Heme activates the heme oxygenase-1 gene in renal epithelialcells by stabilizin



免疫细胞治疗专区	欢迎关注干细胞微信公众号

12 3 4 5 6 7 8 9 10 ... 25 下一页

返回列表

查看: 499163\|回复: 242	go Heme activates the heme oxygenase-1 gene in renal epithelialcells by stabilizin

轻羽

新手上路

Rank: 1

积分: 0
威望: 0
包包: 0

发表于 2009-4-21 13:35 |显示全部帖子

作者：JawedAlam, ErinKilleen, PengfeiGong, RyanNaquin, BinHu, DanielStewart, Julie R.Ingelfinger,  Karl A.Nath作者单位：1 Department of Molecular Genetics, Ochsner ClinicFoundation, New Orleans 70121; Department ofBiochemistry and Molecular Biology, Louisiana State University HealthSciences Center, New Orleans, Louisiana 70112; Pediatric Nephrology Unit, Massachusetts GeneralHospital, Boston, Massachusetts 02114; and # e8 X* d; j" r* d$ d0 Q( _


      7 W% K7 W+ g" G$ K- B  {
                     * ]8 [' O6 R+ q( d- S
         4 U; x6 j; Y2 }$ A, L0 w" v


         1 W# D. @3 p  L5 Y) ]$ g7 G

      + \% ~$ _! `% |
      ) _& b: T( N9 d2 D3 ^
      % C! P; S  i; U; Y
      【摘要】- r) w( z3 `3 w1 a$ N+ k; k
   The mechanism of heme oxygenase-1gene ( ho-1 ) activation by heme in immortalized rat proximaltubular epithelial cells was examined. Analysis of the ho-1 promoter identified the heme-responsive sequences as thestress-response element (StRE), multiple copies of which are present intwo enhancer regions, E1 and E2. Electrophoretic mobility shift assaysidentified Nrf2, MafG, ATF3, and Jun and Fos family members asStRE-binding proteins; binding of Nrf2, MafG, and ATF3 was increased inresponse to heme. Dominant-negative mutants of Nrf2 and Maf, but not ofc-Fos and c-Jun, inhibited basal and heme-induced expression of anE1-controlled luciferase gene. Heme did not affect the transcriptionactivity of Nrf2, dimerization between Nrf2 and MafG, or the level ofMafG, but did stimulate expression of Nrf2. Heme did not influence thelevel of Nrf2 mRNA but increased the half-life of Nrf2 protein from ~10 min to nearly 110 min. These results indicate that heme promotes stabilization of Nrf2, leading to accumulation ofNrf2 · MafG dimers that bind to StREs to activatethe ho-1 gene. . N4 X: q" y4 x. o- @( i, K( W4 k
      【关键词】 Nrf transcription factor stressresponse element proteinstabilization5 C0 ]6 B4 _9 D6 f
   INTRODUCTION2 d; h( y+ X6 w( J0 c
$ w/ i% s: \5 q9 \1 R
HEME, A TETRAPYROLE with a redox active iron center, is a lipophilic molecule of limitedsolubility in aqueousenvironments. 1 Consequently,it is typically associated with (either covalently or noncovalently),and functions as a cofactor for, various proteins such as hemoglobin,myoglobin, mitochondrial and microsomal cytochromes, catalases, nitricoxide synthase, guanylate cyclase, and cyclooxygenases ( 11, 36, 48 ). In this capacity, heme is essential for many biologicalactivities, including oxygen transport, energy production, andxenobiotic detoxification. Tissue damage or cell injury, either inpathological states or in response to noxious stimuli, can destabilizeheme proteins, resulting in altered associations and even release ofthe heme moiety. Due to its lipophilic nature and the reactive ironmolecule, "free" heme can damage cellular components and disruptcellular function through mechanisms that are, at least in part,prooxidant in nature ( 9, 10, 20, 39-41 ). Forinstance, heme is known to promote oxidative degradation of proteins( 2 ) and DNA ( 1 ) and amplify hydrogenperoxide-mediated endothelial cell dysfunction ( 11 ).
, H& h% u  Y8 x, _
Elimination of excess free heme is essential for maintenance ofcellular integrity and is largely the responsibility of heme oxygenases(HOs), enzymes that catalyze the initial and rate-limiting step in hemecatabolism, the oxidative cleavage of the porphyrin ring to generatebiliverdin IX with the release of the heme iron and carbon monoxide.Of the two functional HO isoforms thus far identified (HO-1 and HO-2)( 32 ), HO-1 plays a particularly important role in thisprocess as the expression of this protein, and consequently overall HOactivity is potently stimulated in response to heme. In addition to thesubstrate, a variety of physiological and nonphysiological stimuli suchas inflammatory cytokines, hyperthermia, UV-irradiation, heavy metals,and arsenite, all of which are potentially injurious to cells and canlead to heme protein instability, also induce HO-1 expression.Induction of HO-1 is regulated primarily at the level of ho-1 gene transcription.9 o5 H' @$ w5 C* J# @. g) k2 \

Our previous studies have demonstrated the importance of induction ofHO-1 expression and of ensuing HO activity in the protection againstheme-mediated injury. For instance, in experimental rhabdomyolysis andhemolysis, heme molecules derived from myoglobin and hemoglobin released during injury to skeletal muscle and red blood cells, respectively, readily accumulate in kidney epithelial cells and lead torenal dysfunction. Prior induction of HO-1 in this experimental model,however, protected rats from renal failure and mortality; the oppositeeffect was observed after pharmacological inhibition of HO activity( 38, 53 ). Furthermore, we have directly demonstrated theindispensability of HO-1 induction in protecting against heme proteintoxicity by showing that ho-1 -targeted mice are exquisitely sensitive to such insults ( 41 ).
$ ]0 R  r! M! R4 |3 I' B
Although the cytoprotective function of HO-1 activity duringheme-mediated cellular injury is now readily obvious and has beenexperimentally verified in multiple studies, the mechanism by whichheme activates the ho-1 gene, in the kidney or other organs,is less well understood. Here we examine this mechanism in renalproximal tubular epithelial cells and show that heme-dependent ho-1 gene activation is mediated by the stress-responsiveDNA elements (StREs) and transcription factor Nrf2. Additional studies indicate a novel mechanism for regulation of Nrf2 activity and subsequent ho-1 gene activation, namely posttranscriptionalstabilization of the Nrf2 polypeptide in response to heme.6 [) X: V- Q* c( N

MATERIALS AND METHODS7 ^9 Y* Y5 G  |% O$ ~
6 U7 j6 S4 Q! z4 n
Materials. Tissue culture media were from Life Technologies, and fetal bovineserum was obtained from Mediatech. Restriction endonucleases and otherDNA-modifying enzymes were purchased from either Life Technologies orNew England Biolabs. Oligonucleotides were synthesized by IntegratedDNA Technologies. Radiolabeled nucleotides were obtained fromNEN. Reagents for luciferase assays were purchased from Sigma Chemical.MafG antisera were kindly provided by Dr. V. Blank, and anti-rat HO-1was obtained from StressGen Biotech. All other antibodies werepurchased from Santa Cruz Biotechnology. All other chemicals werereagent grade.

Plasmids. The wild-type and mutant ho-1 promoter/luciferase fusionconstructs have been described previously ( 7 ). pFRluc,containing 5 tandem copies of the Gal-4 binding site, was obtained fromStratagene. Plasmid pCMV/ -gal, encoding the Escherichiacoli -galactosidase gene, was kindly provided by Dr. Ping Wei.Construction of the dominant-negative mutants of Nrf2 and c-Junhas been previously reported ( 6 ). The dominant mutant ofMafK ( 31 ) and the A-Zip mutants derived from c-Fos andcAMP response element binding protein (CREB) sequences ( 3, 44 ) were kindly provided by Drs. Stuart Orkin and CharlesVinson, respectively. Mammalian two-hybrid vectors pEG,containing the DNA-binding domain of Gal-4 (Gdbd), and pAD, containingthe transcription activation domain of Nrf2, have been described( 19 ). Full-length human MafG (kindly provided by Dr. Volker Blank), full-length mouse Nrf2 (amino acid residues 1-597), or truncated Nrf2 ( N, amino acids 329-597) were cloned in-frame with the Gdbd in pEG. Full-length rat HO-1 and MafG sequences were alsocloned into the pAD vectors.$ k' a/ [8 G4 `- F6 S# P
5 Z5 ~: b3 m8 X
Cell culture, transfection, and enzyme assays. Immortalized rat proximal tubular epithelial cells (IRPTCs), developedand characterized as previously described ( 51 ), were cultured in a humidified atmosphere (95% air, 5% CO 2 ) at37°C in DMEM containing 0.1% glucose, 0.1 mM nonessential aminoacids, 5% fetal bovine serum, and 50 µg/ml gentamicin. Transienttransfections were carried out with Fugene6 transfection reagent (RocheMolecular Biochemicals) according to the manufacturer'srecommendations. Briefly, cells were seeded (7 × 10 4 /well) in 12-well plates and, 20 h after plating,cells in each well were transfected with a FuGene6-DNA mixtureconsisting of 50 ng of the luciferase plasmid, 25 ng of pCMV -gal and175 ng of empty vector, or the indicated effector plasmid. Thetransfection media were removed 24 h later, and the cells wereexposed to vehicle [DMSO, final concentration of 0.5% (vol/vol)] or10 µM heme in DMSO for 5 h in serum-free medium. Preparation ofcell extract and measurement of reporter enzyme activities were carriedout as described ( 4 ).2 B& u/ R/ _5 q& }8 S4 A
  z7 l% n, [1 s, e; |0 w! r1 N8 `0 u
Electrophoretic mobility shift assay. IRPTCs were seeded (1 × 10 6 cells/10-cm plate) andcultured for 48-72 h in complete medium and then treated withvehicle or 10 µM heme in serum-free medium for 3 h. Whole cellextracts (WCEs) were prepared as described previously( 16 ). The standard binding reaction mixture (12.5 µl)contained 18 mM HEPES (pH 7.9), 80 mM KCl, 2 mM MgCl 2, 10 mM DTT, 10% glycerol, 0.2 mg/ml bovine serum albumin, 160 µg/mlpoly(dI-dC), 20,000 counts per minute [ - 32 P]ATP-labeled probe, and 3.5-5 µg ofWCE. Reaction mixtures were incubated at 25°C for 20 min andanalyzed by native 5% polyacrylamide gel electrophoresis andautoradiography as described previously ( 16, 37 ). Adouble-stranded oligonucleotide containing the sequence5'-TTTTC TGCTGAGTCA AGGTCCG-3' was used as a probe in EMSA reactions (core StRE sequence is underlined). In supershift assays, 1 µl of preimmune serum or anti-MafG serum and preimmune IgG or anti-transcription factor IgG (2 µg/µl) were added to the reaction mixture and incubated for 20 min at room temperature before electrophoresis.8 g. Z9 x5 j) t6 E8 h8 \

Protein and RNA blot analyses. IRPTCs were plated in 10-cm (1 × 10 6 cells) or insix-well plates (2 × 10 5 cells/well) and cultured for48-72 h. The culture media were removed, and cells were exposed tovehicle or 10 µM heme for 0-4 h in serum-free medium. Wholecell, cytoplasmic, and nuclear protein extracts were prepared aspreviously described ( 15 ). Extracts were electrophoresedon a 4-12% gradient SDS-PAGE gel (Invitrogen), and proteins weretransferred to a polyvinylidene difluoride membrane. The membrane wasblocked overnight in Tris-buffered saline containing 0.1% (vol/vol)Tween 20 and 5% (wt/vol) nonfat dry milk and then incubated with theprimary antibody (1:1,000 dilution) for 3 h. Treatment with thesecondary antibody and antigen detection were carried out using the ECLsystem (Amersham Pharmacia Biotech) according to the manufacturer'srecommendations. Total RNA was isolated by the procedure of Chomczynskiand Sacchi ( 12 ), and RNA dot blot analysis was carried outas previously described ( 7 ). Successive hybridizationswere carried on the same filter using cDNA probes encoding mouse Nrf2,rat HO-1, and rat ribosomal protein S3. Hybridization signals werequantified using a phosphorimager (Packard). Relative mRNA levelswere calculated after correcting for RNA adsorption by normalizing theprimary hybridization signals with the S3 signal.
, H' p: X4 h8 u: S3 E+ L4 g
RESULTS
# d5 K! ]( e6 E. E
StREs within the E1 and E2 enhancers mediate heme-dependentactivation of the ho-1 gene. We have previously isolated and characterized a 15-kbp region of themouse ho-1 gene ( 5 ) and identified two5'-distal enhancer regions, E1 and E2, that mediate gene activation inresponse to nonheme HO-1 inducers such as cadmium ( 7 ) and15-deoxy- 12,14 -prostaglandin J 2 ( 17 ) (Fig. 1 ). Both E1 andE2 contain three copies each of a sequence motif termed theStRE; core StRE consensus sequence = 5'-[(T/C)GCTGAGTCA-3'] thatare essential for induction by these agents. To characterize themechanism of ho-1 gene activation by heme and to determinethe role of StREs in this response, we examined the transcriptionactivity of the wild-type mouse ho-1 promoter andappropriate mutants in reporter gene transfection assays. As shown inFig. 1, expression of pHO15luc, a chimera containing the full-length15-kbp promoter and the firefly luciferase gene, was stimulated~12-fold after treatment of IRPTCs with 10 µM heme, a concentrationthat elicited maximal induction (data not shown). Targeted deletion ofthe E1 enhancer ( E1) inhibited heme responsiveness by ~75%,whereas deletion of the E2 fragment reduced induction by only 25%,indicating a greater importance of E1 in this response. Both enhancers,however, are required for optimal induction since deletion of E1 and E2completely abolished heme responsiveness and also drasticallyattenuated basal luciferase activity. The specificity of this responseis demonstrated by the fact that deletion of sequences between 1.3 kband 3.5 kb ( B) only minimally affected basal and heme-inducedluciferase expression.

Fig. 1. Identification of the ho-1 promoter regionsresponsible for heme-mediated gene induction. Plasmid pHO15luc contains~15 kbp of the 5'-flanking region of the mouse ho-1 genefused to the firefly luciferase (Luc) reporter gene. The relativelocations of the previously identified E1 and E2 enhancers areindicated. Derivatives of pHO15luc, generated by internal deletion ofthe promoter, and other fusion constructs are diagrammed. X, mutationof the stress-response elements (StREs) within the indicated regulatoryregion. Transfection and treatment of immortalized rat proximal tubularepithelial cells (IRPTCs) were carried out as described in MATERIALS AND METHODS. Aliquots representing 8% and 4% ofthe cell extracts were used for luciferase and -galactosidaseassays, respectively. Background luciferase activity (frommock-transfected cells) was subtracted from each experimentalmeasurement, and the resulting value was corrected for variation intransfection efficiency by normalization with background-subtracted -galactosidase activity. Normalized luciferase activity in theabsence of heme (basal) was arbitrarily assigned a value of 100 foreach wild-type plasmid; fold induction by heme is provided for eachconstruct. Each data point represents the mean from at least 3 independent experiments. SE varied from 8-22%.
2 J$ l& x% ^) C
The heme-dependent transcription activities of E1 and E2 were testeddirectly using constructs in which each of the enhancers was fused tothe minimal ho-1 promoter ( 44 to  73). In this context, E1and E2 mediated an ~16- and 20-fold stimulation, respectively, ofluciferase expression. Mutation of all three StREs in either enhancer(E1-M739 or E2-M45) reduced basal activity and abolished hemeinducibility. Furthermore, a synthetic regulatory sequence composed ofthree tandem copies of an individual StRE (3× StRE) was as responsiveas the full-length ho-1 promoter. Mutation of the firstthree residues within the core StRE (3× StRE-M2) completely abrogatedheme-dependent gene activation. From these results, we conclude thatthe StRE is sufficient and necessary for heme-mediated ho-1 gene induction in IRPTCs.

StRE-binding activities in IRPTCs. EMSA reactions using WCEs from IRPTCs were carried out to identifyDNA-binding proteins potentially responsible for heme-mediated ho-1 gene induction. With the use of extracts fromvehicle-treated cells, six SRE-protein complexes of relatively similarintensities were typically observed (Fig. 2, laneb ). Heme treatment of IRPTCs significantly altered the subsequent EMSA profile: some of the "control" complexes (complexes 1, 3, and 4) decayed while anapparently novel complex (Fig. 2, arrow) was formed in a time-dependentmanner. Of course, it is possible that this band represents multipledistinct, but comigrating, complexes. The abundance of other complexes(e.g., 5 and 6) varied among experiments, and their apparent increase in response to heme was not observed consistently (for instance, seeFig. 3 ). The specificity of the complexesgenerated was confirmed using wild-type and mutant StREoligonucleotides in competition experiments (data not shown).& b& k( @& N5 P% v* p
1 ]6 S2 f4 W: W1 e
Fig. 2. StRE-binding activities in IRPTCs: IRPTCs were treatedwith 10 µM heme for the indicated time period ( lanesb-g ). Preparation of cell extracts and EMSA reactions werecarried out as described in MATERIALS AND METHODS. TheDNA-protein complexes typically observed in untreated cells( 1-6 ) and the heme-induced complex (arrow) areindicated. Lane a, no protein extract.* }) m) T9 M/ p* `$ @9 c# b) |
  b8 P+ y3 |$ f
Fig. 3. Identification of SRE-binding proteins using antibodysupershift EMSA. Extracts prepared from IRPTCs treated with vehicle(DMSO) or with 10 µM heme for 3 h were used in EMSA reactions.Supershift analysis using preimmune IgG (pre IgG) or antibodies (Ab)directed against the indicated transcription factors was carried out asdescribed in MATERIALS AND METHODS. * Supershiftedcomplexes. The control complexes (1-6, top ) areidentified, and the corresponding migration positions of some of thesecomplexes are indicated in the heme-treated panel. NS, nonspecificcomplex; lane 1, no protein extract. ATF, activatingtranscription factor.# }5 X( U8 u4 P0 r6 v0 D5 T/ d
- _: v6 s4 Y; C0 R
Multiple basic region/leucine zipper proteins bind to StRE. The consensus StRE resembles the consensus binding sites for the Fosand Jun [TGA(G/C)TCA] ( 27 ), activatingtranscription factor (ATF)/CREB (TGACGTCA) ( 18 ), Maf[TGCTGA(G/C)TCAGCA or TGCTGACGTCAGCA] ( 28 ), and Cap`N' Collar (CNC)-basic region/leucine zipper (bZIP)[(T/C)GCTGA(G/C)TCA(C/T)] ( 8 ) subfamilies of bZIPtranscription factors that function as homo- or heterodimers. Forinstance, Fos family members dimerize most commonly with Jun proteins,and CNC-bZIP factors such as Nrf1 and Nrf2 dimerize most efficientlywith small Maf proteins, including MafG and MafK. "Supershift" EMSAreactions using antibodies directed against bZIP proteins were carriedout to identify specific StRE-binding proteins (StRE-BPs) and potentialheme-responsive transcription factor(s). Of several factors tested,only Fos and Jun proteins were consistently detected in controlcomplexes (Fig. 3, top, lanes 11 and 12 ). These proteins were also detected in the "heme" complexes, and the intensity of the supershifted bands did not changeappreciably in response to heme ( bottom ). Because pan-Fos and pan-Jun antibodies were used, individual family members were notidentified in this analysis. Heme-induced StRE-BPs included Nrf2, MafG,and ATF3 (the latter exhibited a weak signal but was consistentlyobserved in multiple experiments, whereas MafG was observedinconsistently in the control complexes but was routinely detected athigher intensity in the heme complexes). Other transcription factorstested, including Nrf1, MafK, ATF1, ATF2, and ATF4, were not detectedin the absence or presence of heme.1 v0 L" T% L" I. b

Inhibition of gene activity by dominant-negative mutants of Nrf2and Maf. To explore the functional role of the StRE-BPs identified in Multiple basic region/leucine zipper proteins bind toStRE in heme-regulated ho-1 gene expression, weexamined the effect of appropriate dominant-negative mutants (DNMs) onE1 transcription activity. Because Fos family members do nothomodimerize and heterodimerize most efficiently with Jun proteins, theFos DNM would be expected to directly inhibit endogenous Jun factors.Conversely, the Jun DNM would be expected to inhibit Fos familymembers. As shown in Fig. 4,overexpression of the Fos or Jun DNM did not appreciably alter basal orheme-dependent E1 activity, suggesting that such proteins are notinvolved in heme-mediated gene activation. The CREB DNM, which wouldmost efficiently inhibit CREB/ATF-type factors, increased basalluciferase activity but did not influence heme-induced luciferaseexpression. On the other hand, the Nrf2 and MafK DNMs, which would mosteffectively inhibit small Maf and CNC-bZIP proteins, respectively,significantly attenuated both basal and heme-dependent luciferaseactivity. The MafK DNM was more effective in inhibiting heme-dependentluciferase activity than basal activity (13 and 42% of controls,respectively), whereas the Nrf2 DNM exhibited similar inhibitoryactivities toward both heme-induced and basal luciferase expression (8 and 13% of controls, respectively). Although a DNM of MafK was used inthis experiment, an analogous mutant of MafG would be expected tobehave in a similar manner because of the structural and functionalsimilarities of small Maf proteins ( 35 ). Because MafG, butnot MafK, was detected in the EMSA analysis, we conclude thatNrf2 · MafG heterodimers are at least partly responsible for ho-1 gene activation in response to heme.

Fig. 4. Dominant-negative mutants of Nrf2 and MafK inhibit E1activity. IRPTCs were transfected, treated with vehicle or heme, andprocessed as described in MATERIALS AND METHODS and in thelegend to Fig. 1. Data are presented as percent of normalizedluciferase activity in the absence of exogenous dominant-negativemutants, and each data point represents the average ± SE from3-5 independent experiments. CREB, cAMP response element bindingprotein.1 U7 E0 V/ y# K$ M

Heme stimulates Nrf2 expression. To explore the mechanism(s) by which heme regulatesNrf2 · MafG function, we first examined theeffect of this agent on the steady-state level of each of theseproteins. As shown in Fig. 5 A,the abundance of Nrf2, but not of MafG, in WCEs increased in atime-dependent manner after treatment of IRPTCs with 10 µM 10-fold higher than inuntreated cells. The temporal pattern of Nrf2 induction wasqualitatively similar to that observed for HO-1. Previous studies( 13, 22, 25, 46, 55 ) have suggested that Nrf2 issequestered in the cytoplasm in an inactive form, and oxidants andxenobiotics activate Nrf2 in part by permitting transport into thenucleus. To determine whether heme regulates Nrf2 activity in thismanner, we monitored Nrf2 levels in cytoplasmic and nuclear extracts ofcells treated with vehicle (DMSO) or heme for different time periods.The underlying assumption of this experiment is that regulatedtransport would reveal a time-dependent decay in the level ofcytoplasmic Nrf2 with a concomitant increase in nuclear Nrf2. Lowlevels of Nrf2 were observed in the nuclear fraction, and this levelincreased substantially after 3 h of treatment with heme (Fig. 5 B, lanes 6-10 ). However, no Nrf2 was detected inthe cytoplasmic fraction ( lanes 1-5 ) at any time pointtested using this assay. The integrity of the cytoplasmic and nuclear fractions was confirmed by analysis of the compartment-specific -tubulin and histone H1 proteins.2 ^) c& @  R3 S1 F5 \
$ }. R. i- {3 B
Fig. 5. Heme induces expression of Nrf2. IRPTCs were treated withDMSO (D) or heme (H) for the indicated time period. Whole cell extracts( A ) or cytoplasmic (C; B ) and nuclear (N)fractions were prepared and subjected to Western blot analysis usingthe indicated antibodies as described in MATERIALS AND METHODS. Tub, -tubulin; H1, histone H1; HO-1, hemeoxygenase-1.4 n# T( T8 D/ K
- `9 T# {  N7 C) \  h
Heme decreases the rate of Nrf2 degradation. To delineate the mechanism of Nrf2 induction, we first examined theeffect of heme on the level Nrf2 mRNA. As shown in Fig. 6 A, treatment of IRPTCs withheme for up to 4 h did not alter the steady-state level of Nrf2transcripts. As expected, HO-1 40-foldduring this period. This result suggests that heme regulates Nrf2expression by a posttranscriptional mechanism(s). Pulse labelingexperiments indicated that heme does not influence the rate of Nrf2synthesis (data not shown), suggesting it may regulate Nrf2 stability.Nrf2 stability was examined by monitoring the decay of basal Nrf2 inthe absence of heme and of heme-induced Nrf2 in the presence of hemeafter inhibition of protein synthesis by cycloheximide. On the basis ofthese experiments, the half-life ( t 1/2 ) of Nrf2in unstimulated cells was calculated to be ~9.7 min. Heme stimulationincreased the t 1/2 to 107 min (Fig. 6 B ). Similar values were obtained in pulse-chase experiments (data not shown). As control, we monitored the induction and decay ofJunD, both of which were not appreciably altered in response to heme(Fig. 6 C ).2 r$ {% i+ s" l- L( H* @) P, E
2 X, \+ `8 z( ~1 T
Fig. 6. Posttranscriptional regulation of Nrf2 expressionby heme. A : heme does not affect the steady-state level ofNrf2 mRNA. IRPTCs were treated with 10 µM heme for the indicated timeperiod. Total RNA (5-µg portions) was dot blotted onto a nylonmembrane and successively hybridized to the indicated cDNA probes.Hybridization signals were detected and quantified with aphosphorimager. S3-normalized, relative mRNA levels for Nrf2 and HO-1are presented. B and C : heme promotesstabilization of Nrf2 ( B ) but not of JunD ( C ).IRPTCs were treated with DMSO ( Heme) or 10 µM heme in serum-freemedium for 2 h. The treatment media were replaced with the samemedia containing 100 µg/ml of cycloheximide, and cells were incubatedat 37°C for the indicated time period. Twenty-microgram portions ofnuclear extracts were analyzed by Western blotting as described in MATERIALS AND METHODS. Protein extract (2.5 µg) fromHEK-293 cells transfected with expression plasmids encoding mouse Nrf2or mouse JunD were used as positive controls ( ). The migration of themolecular mass (kDa) markers are indicated. Relative Nrf2 levels werequantified by densitometry, and each data point represents the averagevalue from 2 independent experiments.7 \/ Y4 G( z2 U

Heme does not stimulates the transcription activity of Nrf2 nordoes it affect Nrf2 · MafG dimerization. Because heme could potentially regulate Nrf2 and MafG function atadditional levels, we also examined the effects of this agent on thetranscription activity and heterodimerization potential of Nrf2 andMafG by mammalian one-hybrid and two-hybrid assays. For theseexperiments, transcription factor sequences were fused in-frame to theGdbd, and the fusions were tested for their ability to trans -activate a luciferase reporter gene under the control of five copies of the Gal-4 recognition sequence. Additionally, fortwo-hybrid assays, appropriate sequences were cloned into the pAD"activation domain" vector and tested for their ability to interactwith, and potentiate the activity of, Gdbd fusions. The parent Gdbdvector does not encode a transcription activation domain and thus didnot promote luciferase expression (Fig. 7 ). Similarly, MafG also does not containsuch a domain, and the Gdbd-MafG was transcriptionally inactive. Nrf2,on the other hand, encodes a potent activation domain and sponsored ahigh level of luciferase activity. Nrf2-mediated trans -activation, however, was not affected by heme.Deletion of the NH 2 -terminal half of Nrf2 eliminates theactivation domain, resulting in a protein (Nrf2 N) with no transcription activity. Nrf2 N does retain the leucine zipper dimerization domain, and association with the AD-MafG fusionelicited a high level of luciferase activity. AD-MafG also readilydimerized with full-length Nrf2; the rate and/or the extent ofNrf2 · MafG dimerization, however, was notaffected by heme. HO-1 is not expected to associate with Nrf2 and wasused as a negative control for these studies.7 A& a8 [, k- d5 z* E3 V
4 y7 q3 l2 s3 k4 u! d
Fig. 7. Heme does not stimulate the transcription activity ofNrf2 or Nrf2 · MafG dimerization. Transfectionand treatment of IRPTCs, preparation of cell extract, reporter enzymeassays, and data analysis were carried out as described in MATERIALS AND METHODS and in the legend to Fig. 1. PlasmidpFRluc was used as the experimental reporter gene. Equal mass amountsof the parent DNA-binding domain of Gal-4 (Gdbd) and activation domain(AD) plasmids or the fusion plasmids were used in the indicatedmixtures. Luciferase activities were normalized to that obtained withthe Gdbd-Nrf2 fusion. Data are presented as means ± SE; n = 3-4. Nrf2 N, NH 2 -terminaldeleted Nrf2.

DISCUSSION* F. o8 P: D. e
2 M- Q7 d, l* t
On the basis of the data obtained in this study, we propose thefollowing model for ho-1 gene activation by heme in renal epithelial cells (Fig. 8 ). Inunstimulated cells, Nrf2 is expressed at constitutive levels butrapidly degraded, either in the cytoplasm or after transit into thenucleus. Other StRE-BPs (depicted by circles and diamonds) bind to the ho-1 enhancers (squares) and effectively repress, orotherwise permit only low levels of, gene activity. One or more membersof the CREB/ATF family, not all of which were tested by antibodysupershift EMSA, may function in this capacity, as overexpression ofthe CREB DNM stimulates basal transcription activity of the E1 enhancer(Fig. 4 ). Other likely candidates of repressor StRE-BPsinclude heterodimers between Bach1 and small Maf proteins since theBach1 · MafK dimer is known to bind to the ho-1 StREs, and targeted deletion of the bach1 gene leads high constitutive expression of HO-1 mRNA and protein inseveral organs ( 43, 50 ) (studies to determine whether one or more of the 6 specific StRE/protein complexes detected by EMSA inunstimulated IRPTCs contain Bach1 protein are currently in progress).Upon cellular stimulation, heme interferes with the Nrf2 degradationpathway, permitting accumulation of the transcription factor in thenucleus, where it heterodimerizes with MafG.Nrf2 · MafG heterodimers displace some of therepressor StRE-BPs bound to the ho-1 enhancers and promotehigh rates of transcription. Additionally, heme may directly interferewith the binding of repressor StRE-BPs, as was recently demonstratedfor Bach1 · MafK heterodimers ( 43, 50 ). Overall, heme-mediated activation of the ho-1 gene, therefore, likely reflects the net effect of relief of repression(i.e., inhibition of repressor StRE-BPs) and the positive action ofNrf2. Although renal epithelial cells, a relevant target for hemereleased during hemolysis or rhabdomyolysis, were used in the presentstudy, we suspect that the mechanism described herein is generallyoperative in other cell types since heme-mediated activation of the ho-1 gene is an ubiquitous phenomenon.- u8 U9 b. B4 `* |
7 m% k" L7 K( U; T/ x
Fig. 8. Model of ho-1 gene regulation in the absence( A ) or presence ( B ) of heme. N, Nrf2; M, MafG; H,heme. See text for additional details." Z) q. T- p' a2 |$ |

How does heme stabilize Nrf2? We consider it unlikely that the hememolecule directly associates or interferes with a component of the Nrf2degradation pathway as other, structurally distinct HO-1 inducers,including cadmium and arsenite, also promote Nrf2 stability (Stewart,Killeen, and Alam, unpublished observations). All of theseagents have in common the ability to generate cellular oxidativestress, and it is probably this common condition, or more precisely, asignal generated during this state, that triggers the stabilizationprocess. In a prooxidant state, increased Nrf2 levels and activitywould result in activation of select genes encoding proteins, such asNAD(P)H:quinone oxidoreductase, -glutamylcysteine synthase, orglutathione S -transferase, with antioxidant and xenobiotic detoxification activities ( 14, 24 ). Induction of HO-1would represent one component of this homeostatic response. Increased HO-1 activity will lead to elimination of the prooxidant heme molecules, those internalized from the extracellular environment andthose released intracellularly on destabilization of heme proteins. Incertain context, induction of HO-1 is accompanied by increasedsynthesis of apoferritin ( 38 ), which can sequester thereleased heme iron. Additionally, HO-1-catalyzed degradation of hemegenerates the antioxidants biliverdin and bilirubin ( 49 ), promoting further abatement of the oxidative environment. Attenuation of the oxidative stress, and return to a normal reducing environment, will lead to resumption of normal Nrf2 degradation and downmodulation of target genes such as HO-1.6 |& G! q: R2 w+ o6 L

In this study (and as illustrated in Fig. 8 ), we have identified theStREs and transcription factor Nrf2 as the key components of theheme-dependent ho-1 gene regulatory circuit. In this and other respects, the proposed mechanism is remarkably similar to thatfor heme-dependent regulation of the thioredoxin gene ( trx ) in K562 human erythroleukemia cells ( 29 ). Activation ofthe trx gene is also proposed to be mediated by Nrf2 and anantioxidant response element (ARE), 5'-TGCTGAGTAAC-3', that is verysimilar to the ho-1 StREs. Furthermore, as proposed for the ho-1 gene (see Fig. 8 ), activation of the trx gene also involves exchange of factors bound at the ARE. Inunstimulated K562 cells, a dimer composed of a small Maf protein andthe CNC-bZIP factor NF-E2p45 is proposed to occupy the trx ARE. Upon heme stimulation, an Nrf2 · small Mafheterodimer replaces the p45 · Maf factor andpromotes higher levels of transcription. It is unlikely that thep45 · Maf dimer plays a similar role in ho-1 gene regulation in IRPTCs. Normal expression of p45 islimited to the erythroid lineage cells ( 8 ), and we did notdetect this protein in antibody supershift EMSA assays (data notshown). On the other hand, activator protein-1 (Fos/Jun) proteins mayfunction in this capacity since these proteins were identified asStRE-BPs in unstimulated IRPTCs. Interestingly, Fos/Jun factors did notbind to the trx ARE either in unstimulated orheme-stimulated K562 cells, pointing to mechanistic differences between ho-1 and trx gene activation by heme.

The mechanism described here differs from that proposed for trx gene activation in an even more fundamental respect,namely the process by which heme modulates Nrf2 activity. Theprevailing model for the regulation of Nrf2 function stipulates that,under normal conditions, Nrf2 exists in an inactive,cytoplasm-localized state, in part or fully as a consequence of bindingto the cytoskeleton-associated protein Keap1 ( 25, 30 ). Oncellular stimulation by xenobiotics, electrophiles, or oxidativestress-generating agents, the cytoplasmic-retention mechanism isinactivated, and Nrf2 is transported to the nucleus by an as yetuncharacterized mechanism(s) but one that, under certain circumstances,may involve protein kinase C-mediated phosphorylation of Nrf2( 22 ). In the nucleus, Nrf2 dimerizes with other bZIP factors, including Jun ( 52 ), ATF4 ( 19 ), andsmall Maf proteins ( 23, 33 ), and the resultingheterodimers bind to response elements to regulate target gene transcription.
2 Y% a; u: K4 J$ b4 t
In line with the above model, Kim et al. ( 29 ) haveproposed that heme stimulates trx gene activity by promotingtransport of Nrf2 from the cytoplasm to the nucleus. Ideally, for suchregulated nuclear transport, and as documented for other transcriptionfactors ( 26, 45, 47 ), one would expect a decay in thelevel of cytoplasmic Nrf2 concomitant with the increase in nuclearNrf2. In our studies, however, we have been unable to detect anycytoplasmic Nrf2 in unstimulated or stimulated IRPTCs, making itdifficult to implicate the Keap1-dependent pathway in regulation ofNrf2 activity by heme. Indeed, our data are more supportive of amechanism in which Nrf2 is transported into the nucleus by aconstitutive rather than a regulated transport pathway, and heme (andother stimuli) enhances Nrf2 activity by promoting Nrf2 protein stabilization.5 t% l# c) g% f6 `/ \; c

Reconciliation of these divergent mechanisms will require additionalexperimentation, but one possibility is readily obvious. The functionaloperation of the regulated, subcellular trafficking mechanism should becritically dependent on the relative levels of Keap1 and Nrf2, whichare also likely to vary in a cell type-dependent manner. In some cells,Keap1 may be expressed at extremely low levels, and the majority of theNrf2 will be constitutively transported into the nucleus and notdetected in the cytoplasmic fraction. In such cells, stimulation ofNrf2 activity will result primarily from inducer-dependent regulationof Nrf2 turnover, as described in this report. In IRPTCs, we detectvery low levels of Keap1 mRNA; measurement of the amount of Keap1protein, however, has not been possible because of the lack ofanti-Keap1 antibodies (data not shown). In cells expressing higherlevels of Keap1, Nrf2 will associate with Keap1 and presumably bypassthe degradation apparatus, resulting in detectable levels ofcytoplasmic Nrf2, as found in K562 ( 29 ) and other cells( 22, 34 ). In such cells, it is likely that both nucleartransport and inhibition of protein degradation contribute to theoverall induction of Nrf2 activity in response to stimuli.

The findings derived from the mammalian one-hybrid and two-hybridassays are consistent with the view that heme does not modulate thetranscription potential of Nrf2 or its rate of dimerization with MafG.In this regard, we point out that the lack of activation of theGdbd-Nrf2 fusion by heme is not necessarily at odds with a mechanism inwhich heme promotes stability of Nrf2, the latter mechanism possiblyleading to the expectation that heme would also stimulate stabilizationof the fusion protein. For example, it is quite tenable thatstabilization of Nrf2 requires a protein present in sufficientlylimiting amounts such that its phenotypic effects may not be observedin transient transfection assays where the exogenous protein (e.g.,Gdbd-Nrf2) is expressed at artificially high concentrations pertransfected cell. Alternatively, or additionally, stabilization of Nrf2is likely to involve specific cis -acting structural signalsor domains ( 21 ) that may be masked or become inoperativeon fusion to the Gdbd. Resolution of this issue requires furthercharacterization of the mechanism of Nrf2 stabilization, the latterresiding beyond the scope of the present study.

In summary, our studies provide, to the best of our knowledge, thefirst mechanistic analysis of heme-mediated ho-1 geneactivation in renal epithelial cells, in the course of which we haveidentified the key components of this regulatory circuit, namely theStREs and transcription factor Nrf2. In addition, we have uncovered anovel system, heme-dependent stabilization of Nrf2 protein, forregulation of Nrf2 activity; notably, Nrf2 function is regulated primarily by posttranslational processes, both at the level of subcellular compartmentalization and protein turnover. Our findings addto the growing appreciation of the relevance of Nrf2 to mechanisms ofrenal injury, and, in this regard, it is germane that Nrf2-deficient female mice develop lupus-like autoimmune nephritis ( 54 ).In light of our prior observations demonstrating the exaggeration ofrenal inflammation in stressed ho-1 / mutantmice ( 42 ) and our present observations attesting to induction of HO-1 via Nrf2-dependent pathways, we speculate that thedevelopment of lupus-like autoimmune nephritis in Nrf2-deficient micereflects, at least in part, an inability to induce HO-1, and therebyrestrain inflammatory responses in the kidney.$ K6 A8 D6 Q# `' ]

ACKNOWLEDGEMENTS1 F. F9 i& q+ F$ T+ E' e9 m
7 s. _  s, l: x4 s. |
We thank Margaret Overstreet for assistance in preparation of the manuscript.- m. |. u, ^* c# h8 x/ f
      【参考文献】
1. Aft, RL,andMueller GC. Hemin-mediated DNA strand scission. J Biol Chem 258:12069-12072,1983 ." n8 Z1 q4 N  n* G: ]

- i! S$ F" h6 `3 P
2. Aft, RL,andMueller GC. Hemin-mediated oxidative degradation of proteins. J Biol Chem 259:301-305,1984 .# @; g% P. C& [
! `1 o1 v9 M- U3 _# s

. o2 Z& ~+ b# Z+ W1 M( m. H  t
3. Ahn, S,Olive M,Aggarwal S,Krylov D,Ginty DD,andVinson C. A dominant-negative inhibitor of CREB reveals that it is a general mediator of stimulus-dependent transcription of c- fos. Mol Cell Biol 18:967-977,1998 .
2 C; k9 ?- y9 q" M3 I3 A+ Y

4. Alam, J. Functional analysis of the heme oxygenase-1 gene promoter.In: Current Protocols in Toxicology, edited by Maines MD,Costa LG,Reed DJ,Sassa S,and Sipes IG.. New York: Wiley, 2000, p. 9.7.1-9.7.21.
/ m" z& W" P& m6 r' N$ M
: p! r+ u+ T4 ]" H$ a
2 s3 m8 d( P6 E/ V5 R* _9 b
5. Alam, J,Cai J,andSmith A. Isolation and characterization of the mouse heme oxygenase-1 gene. Distal 5' sequences are required for induction by heme or heavy metals. J Biol Chem 269:1001-1009,1994 .( l) e) i, S! [

6. Alam, J,Stewart D,Touchard C,Boinapally S,Choi AM,andCook JL. Nrf2, a Cap'n'Collar transcription factor, regulates induction of the heme oxygenase-1 gene. J Biol Chem 274:26071-26078,1999 .: B; x! e+ D6 c% V
& W0 d2 ?, y. i. w

7. Alam, J,Wicks C,Stewart D,Gong P,Touchard C,Otterbein S,Choi AM,Burow ME,andTou J. Mechanism of heme oxygenase-1 gene activation by cadmium in MCF-7 mammary epithelial cells. Role of p38 kinase and Nrf2 transcription factor. J Biol Chem 275:27694-27702,2000 .
% `# M2 L* H3 i
2 J* S) M' q& q( x7 k( @. g9 o
) B/ N8 [& ]. ?- Y
8. Andrews, NC,Erdjument-Bromage H,Davidson MB,Tempst P,andOrkin SH. Erythroid transcription factor NF-E2 is a haematopoietic-specific basic-leucine zipper protein. Nature 362:722-728,1993 .$ }& t6 Y5 l- [) x$ Q2 y/ L
: ?) L: E; @, `1 `

9. Balla, G,Jacob HS,Eaton JW,Belcher JD,andVercellotti GM. Hemin: a possible physiological mediator of low density lipoprotein oxidation and endothelial injury. Arterioscler Thromb 11:1700-1711,1991 .# }8 z0 c/ \5 F" i8 y* F
6 J" j: Y" m+ e. z8 j: y4 E

; k. ~* s% ~: L) @, q) w1 Z: I
10. Balla, G,Vercellotti GM,Muller-Eberhard U,Eaton J,andJacob HS. Exposure of endothelial cells to free heme potentiates damage mediated by granulocytes and toxic oxygen species. Lab Invest 64:648-655,1991  .1 n# O+ H9 T4 a* v5 v6 K
5 j9 z3 d' k' m  s& C! C) E

11. Balla, J,Jacob HS,Balla G,Nath K,Eaton JW,andVercellotti GM. Endothelial-cell heme uptake from heme proteins: induction of sensitization and desensitization to oxidant damage. Proc Natl Acad Sci USA 90:9285-9289,1993 .
' x& a7 {2 j3 S; Y5 J0 ~6 k
- n$ U  P& x! j9 H( u  i6 K

12. Chomczynski, P,andSacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162:156-159,1987  .  }, k4 F- |( }
4 y! A; i6 Y1 N
! D# ~; ]$ n" L+ n
. G0 |3 S" u+ u. {  `
13. Dhakshinamoorthy, S,andJaiswal AK. Functional characterization and role of INrf2 in antioxidant response element-mediated expression and antioxidant induction of NAD(P)H:quinone oxidoreductase1 gene. Oncogene 20:3906-3917,2001  .0 Y4 \1 S% E% P) F* P

+ F( U& L. r7 a9 D0 {
- i) H8 O5 X* l. f' ~1 D
14. Dhakshinamoorthy, S,Long DJ, II,andJaiswal AK. Antioxidant regulation of genes encoding enzymes that detoxify xenobiotics and carcinogens. Curr Top Cell Regul 36:201-216,2000  .% Q* T  _( A0 {+ X, a# }* n8 n- A
8 d8 u+ Z; x' W8 |; _% R4 o

4 n9 X6 d9 ^% x
15. Gong, P,Hu B,Stewart D,Ellerbe M,Figueroa YG,Blank V,Beckman BS,andAlam J. Cobalt induces heme oxygenase-1 expression by a hypoxia-inducible factor-independent mechanism in Chinese hamster ovary cells: regulation by Nrf2 and MafG transcription factors. J Biol Chem 276:27018-27025,2001 .
. ]0 h8 E1 i- j! c+ N8 @$ r! j
+ L9 u/ L* D2 f8 b) o8 x" T

16. Gong, P,Ogra Y,andKoizumi S. Inhibitory effects of heavy metals on transcription factor Sp1. Ind Health 38:224-227,2000  .

. Y/ B+ x5 i: I0 U3 F) c. E

17. Gong, P,Stewart D,Hu B,Li N,Cook J,Nel A,andAlam J. Activation of the mouse heme oxygenase-1 gene by 15-deoxy-delta12,14-prostaglandin j2 is mediated by the stress response elements and transcription factor nrf2. Antioxid Redox Signal 4:249-257,2002  .; q- p& e+ L# d

9 E* \) r! l6 M3 |& N5 [
0 }& _0 ~4 Z3 ]) C0 Z% O+ x
18. Hai, T,Wolfgang CD,Marsee DK,Allen AE,andSivaprasad U. ATF3 and stress responses. Gene Expr 7:321-335,1999  .6 B4 Q) W8 I. l

/ P7 e6 i2 e5 V- f8 @# N
19. He, CH,Gong P,Hu B,Stewart D,Choi ME,Choi AM,andAlam J. Identification of activating transcription factor 4 (ATF4) as an Nrf2-interacting protein. Implication for heme oxygenase-1 gene regulation. J Biol Chem 276:20858-20865,2001 .4 ^7 @2 G) l6 U1 ~0 M& z
: J8 ?, }0 `5 S7 J: X. a

; u% n" u. P4 v( }3 n* M8 ~$ U
20. Hebbel, RP,andEaton JW. Pathobiology of heme interaction with the erythrocyte membrane. Semin Hematol 26:136-149,1989  .) j" Q. ]- E6 J, W& q% a! q
6 x) B' O: E- R  u6 w) l3 U

1 z  Y4 b. n& i. f! e
21. Hershko, A,andCiechanover A. The ubiquitin system. Annu Rev Biochem 67:425-479,1998  .4 d7 o- X1 _* {) T' x

22. Huang, HC,Nguyen T,andPickett CB. Regulation of the antioxidant response element by protein kinase C-mediated phosphorylation of NF-E2-related factor 2. Proc Natl Acad Sci USA 97:12475-12480,2000 .9 x# U3 Y4 U7 o6 O  O

1 C4 s- g1 _9 ?4 y5 S- S
4 O* V: w& v1 o, w  v( u
23. Itoh, K,Igarashi K,Hayashi N,Nishizawa M,andYamamoto M. Cloning and characterization of a novel erythroid cell-derived CNC family transcription factor heterodimerizing with the small Maf family proteins. Mol Cell Biol 15:4184-4193,1995 .1 X2 J$ u8 J& p, j8 U
5 c: Z  E, p3 l: ~

; D5 r4 L8 R$ d' u, j0 d1 {. l
24. Itoh, K,Ishii T,Wakabayashi N,andYamamoto M. Regulatory mechanisms of cellular response to oxidative stress. Free Radic Res 31:319-324,1999  .
0 i0 k1 P$ Q% }. L. P

25. Itoh, K,Wakabayashi N,Katoh Y,Ishii T,Igarashi K,Engel JD,andYamamoto M. Keap1 represses nuclear activation of antioxidant responsive elements by Nrf2 through binding to the amino-terminal Neh2 domain. Genes Dev 13:76-86,1999 .; a! E, L% X6 N: `5 u& k/ Q
" ^; O9 `* e0 h% Q! k; u( f
, n9 V( Z9 Q  U9 Q0 q
, Z& d' h! H4 W# Z
26. Kallio, PJ,Okamoto K,O'Brien S,Carrero P,Makino Y,Tanaka H,andPoellinger L. Signal transduction in hypoxic cells: inducible nuclear translocation and recruitment of the CBP/p300 coactivator by the hypoxia-inducible factor-1. EMBO J 17:6573-6586,1998  .

8 @/ h3 f# C2 `; F1 p2 S9 C
6 j, w5 S/ ]$ o4 T
27. Karin, M,Liu Z,andZandi E. AP-1 function and regulation. Curr Opin Cell Biol 9:240-246,1997  .

5 T5 g! A/ W. _
28. Kataoka, K,Noda M,andNishizawa M. Maf nuclear oncoprotein recognizes sequences related to an AP-1 site and forms heterodimers with both Fos and Jun. Mol Cell Biol 14:700-712,1994 .5 K  Y9 W) d3 W5 b$ |" D: O) H
/ i" H. h: L6 M; |) q  s
" Y2 }2 A7 M. ^7 W" A# T

29. Kim, YC,Masutani H,Yamaguchi Y,Itoh K,Yamamoto M,andYodoi J. Hemin-induced activation of the thioredoxin gene by Nrf2. A differential regulation of the antioxidant responsive element by a switch of its binding factors. J Biol Chem 276:18399-18406,2001 .
3 H: b+ p; F* O: C3 N4 n

30. Kobayashi, M,Itoh K,Suzuki T,Osanai H,Nishikawa K,Katoh Y,Takagi Y,andYamamoto M. Identification of the interactive interface and phylogenic conservation of the Nrf2-Keap1 system. Genes Cells 7:807-820,2002 .5 }' Z! M4 R9 G, u/ _3 }- f& x+ p
" T- z" W+ K4 W% [

31. Kotkow, KJ,andOrkin SH. Dependence of globin gene expression in mouse erythroleukemia cells on the NF-E2 heterodimer. Mol Cell Biol 15:4640-4647,1995 ./ t, d+ |: S7 C& z6 V3 [( B

0 ^) U! C0 a- G

32. Maines, MD,andPanahian N. The heme oxygenase system and cellular defense mechanisms. Do HO-1 and HO-2 have different functions? Adv Exp Med Biol 502:249-272,2001  .

; S  P$ g6 S- }9 L/ ?6 \  W
33. Marini, MG,Chan K,Casula L,Kan YW,Cao A,andMoi P. hMAF, a small human transcription factor that heterodimerizes specifically with Nrf1 and Nrf2. J Biol Chem 272:16490-16497,1997 .
+ ^5 a; p/ Y2 a/ Y) v" G6 g. Z1 _! p
' k- }7 n  u: ]) v& a" @+ I; N

34. Morimitsu, Y,Nakagawa Y,Hayashi K,Fujii H,Kumagai T,Nakamura Y,Osawa T,Horio F,Itoh K,Iida K,Yamamoto M,andUchida K. A sulforaphane analog that potently activates the Nrf2-dependent detoxification pathway. J Biol Chem 277:3456-3463,2002 .
  |- S' a6 O2 {2 D6 p( D) G
6 o9 v" Q6 U( P; x  `8 s
- B( n9 a! i2 c- k7 ]+ K
35. Motohashi, H,Shavit JA,Igarashi K,Yamamoto M,andEngel JD. The world according to Maf. Nucleic Acids Res 25:2953-2959,1997 .( Y8 w9 O0 {/ i- N* ?3 I
' r; _) E5 q& N* b/ B( x, i: H2 `8 h" o
+ o8 u* S$ Q" R  m
  A! |$ _7 t2 T$ Q5 P: S* ?
36. Muller-Eberhard, U,andFraig M. Bioactivity of heme and its containment. Am J Hematol 42:59-62,1993  .7 M- u* f* J- p2 B4 G, [
4 a- v' I# \' c  ?5 G: D2 O1 @- \
/ `' ?7 o6 n$ [, o# X0 t8 J% ?

37. Murata, M,Gong P,Suzuki K,andKoizumi S. Differential metal response and regulation of human heavy metal-inducible genes. J Cell Physiol 180:105-113,1999  .
* e& [# H$ n6 {9 n* S& N/ ]
; {1 A9 [$ j: b/ r& W/ d# |

38. Nath, KA,Balla G,Vercellotti GM,Balla J,Jacob HS,Levitt MD,andRosenberg ME. Induction of heme oxygenase is a rapid, protective response in rhabdomyolysis in the rat. J Clin Invest 90:267-270,1992  .3 k5 v! J, d, B7 M9 m4 f0 d

39. Nath, KA,Balla J,Croatt AJ,andVercellotti GM. Heme protein-mediated renal injury: a protective role for 21-aminosteroids in vitro and in vivo. Kidney Int 47:592-602,1995  .  y' d: h" R7 K

40. Nath, KA,Grande JP,Croatt AJ,Likely S,Hebbel RP,andEnright H. Intracellular targets in heme protein-induced renal injury. Kidney Int 53:100-111,1998  .
8 w6 d: E" n3 N% s2 @% h
, p. w  X& z( f9 B7 ^2 v8 y

41. Nath, KA,Haggard JJ,Croatt AJ,Grande JP,Poss KD,andAlam J. The indispensability of heme oxygenase-1 in protecting against acute heme protein-induced toxicity in vivo. Am J Pathol 156:1527-1535,2000 ., B" r" t( h! ^' o) W
: T5 C: B5 V0 q3 ]

42. Nath, KA,Vercellotti GM,Grande JP,Miyoshi H,Paya CV,Manivel JC,Haggard JJ,Croatt AJ,Payne WD,andAlam J. Heme protein-induced chronic renal inflammation: suppressive effect of induced heme oxygenase-1. Kidney Int 59:106-117,2001  .
, z/ n1 b/ |  D3 R; I) p1 R

43. Ogawa, K,Sun J,Taketani S,Nakajima O,Nishitani C,Sassa S,Hayashi N,Yamamoto M,Shibahara S,Fujita H,andIgarashi K. Heme mediates derepression of Maf recognition element through direct binding to transcription repressor Bach1. EMBO J 20:2835-2843,2001  .

3 O9 s- V+ _0 b( U
44. Olive, M,Krylov D,Echlin DR,Gardner K,Taparowsky E,andVinson C. A dominant negative to activation protein-1 (AP1) that abolishes DNA binding and inhibits oncogenesis. J Biol Chem 272:18586-18594,1997 .

- x' C# m, N2 H1 h8 [# p
" I+ o  R' \# o% J+ p' Q% A9 R
45. Roberts, BJ,andWhitelaw ML. Degradation of the basic helix-loop-helix/Per-ARNT-Sim homology domain dioxin receptor via the ubiquitin/proteasome pathway. J Biol Chem 274:36351-36356,1999 .. m9 V* [7 H: ]4 `+ r
2 s: J" p$ R' n& D4 A
' Y. W9 e4 H; t

46. Sekhar, KR,Spitz DR,Harris S,Nguyen TT,Meredith MJ,Holt JT,Guis D,Marnett LJ,Summar ML,andFreeman ML. Redox-sensitive interaction between KIAA0132 and Nrf2 mediates indomethacin-induced expression of -glutamylcysteine synthetase. Free Radic Biol Med 32:650-662,2002  .6 u7 a+ @- W+ F4 _! j* b
  h' ~2 z+ y2 [5 B! X9 O! N) o
2 e) B/ ^, b+ {- x/ K( v3 ]
* Y% C! u  F# ]5 K, j0 s4 R
47. Shuai, K,Stark GR,Kerr IM,andDarnell JE, Jr. A single phosphotyrosine residue of Stat91 required for gene activation by interferon-. Science 261:1744-1746,1993 .

48. Stocker, R. Induction of haem oxygenase as a defence against oxidative stress. Free Radic Res Commun 9:101-112,1990  .# g; |3 x" ]5 o2 T- B* [. c
5 G- r; a1 q$ W7 E! r$ {
3 M* `) c$ [  F

49. Stocker, R,McDonagh AF,Glazer AN,andAmes BN. Antioxidant activities of bile pigments: biliverdin and bilirubin. Methods Enzymol 186:301-309,1990 .
: d$ I- d9 m$ N2 G+ s
( j- Z9 ~0 U6 l/ w$ Y4 @

50. Sun, J,Hoshino H,Takaku K,Nakajima O,Muto A,Suzuki H,Tashiro S,Takahashi S,Shibahara S,Alam J,Taketo MM,Yamamoto M,andIgarashi K. Hemoprotein Bach1 regulates enhancer availability of heme oxygenase-1 gene. EMBO J 21:5216-5224,2002  .3 N" ?2 K. U8 E  T/ w

% y2 T7 s1 I  N, f" k' c- a
51. Tang, SS,Jung F,Diamant D,Brown D,Bachinsky D,Hellman P,andIngelfinger JR. Temperature-sensitive SV40 immortalized rat proximal tubule cell line has functional renin-angiotensin system. Am J Physiol Renal Fluid Electrolyte Physiol 268:F435-F446,1995 .+ |; a, M. z. F+ w6 O0 w
# n: W5 D4 R  r) n( {; c" p
+ B( Q/ d, z7 x* q# ~) H

52. Venugopal, R,andJaiswal AK. Nrf2 and Nrf1 in association with Jun proteins regulate antioxidant response element-mediated expression and coordinated induction of genes encoding detoxifying enzymes. Oncogene 17:3145-3156,1998  .0 \, O; U; k# R# b2 h9 q
9 v' ~; z8 O. L* L+ ?8 @

9 g+ h( x6 P, g
53. Vogt, BA,Alam J,Croatt AJ,Vercellotti GM,andNath KA. Acquired resistance to acute oxidative stress. Possible role of heme oxygenase and ferritin. Lab Invest 72:474-483,1995  .: K: q( j- u  H5 n1 G9 l

9 P6 d, U7 K0 {' u" l

54. Yoh, K,Itoh K,Enomoto A,Hirayama A,Yamaguchi N,Kobayashi M,Morito N,Koyama A,Yamamoto M,andTakahashi S. Nrf2-deficient female mice develop lupus-like autoimmune nephritis. Kidney Int 60:1343-1353,2001  .

8 I% l* m8 @1 X/ _

55. Zipper, LM,andMulcahy RT. The Keap1 BTB/POZ dimerization function is required to sequester Nrf2 in cytoplasm. J Biol Chem 277:36544-36552,2002 .