucleotides and then ligated into the pGL4.23[minP] vector following the manufacturer’s instructions. The vectors were engineered by inserting three copies of each of the ARE elements present in the rat Gclc 5′-flanking region. After the plasmids were generated, the DNA sequence of the inserts was verified.

Hepatocyte transfection and luciferase assays

Reporter gene assays were used to determine the transcriptional activities of individual Gclc ARE elements and the ARE-like element in primary hepatocyte cultures from young and old rats. Transient transfections were performed in hepatocytes cultured on six-well collagen-coated plates for at least 48 h using the Effectene Transfection reagent (Qiagen, Valencia, CA, USA). The cells were transfected with 1.6 μg of Gclc-luciferase plasmids. The total amount of plasmid DNA for transfection was adjusted by empty expression vector (pGL4.23). The control plasmid phRL-CMV encoding Renilla luciferase was included in each transfection (0.02 μg) to account for variability in transfection efficiency. In some cases, 0.5 μg of Nrf2 expression plasmid (pcDNA2.1 Nrf2) was co-transfected with luciferase reporter constructs. Thirty-six hours after transfection, cells were harvested with 1× passive lysis buffer (Promega), and the supernatant was collected by brief centrifugation. Transcription activity was determined by the expression of firefly luciferase and was normalized to the Renilla luciferase levels by using a dual luciferase reporter assay kit (Promega) on a Biolumat LB9505 luminometer (Berthold Detection Systems, Pforzheim, Germany). The means of at least three independent experiments, each carried out in duplicate, are shown with the ±SE. Statistical significance is determined by an one-way ANOVA, and individual comparison is calculated by Bonferroni’s multiple comparison test.

Real-Time PCR of GCLC and GCLM mRNA

A portion of each liver was excised and stored in RNALater (Ambion, Austin, TX, USA) at 80 °C and homogenized using a Dounce homogenizer. Total RNA was isolated from both young and old rat livers (n = 4) by using an RNeasy Midi Kit (Qiagen). cDNA was prepared from 1 μg of total RNA per group using SuperScript II (Life Technologies, Gaithersburg, MD, USA). Relative transcript amounts of Gclc and Gclm were determined by quantitative real-time PCR using primers, cycling conditions, and housekeeping genes as described elsewhere (Suh et al., 2004a,b).

Statistical analysis

The data are expressed as the means ±SE. Statistical analysis was performed with the GraphPad Prism software version 3.03 (GraphPad Software Inc., San Diego, CA, USA). We used a two-tailed Student’s t-test to compare the luciferase activity of individual Gclc promoter constructs. A P value < 0.05 was considered to be significant. One-way analysis of variance (ANOVA) was used when multiple comparisons were made, followed by Tukey’s post hoc analysis for multiple comparisons to a control.

Acknowledgments

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. Author contributions
  9. References
  10. Supporting Information

This work was supported by grant P01 AT002034 from the National Center for Complementary and Alternative Medicine (NCCAM) and 2R01 AG017141-06A2 from the National Institute on Aging (NIA). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of NCCAM, NIA, or the National Institutes of Health. We would like to acknowledge Dr. Chrissa Kioussi’s guidance and Mrs. Judy Butler’s technical assistance in performing chromatin immunoprecipitation experiments. We would also like to thank Dr. Anil Jaiswal for providing the Nrf2 expression vector.

Author contributions

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. Author contributions
  9. References
  10. Supporting Information

S.V.S. and E.S. contributed to the acquisition of data in this manuscript. S.V.S., E.S., and T.M.H. all contributed to the conception and study design as well as analysis and interpretation of the data. All three authors were responsible for the writing of the manuscript.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. Author contributions
  9. References
  10. Supporting Information
  • Bandyopadhyay D, Okan NA, Bales E, Nascimento L, Cole PA, Medrano EE (2002) Down-regulation of p300/CBP histone acetyltransferase activates a senescence checkpoint in human melanocytes. Cancer Res. 62, 62316239.
  • Burzynski SR (2005) Aging: gene silencing or gene activation? Med. Hypotheses 64, 201208.
  • Chung YH, Kim EJ, Shin CM, Joo KM, Kim MJ, Woo HW, Cha CI (2002) Age-related changes in CREB binding protein immunoreactivity in the cerebral cortex and hippocampus of rats. Brain Res. 956, 312318.
  • Dhakshinamoorthy S, Jaiswal AK (2000) Small maf (MafG and MafK) proteins negatively regulate antioxidant response element-mediated expression and antioxidant induction of the NAD(P)H:Quinone oxidoreductase1 gene. J. Biol. Chem. 275, 4013440141.
  • Dhakshinamoorthy S, Jain AK, Bloom DA, Jaiswal AK (2005) Bach1 competes with Nrf2 leading to negative regulation of the antioxidant response element (ARE)-mediated NAD(P)H:quinone oxidoreductase 1 gene expression and induction in response to antioxidants. J. Biol. Chem. 280, 1689116900.
  • Erickson AM, Nevarea Z, Gipp JJ, Mulcahy RT (2002) Identification of a variant antioxidant response element in the promoter of the human glutamate-cysteine ligase modifier subunit gene. Revision of the ARE consensus sequence. J. Biol. Chem. 277, 3073030737.
  • Finkel T, Holbrook NJ (2000) Oxidants, oxidative stress and the biology of ageing. Nature 408, 239247.
  • Gupta S, Pathak RU, Kanungo MS (2006) DNA methylation induced changes in chromatin conformation of the promoter of the vitellogenin II gene of Japanese quail during aging. Gene 377, 159168.
  • Hagen TM, Moreau R, Suh JH, Visioli F (2002) Mitochondrial decay in the aging rat heart: evidence for improvement by dietary supplementation with acetyl-L-carnitine and/or lipoic acid. Ann. N. Y. Acad. Sci. 959, 491507.
  • Hall DM, Sattler GL, Sattler CA, Zhang HJ, Oberley LW, Pitot HC, Kregel KC (2001) Aging lowers steady-state antioxidant enzyme and stress protein expression in primary hepatocytes. J. Gerontol. A Biol. Sci. Med. Sci. 56, B259B267.
  • Huang CS, Anderson ME, Meister A (1993a) Amino acid sequence and function of the light subunit of rat kidney gamma-glutamylcysteine synthetase. J. Biol. Chem. 268, 2057820583.
  • Huang CS, Chang LS, Anderson ME, Meister A (1993b) Catalytic and regulatory properties of the heavy subunit of rat kidney gamma-glutamylcysteine synthetase. J. Biol. Chem. 268, 1967519680.
  • Hyun DH, Emerson SS, Jo DG, Mattson MP, de Cabo R (2006) Calorie restriction up-regulates the plasma membrane redox system in brain cells and suppresses oxidative stress during aging. Proc. Natl Acad. Sci. U S A 103, 1990819912.
  • Iakova P, Awad SS, Timchenko NA (2003) Aging reduces proliferative capacities of liver by switching pathways of C/EBP alpha growth arrest. Cell 113, 495506.
  • Igarashi K, Hoshino H, Muto A, Suwabe N, Nishikawa S, Nakauchi H, Yamamoto M (1998) Multivalent DNA binding complex generated by small Maf and Bach1 as a possible biochemical basis for beta-globin locus control region complex. J. Biol. Chem. 273, 1178311790.
  • Jaiswal AK (2004) Nrf2 signaling in coordinated activation of antioxidant gene expression. Free Radic. Biol. Med. 36, 11991207.
  • Kaspar JW, Jaiswal AK (2010) Antioxidant-induced phosphorylation of tyrosine 486 leads to rapid nuclear export of Bach1 that allows Nrf2 to bind to the antioxidant response element and activate defensive gene expression. J. Biol. Chem. 285, 153162.
  • Katoh Y, Itoh K, Yoshida E, Miyagishi M, Fukamizu A, Yamamoto M (2001) Two domains of Nrf2 cooperatively bind CBP, a CREB binding protein, and synergistically activate transcription. Genes Cells 6, 857868.
  • Lee JI, Kang J, Stipanuk MH (2006) Differential regulation of glutamate-cysteine ligase subunit expression and increased holoenzyme formation in response to cysteine deprivation. Biochem. J. 393, 181190.
  • Levy S, Jaiswal AK, Forman HJ (2009) The role of c-Jun phosphorylation in EpRE activation of phase II genes. Free Radic. Biol. Med. 47, 11721179.
  • Li W, Yu SW, Kong AN (2006) Nrf2 possesses a redox-sensitive nuclear exporting signal in the Neh5 transactivation domain. J. Biol. Chem. 281, 2725127263.
  • Li W, Yu S, Liu T, Kim JH, Blank V, Li H, Kong AN (2008) Heterodimerization with small Maf proteins enhances nuclear retention of Nrf2 via masking the NESzip motif. Biochim. Biophys. Acta 1783, 18471856.
  • Lu SC (2009) Regulation of glutathione synthesis. Mol. Aspects Med. 30, 4259.
  • Lu SC, Huang ZZ, Yang H, Tsukamoto H (1999) Effect of thioacetamide on the hepatic expression of gamma-glutamylcysteine synthetase subunits in the Rat. Toxicol. Appl. Pharmacol. 159, 161168.
  • Malhotra D, Portales-Casamar E, Singh A, Srivastava S, Arenillas D, Happel C, Shyr C, Wakabayashi N, Kensler TW, Wasserman WW, Biswal S (2010) Global mapping of binding sites for Nrf2 identifies novel targets in cell survival response through ChIP-Seq profiling and network analysis. Nucleic Acids Res. 38, 57185734.
  • Pearson KJ, Lewis KN, Price NL, Chang JW, Perez E, Cascajo MV, Tamashiro KL, Poosala S, Csiszar A, Ungvari Z (2008) Nrf2 mediates cancer protection but not prolongevity induced by caloric restriction. Proc. Natl. Acad. Sci. 105, 2325.
  • Reichard JF, Motz GT, Puga A (2007) Heme oxygenase-1 induction by NRF2 requires inactivation of the transcriptional repressor BACH1. Nucleic Acids Res. 35, 70747086.
  • Shen G, Hebbar V, Nair S, Xu C, Li W, Lin W, Keum YS, Han J, Gallo MA, Kong AN (2004) Regulation of Nrf2 transactivation domain activity. The differential effects of mitogen-activated protein kinase cascades and synergistic stimulatory effect of Raf and CREB-binding protein. J. Biol. Chem. 279, 2305223060.
  • Shenvi SV, Dixon BM, Petersen Shay K, Hagen TM (2008) A rat primary hepatocyte culture model for aging studies. Curr. Protoc. Toxicol. 14.7, 14.7.114.7.10.
  • Shenvi SV, Smith EJ, Hagen TM (2009) Transcriptional regulation of rat gamma-glutamate cysteine ligase catalytic subunit gene is mediated through a distal antioxidant response element. Pharmacol. Res. 60, 229236.
  • Song IS, Tatebe S, Dai W, Kuo MT (2005) Delayed mechanism for induction of gamma-glutamylcysteine synthetase heavy subunit mRNA stability by oxidative stress involving p38 mitogen-activated protein kinase signaling. J. Biol. Chem. 280, 2823028240.
  • Stio M, Iantomasi T, Favilli F, Marraccini P, Lunghi B, Vincenzini MT, Treves C (1994) Glutathione metabolism in heart and liver of the aging rat. Biochem. Cell Biol. 72, 5861.
  • Strunnikova M, Schagdarsurengin U, Kehlen A, Garbe JC, Stampfer MR, Dammann R (2005) Chromatin inactivation precedes de novo DNA methylation during the progressive epigenetic silencing of the RASSF1A promoter. Mol. Cell. Biol. 25, 39233933.
  • Suh JH, Shenvi SV, Dixon BM, Liu H, Jaiswal AK, Liu RM, Hagen TM (2004a) Decline in transcriptional activity of Nrf2 causes age-related loss of glutathione synthesis, which is reversible with lipoic acid. Proc. Natl. Acad. Sci. U S A 101, 33813386.
  • Suh JH, Wang H, Liu RM, Liu J, Hagen TM (2004b) (R)-alpha-lipoic acid reverses the age-related loss in GSH redox status in post-mitotic tissues: evidence for increased cysteine requirement for GSH synthesis. Arch. Biochem. Biophys. 423, 126135.
  • Sun Z, Chin YE, Zhang DD (2009) Acetylation of Nrf2 by p300/CBP augments promoter-specific DNA binding of Nrf2 during the antioxidant response. Mol. Cell. Biol. 29, 26582672.
  • Theodore M, Kawai Y, Yang J, Kleshchenko Y, Reddy SP, Villalta F, Arinze IJ (2008) Multiple nuclear localization signals function in the nuclear import of the transcription factor Nrf2. J. Biol. Chem. 283, 89848994.
  • Thompson RF, Atzmon G, Gheorghe C, Liang HQ, Lowes C, Greally JM, Barzilai N (2010) Tissue-specific dysregulation of DNA methylation in aging. Aging Cell 9, 506518.
  • Toroser D, Yarian CS, Orr WC, Sohal RS (2006) Mechanisms of gamma-glutamylcysteine ligase regulation. Biochim. Biophys. Acta 1760, 233244.
  • Ungvari Z, Bailey-Downs L, Gautam T, Sosnowska D, Wang M, Monticone RE, Telljohann R, Pinto JT, de Cabo R, Sonntag WE, Lakatta EG, Csiszar A (2011a) Age-associated vascular oxidative stress, Nrf2 dysfunction, and NF-{kappa}B activation in the nonhuman primate Macaca mulatta. J. Gerontol. A Biol. Sci. Med. Sci. 66, 866875.
  • Ungvari ZI, Bailey-Downs L, Sosnowska D, Gautam T, Koncz P, Losonczy G, Ballabh P, de Cabo R, Sonntag WE, Csiszar A (2011b) Vascular oxidative stress in aging: a homeostatic failure due to dysregulation of Nrf2-mediated antioxidant response. Am. J. Physiol. Heart Circ. Physiol. 301, H363372.
  • Venugopal R, Jaiswal AK (1996) Nrf1 and Nrf2 positively and c-Fos and Fra1 negatively regulate the human antioxidant response element-mediated expression of NAD(P)H:quinone oxidoreductase1 gene. Proc. Natl. Acad. Sci. U S A 93, 1496014965.
  • Wasserman WW, Fahl WE (1997) Functional antioxidant responsive elements. Proc. Natl. Acad. Sci. U S A 94, 53615366.
  • Yang H, Magilnick N, Ou X, Lu SC (2005) Tumour necrosis factor alpha induces co-ordinated activation of rat GSH synthetic enzymes via nuclear factor kappaB and activator protein-1. Biochem. J. 391, 399408.
  • Zhu M, Fahl WE (2001) Functional characterization of transcription regulators that interact with the electrophile response element. Biochem. Biophys. Res. Commun. 289, 212219.
  • Zipper LM, Mulcahy RT (2000) Inhibition of ERK and p38 MAP kinases inhibits binding of Nrf2 and induction of GCS genes. Biochem. Biophys. Res. Commun. 278, 484492.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. Author contributions
  9. References
  10. Supporting Information

Data S1 Experimental procedures.

Table S1 Primers used for ChIP analysis of Gclc promoter elements.

Table S2 Sequence of Inserts in the pGL4 minimal promoter vector.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
ACEL_788_sm_SupportingInformation.pdf113KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.