Arsenite down-regulates cytochrome P450 1A1 at the transcriptional and posttranslational levels in human HepG2 cells

Free Radical Biology & Medicine 48 (2010) 1399–1409

Contents lists available at ScienceDirect

Free Radical Biology & Medicine

j ourna l homepage: www.e lsev ie r.com/ locate / f reeradb iomed

Original Contribution

Arsenite down-regulates cytochrome P450 1A1 at the transcriptional andposttranslational levels in human HepG2 cells

Anwar Anwar-Mohamed, Ayman O.S. El-Kadi ⁎Faculty of Pharmacy and Pharmaceutical Sciences, University of Alberta, Edmonton, AB, Canada T6G 2N8

⁎ Corresponding author. Fax: +1 780 492 1217.E-mail address: [email protected] (A.O.

0891-5849/$ – see front matter © 2010 Elsevier Inc. Adoi:10.1016/j.freeradbiomed.2010.02.027

a b s t r a c t

Article history:Received 21 January 2010Revised 18 February 2010Accepted 19 February 2010Available online 25 February 2010

Keywords:ArseniteCYP1A1Gene expressionHeme oxygenase-1Heavy metalsAryl hydrocarbon receptorFree radicals

Aryl hydrocarbon receptor (AhR) ligands, typified by 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), andmetals, typified by arsenite (As(III)), are environmental cocontaminants, and their molecular interactionmay disrupt the coordinated regulation of the carcinogen activating enzyme cytochrome P450 1A1(CYP1A1). Therefore, in this study we examined the effects of coexposure to As(III) and TCDD on theexpression of CYP1A1 in HepG2 cells. Our results showed that As(III) caused a dose-dependent decrease inTCDD-mediated induction of CYP1A1 mRNA, protein, and catalytic activity levels. As(III) significantlyinhibited TCDD-mediated induction of AhR-dependent luciferase reporter gene expression without alteringCYP1A1 mRNA stability. In addition, As(III) increased heme oxygenase-1 (HO-1) mRNA, which coincidedwith a further decrease in the CYP1A1 catalytic activity levels. When a competitive HO-1 inhibitor, tinmesoporphyrin, was applied to HepG2 cells or the cells were transfected with siRNA for HO-1 there was apartial restoration of the inhibition of TCDD-mediated induction of CYP1A1 catalytic activity. Treatment ofcells with heme or hemoglobin partially restored the As(III)-mediated inhibition of CYP1A1 catalytic activity.On the other hand, cobalt protoporphyrin increased HO-1 mRNA, with a concomitant decrease in CYP1A1activity, without affecting CYP1A1 mRNA, which was reversed by HO-1 siRNA transfection. This studydemonstrates that As(III) down-regulates CYP1A1 through transcriptional and posttranslational mechan-isms. In addition, HO-1 is involved in the As(III)-mediated down-regulation of CYP1A1 at the catalyticactivity level.

S. El-Kadi).

ll rights reserved.

© 2010 Elsevier Inc. All rights reserved.

The aryl hydrocarbon receptor (AhR) is a ligand-activatedcytoplasmic transcription factor that belongs to the basic-helix-loop-helix protein family. Inactive AhR resides in the cytoplasmattached to a complex of two heat shock proteins 90, hepatitis B virusX-associated protein 2, and the chaperone protein p23 [1,2]. Uponligand binding, the activated AhR dissociates from the cytoplasmiccomplex and translocates to the nucleus, where it dimerizes with thearyl hydrocarbon nuclear translocator (Arnt) [3]. The ligand/AhR/Arnt complex then acts as a transcription factor that binds to a specificDNA recognition sequence, GCGTG, within the xenobiotic-responsiveelement (XRE), located in the promoter region of a battery of genestermed AhR-regulated genes [4,5]. Among the AhR-regulated genes,cytochrome P450 1A1 (CYP1A1) is the most capable of bioactivatingthe toxic and environmental contaminants polycyclic aromatichydrocarbons (PAHs) and halogenated aromatic hydrocarbons(HAHs) to carcinogenic metabolites. Of interest, it has been shownthat the toxicological effects of PAHs and the more toxic HAHs,typified by 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), are mainly

mediated through the activation of AhR and consequently CYP1A1. Infact, a well-established link between induction of CYP1A1 and cancerhas been previously reported [6].

The toxicological effects of individual AhR ligands have beenextensively studied, yet the combined toxicological effects of theseligands with metals such as arsenic have been a matter of debateover the past decade. Arsenic is widely distributed in theenvironment and can be found as organic and inorganic compoundsin the trivalent (arsenite, As(III)) or the pentavalent (arsenate, As(V)) state [7]. Inorganic As(III) is released into the environmentfrom primary copper, zinc, and lead smelters and glass, pesticide,and herbicide manufacturing plants [7]. It is also found in cigarettesmoke, arsenic-treated wood, and agricultural fertilizers [8].Despite its toxicity, As(III) has been used as a therapeutic agentfor more than 2400 years in traditional Chinese medicine and inancient Greek and Roman times [9]. Interestingly, As(III), as arsenictrioxide, is being used for the treatment of acute promyelocyticleukemia and multiple myeloma [9–11].

Previous reports from our laboratory and others have shown thatAs(III) is capable of altering the expression of the carcinogenactivating enzyme CYP1A1 at various levels of its signaling pathway.Therefore, the objective of this study was to determine the possibleeffects of As(III) on the TCDD-mediated induction of CYP1A1, and

1400 A. Anwar-Mohamed, A.O.S. El-Kadi / Free Radical Biology & Medicine 48 (2010) 1399–1409

further to investigate the underlying molecular mechanisms involvedin this alteration.

We provide here the first evidence that As(III) down-regulates theexpression of CYP1A1 by affecting its transcriptional level with afurther decrease in its catalytic activity levels. The increased inhibitoryeffect of As(III) on the TCDD-mediated induction of CYP1A1 catalyticactivity is merely attributed to the As(III)-mediated induction of hemeoxygenase-1 (HO-1), which subsequently leads to the formation of ahollow functionless CYP1A1 protein.

Materials and methods

Materials

3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide(MTT), 2,6-dichlorophenolindophenol, 7-ethoxyresorufin, 7-methox-yresorufin, fluorescamine, anti-goat IgG peroxidase secondary anti-body, hemin, human hemoglobin, reduced nicotinamide adeninedinucleotide (NADPH), protease inhibitor cocktail, and sodiumarsenite (NaAsO2) were purchased from Sigma Chemical Co. (St.Louis, MO, USA). Tin mesoporphyrin (SnMP) and cobalt protopor-phyrin (CoPP) were purchased from Frontier Scientific (Logan, UT,USA). 2,3,7,8-Tetrachlorodibenzo-p-dioxin, N99% pure, was pur-chased from Cambridge Isotope Laboratories (Woburn, MA, USA).TRIzol reagent and Lipofectamine 2000 reagents were purchased fromInvitrogen (San Diego, CA, USA). The High-Capacity cDNA reversetranscription kit, SYBR green PCR master mix, human Hmox1 (HO-1)validated siRNA, and Silencer Select Negative Control No. 2 siRNAwere purchased from Applied Biosystems (Foster City, CA, USA).INTERFERin siRNA transfecting reagent was supplied by PolyplusTransfection (Illkirch, France). Actinomycin-D (Act-D) was purchasedfrom Calbiochem (San Diego, CA, USA). Chemiluminescence Westernblotting detection reagents were from GE Healthcare Life Sciences(Piscataway, NJ, USA). Nitrocellulose membrane was purchased fromBio-Rad Laboratories (Hercules, CA, USA). CYP1A1 goat polyclonalprimary antibody, GAPDH rabbit polyclonal antibody, and anti-rabbitIgG peroxidase secondary antibody were purchased from Santa CruzBiotechnology (Santa Cruz, CA, USA). HO-1 mouse monoclonalprimary antibody was purchased from Abcam (Cambridge, MA,USA). Anti-mouse IgG peroxidase secondary antibody was purchasedfrom R&D Systems (Minneapolis, MN, USA). Luciferase assay reagentswere obtained from Promega (Madison, WI, USA). All other chemicalswere purchased from Fisher Scientific (Toronto, ON, Canada).

Cell culture

The HepG2 cell line, ATCC HB-8065 (Manassas, VA, USA), wasmaintained in Dulbecco's modified Eagle's medium (DMEM) withphenol red, supplemented with 10% heat-inactivated fetal bovineserum, 20 μM L-glutamine, 50 μg/ml amikacin, 100 IU/ml penicillin,10 μg/ml streptomycin, 25 ng/ml amphotericin B, 0.1 mM nonessen-tial amino acids, and vitamin supplement solution. Cells were grownin 75-cm2 cell culture flasks at 37 °C in a 5% CO2 humidified incubator.

Chemical treatments

Cells were treated in serum-free medium with various concen-trations of As(III) (1–10 μM) in the absence and presence of 1 nMTCDD and/or 5 μM SnMP, 1 µM CoPP, 80 μM hemin, and 100 μMhemoglobin as described in the figure legends. TCDD, SnMP, andCoPP were dissolved in dimethyl sulfoxide (DMSO) and maintainedin DMSO at−20 °C until use. As(III) and hemin (10 mM stocks) wereprepared freshly in double deionized water. Human hemoglobin wasdissolved directly in DMEM to a final concentration of 100 μM. In alltreatments, the DMSO concentration did not exceed 0.05% (v/v).

Effect of As(III) on cell viability

The effect of As(III) on cell viability was determined using the MTTassay as described previously [12]. The MTT assay measures theconversion of MTT to formazan in living cells via mitochondrialenzymes of viable cells. In brief, HepG2 cells were seeded onto 96-wellmicrotiter cell culture plates and incubated for 24 h at 37 °C in a 5%CO2 humidified incubator. Cells were treated with various concentra-tions of As(III) (1–10 μM) in the absence and presence of 1 nM TCDD.After 24 h incubation, the medium was removed and replaced withcell culture medium containing 1.2 mM MTT dissolved in phosphate-buffered saline (PBS; pH 7.4). After 2 h of incubation, the crystalsformed were dissolved in isopropanol. The intensity of the color ineach well was measured at a wavelength of 550 nm using the Bio-TekEL 312e microplate reader (Bio-Tek Instruments, Winooski, VT, USA).

RNA extraction and quantitative real-time PCR

After incubation with the test compounds for the specified timeperiods, total cellular RNA was isolated using TRIzol reagent,according to the manufacturer's instructions (Invitrogen), andquantified by measuring the absorbance at 260 nm. For reversetranscription–polymerase chain reaction (RT-PCR), first-strand cDNAwas synthesized from 1.0 µg of total RNA using the High-CapacitycDNA reverse transcription kit with random primers. Real-time PCRswere performed on an ABI 7500 real-time PCR system (AppliedBiosystems), using SYBR green PCR master mix (Applied Biosystems).The amplification reactions were performed as follows: 10 min at95 °C and 40 cycles of 94 °C for 15 s and 60 °C for 1 min. The primersand probes used were the following: human CYP1A1, forward primer5′-CTATCTGGGCTGTGGGCAA-3′, reverse primer 5′-CTGGCTCAAGCA-CAACTTGG-3′; HO-1, forward primer 5′-ATGGCCTCCCTGTACCACATC-3′, reverse primer 5′-TGTTGCGCTCAATCTCCTCCT-3′; and β-actin,forward primer 5′-CTGGCACCCAGGACAATG-3′, reverse primer 5′-GCCGATCCACACGGAGTA-3′. Primers and probes were purchasedfrom Integrated DNA technologies (Coralville, IA, USA). The foldchange in the level of CYP1A1 or HO-1 (target genes) between treatedand untreated cells, corrected by the level of β-actin, was determinedusing the equation fold change=2−Δ(ΔCt), where ΔCt=Ct (target)−Ct (β-actin) and Δ(ΔCt)=ΔCt (treated)−ΔCt (untreated).

Protein extraction and Western blot analysis

Twenty-four hours after incubation with the test compounds, cellswere collected in lysis buffer containing 50 mM Hepes, 0.5 M sodiumchloride, 1.5 mM magnesium chloride, 1 mM EDTA, 10% (v/v)glycerol, 1% Triton X-100, and 5 μl/ml protease inhibitor cocktail.The cell homogenates were obtained by incubating the cell lysates onice for 1 h, with intermittent vortexing every 10 min, followed bycentrifugation at 12,000 g for 10 min at 4 °C. Proteins (50 μg) wereresolved by denaturing electrophoresis, as described previously [13].Briefly, the cell homogenates were dissolved in 1× sample buffer,boiled for 5 min, separated by 10% SDS–PAGE, and electrophoreticallytransferred to a nitrocellulose membrane. Protein blots were blockedfor 24 h at 4 °C in blocking buffer containing 5% skim milk powder, 2%bovine serum albumin, and 0.05% (v/v) Tween 20 in Tris-bufferedsaline solution (TBS; 0.15 M sodium chloride, 3 mM potassiumchloride, 25 mM Tris–base). After being blocked, the blots wereincubated with a primary polyclonal goat anti-rat CYP1A1 antibodyfor 2 h at room temperature, HO-1 primary monoclonal mouse anti-human antibody overnight at 4 °C, or primary polyclonal rabbit anti-human GAPDH antibody overnight at 4 °C in TBS containing 0.05%(v/v) Tween 20 and 0.02% sodium azide. Incubation with aperoxidase-conjugated rabbit anti-goat IgG secondary antibody forCYP1A1 and GAPDH, or peroxidase-conjugated rabbit anti-mouse IgGsecondary antibody for HO-1, was carried out in blocking buffer for

Fig. 1. Effect of As(III) on cell viability. HepG2 cells were treated for 24 h with As(III) (0,1, 5, and 10 μM) in the absence and presence of 1 nM TCDD. Cell cytotoxicity wasdetermined using the MTT assay. Data are expressed as the percentage of untreatedcontrol (set at 100%)±SE (n=8).

1401A. Anwar-Mohamed, A.O.S. El-Kadi / Free Radical Biology & Medicine 48 (2010) 1399–1409

2 h at room temperature. The bands were visualized with theenhanced chemiluminescence method according to the manufac-turer's instructions (Amersham, Arlington Heights, IL, USA). Theintensity of the CYP1A1 and HO-1 protein bands was quantified,relative to the signals obtained for GAPDH protein, using ImageJsoftware.

Determination of CYP1A1 and CYP1A2 enzymatic activities

CYP1A1-dependent 7-ethoxyresorufin-O-deethylase and CYP1A2-dependent 7-methoxyresorufin-O-deethylase activities were assessedon intact living cells using 7-ethoxyresorufin and 7-methoxyresorufinas substrates, respectively, as previously described [14]. Enzymaticactivity was normalized for cellular protein content, which wasdetermined using a modified fluorescence assay [15].

Determination of CYP1A1 activity using total cell lysate

CYP1A1 activity was determined according to the methoddescribed by Kennedy et al. [16]. Subconfluent HepG2 cells (100-mm plates) were treated with As(III) (5 μM) in the absence andpresence of 1 nM TCDD for 24 h. Thereafter, cells were harvested,washed in PBS, and recovered by scraping in 500 μl of 25 mM Hepes,1.5 mM EDTA, and 10% (v/v) glycerol, pH 7.5. Cells were lysed byfreeze–thawing at−80 °C, and 50 μl of lysate (1–2 mg/ml of protein)was mixed with 20 μM 7-ethoxyresorufin (25 μl) in 100 mM sodiumphosphate, pH 7.8, and incubated at 37 °C for 15 min in 96-well plates.Reactions were started by the addition of 25 μl of 4 mM NADPH andstopped after 5 min with 150 μl of acetonitrile (resorufin productionwas linear with respect to time over this period). Bovine serumalbumin was substituted for cell lysates in blank reactions. Enzymaticactivity was normalized for cellular protein content, which wasdetermined using a modified fluorescence assay [15].

Transient transfection and luciferase assay

HepG2 cells were plated onto 12-well cell culture plates. Each wellof cells was transfected with 1.6 μg of the XRE-driven luciferasereporter plasmid pGudLuc 6.1, generously provided by Dr. M.S.Denison (University of California at Davis), using Lipofectamine 2000reagent according to the manufacturer's instructions (Invitrogen).Luciferase assay was performed according to the manufacturer'sinstructions (Promega) as described previously [17]. Briefly, afterincubation with test compounds for 24 h, the cells were washed withPBS and 200 μl of 1× lysis buffer was added to each well withcontinuous shaking for at least 20 min, then the content of each wellwas collected separately in 1.5-ml microcentrifuge tubes. The tubeswere then centrifuged to precipitate cellular waste, 100 μl of celllysate was incubated with 100 μl of stabilized luciferase reagent, andluciferase activity was quantified using a TD-20/20 luminometer(Turner BioSystems).

CYP1A1 mRNA stability

The half-life of CYP1A1 mRNA was analyzed by an Act-D-chaseassay. Cells were pretreated with 1 nM TCDD for 12 h. The cells werethenwashed and incubatedwith 5 μg/ml Act-D, to inhibit further RNAsynthesis, immediately before treatment with 5 µM As(III). Total RNAwas extracted at 0, 1, 3, 6, and 12 h after incubation with As(III). Real-time PCRswere performed using SYBR green PCRmastermix (AppliedBiosystems). The fold change in the level of CYP1A1 (target gene)between treated and untreated cells, corrected by the level of β-actin,was determined using the equation fold change=2−Δ(ΔCt), whereΔCt=Ct (target)−Ct (β-actin) and Δ(ΔCt)=ΔCt (treated)−ΔCt (untreated).

Determination of total cellular heme content

Cellular heme content was determined using the method of Wardet al. [18], with some modifications. After a 24-h incubation periodwith test compounds, cells were pelleted and boiled in 2.0 M oxalicacid for 30 min followed by resuspension in cold PBS and centrifu-gation at 14,000 g for 15 min. The supernatant was then removed andthe fluorescence of protoporphyrin IX was assayed using the Eclipsefluorescence spectrophotometer (Varian Australia Pty Ltd., Mulgrave,Victoria Australia) using an excitation wavelength of 405 nm and anemission wavelength of 600 nm. Background was determined bymeasuring the fluorescence in the absence of cells. Cellular hemecontent was calculated as a percentage of serum-freemedium-treatedcells after normalization of cellular heme content with cellularprotein, which was determined using the method of Lowry et al. [19].

Transfecting HepG2 cells with HO-1 siRNA

HepG2 cells were plated onto 24-well cell culture plates. Each wellof cells was transfected with HO-1 siRNA or Silencer Select NegativeControl No. 2 siRNA at the concentration of 20 nM using INTERFERinreagent according to the manufacturer's instructions (Polyplus). HO-1siRNA sequences were sense, CAAUGCAGUAUUUUUGUtt, and anti-sense, AACAAAAAUACUGCAUUUGag. Transfection efficiency wasdetermined using real-time PCR to detect HO-1 mRNA posttransfec-tion at 6, 12, and 24 h. Therefore, the cells were treated 6 hposttransfection with TCDD in the absence or presence of 5 μM As(III) for 6 h, to determine HO-1 and CYP1A1 mRNA levels, or 24 h todetermine CYP1A1 catalytic activity levels.

Statistical analysis

The comparative analysis of the results from the variousexperimental groups with their corresponding controls was per-formed using SigmaStat for Windows (Systat Software, Chicago, IL,USA). A one-way analysis of variance followed by the Student–Newman–Keuls test was carried out to assess statistical significance.The differences were considered significant when Pb0.05.

Results

Effect of coexposure to As(III) and TCDD on cell viability

To determine the nontoxic concentrations of As(III) to be utilizedin this study, HepG2 cells were exposed for 24 h with increasingconcentrations of As(III) (1–10 µM) in the absence and presence of1 nM TCDD; thereafter cytotoxicity was assessed using the MTT assay.Fig. 1 shows that As(III) at concentrations of 1–10 µM in the presence

Fig. 2. Effect of As(III) on CYP1A1 mRNA, protein, and catalytic activity in HepG2 cells.HepG2 cells were treated with increasing concentrations of As(III) in the presence of1 nM TCDD for 6 h for mRNA or 24 h for protein and catalytic activity. (A) First-strandcDNA was synthesized from total RNA (1 μg) extracted from HepG2 cells. cDNAfragments were amplified and quantitated using the ABI 7500 real-time PCR system asdescribed under Materials and methods. Duplicate reactions were performed for eachexperiment, and the values presented are themeans of three independent experiments.(B) Protein (50 µg) was separated by 10% SDS–PAGE and transferred to nitrocellulosemembrane. Protein blots were blocked overnight at 4 °C and then incubated with aprimary CYP1A1 antibody for 24 h at 4 °C, followed by 1 h incubation with secondaryantibody at room temperature. CYP1A1 protein was detected using the enhancedchemiluminescence method. The intensity of the bands was normalized to GAPDHsignal, which was used as the loading control. One of three representative experimentsis shown. (C) CYP1A1 activity was measured in intact living cells treated withincreasing concentrations of As(III), in the absence and presence of 1 nM TCDD, for 24 h.CYP1A1 activity was measured using 7-ethoxyresorufin as a substrate. Values arepresented as means±SE (n=6). +Pb0.05, compared to control (C; concentration0 µM); *Pb0.05, compared to respective TCDD (T) treatment.

1402 A. Anwar-Mohamed, A.O.S. El-Kadi / Free Radical Biology & Medicine 48 (2010) 1399–1409

and absence of 1 nM TCDD did not affect cell viability. Therefore, allsubsequent studies were conducted using the concentrations of 1–10 µM.

Concentration-dependent effect of coexposure to As(III) and TCDD oninducible CYP1A1 mRNA

To examine the effect of coexposure to As(III) and TCDD onCYP1A1 mRNA, HepG2 cells were treated with various concentrationsof As(III) in the presence of 1 nM TCDD (Fig. 2A). Thereafter, CYP1A1mRNAwas assessed using real-time PCR. TCDD alone caused a 40-foldincrease in CYP1A1 mRNA that was inhibited in a dose-dependentmanner by As(III). Initially, As(III) at the concentration of 1 μM causeda significant decrease in TCDD-mediated induction of CYP1A1 mRNAby 10-fold. The maximum inhibition took place at the highestconcentration tested, 10 μM, which caused a decrease in the TCDD-mediated induction of CYP1A1 mRNA by 25-fold (Fig. 2A).

Concentration-dependent effect of coexposure to As(III) and TCDD onCYP1A1 protein and catalytic activity

To examine whether the observed inhibition of the TCDD-mediated induction of CYP1A1 mRNA by As(III) is further translatedto the protein and activity levels, HepG2 cells were treated for 24 hwith increasing concentrations of As(III) in the presence of 1 nMTCDD. Figs. 2B and C show that TCDD alone caused 80- and 40-foldincreases in CYP1A1 protein and catalytic activity, respectively. Ofinterest, As(III) decreased the TCDD-mediated induction of CYP1A1protein and catalytic activity levels in a dose-dependent manner. Thisinhibitory effect of As(III) on the CYP1A1 protein and catalytic activitylevels is in concordance with the observed effect at the mRNA level, inwhich the initial significant inhibition took place at 1 μM As(III) andreached the maximal inhibition at 10 μM (Figs. 2B and C).

Transcriptional inhibition of CYP1A1 gene by As(III)

HepG2 cells were transiently transfected with the XRE-drivenluciferase reporter gene to study the effect of As(III) on the AhR-dependent transcriptional activation. Luciferase activity resultsshowed that 5 µM As(III) alone inhibited the constitutive expressionof the luciferase activity (Fig. 3). On the other hand, 1 nM TCDD alonecaused a significant increase in luciferase activity, by 280-foldcompared to control. Interestingly, cotreatment with As(III) andTCDD significantly decreased the TCDD-mediated induction ofluciferase activity by 65% (Fig. 3).

Posttranscriptional modification of CYP1A1 mRNA by As(III)

To examine the effect of As(III) on the CYP1A1 mRNA stability, weperformed the Act-D-chase experiment on intact viable HepG2 cells.The level of mRNA expression is not only a function of thetranscription rate, but is also dependent on the elimination rate,through processing or degradation. If As(III) decreases CYP1A1 mRNAby decreasing its stability, a decrease in half-life would be expected totake place. Fig. 4 shows that CYP1A1 mRNA decayed with a half-life of4.6±0.6 h. Furthermore, As(III) did not alter the CYP1A1 mRNA half-life, which was 4.9±0.8 h (Fig. 4).

The effect of supplementing NADPH on As(III)-mediated decrease inCYP1A1 activity

To investigate whether the As(III)-mediated decrease in CYP1A1activity is a result of decreased intracellular NADPH, HepG2 cells weretreated for 24 h with 5 μM As(III) with or without 1 nM TCDD and theCYP1A1 activity was assessed in total cellular lysates in the presenceof NADPH. Fig. 5 shows that TCDD alone caused a 180-fold increase in

Fig. 3. Effect of As(III) on luciferase activity. HepG2 cells were transiently transfectedwith the XRE-luciferase transporter plasmid pGudLuc 6.1. Cells were treated withvehicle, TCDD (1 nM), As(III) (5 μM), or TCDD (1 nM)+As(III) (5 μM) for 24 h. Cellswere lysed and luciferase activity was measured according to the manufacturer'sinstructions. Luciferase activity is reported as relative light units. Values are presentedas means±SE (n=6). +Pb0.05, compared to control (C); *Pb0.05, compared torespective TCDD (T) treatment.

Fig. 5. Effect of supplementing NADPH on As(III)-mediated decrease in CYP1A1 activity.HepG2 cells were treated with 5 μM As(III) in the absence and presence of 1 nM TCDD for24 h. Thereafter, total cell lysateswere prepared as described underMaterials andmethods.CYP1A1 activity was measured in total cell lysate supplemented with NADPH using 7-ethoxyresorufin as a substrate. Values are presented as means±SE (n=8). +Pb0.05,compared to control (C); *Pb0.05, compared to respective TCDD (T) treatment.

1403A. Anwar-Mohamed, A.O.S. El-Kadi / Free Radical Biology & Medicine 48 (2010) 1399–1409

CYP1A1 catalytic activity. On the other hand, As(III) decreased theTCDD-mediated induction of CYP1A1 catalytic activity by 3.5-fold intotal cell lysate supplemented with NADPH (Fig. 5). This inhibitoryeffect of As(III) on the CYP1A1 catalytic activity level in total cell lysate

Fig. 4. Effect of As(III) on CYP1A1 mRNA half-life using real-time PCR. HepG2 cells weregrown to 90% confluence in six-well cell culture plates and then treated with 1 nMTCDD for 12 h. The cells were then washed and incubated in fresh medium containing5 μM As(III) plus 5 µg/ml Act-D, an RNA synthesis inhibitor. First-strand cDNA wassynthesized from total RNA (1 μg) extracted from HepG2 cells. cDNA fragments wereamplified and quantitated using the ABI 7500 real-time PCR system as described underMaterials and methods. Duplicate reactions were performed for each experiment, andthe values presented are the means of three independent experiments. mRNA decaycurves were analyzed individually, and the half-life was estimated from the slope of astraight line fitted by linear regression analysis (r2≥0.90) to a semilog plot of mRNAamount, expressed as a percentage of treatment at the time 0 h (maximum, 100%) levelversus time. The half-lives obtained from three independent experiments were thenused to calculate the mean half-life (mean±SE, n=3).

was similar to that observed at the catalytic activity levels in intactcells.

The effect of SnMP on the posttranslational modification of CYP1A1catalytic activity by As(III)

The fact that As(III) inhibited the TCDD-mediated induction ofCYP1A1 at the catalytic activity level more than what is observed onthe mRNA or protein levels prompted us to investigate the possiblerole of HO-1 in this inhibitory effect. For this purpose HepG2 cellswere coexposed to 5 μM As(III) and 1 nM TCDD in the presence andabsence of 5 μM SnMP. Our results showed that As(III) in the presenceof 1 nM TCDD was able to increase HO-1 mRNA by 15-fold compared

Fig. 6. Effect of SnMP on HO-1 and CYP1A1 mRNA levels in the presence of As(III). HepG2cells were treated with 5 μM As(III) and 1 nM TCDD in the presence and absence of 5 μMSnMP for 6 h and assayed for HO-1 and CYP1A1 mRNA. (A and B) First-strand cDNA wassynthesized from total RNA (1 μg) extracted from HepG2 cells. cDNA fragments wereamplified and quantitated using the ABI 7500 real-time PCR system as described underMaterials and methods. Duplicate reactions were performed for each experiment, and thevalues presented are the means of three independent experiments. +Pb0.05, compared tocontrol; *Pb0.05, compared to respective TCDD treatment.

Fig. 7. Effect of SnMP as a competitive inhibitor of HO-1 on CYP1A1 and CYP1A2 catalyticactivities and total cellular heme content in the presence of As(III). HepG2 cells weretreated with 5 μM As(III) and 1 nM TCDD in the presence and absence of 5 μM SnMP for24 h. (A and B) CYP1A1 and CYP1A2 activities were measured in intact living cells treatedwith 5 μM As(III) and 1 nM TCDD in the presence and absence of 5 μM SnMP for 24 h.CYP1A1 and CYP1A2 activities were measured using 7-ethoxyresorufin and 7-methoxyresorufin as a substrates, respectively. Values are presented as means±SE(n=8). +Pb0.05, compared to control; *Pb0.05, compared to respective TCDDtreatment; †Pb0.05, compared to respective As(III)+TCDD treatment. (C) Total hemecontent in HepG2 cells treated with 5 μM As(III) and 1 nM TCDD in the presence andabsence of 5 μM SnMP for 24 h. The conversion of heme to protoporphyrin IX by oxalicacid wasmeasured fluorimetrically. Cellular heme content was calculated as a percentageof serum-free medium-treated cells after normalization with cellular protein. *Pb0.05,compared to respective TCDD treatment; †Pb0.05, compared to respective As(III)+TCDDtreatment.

1404 A. Anwar-Mohamed, A.O.S. El-Kadi / Free Radical Biology & Medicine 48 (2010) 1399–1409

to control (Fig. 6A). TCDD alone did not significantly alter the HO-1mRNA. As expected, SnMP alone increased HO-1 mRNA by ∼0.5-fold;however, when the cells were coexposed to As(III) and TCDD in thepresence of SnMP, the As(III)-mediated induction of HO-1 mRNA wasnot altered by the SnMP treatment.

Although SnMP did not significantly alter the HO-1 mRNA whenused concomitant with As(III) and TCDD, there was still a possibilitythat SnMP might reverse the inhibitory effect of As(III) on CYP1A1catalytic activity by inducing CYP1A1 mRNA levels. Therefore, weexamined the effect of SnMP on CYP1A1 mRNA levels. SnMP alone orin the presence of TCDD or As(III) plus TCDD did not affect CYP1A1mRNA levels with all treatments, thus eliminating this possibility(Fig. 6B). SnMP alone caused no effect on the CYP1A1 catalyticactivity. Similarly, the TCDD-mediated induction of CYP1A1 catalyticactivity was not affected by SnMP treatment. On the other hand,As(III) at the concentration of 5 μM decreased the TCDD-mediatedinduction of CYP1A1 catalytic activity. Intriguingly, SnMP partiallyreversed the As(III)-mediated decrease in CYP1A1 activity. Upontreatment of the cells with SnMP in the presence of both As(III) andTCDD, there was a partial restoration of the As(III)-mediated down-regulation of CYP1A1 catalytic activity induced by TCDD (Fig. 7A).Despite being successful in partially reversing the As(III)-mediateddecrease in CYP1A1 activity by inhibiting HO-1, SnMP was unable tocompletely restore the CYP1A1 activity.

To examine whether the effect of As(III) is specific for CYP1A1 weinvestigated the effect of As(III) on CYP1A2 activity. In addition, wedetermined if the effect mediated by As(III) could be reversed bySnMP. Our results demonstrate that TCDD alone increased the CYP1A2catalytic activity levels by 4-fold, whereas SnMP alone or in thepresence of TCDD did not affect CYP1A2 activity. In contrast, 5 μMAs(III) was able to decrease the TCDD-mediated induction of CYP1A2catalytic activity by 2.5-fold. On the other hand, when the cells werecoexposed to SnMP in the presence of TCDD plus As(III), there was apartial restoration of the As(III)-mediated decrease in the TCDD-mediated induction of CYP1A2 catalytic activity (Fig. 7B).

The ability of As(III) to inhibit CYP1A1 activity more than its mRNAprompted further investigation. Therefore, we hypothesized that thedecrease in TCDD-mediated induction of CYP1A1 activity is attribut-able to decreased total cellular heme content. Our results showed thatthe treatments of TCDD alone, SnMP alone, or TCDD plus SnMP did notaffect total cellular heme content. On the other hand, 5 μM As(III) inthe presence of 1 nM TCDD caused a 0.4-fold decrease in the totalcellular heme content (Fig. 7C). Incongruously, when the cells weretreated with SnMP in the presence of TCDD plus As(III) there wasalmost a complete restoration of total cellular heme content levelsthat were decreased by As(III) (Fig. 7C). Importantly, with SnMP aloneor in the presence of TCDD, fluorescence readings were similar tothose of untreated cells, eliminating the possibility that SnMP mightact to restore the total heme content through a false positive reading.

The effect of exogenous heme and the carbon monoxide (CO) scavengerhemoglobin (Hb) on As(III)-mediated decrease in CYP1A1 activity

In an attempt to examine whether the presence of external hemewould restore the As(III)-mediated decrease in CYP1A1 activity,HepG2 cells were coexposed to 5 μM As(III) and 1 nM TCDD in thepresence and absence of 80 μM hemin, a precursor of heme. Ourresults showed that hemin alone did not affect CYP1A1 activity. Onthe other hand, hemin significantly potentiated the TCDD-mediatedinduction of CYP1A1 activity. Interestingly, the addition of heminpartially restored the As(III)-mediated decrease in CYP1A1 activity bytwofold compared to As(III) plus TCDD (Fig. 8A).

To examine the possibility that As(III) acts to decrease the TCDD-mediated induction of CYP1A1 activity by liberating CO as a by-product of heme degradation, HepG2 cells were coexposed to 5 μMAs(III) and 1 nM TCDD in the presence and absence of 100 μMHb. Our

results showed that Hb alone was able to induce CYP1A1 activity by 9-fold compared to control. Similarly, TCDD alone induced CYP1A1catalytic activity by 47-fold, which was further induced by Hb to 60-fold compared to control (Fig. 8B). Interestingly, the treatment withHb partially restored the As(III)-mediated decrease in CYP1A1 activity(Fig. 8B).

The effect of HO-1 siRNA on As(III)-mediated inhibition of CYP1A1catalytic activity

We took a genetic approach to confirm whether HO-1 is involvedin the As(III)-mediated decrease in the TCDD-mediated induction ofCYP1A1 catalytic activity. For this purpose, HepG2 cells weretransfected with human HO-1 siRNA or Silencer negative controlsiRNA for 6 h, and then the cells were treated with 5 μM As(III) in thepresence and absence of 1 nM TCDD. Our results showed that HO-1siRNA significantly decreased HO-1 mRNA by 0.8-fold compared to

Fig. 8. Effect of supplementing external heme and the CO scavenger Hb on As(III)-mediated decrease in CYP1A1 activity. (A) CYP1A1 activity was measured in intact livingcells treatedwith 5 μMAs(III) and 1 nMTCDD in the presence and absence of 80 μMheminfor 24 h. CYP1A1 activitywasmeasuredusing 7-ethoxyresorufin as a substrate. (B) CYP1A1activity was measured in intact living cells treated with 5 μMAs(III) and 1 nM TCDD in thepresence and absence of 100 μM Hb for 24 h. CYP1A1 activity was measured using 7-ethoxyresorufin as a substrate. Values are presented as means±SE (n=8). +Pb0.05,compared to control; *Pb0.05, compared to respective TCDD treatment; †Pb0.05,compared to respective As(III)+TCDD treatment.

1405A. Anwar-Mohamed, A.O.S. El-Kadi / Free Radical Biology & Medicine 48 (2010) 1399–1409

control (Fig. 9A). On the other hand, As(III) was able to increase HO-1mRNA levels, in the absence and presence of TCDD, to reach 25-foldcompared to control. When the cells were transfected with HO-1siRNA and then treated with As(III) in the presence or absence ofTCDD there was a statistically significant decrease in HO-1 mRNA by20-fold. Furthermore, the Silencer select negative control siRNA didnot affect the inducible level of HO-1 mRNA by As(III), eliminating thepossibility that the inhibitory effects of HO-1 siRNA might have beendue to any toxicity. To test the selectivity of the siRNA for HO-1, wedetermined the CYP1A1 mRNA levels in cells transfected with siRNAfor HO-1. Fig. 10A shows that CYP1A1 mRNA levels were not alteredby the HO-1 siRNA or Silencer select negative control siRNAtransfection. Thus, the observed effects on the CYP1A1 catalyticactivity levels are solely due to the knockdown of HO-1.

To confirm whether the knockdown of HO-1 mRNA is furthertranslated to functional protein, we examined HO-1 protein expres-sion levels. As(III) at the concentration of 5 μM was able to increasethe HO-1 protein by 22-fold. In contrast, HO-1 siRNA completelyabolished the As(III)-mediated induction of HO-1 protein (Fig. 9B).

Looking at CYP1A1 catalytic activity, As(III) alone or in thepresence of HO-1 siRNA did not affect CYP1A1 catalytic activity(Fig. 10B). TCDD alone increased the CYP1A1 catalytic activity by 50-fold, whereas As(III) significantly decreased the TCDD-mediatedinduction of CYP1A1 catalytic activity by ∼0.8-fold. Interestingly,when HepG2 cells were transfected with HO-1 siRNA and thencoexposed to As(III) and TCDD, As(III) decreased the TCDD-inducedcatalytic activity by ∼0.2-fold, compared to control, and was unable tomaintain the same inhibitory effect on CYP1A1 catalytic activitycompared to nontransfected cells (Fig. 10B). Moreover, when HepG2

cells were transfected with Silencer select negative control siRNA andthen coexposed to As(III) and TCDD, As(III) decreased the TCDD-mediated induction of CYP1A1 catalytic activity by ∼0.8-fold, similarto that of nontransfected cells.

The effect of HO-1 inducer CoPP on total cellular heme content andCYP1A1 mRNA and catalytic activity

To confirm the role of HO-1 in the down-regulation of CYP1A1 atthe catalytic activity level, we used CoPP, which is a potent HO-1inducer, to simulate the As(III)-mediated effect on CYP1A1 activity.Our results showed that 1 μM CoPP in the absence and presence of1 nM TCDD was able to significantly induce HO-1 mRNA by 37-foldcompared to control (Fig. 11A). In contrast, when the cells weretransfected with HO-1 siRNA before the treatment with CoPP in thepresence of 1 nM TCDD, the HO-1 mRNA levels were significantlydecreased by ∼9-fold compared to 1 nM TCDD+1 μMCoPP treatment(Fig. 11A). In agreement with HO-1 mRNA data, treatment with 1 µMCoPP in the absence and presence of 1 nM TCDD was able to decreasethe total cellular heme content to 0.6-fold, compared to control(Fig. 11B). Thus, these results imply that CoPP, by inducing HO-1,decreases the total cellular heme content. The treatment with CoPP(1 µM) did not alter the TCDD-mediated induction of CYP1A1 mRNAlevels. Moreover, the knockdown of HO-1 mRNA did not affect theTCDD-mediated induction of CYP1A1 mRNA (Fig. 12A).

Interestingly, CoPP at the concentration of 1 μM was able tosignificantly decrease the TCDD-mediated induction of CYP1A1activity by threefold (Fig. 12B). On the other hand when the cellswere transfected with HO-1 siRNA and then treated with TCDD in thepresence of 1 μM CoPP there was a complete restoration of the TCDD-mediated induction of CYP1A1 catalytic activity (Fig. 12B).

Discussion

Themechanisms by which As(III) exhibits its anticancer effects arestill at large and undeniably multiple. One of the proposed mechan-isms by which As(III) mediates its anticancer effects is inhibition ofCYP1A1 gene expression.

In this study we have shown that As(III) inhibits the TCDD-mediated induction of CYP1A1 at the mRNA, protein, and catalyticactivity levels in a dose-dependent manner. Our findings are inagreement with previous studies showing that As(III)-mediateddecrease in CYP1A1 activity was accompanied by either a decreaseor no change in the CYP1A1 mRNA levels. For example, it was shownthat As(III) was able to reduce benzo[k]fluoranthene- and TCDD-induced CYP1A1mRNA in HepG2 and Huh7 cells, respectively [20,21].An explanation offered for this phenomenon is that As(III) blocks therecruitment of polymerase II to the CYP1A1 promoter, thus inhibitingits transcription [22]. In contrast, data from our laboratory haveshown that As(III) differentially up-regulates Cyp1a1 gene expressionand causes further potentiation of the TCDD-mediated induction ofCyp1a1 mRNA in murine hepatoma Hepa 1c1c7 cells [17]. Thedifference between the effect of As(III) on the human HepG2 and themouse Hepa 1c1c7 cells could be attributed to the mechanisticdifferences in the regulation of CYP1A1 gene expression upontreatment by TCDD [23,24]. Factors that could be responsible forthese species-specific characteristics of AhR functions, and subse-quently CYP1A1 inducibility, could be summarized in three majorcomponents: the nuclear translocation, transcription initiation viaremodeling of chromatin, and finally proteasomal degradation of theAhR [25]. For example, it has been shown that in Hepa 1c1c7 cells thecoactivator CREB-binding protein (CBP) is recruited to the CYP1A1promoter region after treatment with TCDD, reaching its peak at 4 h,and this coincides with the recruitment of AhR and polymerase II,whereas there is no recruitment of p300 [26]. In contrast, in HepG2,

Fig. 9. Effect of HO-1 siRNA on As(III)-mediated induction of HO-1 mRNA and protein. HepG2 cells were transiently transfected with 20 nM HO-1 siRNA (siRNA) or 20 nM Silencerselect negative control siRNA (−ve siRNA) for 6 h; thereafter cells were treated with vehicle, TCDD (1 nM), As(III) (5 μM), or TCDD (1 nM)+As(III) (5 μM) for 6 h for HO-1 mRNAassay or 24 h for HO-1 protein assay. (A) First-strand cDNAwas synthesized from total RNA (1 μg) extracted fromHepG2 cells. cDNA fragments were amplified and quantitated usingan ABI 7500 real-time PCR system as described under Materials and methods. Duplicate reactions were performed for each experiment, and the values presented are the means ofthree independent experiments. +Pb0.05, compared to control; *Pb0.05, compared to respective As(III) treatment; †Pb0.05, compared to respective As(III)+TCDD treatment. (B)Protein (50 µg) was separated by 10% SDS–PAGE and transferred to nitrocellulose membrane. Protein blots were blocked overnight at 4 °C and then incubated with a primary HO-1antibody overnight at 4 °C, followed by 4 h incubation with secondary antibody at room temperature. HO-1 protein was detected using the enhanced chemiluminescence method.The intensity of the bands was normalized to GAPDH signal, which was used as loading control. One of three representative experiments is shown. +Pb0.05, compared to control;*Pb0.05, compared to respective As(III) treatment.

1406 A. Anwar-Mohamed, A.O.S. El-Kadi / Free Radical Biology & Medicine 48 (2010) 1399–1409

p300 recruitment is increased in response to TCDD to reach its peakbetween 4 and 12 h, whereas CBP recruitment is unaffected [26].

The transcriptional regulation of CYP1A1 gene expression by As(III)was also investigated. In this regard,we have shown that As(III) aloneorin the presence of TCDD was able to significantly decrease the AhR-dependent, XRE-driven luciferase reporter gene expression. Ourfindings are in agreement with previous findings showing that As(III)decreases luciferase activity in cells transfected with the XRE-drivenluciferase reporter gene in HepG2 and Huh7 cells [20,21]. Being able todecrease the CYP1A1 gene expression through a transcriptionalmechanism, As(III) was also suspected of participating in decreasingCYP1A1 mRNA stability. Our results showed that As(III) was unable tosignificantly alter CYP1A1half-life. These results suggest that As(III) actsthrough a transcriptionalmechanismto inhibitCYP1A1geneexpression.

The ability of As(III) to inhibit CYP1A1 at the activity levelmore thanthe mRNA or protein expression levels raised the question of whetherthere is a posttranslationalmodification thatmighthaveoccurredon theCYP1A1 apoprotein. It has beenpreviously reported thatAs(III) is able todecrease glucose-6-phosphate dehydrogenase (G6PD) in various brainregions of male Wistar rats [27]. The deficiency in G6PD decreases the

ability of the cells to generate NADPH [28], which is required for CYP450activity. Therefore, we examined whether As(III) decreases the cellularlevel of NADPH. Our results showed that As(III) preserved its inhibitoryeffect on the TCDD-mediated induction of CYP1A1 despite supplemen-tation of the enzymatic system with excessive NADPH. Therefore, theeffect of As(III) on theTCDD-mediated inductionof CYP1A1activitydoesnot involve an effect on intracellular NADPH levels. Furthermore,mounting evidence suggests a role for HO-1 in the As(III)-mediateddecrease in CYP1A1 catalytic activity levels [29–31]. In this study wehave shown that As(III) increases HO-1mRNA and protein levels. HO-1,an enzyme of 32 kDa, catalyzes the oxidative conversion of heme intobiliverdin, which serves an important role in protecting cells fromoxidative damage, such as that from free radicals [32]. HO-1 anchors tothe endoplasmic reticulum membrane via a stretch of hydrophobicresidues at the C-terminus [33]. Thus, it is expected to interact withCYP450s, which are also endoplasmic reticulum-bound enzymes. Thefact that As(III) induces HO-1 with a consequent decrease in the hemepool could result in the failure to form a functioning CYP1A1 protein.Moreover, the apoprotein would be more susceptible to proteasomaldegradation [25].

Fig. 10. Effect of HO-1 siRNA on As(III)-mediated inhibition of CYP1A1 mRNA and catalytic activity. HepG2 cells were transiently transfected with 20 nM HO-1 siRNA (siRNA) or20 nM Silencer select negative control siRNA (−ve siRNA) for 6 h; thereafter the cells were treated with vehicle, TCDD (1 nM), As(III) (5 μM), or TCDD (1 nM)+As(III) (5 μM) for 6 hfor CYP1A1 mRNA assay or 24 h for CYP1A1 catalytic activity assay. (A) First-strand cDNA was synthesized from total RNA (1 μg) extracted from HepG2 cells. cDNA fragments wereamplified and quantitated using an ABI 7500 real-time PCR system as described under Materials and methods. Duplicate reactions were performed for each experiment, and thevalues presented are the means of three independent experiments. +Pb0.05, compared to control; *Pb0.05, compared to respective As(III) treatment. (B) CYP1A1 activity wasmeasured in intact living cells transiently transfected with 20 nM HO-1 siRNA (siRNA) or 20 nM Silencer select negative control siRNA (−ve siRNA) for 6 h and thereafter treatedwith 5 μM As(III), 1 nM TCDD, or 5 μM As(III) + 1 nM TCDD for 24 h. CYP1A1 activity was measured using 7-ethoxyresorufin as a substrate. Values are presented as means±SE(n=6). +Pb0.05, compared to control; *Pb0.05, compared to respective TCDD treatment; †Pb0.05, compared to respective As(III)+TCDD treatment.

1407A. Anwar-Mohamed, A.O.S. El-Kadi / Free Radical Biology & Medicine 48 (2010) 1399–1409

The role of HO-1 in the down-regulation of CYP1A1 at the catalyticactivity level was supported by a series of observations. At first, thisinhibitory effect of As(III) was also observed on the CYP1A2 catalyticactivity. Second, As(III) decreased the total cellular heme content inHepG2 cells. These results are in agreementwith previously publishedstudies demonstrating that As(III) decreases the total cellular hemecontent in HepG2 and other cell lines [29–31,34,35]. Thus, the As(III)-mediated decrease of CYP1A1 is not exclusive; rather it affects thetotal heme pool causing a general inhibitory effect to all CYP450s.Third, the HO-1 inhibitor SnMP partially restored the As(III)-mediateddown-regulation of CYP1A1 and CYP1A2 catalytic activities and totalcellular heme content. The observed effect of SnMP on the As(III)-mediated decrease in CYP1A1 and CYP1A2 catalytic activities and totalcellular heme content levels was solely from competitively inhibitingHO-1 protein and not from altering HO-1 or CYP1A1 mRNAs.

Despite the fact that As(III) decreased CYP1A1 gene expressionthrough a transcriptional mechanism, there was still a persistentinhibition in the CYP1A1 catalytic activity that was in concordancewith the inhibition occurring at themRNA and protein expression levels.Our result is in agreement with previous studies showing that As(III),independent of its action on CYP1A1 mRNA levels, decreases CYP1A1catalytic activity, total CYP450 content, and total cellular heme content[29–31,34,35]. In contrast, a previous study using primary cultured chickhepatocytes has shown that SnMP was unable to reverse the action ofAs(III) on CYP1A1 catalytic activity [30]. These contradictory results maybe explained by cell and species differences.

To decipher the role of HO-1 in the As(III)-mediated decrease inCYP1A1 activity we used hemin, a precursor of heme. If As(III) decreases

the TCDD-mediated induction of CYP1A1 activity through degrading itsheme, then supplying hemewill restore the TCDD-mediated induction ofCYP1A1 activity. Our results demonstrated that treatment with hemincaused partial restoration of the As(III)-mediated decrease in CYP1A1activity. In contrast to our results, previous reports have shown thatexogenous heme was incapable of restoring the As(III)-mediateddecrease in CYP1A1 activity in primary cultures of chick and rathepatocytes [30,34]. These contradictions may be explained by thedifferences in the cells used and the differences in the concentrations ofhemeused; in our studyweused80 μMhemin [36],whereas in the otherstudies 2–5 μMwas used [30,34].

HO-1 catalyzes the degradation of heme to CO, biliverdin, and freeiron [37]. CO is known to bind and inhibit heme-containing proteinssuch as CYP1A1 [38]. Therefore, it was of importance to examinewhether the As(III)-mediated decrease in CYP1A1 activity is due to COgeneration. In this study, we have shown that Hb was able to restorethe As(III)-mediated decrease in CYP1A1 activity, suggesting that COproduced by HO-1 activity participated in lowering CYP1A1 activity.In addition, Hb is a heme-containing protein and may protect theCYP1A1 degradation by serving as a substrate for HO-1. In agreementwith our results, previous studies have shown that Hb, a CO scavenger,was able to induce CYP2E1 activity in primary human hepatocytes[39] and E47 cells without altering protein level [40].

Because of the potential nonspecific effects of pharmacologicalinhibitors, siRNA was used to confirm the role of HO-1 in the down-regulation of CYP1A1 catalytic activity by As(III). Our results showed thatHO-1mRNAandproteinwere successfully knockeddown inHepG2cells.Interestingly, the As(III)-mediated decrease in CYP1A1 catalytic activity

Fig. 12. Effect of HO-1 siRNA on CoPP-mediated inhibition of CYP1A1 mRNA andcatalytic activity levels. HepG2 cells were transiently transfected with 20 nM HO-1siRNA (siRNA) for 6 h; thereafter cells were treated with vehicle, TCDD (1 nM), CoPP(1 μM), or TCDD (1 nM)+CoPP (1 μM) for 6 h for CYP1A1 mRNA assay or 24 h forCYP1A1 catalytic activity assay. (A) First-strand cDNA was synthesized from total RNA(1 μg) extracted from HepG2 cells. cDNA fragments were amplified and quantitatedusing an ABI 7500 real-time PCR system as described under Materials and methods.Duplicate reactions were performed for each experiment, and the values presented arethe means of three independent experiments. +Pb0.05, compared to control. (B)CYP1A1 activity was measured in intact living cells transiently transfected with 20 nMHO-1 siRNA (siRNA) for 6 h and thereafter treated with vehicle, TCDD (1 nM), CoPP(1 μM), or TCDD (1 nM)+CoPP (1 μM) for 24 h. CYP1A1 activity wasmeasured using 7-ethoxyresorufin as a substrate. Values are presented as means±SE (n=8). +Pb0.05,compared to control; *Pb0.05, compared to respective TCDD treatment; †Pb0.05,compared to respective CoPP+TCDD treatment.

Fig. 11. Effect of HO-1 siRNA on CoPP-mediated induction of HO-1 mRNA and COPP-mediated inhibition of total cellular heme content. HepG2 cells were transientlytransfected with 20 nM HO-1 siRNA (siRNA) for 6 h; thereafter cells were treated withvehicle, TCDD (1 nM), CoPP (1 μM), or TCDD (1 nM)+CoPP (1 μM) for 6 h for HO-1mRNA assay. (A) First-strand cDNA was synthesized from total RNA (1 μg) extractedfrom HepG2 cells. cDNA fragments were amplified and quantitated using an ABI 7500real-time PCR system as described under Materials and methods. Duplicate reactionswere performed for each experiment, and the values presented are the means of threeindependent experiments. +Pb0.05, compared to control; *Pb0.05, compared torespective CoPP+TCDD treatment. (B) Total heme content in HepG2 cells treated with1 μM CoPP in the absence and presence of 1 nM TCDD for 24 h. The conversion of hemeto protoporphyrin IX by oxalic acid was measured fluorimetrically. Cellular hemecontent was calculated as a percentage of serum-free medium-treated cells afternormalization with cellular protein. +Pb0.05, compared to control; *Pb0.05,compared to respective CoPP+TCDD treatment.

1408 A. Anwar-Mohamed, A.O.S. El-Kadi / Free Radical Biology & Medicine 48 (2010) 1399–1409

was partially reversed upon transfectionwith HO-1 siRNA. In agreementwith our results, it has been recently shown that in HepG2 cells, HO-1knockdown partially reverses the effect of As(III) on the benzo[k]fluoranthene-induced CYP1A1 catalytic activity level [31].

Despite the fact that As(III)-mediated inhibition of CYP1A1 catalyticactivity occurs in part through degrading its heme content, wedetermined whether CoPP, a known HO-1 inducer, would behavesimilarly. Our results demonstrated that CoPP caused an induction ofHO-1 and a decrease in the total cellular heme content. Looking at theeffect of CoPP on theCYP1A1mRNA, our results demonstrated that 1 µMCoPP was unable to cause any significant alteration of CYP1A1 mRNA,whereas it decreased the TCDD-mediated induction of CYP1A1 activity.In addition, transfection with HO-1 siRNA completely restored theTCDD-mediated induction of CYP1A1 catalytic activity that was initiallyinhibited by CoPP. Taken together, these results further confirm the roleof HO-1 in the modulation of CYP1A1 activity.

In conclusion, this study demonstrated that As(III) down-regulatesCYP1A1 through transcriptional and posttranslational mechanisms.Moreover, HO-1 is partially involved in the As(III)-mediated down-regulation of CYP1A1 at the catalytic activity level, as treatment withSnMP, heme, and hemoglobin and knockdown of HO-1 caused partialrestoration of the CYP1A1 activity.

Acknowledgments

This work was supported by Natural Sciences and EngineeringResearch Council of Canada Discovery Grant RGPIN 250139-07 to A.O.S.

A.A.-M. is the recipient of an Alberta Ingenuity Graduate Scholarshipaward. We are grateful to Dr. M.S. Denison (University of California,Davis, CA, USA) for providing us with the XRE-luciferase reporterplasmid pGudLuc 6.1.

References

[1] Hankinson, O. The aryl hydrocarbon receptor complex. Annu. Rev. Pharmacol.Toxicol. 35:307–340; 1995.

[2] Meyer, B. K.; Pray-Grant, M. G.; Vanden Heuvel, J. P.; Perdew, G. H. Hepatitis Bvirus X-associated protein 2 is a subunit of the unliganded aryl hydrocarbonreceptor core complex and exhibits transcriptional enhancer activity. Mol. Cell.Biol. 18:978–988; 1998.

[3] Whitelaw, M. L.; Gustafsson, J. A.; Poellinger, L. Identification of transactivationand repression functions of the dioxin receptor and its basic helix-loop-helix/PASpartner factor Arnt: inducible versus constitutive modes of regulation. Mol. Cell.Biol. 14:8343–8355; 1994.

[4] Denison, M. S.; Fisher, J. M.; Whitlock Jr., J. P. Protein–DNA interactions atrecognition sites for the dioxin–Ah receptor complex. J. Biol. Chem. 264:16478–16482; 1989.

[5] Nebert, D. W.; Dalton, T. P.; Okey, A. B.; Gonzalez, F. J. Role of aryl hydrocarbonreceptor-mediated induction of the CYP1 enzymes in environmental toxicity andcancer. J. Biol. Chem. 279:23847–23850; 2004.

[6] McLemore, T. L.; Adelberg, S.; Liu, M. C.; McMahon, N. A.; Yu, S. J.; Hubbard, W. C.;Czerwinski, M.; Wood, T. G.; Storeng, R.; Lubet, R. A., et al. Expression of CYP1A1gene in patients with lung cancer: evidence for cigarette smoke-induced geneexpression in normal lung tissue and for altered gene regulation in primarypulmonary carcinomas. J. Natl. Cancer Inst. 82:1333–1339; 1990.

[7] Klaassen, C. D. Cassarett and Doull's Toxicology: the Basic Science of Poisons.McGraw–Hill, New York; 2001.

[8] Patrick, L. Toxic metals and antioxidants. Part II. The role of antioxidants in arsenicand cadmium toxicity. Altern. Med. Rev. 8:106–128; 2003.

[9] Shen, Z. X.; Chen, G. Q.; Ni, J. H.; Li, X. S.; Xiong, S. M.; Qiu, Q. Y.; Zhu, J.; Tang, W.;Sun, G. L.; Yang, K. Q.; Chen, Y.; Zhou, L.; Fang, Z. W.; Wang, Y. T.; Ma, J.; Zhang, P.;

1409A. Anwar-Mohamed, A.O.S. El-Kadi / Free Radical Biology & Medicine 48 (2010) 1399–1409

Zhang, T. D.; Chen, S. J.; Chen, Z.; Wang, Z. Y. Use of arsenic trioxide (As2O3) in thetreatment of acute promyelocytic leukemia (APL). II. Clinical efficacy andpharmacokinetics in relapsed patients. Blood 89:3354–3360; 1997.

[10] Hussein, M. A. Arsenic trioxide: a new immunomodulatory agent in themanagement of multiple myeloma. Med. Oncol. 18:239–242; 2001.

[11] Murgo, A. J. Clinical trials of arsenic trioxide in hematologic and solid tumors:overview of the National Cancer Institute Cooperative Research and DevelopmentStudies. Oncologist 6 (Suppl. 2):22–28; 2001.

[12] Anwar-Mohamed, A.; El-Kadi, A. O. Sulforaphane induces CYP1A1 mRNA, protein,and catalytic activity levels via an AhR-dependent pathway in murine hepatomaHepa 1c1c7 and human HepG2 cells. Cancer Lett. 275:93–101; 2009.

[13] Elbekai, R. H.; El-Kadi, A. O.Modulation of aryl hydrocarbon receptor-regulated geneexpression by arsenite, cadmium, and chromium. Toxicology 202:249–269; 2004.

[14] Anwar-Mohamed, A.; Elbekai, R. H.; El-Kadi, A. O. MG-132 inhibits the TCDD-mediated induction of Cyp1a1 at the catalytic activity but not the mRNA orprotein levels in Hepa 1c1c7 cells. Toxicol. Lett. 182:121–126; 2008.

[15] Lorenzen, A.; Kennedy, S. W. A fluorescence-based protein assay for use with amicroplate reader. Anal. Biochem. 214:346–348; 1993.

[16] Kennedy, S. W.; Lorenzen, A.; James, C. A.; Collins, B. T. Ethoxyresorufin-O-deethylase and porphyrin analysis in chicken embryo hepatocyte cultures with afluorescence multiwell plate reader. Anal. Biochem. 211:102–112; 1993.

[17] Elbekai, R. H.; El-Kadi, A. O. Transcriptional activation and posttranscriptionalmodification of Cyp1a1 by arsenite, cadmium, and chromium. Toxicol. Lett. 172:106–119; 2007.

[18] Ward, J. H.; Jordan, I.; Kushner, J. P.; Kaplan, J. Heme regulation of HeLa celltransferrin receptor number. J. Biol. Chem. 259:13235–13240; 1984.

[19] Lowry, O. H.; Rosebrough, N. J.; Farr, A. L.; Randall, R. J. Protein measurement withthe Folin phenol reagent. J. Biol. Chem. 193:265–275; 1951.

[20] Chao, H. R.; Tsou, T. C.; Li, L. A.; Tsai, F. Y.; Wang, Y. F.; Tsai, C. H.; Chang, E. E.; Miao,Z. F.; Wu, C. H.; Lee, W. J. Arsenic inhibits induction of cytochrome P450 1A1 by 2,3, 7, 8-tetrachlorodibenzo-p-dioxin in human hepatoma cells. J. Hazard. Mater.137:716–722; 2006.

[21] Bessette, E. E.; Fasco, M. J.; Pentecost, B. T.; Kaminsky, L. S. Mechanisms ofarsenite-mediated decreases in benzo[k]fluoranthene-induced human cyto-chrome P4501A1 levels in HepG2 cells. Drug Metab. Dispos. 33:312–320; 2005.

[22] Bonzo, J. A.; Chen, S.; Galijatovic, A.; Tukey, R. H. Arsenite inhibition of CYP1A1induction by 2, 3, 7, 8-tetrachlorodibenzo-p-dioxin is independent of cell cyclearrest. Mol. Pharmacol. 67:1247–1256; 2005.

[23] Ramadoss, P.; Petrulis, J. R.; Hollingshead, B. D.; Kusnadi, A.; Perdew, G. H.Divergent roles of hepatitis B virus X-associated protein 2 (XAP2) in humanversus mouse Ah receptor complexes. Biochemistry 43:700–709; 2004.

[24] Ramadoss, P.; Perdew, G. H. The transactivation domain of the Ah receptor is a keydeterminant of cellular localization and ligand-independent nucleocytoplasmicshuttling properties. Biochemistry 44:11148–11159; 2005.

[25] Anwar-Mohamed, A.; Elbekai, R. H.; El-Kadi, A. O. Regulation of CYP1A1 by heavymetals and consequences for drugmetabolism. Expert Opin. Drug Metab. Toxicol. 5:501–521; 2009.

[26] Suzuki, T.; Nohara, K. Regulatory factors involved in species-specific modulationof arylhydrocarbon receptor (AhR)-dependent gene expression in humans andmice. J. Biochem. 142:443–452; 2007.

[27] Mishra, D.; Flora, S. J. Differential oxidative stress and DNA damage in rat brainregions and blood following chronic arsenic exposure. Toxicol. Ind. Health 24:247–256; 2008.

[28] Scott, M. D.; Zuo, L.; Lubin, B. H.; Chiu, D. T. NADPH, not glutathione, statusmodulates oxidant sensitivity in normal and glucose-6-phosphate dehydroge-nase-deficient erythrocytes. Blood 77:2059–2064; 1991.

[29] Falkner, K. C.; McCallum, G. P.; Cherian, M. G.; Bend, J. R. Effects of acute sodiumarsenite administration on the pulmonary chemical metabolizing enzymes,cytochrome P-450 monooxygenase, NAD(P)H:quinone acceptor oxidoreduc-tase and glutathione S-transferase in guinea pig: comparison with effects inliver and kidney. Chem. Biol. Interact. 86:51–68; 1993.

[30] Jacobs, J. M.; Nichols, C. E.; Andrew, A. S.; Marek, D. E.; Wood, S. G.; Sinclair, P. R.;Wrighton, S. A.; Kostrubsky, V. E.; Sinclair, J. F. Effect of arsenite on induction ofCYP1A, CYP2B, and CYP3A in primary cultures of rat hepatocytes. Toxicol. Appl.Pharmacol. 157:51–59; 1999.

[31] Bessette, E. E.; Fasco, M. J.; Pentecost, B. T.; Reilly, A.; Kaminsky, L. S. Investigationsof the posttranslational mechanism of arsenite-mediated downregulation ofhuman cytochrome P4501A1 levels: the role of heme oxygenase-1. J. Biochem.Mol. Toxicol. 23:222–232; 2009.

[32] Marilena, G. New physiological importance of two classic residual products:carbon monoxide and bilirubin. Biochem. Mol. Med. 61:136–142; 1997.

[33] Schuller, D. J.; Wilks, A. Ortiz de Montellano, P.; Poulos, T. L. Crystallization ofrecombinant human heme oxygenase-1. Protein Sci. 7:1836–1838; 1998.

[34] Jacobs, J.; Roussel, R.; Roberts,M.;Marek,D.;Wood, S.;Walton,H.; Dwyer, B.; Sinclair,P.; Sinclair, J. Effect of arsenite on induction of CYP1A and CYP2H in primary culturesof chick hepatocytes. Toxicol. Appl. Pharmacol. 150:376–382; 1998.

[35] Vakharia, D. D.; Liu, N.; Pause, R.; Fasco, M.; Bessette, E.; Zhang, Q. Y.; Kaminsky, L. S.Effect of metals on polycyclic aromatic hydrocarbon induction of CYP1A1 andCYP1A2 in human hepatocyte cultures. Toxicol. Appl. Pharmacol. 170:93–103; 2001.

[36] Hintze, K. J.; Theil, E. C. DNA and mRNA elements with complementary responsesto hemin, antioxidant inducers, and iron control ferritin-L expression. Proc. Natl.Acad. Sci. USA 102:15048–15052; 2005.

[37] Abraham, N. G.; Lin, J. H.; Mitrione, S. M.; Schwartzman, M. L.; Levere, R. D.;Shibahara, S. Expression of heme oxygenase gene in rat and human liver. Biochem.Biophys. Res. Commun. 150:717–722; 1988.

[38] Dulak, J.; Jozkowicz, A. Carbon monoxide—a “new” gaseous modulator of geneexpression. Acta Biochim. Pol. 50:31–47; 2003.

[39] Yao, P.; Hao, L.; Nussler, N.; Lehmann, A.; Song, F.; Zhao, J.; Neuhaus, P.; Liu, L.;Nussler, A. The protective role of HO-1 and its generated products (CO, bilirubin,and Fe) in ethanol-induced human hepatocyte damage. Am. J. Physiol. Gastrointest.Liver Physiol. 296:G1318–G1323; 2009.

[40] Gong, P.; Cederbaum, A. I.; Nieto, N. Heme oxygenase-1 protects HepG2 cellsagainst cytochrome P450 2E1-dependent toxicity. Free Radic. Biol. Med. 36:307–318; 2004.