Advertisement

Regulation of Heme Oxygenase-1 Expression through the Phosphatidylinositol 3-Kinase/Akt Pathway and the Nrf2 Transcription Factor in Response to the Antioxidant Phytochemical Carnosol*

  • Daniel Martin
    Affiliations
    Instituto de Investigaciones Biomédicas and the Departamento de Bioquímica, Facultad de Medicina, Universidad Autónoma de Madrid, 28029 Madrid, Spain
    Search for articles by this author
  • Ana I. Rojo
    Affiliations
    Instituto de Investigaciones Biomédicas and the Departamento de Bioquímica, Facultad de Medicina, Universidad Autónoma de Madrid, 28029 Madrid, Spain
    Search for articles by this author
  • Marta Salinas
    Affiliations
    Instituto de Investigaciones Biomédicas and the Departamento de Bioquímica, Facultad de Medicina, Universidad Autónoma de Madrid, 28029 Madrid, Spain
    Search for articles by this author
  • Raquel Diaz
    Affiliations
    Instituto de Investigaciones Biomédicas and the Departamento de Bioquímica, Facultad de Medicina, Universidad Autónoma de Madrid, 28029 Madrid, Spain

    Department of Biochemistry and Molecular Biology, Louisiana State University Health Science Center, New Orleans, Louisiana 70112

    Departamento de Bioquímica, Facultad de Medicina, Universidad de las Palmas de Gran Canaria, 35016 Gran Canaria, Spain
    Search for articles by this author
  • German Gallardo
    Affiliations
    Departamento de Bioquímica, Facultad de Medicina, Universidad de las Palmas de Gran Canaria, 35016 Gran Canaria, Spain
    Search for articles by this author
  • Jawed Alam
    Affiliations
    Department of Biochemistry and Molecular Biology, Louisiana State University Health Science Center, New Orleans, Louisiana 70112
    Search for articles by this author
  • Carlos M. Ruiz de Galarreta
    Affiliations
    Departamento de Bioquímica, Facultad de Medicina, Universidad de las Palmas de Gran Canaria, 35016 Gran Canaria, Spain
    Search for articles by this author
  • Antonio Cuadrado
    Correspondence
    To whom correspondence should addressed: Dept. de Bioquímica, Universidad Autónoma de Madrid, Arzobispo Morcillo 4, 28029 Madrid, Spain. Tel.: 34-91-497-5327; Fax: 34-91-585-4401
    Affiliations
    Instituto de Investigaciones Biomédicas and the Departamento de Bioquímica, Facultad de Medicina, Universidad Autónoma de Madrid, 28029 Madrid, Spain
    Search for articles by this author
  • Author Footnotes
    * This work was supported by Grant SAF2001-0545 from the Ministerio de Ciencia y Tecnología and Grant 08.5/0048/2001 from the Comunidad Autónoma de Madrid (to A. C.) and by Grants FEDER-97/0602 and PI-2002/168 from the Comunidad Autónoma de Canarias (to C. M. R. d. G.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Open AccessPublished:December 19, 2003DOI:https://doi.org/10.1074/jbc.M309660200
      The phosphatidylinositol 3-kinase (PI3K)/Akt pathway elicits a survival signal against multiple apoptotic insults. In addition, phase II enzymes such as heme oxygenase-1 (HO-1) protect cells against diverse toxins and oxidative stress. In this work, we describe a link between these defense systems at the level of transcriptional regulation of the antioxidant enzyme HO-1. The herb-derived phenol carnosol induced HO-1 expression at both mRNA and protein levels. Luciferase reporter assays indicated that carnosol targeted the mouse ho1 promoter at two enhancer regions comprising the antioxidant response elements (AREs). Moreover, carnosol increased the nuclear levels of Nrf2, a transcription factor governing AREs. Electrophoretic mobility shift assays and luciferase reporter assays with a dominant-negative Nrf2 mutant indicated that carnosol increased the binding of Nrf2 to ARE and induced Nrf2-dependent activation of the ho1 promoter. While investigating the signaling pathways responsible for HO-1 induction, we observed that carnosol activated the ERK, p38, and JNK pathways as well as the survival pathway driven by PI3K. Inhibition of PI3K reduced the increase in Nrf2 protein levels and activation of the ho1 promoter. Expression of active PI3K-CAAX (where A is aliphatic amino acid) was sufficient to activate AREs. The use of dominant-negative mutants of protein kinase Cζ and Akt1, two kinases downstream from PI3K, demonstrated a requirement for active Akt1, but not protein kinase Cζ. Moreover, the long-term antioxidant effect of carnosol was partially blocked by PI3K or HO-1 inhibitors, further demonstrating that carnosol attenuates oxidative stress through a pathway that involves PI3K and HO-1.
      High levels of reactive oxygen species cause damage to cells and are involved in several human pathologies, including neurodegenerative disorders and cancer (
      • Klein J.A.
      • Ackerman S.L.
      ,
      • Cooke M.S.
      • Evans M.D.
      • Dizdaroglu M.
      • Lunec J.
      ). Therefore, the use of compounds with antioxidant properties may help prevent or alleviate diseases in which oxidative stress is a primary cause (
      • Lim G.P.
      • Chu T.
      • Yang F.
      • Beech W.
      • Frautschy S.A.
      • Cole G.M.
      ). Carnosol, a diterpene derived from the herb rosemary, is a representative member of a family of plant-derived phenols, which also include curcumin, carnosic acid, phenylethyl isothiocyanate, epigallocatechin gallate, and other green tea polyphenols. These bioactive phytochemicals exhibit Michael acceptor function and therefore behave as antioxidants (
      • Lo A.H.
      • Liang Y.C.
      • Lin-Shiau S.Y.
      • Ho C.T.
      • Lin J.K.
      ). In addition, being themselves xenobiotic compounds, they activate a xenobiotic response in the target cells affecting the expression of phase II enzymes such as NAD(P)H:quinone oxidoreductase, aldoketoreductase, glutathione S-transferase, γ-glutamylcysteine synthetase, glutathione synthetase, and heme oxygenase-1 (HO-1)
      The abbreviations used are: HO-1, heme oxygenase-1; MAPK, mitogen-activated protein kinase; ERK, extracellular signal-regulated kinase; JNK, c-Jun N-terminal kinase; PI3K, phosphatidylinositol 3-kinase; ARE, antioxidant response element; EMSA, electrophoretic mobility shift assay; H2DCFDA, 2′,7′-dichlorodihydrofluorescein diacetate; ZnPPIX, zinc-protoporphyrin IX; MEK, mitogen-activated protein kinase/extracellular signal-regulated kinase kinase; HA, hemagglutinin; PKC, protein kinase C.
      1The abbreviations used are: HO-1, heme oxygenase-1; MAPK, mitogen-activated protein kinase; ERK, extracellular signal-regulated kinase; JNK, c-Jun N-terminal kinase; PI3K, phosphatidylinositol 3-kinase; ARE, antioxidant response element; EMSA, electrophoretic mobility shift assay; H2DCFDA, 2′,7′-dichlorodihydrofluorescein diacetate; ZnPPIX, zinc-protoporphyrin IX; MEK, mitogen-activated protein kinase/extracellular signal-regulated kinase kinase; HA, hemagglutinin; PKC, protein kinase C.
      (
      • Dinkova-Kostova A.T.
      • Massiah M.A.
      • Bozak R.E.
      • Hicks R.J.
      • Talalay P.
      • Dinkova-Kostova A.T.
      • Iqbal M.
      • Sharma S.D.
      • Okazaki Y.
      • Fujisawa M.
      • Okada S.
      ).
      Heme oxygenase isozymes (HO-1 and HO-2) catalyze the stepwise degradation of heme to release free iron and equimolar concentrations of carbon monoxide and the linear tetrapyrrole biliverdin, which is converted to bilirubin by the enzyme biliverdin reductase (
      • Maines M.D.
      ). The HO-1 isozyme is a phase II enzyme that is transcriptionally regulated by a large variety of stimuli. These include its substrate, heme (
      • Immenschuh S.
      • Hinke V.
      • Ohlmann A.
      • Gifhorn-Katz S.
      • Katz N.
      • Jungermann K.
      • Kietzmann T.
      ,
      • Alam J.
      • Killeen E.
      • Gong P.
      • Naquin R.
      • Hu B.
      • Stewart D.
      • Ingelfinger J.R.
      • Nath K.A.
      ); oxidative stress (
      • Bauer M.
      • Bauer I.
      ,
      • Ryter S.W.
      • Choi A.M.
      ); the signaling proteins nerve growth factor, tumor necrosis factor-α, interleukin-1β, and interferon-γ (
      • Salinas M.
      • Diaz R.
      • Abraham N.G.
      • Ruiz de Galarreta C.M.
      • Cuadrado A.
      • Rizzardini M.
      • Terao M.
      • Falciani F.
      • Cantoni L.
      • Terry C.M.
      • Clikeman J.A.
      • Hoidal J.R.
      • Callahan K.S.
      ); and phenolic compounds such as curcumin and caffeic acid (
      • Scapagnini G.
      • Foresti R.
      • Calabrese V.
      • Giuffrida Stella A.M.
      • Green C.J.
      • Motterlini R.
      ). Many reports have demonstrated the potent antioxidant activity of the heme-derived metabolites generated by HO-1 catalysis (biliverdin and bilirubin) and the cytoprotective actions of carbon monoxide on vascular endothelium and nerve cells (
      • Chen K.
      • Gunter K.
      • Maines M.D.
      • Mazza F.
      • Goodman A.
      • Lombardo G.
      • Vanella A.
      • Abraham N.G.
      • Maines M.D.
      • Chen J.
      • Tu Y.
      • Moon C.
      • Nagata E.
      • Ronnett G.V.
      ). Moreover, both HO-1-deficient mice (
      • Poss K.D.
      • Tonegawa S.
      ) and a human case of genetic HO-1 deficiency (
      • Saikawa Y.
      • Kaneda H.
      • Yue L.
      • Shimura S.
      • Toma T.
      • Kasahara Y.
      • Yachie A.
      • Koizumi S.
      ) exhibit a serious impairment of iron metabolism, leading to liver and kidney oxidative damage and inflammation. Therefore, it is now widely accepted that induction of HO-1 expression represents an adaptive response that increases cell resistance to oxidative injury.
      The signaling mechanisms used to activate transcription of phase II genes such as HO-1 are poorly defined. Most studies have focused on the activation of the mitogen-activated protein kinases (MAPKs) related to cell growth and stress response. In vertebrates, the three major MAPK pathways are represented by kinase cascades leading to activation of extracellular signal-regulated kinase (ERK), c-Jun N-terminal kinase (JNK), and p38 (for a review, see Ref.
      • Johnson G.L.
      • Lapadat R.
      ). All three pathways appear to be involved to some extent in the up-regulation of HO-1 expression in response to diverse stimuli. For instance, in rat hepatocytes, sodium arsenite appears to regulate HO-1 expression via activation of Ras and the JNK pathway, but not the ERK pathway (
      • Kietzmann T.
      • Samoylenko A.
      • Immenschuh S.
      ). On the other hand, arsenite induces HO-1 expression via the ERK and p38 pathways in LMH chicken hepatoma cells (
      • Shan Y.
      • Pepe J.
      • Lu T.H.
      • Elbirt K.K.
      • Lambrecht R.W.
      • Bonkovsky H.L.
      ). Ischemia-reperfusion in the lung activates HO-1 through JNK and p38 (
      • Ohlmann A.
      • Giffhorn-Katz S.
      • Becker I.
      • Katz N.
      • Immenschuh S.
      ). Finally, regarding phytophenols, curcumin up-regulates HO-1 through the p38 pathway in porcine renal epithelial cells (
      • Balogun E.
      • Hoque M.
      • Gong P.
      • Killeen E.
      • Green C.J.
      • Foresti R.
      • Alam J.
      • Motterlini R.
      ).
      However, the role of the survival signaling pathway represented by phosphatidylinositol 3-kinase (PI3K)/Akt in the regulation of HO-1 expression is still not defined. An initial study by Lee et al. (
      • Lee J.M.
      • Hanson J.M.
      • Chu W.A.
      • Johnson J.A.
      ) suggested a role for PI3K in the regulation of the phase II enzyme NAD(P)H:quinone oxidoreductase. In addition, we previously showed that nerve growth factor induces the expression of HO-1 by a still unknown PI3K-dependent mechanism (
      • Salinas M.
      • Diaz R.
      • Abraham N.G.
      • Ruiz de Galarreta C.M.
      • Cuadrado A.
      ).
      It is accepted that induction of HO-1 expression is mediated through cis-regulatory DNA sequences located in their promoter region. Among the large variety of putative regulatory sites found in the promoter 5′-flanking region of human, rat, mouse, and chicken HO-1 genes (
      • Elbirt K.K.
      • Bonkovsky H.L.
      ), most investigations have focused on those that exhibit inducible responses to oxidative signals (
      • Hill-Kapturczak N.
      • Thamilselvan V.
      • Liu F.
      • Nick H.S.
      • Agarwal A.
      ). Several AP-1-binding sites have been shown to be involved in the transcriptional induction of HO-1 by MAPKs, cyclic nucleotides, protein phosphatase inhibitors, and the polyphenol curcumin (
      • Gong P.
      • Stewart D.
      • Hu B.
      • Vinson C.
      • Alam J.
      • Maeshima H.
      • Sato M.
      • Ishikawa K.
      • Katagata Y.
      • Yoshida T.
      • Kiemer A.K.
      • Bildner N.
      • Weber N.C.
      • Vollmar A.M.
      ).
      More recently, a new class of AP-1-related sites has been shown to mediate oxidative stress responsiveness of phase II genes. These regions are termed stress response elements or antioxidant response elements (AREs) and are tightly regulated by the redox-sensitive transcription factor Nrf2 (NF-E2-related factor-2) (
      • Nguyen T.
      • Sherratt P.J.
      • Pickett C.B.
      ). Nrf2 is a member of the Cap’n’Collar family of transcription factors (
      • Moi P.
      • Chan K.
      • Asunis I.
      • Cao A.
      • Kan Y.W.
      ). Under non-stimulated conditions, Nrf2 is sequestered in the cytoplasm by binding to Keap1, an actin-binding protein (
      • Itoh K.
      • Wakabayashi N.
      • Katoh Y.
      • Ishii T.
      • Igarashi K.
      • Engel J.D.
      • Yamamoto M.
      ). Several stimuli, including oxidative stress, lead to the disruption of this complex, freeing Nrf2 for translocation to the nucleus and dimerization with basic leucine zipper transcription factors such as Maf and Jun family members (
      • Hayes J.D.
      • McMahon M.
      ). The mechanism by which Nrf2 is liberated from the Nrf2-Keap1 complex remains controversial, but probably involves phosphorylation and stabilization of Nrf2 protein (
      • Huang H.C.
      • Nguyen T.
      • Pickett C.B.
      ,
      • Nguyen T.
      • Sherratt P.J.
      • Huang H.C.
      • Yang C.S.
      • Pickett C.B.
      ), which otherwise has a half-life of just 30 min and is degraded via the ubiquitin-proteasome pathway (
      • Nguyen T.
      • Sherratt P.J.
      • Huang H.C.
      • Yang C.S.
      • Pickett C.B.
      • Stewart D.
      • Killeen E.
      • Naquin R.
      • Alam S.
      • Alam J.
      • McMahon M.
      • Itoh K.
      • Yamamoto M.
      • Hayes J.D.
      ).
      In this study, we analyzed the effect of carnosol on expression of the antioxidant enzyme HO-1, and we dissected the signaling pathways leading to HO-1 regulation. Our results indicate that carnosol activates HO-1 expression probably by increasing Nrf2 protein levels in a PI3K/Akt-dependent manner. Therefore, in addition to its intrinsic antioxidant nature, carnosol activates the PI3K/Akt survival pathway and up-regulates expression of HO-1.

      EXPERIMENTAL PROCEDURES

      supplemented with 1% fetal bovine serum and 1% horse serum, an experimental condition that does not compromise viability in our PC12 cells (
      • Salinas M.
      • Diaz R.
      • Abraham N.G.
      • Ruiz de Galarreta C.M.
      • Cuadrado A.
      ,
      • Martin D.
      • Salinas M.
      • Lopez-Valdaliso R.
      • Serrano E.
      • Recuero M.
      • Cuadrado A.
      ). For luciferase assays, transient transfections of PC12 cells were performed wCell Culture, Transfections, and Reagents—Rat pheochromocytoma PC12 cells were seeded on poly-d-lysine-coated plates and grown in Dulbecco’s modified Eagle’s medium supplemented with 7.5% fetal bovine serum, 7.5% heat-inactivated horse serum, and 80 μg/ml gentamicin. Subconfluent cell cultures were maintained overnight (16 h) in Dulbecco’s modified Eagle’s mediumith the expression vectors pGL3bv, pHO1–15-Luc, pHO1–15-LucΔE1, pHO1–15-LucΔE2, pHO1–15-Luc(ΔE1+ΔE2), pEF-Nrf2(DN), pEF18-MafG(DN) (
      • Alam J.
      • Wicks C.
      • Stewart D.
      • Gong P.
      • Touchard C.
      • Otterbein S.
      • Choi A.M.
      • Burow M.E.
      • Tou J.
      ), pCEFL(X)Akt1(K179M) (
      • Martin D.
      • Salinas M.
      • Lopez-Valdaliso R.
      • Serrano E.
      • Recuero M.
      • Cuadrado A.
      ), pcDNA3-HA-PKCζ(DN) (kindly provided by Dr. J. Moscat, Centro de Biología Molecular, Madrid), pcDNA3, and pcDNA3-β-gal. LipofectAMINE reagent (Invitrogen) was used according to the manufacturer’s instructions. Nerve growth factor and carnosol were purchased from Calbiochem, and hemin, LY294002, PD98059, SB203580, SP600125, and MG132 were from Sigma. These drugs were dissolved in Me2SO and used at a final vehicle concentration of <0.2%.
      Analysis of mRNA Levels by Semiquantitative Reverse Transcription-PCR—Cells were plated on 60-mm dishes, and total cellular RNA was extracted using TRIzol reagent (Invitrogen). Equal amounts (1 μg) of RNA from the different treatments were reverse-transcribed for 75 min at 42 °C using 5 units of avian myeloblastosis virus reverse transcriptase (Promega, Madison, WI) in the presence of 20 units of RNasin (Promega). Amplification of cDNA was performed in 25 μl of PCR buffer (10 mm Tris-HCl, 50 mm KCl, 5 mm MgCl2, and 0.1% Triton X-100, pH 9.0) containing 0.6 units of Taq DNA polymerase (Promega) and 30 pmol of synthetic gene-specific primers for HO-1 (forward, 5′-AAGGCTTTAAGCTGGTGATGG-3′; and reverse, 5′-AGCGGTGTCTGGGATGAACTA-3′). To ensure that equal amounts of cDNA were added to the PCR, the β-actin gene was amplified using synthetic oligonucleotides (forward, 5′-TGTTTGAGACCTTCAACACC-3′; and reverse, 5′-CGCTCATTGCCGATAGTGAT-3′). After an initial denaturation step for 4 min at 94 °C, amplification of each cDNA was performed for the minimal number of cycles that allowed detection of basal HO-1 mRNA levels in control cells (data not shown), 19 cycles for β-actin and 24 cycles for HO-1, using a thermal profile of 1 min at 94 °C (denaturalization), 1 min at 58 °C (annealing), and 1 min at 72 °C (elongation). The amplified PCR products were resolved by 1.8% agarose gel electrophoresis and stained with ethidium bromide.
      Real-time Quantitative PCR—Quantitative PCR was performed with 20 ng of cDNA obtained as described above in 25 μl containing 0.5 μm primers (HO-1, 5′-GCCTGCTAGCCTGGTTCAAG-3′ (forward) and 5′-AGCGGTGTCTGGGATGAACTA-3′ (reverse); and β-actin, 5′-GCCTCACTGTCCACCTTCCA-3′ (forward) and 5′-CCCGGCCTGAGTAGCATGA-3′ (reverse)) and the nucleotides, buffer, and Taq polymerase included in SYBR Green I Mastermix (PE Applied Biosystems). Amplification was conducted in an ABI Prism 7700 sequence detection system. PCR conditions were as follows: 50 °C for 2 min, 95 °C for 10 min, 40 cycles at 94 °C for 15 s, and 60 °C for 1 min. HO-1 expression was estimated in duplicate samples and was normalized to β-actin expression levels.
      Preparation of Nuclear Extracts—PC12 cells (5 × 106) were washed three times with cold phosphate-buffered saline and harvested by centrifugation at 1100 rpm for 10 min. The cell pellet was carefully resuspended in 3 pellet volumes of cold buffer containing 20 mm HEPES, pH 7.0, 0.15 mm EDTA, 0.015 mm EGTA, 10 mm KCl, 1% Nonidet-40, 1 μg/ml leupeptin, 1 mm phenylmethylsulfonyl fluoride, 20 mm NaF, 1 mm sodium pyrophosphate, and 1 mm Na3VO4. The homogenate was then centrifuged at 500 × g for 20 min, and the nuclear pellet was resuspended in 5 pellet volumes of cold buffer containing 10 mm HEPES, pH 8.0, 25% glycerol, 0.1 m NaCl, 0.1 mm EDTA, 1 μg/ml leupeptin, 1 mm phenylmethylsulfonyl fluoride, 20 mm NaF, 1 mm sodium pyrophosphate, and 1 mm Na3VO4. After centrifugation at 500 × g for 20 min, nuclei were resuspended in 2 pellet volumes of hypertonic cold buffer containing 10 mm HEPES, pH 8.0, 25% glycerol, 0.4 m NaCl, 0.1 mm EDTA, 1 μg/ml leupeptin, 1 mm phenylmethylsulfonyl fluoride, 20 mm NaF, 1 mm sodium pyrophosphate, and 1 mm Na3VO4 and incubated for 30 min at 4 °C on a rotating wheel. Nuclear debris was removed by centrifugation at 900 × g for 20 min at 4 °C; one part of the supernatant was resolved by SDS-PAGE and submitted to immunoblot analysis using anti-Nrf2 and anti-Sp1 antibodies. The rest of the supernatant was used for electrophoretic mobility shift assay (EMSA). Protein concentration was determined with the Dc protein assay kit (Bio-Rad).
      EMSAs—The double-stranded wild-type oligonucleotide used as the Nrf2 probe was 5′-TTTTATGCTGTGTCATGGTT-3′. The complementary strand was annealed by incubation at a concentration of 250 ng/μl in STE buffer (100 mm NaCl, 10 mm Tris, pH 8.0, and 1 mm EDTA) at 80 °C for 2 min. The mixture was then slowly cooled to 4 °C with a thermal profile of 1 °C/min in a thermal incubator. Annealed oligonucleotides were diluted to 25 ng/μl in STE buffer. The 5′-end labeling was performed with T4 polynucleotide kinase (Promega) using 25 ng of double-stranded oligonucleotide and 25 μCi of [γ-32P]ATP (3000 Ci/mmol; Amersham Biosciences, Little Chalfont, Buckinghamshire, United Kingdom). The labeled probes were purified using a Sephadex G-25 spin column (Amersham Biosciences). The binding reaction mixture contained 5 μg of nuclear extract in buffer containing 40 mm HEPES, pH 8.0, 50 mm KCl, 0.05% Nonidet-40, 1% dithiothreitol, and 10 μg/ml poly(dI)·poly(dC) in a total volume of 20 μl. Anti-Nrf2 antibody (2 μg) or unlabeled competitor probe (25 ng) was added as indicated, and the reaction was incubated for 1 h at 25 °C. Labeled DNA (0.25 ng) was added to the mixture and submitted to an additional 20-min incubation at 25 °C. Samples were resolved at 4 °C on a 5% nondenaturing polyacrylamide gel in 0.5× Tris borate/EDTA. After electrophoresis, the gel was dried and autoradiographed.
      Immunoblotting—30 μg of cell lysate were resolved by SDS-PAGE and transferred to Immobilon-P membranes (Millipore Corp.). Blots were analyzed with the appropriate antibodies: anti-Cys20 Akt1 (1:500), anti-ERK2 (1:1000), anti-p38 (1:1000), anti-Nrf2 (1:16000), and anti-Sp1 (1:1000) (Santa Cruz Biotechnology, Santa Cruz, CA); anti-phospho-Ser473 Akt1 (1:1000), anti-phospho-ERK2 (1:1000), and anti-phospho-p38 (1:1000) (Cell Signaling Technology, Inc., Beverly, MA); anti-phospho-JNK (1:1000) (Upstate Biotechnology, Inc., Lake Placid, NY); anti-protein-disulfide isomerase (kind gift of Dr. González Castaño, Universidad Autónoma de Madrid); and anti-HO-1 (1:1000) and anti-HO-2 (1:1000) (Stressgen Biotech Corp., Victoria, British Columbia, Canada). Appropriate peroxidase-conjugated secondary antibodies (1:10,000) were used to detect the proteins of interest by enhanced chemiluminescence.
      Flow Cytometry—A FACScan flow cytometer (BD Biosciences) was used to analyze intracellular reactive oxygen species with the fluorescence probe 2′,7′-dichlorodihydrofluorescein diacetate (H2DCFDA) (Molecular Probes, Inc., Leiden, The Netherlands), which passively diffuses into the cell and is cleaved and oxidized to 2′,7′-dichlorofluorescein (band pass 530/25 nm). Cells were submitted to carnosol, zinc-protoporphyrin IX (ZnPPIX), LY294002 (as indicated for each experiment), and H2O2 (5 min); detached mechanically from the plates; washed once with cold phosphate-buffered saline; and analyzed immediately. Three independent samples of 10,000 cells were analyzed for each experimental condition.
      Statistics—Student’s t test was used to assess differences between groups. A p value <0.05 was considered significant. All luciferase and flow cytometry experiments were performed at least three times using triplicate or quadruplicate samples per group. The values in graphs correspond to the mean ± S.E. of at least three samples. Densitometric analyses were performed with NIH Image Version 2.51 software on representative immunoblots from experiments that were repeated two to five times.

      RESULTS

      Carnosol Induces Expression of HO-1—We investigated the possibility that carnosol, a natural diterpene derived from the herb rosemary, might alter the expression of the antioxidant enzyme HO-1. PC12 cells were maintained for 16 h in Dulbecco’s modified Eagle’s medium supplemented with 7.5% fetal bovine serum and 7.5% horse serum (high serum medium) or with 1% fetal bovine serum and 1% horse serum (low serum medium). Under both growing conditions, cells retained a similar viability as determined by cytometric analysis of propidium iodide incorporation and annexin V-phyroerythrin staining, at least within the time frame of these experiments (data not shown and Ref.
      • Martin D.
      • Salinas M.
      • Lopez-Valdaliso R.
      • Serrano E.
      • Recuero M.
      • Cuadrado A.
      ). Cells were stimulated with 10 μm carnosol for 6 h or with an equivalent volume of Me2SO vehicle (final concentration of <0.2%) and analyzed for the amount of HO-1 protein. In agreement with a recent study (
      • Liu H.
      • Nowak R.
      • Chao W.
      • Bloch K.D.
      ), cells grown in low serum medium exhibited a slight decrease in HO-1 protein levels compared with cells grown in high serum medium (Fig. 1A). Interestingly, carnosol induced a significant increase in HO-1 protein levels that was more evident in cells submitted to low serum conditions. To further study the effect of carnosol on induction of HO-1 expression, PC12 cells were maintained in low serum medium for 16 h and then treated for 6 h with carnosol concentrations ranging from 1 to 100 μm. Real-time PCR (Fig. 1B) indicated that carnosol induced a maximal increase (>200-fold) in HO-1 mRNA at 10 μm. Moreover, as shown in Fig. 1C, semiquantitative PCR also evidenced a dose-dependent increase in HO-1 mRNA levels. This increase was not as high as that found with real-time PCR because amplicon size and primers were different (see “Experimental Procedures”) and because, to detect basal mRNA levels in the carnosol-untreated cells, it was necessary to increase PCR cycles over the linear range in the carnosol-induced cells. In addition, carnosol also produced a parallel increase in HO-1 protein levels, whereas HO-2 levels corresponding to the other isozyme constitutively expressed in these cells remained without significant alteration (Fig. 1D). Consistent with these data, 10 μm carnosol also induced a time-dependent increase in HO-1 mRNA levels that was apparent as early as 2 h after treatment and was followed by a robust increase in HO-1 protein levels at 4 h (Fig. 1, E and F). Similar results were obtained in other cell lines of human and mouse origin (data not shown). Therefore, these results suggest that carnosol specifically up-regulates expression of the inducible HO-1 isozyme.
      Figure thumbnail gr1
      Fig. 1Carnosol induces HO-1 expression.A, effects of serum and carnosol on the protein levels of HO-1 and HO-2 isozymes. PC12 cells were maintained for 16 h in low or high serum medium and then stimulated with vehicle (Me2SO) or 10 μm carnosol for 6 h. Upper panel, immunoblot with anti-HO-1 antibody; middle panel, immunoblot with anti-HO-2 antibody; lower panel, densitometric quantification of relative HO-1 protein levels. B, real-time PCR determination of HO-1 mRNA levels in the presence of carnosol. PC12 cells maintained for 16 h under low serum conditions were incubated with the indicated concentrations of carnosol and analyzed as described under “Experimental Procedures.” C, semiquantitative reverse transcription-PCR showing a dose-dependent induction of HO-1 mRNA by carnosol. Upper panel, HO-1 mRNA; middle panel, β-actin mRNA used for normalization; lower panel, densitometric quantification of relative HO-1 mRNA levels. D, dose-dependent induction of HO-1 protein by carnosol. Upper panel, immunoblot with anti-HO-1 antibody; middle panel, immunoblot with anti-HO-2 antibody; lower panel, densitometric quantification of relative HO-1 protein levels. For C and D, PC12 cells maintained in low serum medium for 16 h were treated with the indicated carnosol concentrations for 6 h or with 50 μm hemin, a well established induced of HO-1, and then analyzed for HO-1 mRNA and protein levels. E, time course of HO-1 mRNA expression by carnosol. Upper panel, HO-1 mRNA; middle panel, β-actin mRNA used for normalization; lower panel, densitometric quantification of relative HO-1 mRNA levels. F, time course of HO-1 protein expression by carnosol. Upper panel, immunoblot with anti-HO-1 antibody; middle panel, immunoblot with anti-HO-2 antibody; lower panel, densitometric quantification of relative HO-1 protein levels. For E and F, PC12 cells were maintained in low serum medium for 16 h and then submitted to 10 μm carnosol for the indicated times or with 50 μm hemin for 6 h.
      Carnosol Activates HO-1 through the AREs—The mouse ho1 promoter contains two AREs that might represent potential targets of regulation by carnosol. To investigate this possibility, PC12 cells were transfected with luciferase expression vectors carrying the wild-type 15-kb mouse promoter or deletion mutant versions lacking either the proximal (ΔE1) or distal (ΔE2) ARE regulatory sequence or both regulatory sequences (ΔE1+ΔE2). Cells were maintained in low serum medium for 16 h and then stimulated with 10 μm carnosol for 16 h. As shown in Fig. 2, deletion of ARE sequence E1 or E2 resulted in 30 and 36% decreases in carnosol induction, respectively. Interestingly, deletion of both AREs abolished the response to carnosol. Therefore, these results suggest that carnosol activates HO-1 expression through both AREs.
      Figure thumbnail gr2
      Fig. 2Carnosol targets the mouse ho1 gene promoter at the two AREs. PC12 cells were transfected with expression vectors for luciferase under the control of the 15-kb 5′-promoter region of mouse wild-type ho1 (WT) or the same promoter with deletions of E1 (ΔE1), E2 (ΔE2), or both regions (ΔE1+ΔE2) comprising the AREs. In addition, cells were cotransfected with pcDNA3-β-gal for normalization. 16 h after transfection, cells were maintained in low serum medium for 16 h and then stimulated with 10 μm carnosol for an additional 16 h. Cells were lysed and analyzed for luciferase and β-galactosidase activities.
      Carnosol Increases the Levels of the Nrf2 Transcription Factor—Several transcription factors have been reported to interact with AREs. However, considering that induction of HO-1 expression by several inducers requires de novo protein synthesis (
      • Chen K.
      • Maines M.D.
      ,
      • Hill-Kapturczak N.
      • Truong L.
      • Thamilselvan V.
      • Visner G.A.
      • Nick H.S.
      • Agarwal A.
      ), we analyzed the possibility that this molecule would operate on Nrf2, a transcription factor with a very short half-life (
      • Nguyen T.
      • Sherratt P.J.
      • Huang H.C.
      • Yang C.S.
      • Pickett C.B.
      • Stewart D.
      • Killeen E.
      • Naquin R.
      • Alam S.
      • Alam J.
      • McMahon M.
      • Itoh K.
      • Yamamoto M.
      • Hayes J.D.
      ) that regulates AREs. PC12 cells were maintained in low serum medium for 16 h and then stimulated for 3 h with the carnosol concentrations indicated in Fig. 3A. As a control, cells were also treated with the proteasome inhibitor MG132. In agreement with previous studies (
      • Nguyen T.
      • Sherratt P.J.
      • Huang H.C.
      • Yang C.S.
      • Pickett C.B.
      • Stewart D.
      • Killeen E.
      • Naquin R.
      • Alam S.
      • Alam J.
      • McMahon M.
      • Itoh K.
      • Yamamoto M.
      • Hayes J.D.
      ), blockage of proteolysis with MG132 resulted in a large accumulation of Nrf2 protein within 3 h, even in the absence of carnosol. Interestingly, carnosol increased Nrf2 levels in a dose-dependent manner, with the maximal effect at 10 μm in our PC12 cells. Additionally, we analyzed the time course of Nrf2 accumulation in the presence of 10 μm carnosol. As shown in Fig. 3B, the accumulation of Nrf2 protein was already detectable after 30 min in the presence of 10 μm carnosol and increased for at least 3 h during treatment. Of note, the combined treatment with the proteasome inhibitor MG132 and carnosol did not significantly increase Nrf2 levels over those obtained with each drug separately, suggesting that carnosol prevents degradation of Nrf2 via the ubiquitin-proteasome pathway.
      Figure thumbnail gr3
      Fig. 3Carnosol increases the levels of the Nrf2 transcription factor. Shown are the dose effects (A) and time course (B) of carnosol-induced increases in Nrf2 protein levels. Upper panels, blots with anti-Nrf2 antibody; middle panels, blots with anti-protein-disulfide isomerase (PDI) antibody used for normalization; lower panels, densitometric quantification of relative Nrf2 levels. In A, PC12 cells grown in low serum medium were treated for 3 h with the indicated carnosol concentrations, with MG132 (20 μm), or with both MG132 (20 μm) and carnosol (10 μm) as indicated. In B, PC12 cells maintained in low serum medium for 16 h were treated with 10 μm carnosol for the indicated times, with MG132, or with MG132 plus carnosol for 3 h.
      Next, we analyzed the nuclear levels of carnosol-induced Nrf2 protein. As shown in Fig. 4A, both MG132 and carnosol induced a strong accumulation of Nrf2 in the nucleus, whereas the levels of the housekeeping transcription factor Sp1 were stable. Moreover, we investigated whether the carnosol-induced increase in Nrf2 in the nucleus was sufficient to promote binding of protein complexes containing this transcription factor to the ARE (E1 sequence) corresponding to the mouse ho1 promoter. EMSA reactions were performed with nuclear extracts from cells exposed to carnosol or MG132. As shown in Fig. 4B, one of the several retarded complexes formed was significantly increased in intensity upon treatment with carnosol or MG132. Moreover, when the EMSA reaction was conducted in the presence of anti-Nrf2 antibody, the intensity of this complex dropped dramatically, indicating the presence of Nrf2 in the complex. Taken together, these results indicate that carnosol increases both total and nuclear protein levels of Nrf2, resulting in increased binding to the ARE promoter sequence of HO-1.
      Figure thumbnail gr4
      Fig. 4Carnosol stimulates activation of ho1 promoter AREs by Nrf2.A, increases in Nrf2 protein levels in nuclear extracts of cells treated with 10 μm carnosol or 20 μm MG132 for 3 h. B, EMSA reactions using a double-stranded oligonucleotide corresponding to the mouse ho1 ARE and nuclear extracts from cells treated with 10 μm carnosol or 20 μm MG132. The last two lanes correspond to cell extracts from carnosol-treated cells preincubated without (–) or with (+) anti-Nrf2 antibody. The arrowhead points to the position of the complex that was not formed in the presence of anti-Nrf2 antibody. C, requirement of Nrf2 and MafG for activation of the mouse ho1 promoter by carnosol. PC12 cells were cotransfected with pHO1–15-Luc, pcDNA-β-gal, and expression vectors for dominant-negative (DN) versions of Nrf2 or MafG as indicated. After 16 h of transfection, cells were kept in low serum medium for 16 h and then stimulated with 10 μm carnosol. After 6 h, cells were lysed and analyzed for luciferase and β-galactosidase activities.
      To verify the functional relevance of Nrf2 binding to the ho1 ARE, PC12 cells were cotransfected with pHO1–15-Luc and expression vectors for dominant-negative mutants of either Nrf2 or one of its dimerization partners, MafG. As shown in Fig. 4C, the carnosol-induced activation of the ho1 promoter was significantly decreased in the presence of dominant-negative Nrf2 or MafG or both mutants. These results further suggest that Nrf2 mediates the carnosol-induced activation of the ho1 promoter.
      Carnosol Activates the Canonical MAPK Cascades and the PI3K/Akt Pathway—To identify the signaling pathways used by carnosol to induce HO-1 expression, we first analyzed the effect of this phytophenol on the three MAPK cascades leading to activation of ERK, p38 and JNK and on the survival pathway represented by the PI3K and Akt kinases. The activation of these pathways was analyzed with activation-specific antibodies that selectively recognize the active and phosphorylated forms of ERK, p38, JNK, and Akt (see “Experimental Procedures”). Cells, maintained under low serum conditions, were treated with 10 μm carnosol for the indicated times, and cell lysates were immunoblotted with phospho-specific antibodies. As shown in Fig. 5, carnosol activated the three MAPK cascades as well as the PI3K/Akt pathway with different kinetics. Maximal activation of ERK was observed after 30 min of stimulation, whereas maximal activation of Akt, JNK, and p38 required at least 1 h. Moreover, ERK activation was more transient. Therefore, we analyzed which of these pathways might be involved in channeling the stimulus for HO-1 upregulation. Cells grown in low serum medium were preincubated for 15 min with inhibitors of each pathway (LY294002 (PI3K), PD98059 (MEK), SB203580 (p38), and SP600125 (JNK)) and then stimulated with 10 μm carnosol for 6 h. The effectiveness of these inhibitors was confirmed in carnosol-treated cells with antibodies specific for phospho-ERK, phospho-Jun, phospho-activating transcription factor-2, and phospho-Akt1 (data not shown). As shown in Fig. 6A, HO-1 protein levels increased following incubation with 10 μm carnosol, as expected. Inhibition of the ERK and JNK pathways had little or no effect on HO-1 protein levels, suggesting that these MAPKs are not required for HO-1 up-regulation by carnosol. Curiously, inhibition of both PI3K and, to a lower extent, p38 pathways dramatically reduced the capacity of carnosol to increase HO-1 protein levels.
      Figure thumbnail gr5
      Fig. 5Carnosol activates the MAPK and PI3K pathways. PC12 cells in low serum medium were stimulated with 10 μm carnosol for the indicated times and then immunoblotted with activation-specific antibodies that recognize phospho-Akt1 (P-Akt1), phospho-ERK2 (P-ERK2), phospho-JNK (P-JNK), and phospho-p38 (P-p38). Parallel immunoblots were analyzed for total kinase levels with anti-Akt1, anti-ERK, anti-JNK, and anti-p38 antibodies, respectively. The graphs correspond to densitometric quantification of phospho-Akt1, phospho-ERK2, phospho-JNK, and phospho-p38 levels relative to total kinase levels.
      Figure thumbnail gr6
      Fig. 6Effect of MAPK and PI3K inhibitors on induction of HO-1 by carnosol.A, upper panel, immunoblot with anti-HO-1 antibody; lower panel, immunoblot with anti-protein-disulfide isomerase (PDI) antibody used for normalization. PC12 cells were maintained in low serum medium for 16 h and then were preincubated with 40 μm LY294002 (LY), 50 μm PD98059 (PD), 5 μm SB203580 (SB), or 10 μm SP600125 (SP) for 15 min. Cells were submitted to 10 μm carnosol for 6 h. B, carnosol induces luciferase activity in the mouse ho1 promoter in the presence of the MAPK and PI3K inhibitors. PC12 cells were cotransfected with pHO1–15-Luc and pcDNA3-β-gal vectors. After 16 h, cells were transferred to low serum medium in the presence of the inhibitors and then stimulated with 10 μm carnosol for 16 h. Cells were lysed and analyzed for luciferase and β-galactosidase activities.
      To further confirm these observations, we analyzed the effect of carnosol on the upstream regulatory sequences of the mouse ho1 gene. PC12 cells were cotransfected with expression vectors for β-galactosidase (used for normalization) and pHO1–15-Luc. After serum starvation, cells were pretreated with the MAPK and PI3K inhibitors for 15 min and then incubated with 10 μm carnosol for 16 h. As shown in Fig. 6B, carnosol increased luciferase activity, consistent with the notion that it activates HO-1 expression. Moreover, the stimulation of this promoter fragment was significantly reduced in the presence of the p38 and PI3K inhibitors and remained unaltered in the presence of the ERK and JNK inhibitors. Therefore, these observations on the mouse promoter correlate with the induction of the endogenous rat HO-1 protein and strongly suggest that carnosol requires PI3K and, to a lower extent, p38 kinases to induce HO-1 expression in PC12 cells.
      Because other groups have described the involvement of p38 in activation of HO-1 in the presence of polyphenols (
      • Balogun E.
      • Hoque M.
      • Gong P.
      • Killeen E.
      • Green C.J.
      • Foresti R.
      • Alam J.
      • Motterlini R.
      ), we focused our study on the role of PI3K. Cells maintained in low serum medium were pretreated with several concentrations of the PI3K inhibitor LY294002 and then submitted to 10 μm carnosol for 6 h. As a control for the effectiveness and specificity of the inhibitor, we monitored the activation of the PI3K downstream effector Akt1 after 30 min of carnosol addition with the activation-specific anti-phospho-Akt antibody. As shown Fig. 7A, LY294002 concentrations in the low micromolar range successfully inhibited the carnosol-induced activation of Akt and also blocked, at least partially, the increase in HO-1 and Nrf2 protein levels. Therefore, these results indicate that PI3K is required to induce HO-1 expression and suggest that the Nrf2 transcription factor may mediate, at least in part, this response.
      Figure thumbnail gr7
      Fig. 7The PI3K/Akt axis is involved in the carnosol-induced increase in Nrf2 protein levels and in ho1 promoter activation.A, inhibition of PI3K prevents the increase in HO-1 and Nrf2 protein levels in response to carnosol. PC12 cells maintained in low serum medium for 16 h were preincubated with the indicated LY294002 (LY) concentrations for 15 min and then with 10 μm carnosol for either 30 min (immunoblot with anti-phospho-Akt1 antibodies (P-Akt1)) or 6 h (immunoblots with anti-HO-1, anti-Nrf2, and anti-protein-disulfide isomerase (PDI) antibodies). B, PI3K induces HO-1 expression through regulation of AREs. PC12 cells were transfected with pHO1–15-Luc, pHO1–15-LucΔE1, pHO1–15-LucΔE2, or pHO1–15-Luc(ΔE1+ΔE2) together with either empty vector or the expression vector for active PI3K (PI3K-CAAX). In addition, cells were cotransfected with pcDNA3-β-gal for normalization. After 16 h of transfection, cells were transferred to low serum medium for 16 h and analyzed for luciferase and β-galactosidase activities. C, induction of HO-1 expression by PI3K is attenuated by interfering with Nrf2 function. PC12 cells were cotransfected with pHO1–15-Luc, empty vector (pcDNA3), PI3K-CAAX, pcDNA-β-gal, and expression vectors for the dominant-negative (DN) versions of Nrf2 and MafG as indicated. After 16 h of transfection, cells were kept in low serum medium for 16 h and then stimulated with 10 μm carnosol. After 6 h, cells were lysed and analyzed for luciferase and β-galactosidase activities. WT, wild-type.
      Next, we analyzed whether PI3K alone is sufficient to activate the ho1 promoter and the role of AREs in such a case. We cotransfected PC12 cells with pcDNA3-β-gal, an expression vector for constitutively active PI3K (PI3K-CAAX, where A is aliphatic amino acid), and a luciferase expression vector (pHO1–15-Luc, pHO1–15-LucΔE1, pHO1–15-LucΔE2, or pHO1–15-Luc(ΔE1+ΔE2)). As shown in Fig. 7B, active PI3K was sufficient to induce the activity of the wild-type ho1 promoter. However, this activation was significantly decreased in the deletion mutants lacking the E1 or E2 site and was drastically reduced in the double mutant lacking both sites. In addition, we analyzed the requirement of Nrf2 for PI3K-induced activation of the ho1 promoter. PC12 cells were cotransfected with pHO1–15-Luc, pcDNA3-β-gal, and PI3K-CAAX together with the expression vectors for dominant-negative Nrf2 and MafG mutants. In several independent experiments, the stimulatory effect of PI3K-CAAX on the ho1 promoter was smaller in this experimental setting, implying cotransfection of four plasmids, compared with the previous experiments performed by cotransfection of two or three plasmids. Nevertheless, as shown in Fig. 7C, dominant-negative Nrf2 and MafG alone or in combination significantly blocked PI3K-induced activation of the ho1 promoter. These results indicate that PI3K is sufficient to activate the ho1 promoter through the AREs in an Nrf2-dependent manner.
      We also analyzed the contribution of kinases regulated by PI3K to the carnosol-induced up-regulation of HO-1. PC12 cells were transfected with the expression vector pHO1–15-Luc and hemagglutinin (HA)-tagged dominant-negative versions of Akt1 and protein kinase Cζ (PKCζ), two well known effectors of PI3K. Overexpression of these proteins was determined by immunoblot analysis with anti-HA antibody (Fig. 8C). As shown in Fig. 8A, transfection with dominant-negative PKCζ(K281W) did not have a significant effect on the response to carnosol. By contrast, as shown in Fig. 8B, dominant-negative Akt1(K179M) significantly blocked the induction of HO-1. Moreover, cotransfection of both expression vectors inhibited carnosol induction to the same extent as dominant-negative Akt1 alone (data not shown). These results suggest that Akt1 (rather than PKCζ) is responsible for at least part of the carnosol-induced up-regulation of HO-1. Moreover, as shown in Fig. 8D, PC12 cells stably transfected with active PI3K and Akt1 versions displayed increased HO-1 protein levels, further indicating that the PI3K/Akt axis is sufficient to up-regulate HO-1 expression.
      Figure thumbnail gr8
      Fig. 8Carnosol requires Akt1 (but not PKCζ) to activate the ho1 promoter. Cells were transfected with pHO1–15-Luc and the indicated vectors for expression of dominant-negative (DN) Nrf2 or the indicated amounts of dominant-negative PKCζ (A) or of dominant-negative Akt1 (B). Cells were also transfected with pcDNA3-β-gal for normalization. After 16 h of transfection, cells were maintained in low serum medium for 16 h and analyzed for luciferase and β-galactosidase activities. In C, HA-tagged dominant-negative versions of PKCζ and Akt1 were detected by immunoblotting with anti-HA antibody. Cells were treated in parallel with the luciferase experiments indicated for A and B, lysed, and immunoprecipitated with anti-HA antibody. The immune complexes were resolved by SDS-PAGE and immunoblotted with anti-HA antibody. The lower band corresponds to the immunoglobulin heavy chain (IgG H) of anti-HA antibody used during immunoprecipitation. In D, the immunoblots show HO-1 protein levels in cells stably transfected with empty vector (pcDNA3), membrane-anchored active PI3K (PI3K-CAAX), and membrane-anchored active myristoylated (myr) Akt1. Upper panel, immunoblot with anti-HO-1 antibody; lower panel, immunoblot with anti-protein-disulfide isomerase (PDI) antibody.
      Finally, we studied whether carnosol attenuates oxidative stress through the induction of HO-1 expression and the relevance of the PI3K pathway to such an induction. PC12 cells maintained in low serum medium for 16 h were preloaded for 1 h at 37 °C with 8 μm H2DCFDA, a probe that passively diffuses into cells and is cleaved and oxidized in the intracellular environment to the green fluorescence-emitting compound 2′,7′-dichlorofluorescein (
      • Ischiropoulos H.
      • Gow A.
      • Thom S.R.
      • Kooy N.W.
      • Royall J.A.
      • Crow J.P.
      ). Next, cells were treated with Me2SO vehicle (control) or carnosol for 30 min and finally exposed to 250 μm H2O2 for 5 min. Cells were immediately analyzed by flow cytometry. As shown in Fig. 9 (A and B), micromolar concentrations of carnosol dose-dependently prevented the intracellular oxidation of the fluorescent probe despite the presence of H2O2. Moreover, as shown in Fig. 9C, 10 μm carnosol fully prevented the oxidation of H2DCFDA by H2O2 concentrations ranging from 50 to 500 μm. Because the length of the carnosol incubation (just 30 min) was probably insufficient to elicit a significant effect on gene expression, these results support the intrinsic antioxidant nature of this phytochemical, in agreement with other studies (
      • Zeng H.H.
      • Tu P.F.
      • Zhou K.
      • Wang H.
      • Wang B.H.
      • Lu J.F.
      ).
      Figure thumbnail gr9
      Fig. 9Carnosol prevents oxidative stress. PC12 cells were maintained in low serum medium for 16 h and then loaded for 1 h with 8 μm H2DCFDA and simultaneously treated with vehicle (control) or 10 μm carnosol for 30 min. Finally, cells were exposed to 250 μm H2O2 for 5 min and immediately analyzed by flow cytometry. A, a representative sample of 10,000 cells shown for vehicle-treated cells and cells treated with H2O2 alone or in combination with carnosol. B, dose-dependent antioxidant effect of carnosol. PC12 cells were loaded with H2DCFDA and treated with Me2SO vehicle or the indicated carnosol concentrations for 30 min. During the last 5 min, cells were exposed to 250 μm H2O2 as indicated and rapidly used for flow cytometry analysis. C, carnosol-mediated antioxidant protection against micromolar concentrations of H2O2. PC12 cells were loaded with H2DCFDA as described above, incubated for 30 min with vehicle or 10 μm carnosol, and treated for the last 5 min with increasing concentrations of H2O2 as shown.
      Next, we analyzed the long-term effect of carnosol after a 6-h incubation, a time consistent with the induction of HO-1 expression through the PI3K/Akt pathway described in this study. The antioxidant effect of carnosol was compared in the presence of the HO-1 inhibitor ZnPPIX and the PI3K inhibitor LY294002. PC12 cells maintained in low serum medium for 16 h were pretreated with vehicle, 10 μm LY294002, or 100 μm ZnPPIX as indicated in Fig. 10 (A and B) and then incubated with 10 μm carnosol for 30 min or 6 h. Finally, cells were challenged with 250 μm H2O2 for 5 min and immediately analyzed by flow cytometry. As shown in Fig. 10A, ZnPPIX did not significantly alter the antioxidant effect of carnosol after a 30-min incubation. By contrast, ZnPPIX partially (but not completely) blocked the long-term antioxidant effect of carnosol after a 6-h incubation. Similarly, as shown in Fig. 10B, inhibition of PI3K by 10 μm LY294002 did not significantly alter the antioxidant effect of carnosol after a 30-min incubation, but partially reduced its antioxidant effect after a 6-h incubation. These results indicate that induction of HO-1 by carnosol through the PI3K pathway is essential to elicit at least part of its long-term antioxidant effects (but see “Discussion”).
      Figure thumbnail gr10
      Fig. 10HO-1 and PI3K inhibitors attenuate long-term (but not short-term) antioxidant protection of carnosol.A, effect of the HO-1 inhibitor ZnPPIX on the antioxidant activity of carnosol; B, effect of the PI3K inhibitor LY294002 on the antioxidant activity of carnosol. PC12 cells were maintained in low serum medium for 16 h and then loaded for 1 h with H2DCFDA. Cells were pretreated with either 100 μm ZnPPIX (A) or 10 μm LY294002 (B) for 15 min and submitted to carnosol for 30 min and 6 h as indicated. During the last 5 min of carnosol incubation, cells were challenged with 250 μm H2O2 and then analyzed by flow cytometry.

      DISCUSSION

      In this study, we have shown the involvement of the PI3K/Akt survival pathway in the up-regulation of HO-1, a prototypical phase II enzyme, in response to the herb-derived diterpene carnosol. We found that carnosol up-regulated HO-1 expression by targeting the ARE sequences found in the mouse ho1 gene promoter. The regulation of AREs by carnosol was mediated, at least in part, through an increase in Nrf2 protein levels in a PI3K- and Akt-dependent manner. To our knowledge, this is the first study reporting the activation of the PI3K/Akt1 pathway and its relevance to the induction of phase II enzymes by food phytophenols with therapeutic potential.
      Most studies on the regulation of phase II gene expression have focused on the role of the MAPK pathways. Our results indicate that, although the ERK, p38, and JNK pathways are activated by carnosol, they do not participate equally in the induction of HO-1 in PC12 cells. Thus, ERK and JNK are dispensable for HO-1 up-regulation, whereas inhibition of p38 significantly reduces the response to carnosol. The role of MAPKs in ARE activation remains controversial. For instance, in agreement with our data, dominant-negative mutants of JNK do not prevent cadmium induction of the mouse ho1 promoter (
      • Alam J.
      • Wicks C.
      • Stewart D.
      • Gong P.
      • Touchard C.
      • Otterbein S.
      • Choi A.M.
      • Burow M.E.
      • Tou J.
      ). However, in rat hepatocytes, the glutathione depletor pherone and arsenite promote a JNK-dependent induction of ho1 gene expression (
      • Kietzmann T.
      • Samoylenko A.
      • Immenschuh S.
      ,
      • Oguro T.
      • Hayashi M.
      • Nakajo S.
      • Numazawa S.
      • Yoshida T.
      ). Balogun et al. (
      • Balogun E.
      • Hoque M.
      • Gong P.
      • Killeen E.
      • Green C.J.
      • Foresti R.
      • Alam J.
      • Motterlini R.
      ) described the effect of curcumin on induction of HO-1 expression by promoting inactivation of the Nrf2-Keap1 complex in a p38-dependent manner and subsequent binding of Nrf2 to the AREs. Accordingly, arsenate induces a p38-dependent activation of the ho1 promoter in the LMH chicken hepatoma cell line (
      • Elbirt K.K.
      • Whitmarsh A.J.
      • Davis R.J.
      • Bonkovsky H.L.
      ), but inhibition of p38 has no effect on cadmium-, arsenate-, and hemin-dependent HO-1 mRNA induction in HeLa cells (
      • Masuya Y.
      • Hioki K.
      • Tokunaga R.
      • Taketani S.
      ). Our results are consistent with the requirement of p38 for full activation of the ho1 promoter by carnosol because a generic inhibitor of p38 significantly reduced Nrf2 protein levels (data not shown) and HO-1 protein levels and promoter activity in response to carnosol. Finally, the ERK pathway is dispensable for HO-1 induction in our system and in rat hepatoma cells challenged with arsenite (
      • Kietzmann T.
      • Samoylenko A.
      • Immenschuh S.
      ), but it is necessary in the arsenite-induced response in LMH cells (
      • Elbirt K.K.
      • Whitmarsh A.J.
      • Davis R.J.
      • Bonkovsky H.L.
      ). One possible interpretation for these diverging observations may stem from the diverse assortment and intensity of the signaling pathways activated by different inducers in different cell types. In this regard, it is interesting to note that carnosol failed to induce HO-1 at the highest concentration tested (100 μm) despite the fact that this concentration yielded the strongest activation of the MAPK pathways (data not shown), therefore suggesting that specificity may rely not only on the nature of the inducer and cell type, but also on the relative potencies of these pathways.
      We found that carnosol increased the levels of the Nrf2 transcription factor in the nucleus and its binding to the AREs. The mechanisms leading to nuclear translocation of Nrf2 are poorly defined, but certainly include the release from Keap1 in the cytosol. However, because the half-life of this transcription factor is very short (
      • Nguyen T.
      • Sherratt P.J.
      • Huang H.C.
      • Yang C.S.
      • Pickett C.B.
      • Stewart D.
      • Killeen E.
      • Naquin R.
      • Alam S.
      • Alam J.
      • McMahon M.
      • Itoh K.
      • Yamamoto M.
      • Hayes J.D.
      ), these mechanisms must necessarily rely on the stabilization of Nrf2 protein. Our results demonstrate that carnosol indeed increased the levels of Nrf2, as did the proteasome inhibitor MG132, suggesting that this phytochemical blocks Nrf2 degradation. The increase in Nrf2 protein levels induced by carnosol was partially dependent on the activation of PI3K because LY294002 concentrations that inhibited PI3K and the subsequent phosphorylation and activation of Akt1 decreased the carnosol-induced accumulation of Nrf2 protein. However, inhibition of PI3K did not fully prevent the carnosol-induced increase in Nrf2 and HO-1 protein levels, suggesting that other elements, probably p38 and other PI3K-independent pathways, are also involved in carnosol signaling to Nrf2 and HO-1. Interestingly, active PI3K was sufficient to activate the ho1 promoter through AREs, and constitutive expression of active PI3K or Akt yielded cells with increased HO-1 levels.
      In searching for the downstream kinases that might be responsible for the increased Nrf2 protein levels, we analyzed the effect of PKCζ. A dominant-negative version of this kinase produced only a slight decrease in the carnosol-induced activation of the ho1 promoter, suggesting a very minor role for this atypical PKC in HO-1 transcriptional regulation. However, a recent study by Numazawa et al. (
      • Numazawa S.
      • Ishikawa M.
      • Yoshida A.
      • Tanaka S.
      • Yoshida T.
      ) indicated that PKCι, another member of the atypical PKC family, phosphorylates Nrf2 at Ser40, leading to its release from Keap1 and nuclear translocation. Whether the dispensability of PKCζ is due to a lack of involvement in the regulation of Nrf2 or to redundancy with other atypical PKCs such as PKCι remains to be determined.
      On the other hand, a dominant-negative mutant of Akt1 significantly reversed the carnosol-induced activation of the ho1 gene promoter. Contrary to PKCι, direct phosphorylation of Nrf2 or Keap1 by Akt seems unlikely because these proteins do not have a consensus phosphorylation sequence for Akt. Because Nrf2 is regulated primarily at the level of stability, we propose that Akt might be acting on the ubiquitin ligases involved in its tagging for degradation. In fact, at least two ubiquitin ligases, Mdm2 (
      • Mayo L.D.
      • Donner D.B.
      ) and BRCA1/BARD1 (
      • Altiok S.
      • Batt D.
      • Altiok N.
      • Papautsky A.
      • Downward J.
      • Roberts T.M.
      • Avraham H.
      ), have been reported to be substrates of Akt.
      Interestingly, the activation of Akt by carnosol was a transient event that subsided by 2 h. However, Nrf2 protein levels increased beyond 3 h. Therefore, transient activation of Akt must result in persistent activation/inhibition of effector substrates that regulate Nrf2, resulting in increased protein levels. Thus, phosphorylation of the Nrf2-specific ubiquitin ligase might result in its persistent inhibition, leading to a sustained increase in Nrf2 protein levels. Alternatively, the transient activation of Akt might activate transcription of stable proteins that contribute to increased Nrf2 protein levels.
      Plant-derived phenols exhibit strong antioxidant properties due to Michael reaction acceptor function. Indeed, we have observed the intrinsic antioxidant nature of carnosol because a 30-min pretreatment with this drug completely abolished the oxidant effect of as much as 500 μm H2O2, as determined by the blockage in the conversion of H2DCFDA to its oxidized green fluorescent form, 2′,7′-dichlorofluorescein. More importantly for this study, inhibition of the PI3K pathway leading to activation of Akt and up-regulation of HO-1 expression impaired the long-term (but not short-term) antioxidant function of carnosol, therefore evidencing the relevance of this pathway for carnosol in conferring persistent antioxidant protection. Interestingly, a fraction of about half of the antioxidant effect of carnosol was insensitive to inhibition of PI3K and HO-1 at 6 h. This observation is consistent with the possibility that, at this time, there is still enough non-metabolized carnosol to maintain its intrinsic antioxidant effect. In fact, the PI3K and HO-1 inhibitors decreased H2DCFDA oxidation to levels similar to those observed at 30 min. However, we favor the non-exclusive hypothesis that carnosol also up-regulates other intracellular antioxidant systems that do not belong to the PI3K/HO-1 module and that also contribute to the long-term antioxidant properties of this compound. In fact, preliminary microarray data from mouse neuroblastoma cells indicate that carnosol also up-regulates the expression of mitochondrial manganese-dependent superoxide dismutase and glutathione peroxidase (data not shown). A corollary of this study is that pharmacological activation of the PI3K/Akt pathway by this food-related compound (leading to increases in Nrf2 and HO-1) efficiently protects cells from oxidative stress and should be evaluated as a new therapeutic approach in degenerative processes such as Alzheimer’s and Parkinson diseases that correlate with oxidative damage.

      Acknowledgments

      We thank Beatriz Palacios for excellent technical assistance.

      References

        • Klein J.A.
        • Ackerman S.L.
        J. Clin. Investig. 2003; 111: 785-793
        • Cooke M.S.
        • Evans M.D.
        • Dizdaroglu M.
        • Lunec J.
        FASEB J. 2003; 17: 1195-1214
        • Lim G.P.
        • Chu T.
        • Yang F.
        • Beech W.
        • Frautschy S.A.
        • Cole G.M.
        J Neurosci. 2001; 121: 8370-8377
        • Lo A.H.
        • Liang Y.C.
        • Lin-Shiau S.Y.
        • Ho C.T.
        • Lin J.K.
        Carcinogenesis. 2002; 23: 983-991
        • Dinkova-Kostova A.T.
        • Massiah M.A.
        • Bozak R.E.
        • Hicks R.J.
        • Talalay P.
        Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 3404-3409
        • Dinkova-Kostova A.T.
        Mini. Rev. Med. Chem. 2002; 2: 595-610
        • Iqbal M.
        • Sharma S.D.
        • Okazaki Y.
        • Fujisawa M.
        • Okada S.
        Pharmacol. Toxicol. 2003; 92: 33-38
        • Maines M.D.
        Cell. Mol. Biol. (Noisy-le-Grand). 2000; 46: 573-585
        • Immenschuh S.
        • Hinke V.
        • Ohlmann A.
        • Gifhorn-Katz S.
        • Katz N.
        • Jungermann K.
        • Kietzmann T.
        Biochem. J. 1998; 334: 141-146
        • Alam J.
        • Killeen E.
        • Gong P.
        • Naquin R.
        • Hu B.
        • Stewart D.
        • Ingelfinger J.R.
        • Nath K.A.
        Am. J. Physiol. 2003; 284: F743-F752
        • Bauer M.
        • Bauer I.
        Antioxid. Redox Signal. 2002; 4: 749-758
        • Ryter S.W.
        • Choi A.M.
        Antioxid. Redox Signal. 2002; 4: 625-632
        • Salinas M.
        • Diaz R.
        • Abraham N.G.
        • Ruiz de Galarreta C.M.
        • Cuadrado A.
        J. Biol. Chem. 2003; 278: 13898-13904
        • Rizzardini M.
        • Terao M.
        • Falciani F.
        • Cantoni L.
        Biochem. J. 1993; 290: 343-347
        • Terry C.M.
        • Clikeman J.A.
        • Hoidal J.R.
        • Callahan K.S.
        Am. J. Physiol. 1999; 276: H1493-H1501
        • Scapagnini G.
        • Foresti R.
        • Calabrese V.
        • Giuffrida Stella A.M.
        • Green C.J.
        • Motterlini R.
        Mol. Pharmacol. 2002; 61: 554-561
        • Chen K.
        • Gunter K.
        • Maines M.D.
        J. Neurochem. 2000; 75: 304-313
        • Mazza F.
        • Goodman A.
        • Lombardo G.
        • Vanella A.
        • Abraham N.G.
        Exp. Biol. Med. 2003; 228: 576-583
        • Maines M.D.
        Toxicol. Sci. 2003; 71: 9-10
        • Chen J.
        • Tu Y.
        • Moon C.
        • Nagata E.
        • Ronnett G.V.
        J. Neurochem. 2003; 85: 1247-1261
        • Poss K.D.
        • Tonegawa S.
        Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 10925-10930
        • Saikawa Y.
        • Kaneda H.
        • Yue L.
        • Shimura S.
        • Toma T.
        • Kasahara Y.
        • Yachie A.
        • Koizumi S.
        Hum. Mutat. 2000; 16: 178-179
        • Johnson G.L.
        • Lapadat R.
        Science. 2002; 298: 1911-1912
        • Kietzmann T.
        • Samoylenko A.
        • Immenschuh S.
        J. Biol. Chem. 2003; 278: 17927-17936
        • Shan Y.
        • Pepe J.
        • Lu T.H.
        • Elbirt K.K.
        • Lambrecht R.W.
        • Bonkovsky H.L.
        Arch. Biochem. Biophys. 2000; 380: 219-227
        • Ohlmann A.
        • Giffhorn-Katz S.
        • Becker I.
        • Katz N.
        • Immenschuh S.
        Exp. Biol. Med. 2003; 228: 584-589
        • Balogun E.
        • Hoque M.
        • Gong P.
        • Killeen E.
        • Green C.J.
        • Foresti R.
        • Alam J.
        • Motterlini R.
        Biochem. J. 2003; 371: 887-895
        • Lee J.M.
        • Hanson J.M.
        • Chu W.A.
        • Johnson J.A.
        J. Biol. Chem. 2001; 276: 20011-20016
        • Elbirt K.K.
        • Bonkovsky H.L.
        Proc. Assoc. Am. Physicians. 1999; 111: 438-447
        • Hill-Kapturczak N.
        • Thamilselvan V.
        • Liu F.
        • Nick H.S.
        • Agarwal A.
        Am. J. Physiol. 2001; 281: F851-F859
        • Gong P.
        • Stewart D.
        • Hu B.
        • Vinson C.
        • Alam J.
        Arch. Biochem. Biophys. 2002; 405: 265-274
        • Maeshima H.
        • Sato M.
        • Ishikawa K.
        • Katagata Y.
        • Yoshida T.
        Nucleic Acids Res. 1996; 24: 2959-2965
        • Kiemer A.K.
        • Bildner N.
        • Weber N.C.
        • Vollmar A.M.
        Endocrinology. 2003; 144: 802-812
        • Nguyen T.
        • Sherratt P.J.
        • Pickett C.B.
        Annu. Rev. Pharmacol. Toxicol. 2002; 43: 233-260
        • Moi P.
        • Chan K.
        • Asunis I.
        • Cao A.
        • Kan Y.W.
        Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 9926-9930
        • Itoh K.
        • Wakabayashi N.
        • Katoh Y.
        • Ishii T.
        • Igarashi K.
        • Engel J.D.
        • Yamamoto M.
        Genes Dev. 1999; 13: 76-86
        • Hayes J.D.
        • McMahon M.
        Cancer Lett. 2001; 174: 103-113
        • Huang H.C.
        • Nguyen T.
        • Pickett C.B.
        J. Biol. Chem. 2002; 277: 42769-42774
        • Nguyen T.
        • Sherratt P.J.
        • Huang H.C.
        • Yang C.S.
        • Pickett C.B.
        J. Biol. Chem. 2003; 278: 4536-4541
        • Stewart D.
        • Killeen E.
        • Naquin R.
        • Alam S.
        • Alam J.
        J. Biol. Chem. 2002; 278: 2396-2402
        • McMahon M.
        • Itoh K.
        • Yamamoto M.
        • Hayes J.D.
        J. Biol. Chem. 2003; 278: 21592-21600
        • Martin D.
        • Salinas M.
        • Lopez-Valdaliso R.
        • Serrano E.
        • Recuero M.
        • Cuadrado A.
        J. Neurochem. 2001; 78: 1000-1008
        • Alam J.
        • Wicks C.
        • Stewart D.
        • Gong P.
        • Touchard C.
        • Otterbein S.
        • Choi A.M.
        • Burow M.E.
        • Tou J.
        J. Biol. Chem. 2000; 275: 27694-27702
        • Liu H.
        • Nowak R.
        • Chao W.
        • Bloch K.D.
        J. Neurochem. 2003; 86: 1553-1563
        • Chen K.
        • Maines M.D.
        Cell. Mol. Biol. (Noisy-le-Grand). 2000; 46: 609-617
        • Hill-Kapturczak N.
        • Truong L.
        • Thamilselvan V.
        • Visner G.A.
        • Nick H.S.
        • Agarwal A.
        J. Biol. Chem. 2000; 275: 40904-40909
        • Ischiropoulos H.
        • Gow A.
        • Thom S.R.
        • Kooy N.W.
        • Royall J.A.
        • Crow J.P.
        Methods Enzymol. 1999; 301: 367-373
        • Zeng H.H.
        • Tu P.F.
        • Zhou K.
        • Wang H.
        • Wang B.H.
        • Lu J.F.
        Acta Pharmacol. Sin. 2001; 22: 1094-1098
        • Oguro T.
        • Hayashi M.
        • Nakajo S.
        • Numazawa S.
        • Yoshida T.
        J. Pharmacol. Exp. Ther. 1998; 287: 773-778
        • Elbirt K.K.
        • Whitmarsh A.J.
        • Davis R.J.
        • Bonkovsky H.L.
        J. Biol. Chem. 1998; 273: 8922-8931
        • Masuya Y.
        • Hioki K.
        • Tokunaga R.
        • Taketani S.
        J. Biochem. (Tokyo). 1998; 124: 628-633
        • Numazawa S.
        • Ishikawa M.
        • Yoshida A.
        • Tanaka S.
        • Yoshida T.
        Am. J. Physiol. 2003; 285: C334-C342
        • Mayo L.D.
        • Donner D.B.
        Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 11598-11603
        • Altiok S.
        • Batt D.
        • Altiok N.
        • Papautsky A.
        • Downward J.
        • Roberts T.M.
        • Avraham H.
        J. Biol. Chem. 1999; 274: 32274-32278