J Nutr Sci Vitaminol, 60, 122–128, 2014

Resveratrol Reduces the Hypoxia-Induced Resistance to in Breast Cancer Cells

Takakazu Mitani1, Yuta Ito1, Naoki Harada1, Yoshihisa Nakano2, Hiroshi Inui1, Hitoshi Ashida3 and Ryoichi Yamaji1,*

1 Division of Applied Life Sciences, Graduate School of Life and Environmental Sciences, Osaka Prefecture University, Sakai, Osaka 599–8531, Japan 2 Osaka Women’s Junior College, Fujiidera, Osaka 583–8558, Japan 3 Department of Agrobioscience, Graduate School of Agricultural Science, Kobe University, Kobe, Hyogo 657–8501, Japan (Received August 21, 2013)

Summary (3,4′,5-trihydroxy-trans-stilbene) is known to enhance the cyto- toxicity of the anticancer drug doxorubicin. On the other hand, breast cancer MCF-7 cells acquire resistance to doxorubicin under hypoxic conditions. In this study, we investigated the effect of resveratrol on hypoxia-induced resistance to doxorubicin in MCF-7 cells. Resver- atrol and its derivative 3,5-dihydroxy-4′-methoxy-trans-stilbene, but not 3,5-dimethoxy-4′- hydroxy-trans-stilbene, cancelled hypoxia-induced resistance to doxorubicin at a concentra- tion of 10 mm. Carbonyl reductase 1 (CBR1) catalyzes the conversion of doxorubicin to its metabolite doxorubicinol, which is much less effective than doxorubicin. Hypoxia increased the expression of CBR1 at both mRNA and protein levels, and knockdown of CBR1 inhib- ited hypoxia-induced resistance to doxorubicin in MCF-7 cells. Knockdown of hypoxia- inducible factor (HIF)-1a repressed the hypoxia-induced expression of CBR1. Resveratrol repressed the expression of HIF-1a protein, but not HIF-1a mRNA, and decreased hypoxia- activated HIF-1 activity. Resveratrol repressed the hypoxia-induced expression of CBR1 at both mRNA and protein levels. Likewise, 3,5-dihydroxy-4′-methoxy-trans-stilbene decreased the hypoxia-induced expression of CBR1 protein, but not 3,5-dimethoxy-4′-hydroxy-trans- stilbene. Furthermore, resveratrol decreased the expression of HIF-1a protein even in the presence of the proteasome inhibitor MG132 in hypoxia. Theses results indicate that in MCF-7 cells, HIF-1a-increased CBR1 expression plays an important role in hypoxia-induced resistance to doxorubicin and that resveratrol and 3,5-dihydroxy-4′-methoxy-trans-stilbene decrease CBR1 expression by decreasing HIF-1a protein expression, perhaps through a proteasome-independent pathway, and consequently repress hypoxia-induced resistance to doxorubicin. Key Words resveratrol, doxorubicin, carbonyl reductase 1, hypoxia-inducible factor-1a, breast cancer cells

Breast cancer is the most frequently diagnosed car- phytopolyphenol and is present in a variety of edible cinoma and the second leading cause of cancer death plant products including peanuts, berries, and grape in women (1). Solid tumors grow in a hypoxic environ- skin (7). Resveratrol exerts preventive effects against ment resulting from insufficient oxygen supply due to pathological processes such as inflammation (8), ath- dysfunctional tumor vessels. Hypoxia not only promotes erosclerosis (9), cardiovascular disease (10), type 2 the growth of a tumor by enhancing glycolytic flux diabetes (11), and tumorigenesis including breast can- and increasing angiogenesis, but also allows tumors to cer (12). Furthermore, resveratrol enhances the cyto- acquire resistance to anticancer drugs (2, 3). Recently, toxic effect of the anticancer drug doxorubicin on the the uses of functional foods with antioxidants and anti- growth of breast cancer MCF-7 cells (13, 14). However, cancer activities have emerged as a promising strategy its molecular mechanisms remain unclear. Recently, we for the treatment of various diseases including cancer found that resveratrol inhibits enzymatic activity of car- (4–6). However, little is known about the possible inter- bonyl reductase 1 (CBR1), a doxorubicin-metabolizing actions between functional foods and chemotherapeutic , by directly binding to CBR1 (14). agents. CBR1 is an NADPH-dependent enzyme belonging Resveratrol (3,4′,5-trihydroxy-trans-stilbene) is a to the short chain dehydrogenase family and catalyzes a large number of biologically and pharmacologically E-mail: [email protected] active substrates, including endogenous substrates Abbreviations: CBR1, carbonyl reductase 1; DAPI, 4′,6- such as E2 and S-nitrosogluthathione diamidino-2-phenylindole; HIF-1, hypoxia-inducible factor-1. and xenobiotic substrates such as anthracycline anti-

122 Resveratrol and Doxorubicin 123 cancer drugs (doxorubicin and daunorubicin) (15–17). rubicin (50 nm) in normoxia or hypoxia for 72 h. The Since doxorubicin causes DNA damage by intercalating cells were lysed with phosphate-buffered saline, and cell itself into the DNA base pairs of the double helix, it is lysates were incubated with Hoechst 33342 (0.4 ng/mL) used for the treatment of broad range of solid tumors, at room temperature for 10 min. Fluorescence intensity including breast cancer and hepatocellular carcinoma of Hoechst 33342 was measured by fluorometer with cells (18, 19). However, CBR1 metabolizes the reduction 355/460 nm filter excitation and emission set. of doxorubicin to doxorubicinol, which is less potent Western blot. MCF-7 cells were lysed in lysis buffer at killing tumor cells and which induces cardiotoxicity (20 mm Hepes-NaOH, pH 7.4, containing 150 mm NaCl, (20). The CBR1 activities in lung and breast cancer tis- 10% glycerol, 1 mm EDTA, 1 mm phenylmethylsulfonyl sues are higher than those in the corresponding normal fluoride, 10 mg/mL leupeptin, and 1 mg/mL aprotinin). tissues (21). Furthermore, hypoxia enhances resistance The cell lysates were subjected to SDS-PAGE and ana- to doxorubicin in tumor cells and increases the expres- lyzed by Western blotting using rabbit polyclonal anti- sion of CBR1 through hypoxia-inducible factor (HIF)- CBR1 (Aviva Systems Biology, San Diego, CA) antibody, 1a in hepatocellular carcinoma cells (22). These results rat monoclonal anti-HSP90 (Clone: 16F1, StressGen indicate that CBR1 plays a role in the acquisition of Biotechnologies, Victoria, Canada) antibody, and mouse resistance to doxorubicin and that the clinical applica- monoclonal anti-HIF-1a (Clone: mgc3, Affinity BioRe- tion of doxorubicin is limited because it has a risk of agents, Golden, CO) and anti-a-tubulin (Clone: DM 1A, cardiotoxicity. Sigma-Aldrich) antibodies. Immunoreactive proteins In this study, we assessed the combined effects of res- were incubated with horseradish peroxidase-conjugated veratrol and doxorubicin on the growth of MCF-7 cells secondary antibodies and reacted with Immobilon in hypoxia. We demonstrate that resveratrol suppresses Western Chemiluminescent HRP-Substrate (Millipore, hypoxia-induced resistance to cytotoxicity of doxorubi- Billerica, MA), followed by detection with an LAS4000 cin and represses the expression of CBR1 in MCF-7 cells. imaging system (GE Healthcare, Uppsala, Sweden). Furthermore, we report that resveratrol inhibits CBR1- Semi-quantitative RT-PCR. Total RNAs were extracted mediated resistance to doxorubicin by decreasing the from MCF-7 cells with Sepasol-RNA Super (Nacalai expression of HIF-1a in hypoxia. Tesque, Inc., Kyoto, Japan), and cDNAs were synthe- sized by reverse transcriptase. The nucleotide sequences MATERIALS AND METHODS of primers for HIF-1a and b-actin were described pre- Reagents. Resveratrol and 3,5-dimethoxy-4′-hydroxy- viously (24). The primers for CBR1 were designed trans-stilbene were purchased from Tokyo Chemical (forward primer; 5′-TCGTCCGGCATCCATGTA-3′ and Industry Co., Ltd. (Tokyo, Japan), and 3,5-dihydroxy- reverse primer; 5′-CCACTGTTCAACTCTCTTCT-3′). 4′-methoxy-trans-stilbene was synthesized according to Luciferase reporter assay. MCF-7 cells were grown the method of Paul et al. (23). Doxorubicin hydrochlo- on 48-well plates in DMEM-HG and transiently trans- ride and doxorubicinol hydrochloride were purchased fected with reporter vector (pEpo-HRE-Luc (24) and from Enzo Life Sciences (Farmingdale, NY) and Toronto pRL-SV40 (Promega, Madison, WI)) using HilyMax Research Chemicals (Toronto, Canada), respectively. reagent (Dojindo Laboratories) for 24 h. The medium Hoechst 33342 was obtained from Dojindo Laboratories was replaced with fresh DMEM-HG, and the cells were (Kumamoto, Japan). incubated in the presence of various concentrations Cell culture. Human breast cancer MCF-7 cells were (0–25 mm) of resveratrol in normoxia or hypoxia for obtained from the RIKEN BioResource center (Ibaraki, 24 h. Luciferase reporter activities were determined Japan). The cells were cultured in Dulbecco’s modified using a dual-luciferase reporter assay system (Promega). Eagle’s medium-high glucose (4.5 g/L glucose) supple- Data are expressed as relative light units (RLU, firefly mented with 10% fetal bovine serum, 100 U/mL penicil- luciferase activity divided by Renilla luciferase activity). lin, and 100 mg/mL streptomycin (DMEM-HG) at 37˚C Statistical analysis. Data were assessed by two-way in a 5% CO2/95% air atmosphere at 100% humidity. analysis of variance with Tukey’s post hoc testing. Sta- siRNA. Human HIF-1a siRNA (siHIF-1a) and con- tistical analysis was performed using JMP statistical trol siRNA (siControl) were described previously (24). software version 8.0.1 (SAS Institute, Cary, NC). Experi- Double-strand siRNA for human CBR1 was chemically mental values are expressed as mean6SD, and statisti- synthesized. Target sequences for siRNA duplexes were cally significant differences (p,0.05) are indicated by as follows: siCBR1, 5′-CACAGAATTACTCCCTCTA-3′ different letters. (Sigma-Aldrich, Saint Louis, MO). The duplexes (10 nm) RESULTS were introduced into MCF-7 cells using Lipofectamine RNAiMAX reagent (Invitrogen, Carlsbad, CA) for 48 h, To examine the effect of doxorubicin on cell viability and the cells were incubated in fresh DMEM-HG medium under hypoxic conditions in breast cancer cells, MCF-7 in normoxia or hypoxia for an additional 24 h. cells were incubated with various concentrations of Cell viability assay. Cell survival was quantified using doxorubicin. Doxorubicin decreased cell viability of Hoechst 33342. MCF-7 cells (1.23104 cells/well) were MCF-7 cells in a dose-dependent manner, and hypoxia cultured in DMEM-HG on 24-well plates for 24 h, fol- attenuated the cytotoxicity of doxorubicin at concen- lowed by incubation with various concentrations of res- trations above 10 nm (Fig. 1A). We assessed the inhibi- veratrol (0–10 mm) in the presence or absence of doxo- tory effect of resveratrol on hypoxia-induced resistance 124 Mitani T et al.

(A) (B) 125 a a a a a a a

) ab ab ) 100 l l b 100 bc cd 75 c d on tr o 75 cd on tr o ab ilit y ab ilit y c c i de i v v f f 50 o o ll ll 50 ef

e e e

% % e C C ( ( 25 e e 25 f 0 0 Res ( M) 0 1 10 0 1 10 DOX (nM) 0 5 10 25 50 DOX +

(C) OH OMe OH HO HO MeO

OH (Res) OH (4´-OMe) OMe (3,5-OMe)

(D) (E) a a

a a ) a a a l 100 ) ab l 100 bc c 75 b on tr o b

75 ab ilit y c i on tr o ab ilit y v f c i 50 o v f ll 50 o e c

ll c c c % c c e C (

% 25 C ( 25 0 0 e e eh es M M DOX + + + + V R O O e ´- 5- h h es e M 4 3, Ve Ve R M O ´-O 5- 4 3, Fig. 1. Inhibitory effect of resveratrol on hypoxia-induced resistance to doxorubicin in MCF-7 cells. (A) MCF-7 cells were cultured with various concentrations of doxorubicin (DOX) (0–50 nm) in normoxia (white bars) or hypoxia (black bars) for 72 h, and cell viabilities were determined. (B) MCF-7 cells were incubated with various concentrations of resveratrol (0–10 mm) in the presence or absence of doxorubicin (50 nm) in normoxia (white bars) or hypoxia (black bars) for 72 h, and cell viabilities were determined. (C) Chemical structures. (D) MCF-7 cells were incubated in the presence of vehi- cle (Veh), 3,4′,5-trihydroxy-trans-stilbene (Res), 3,5-dihydroxy-4′-methoxy-trans-stilbene (4′-OMe), and 3,5-dimethoxy- 4′-hydroxy-trans-stilbene (3,5-OMe) (each 10 mm) in normoxia (white bars) or hypoxia (black bars) for 72 h, and cell viabilities were determined. (E) MCF-7 cells were incubated with resveratrol or its derivatives (10 mm) in the presence of doxorubicin (DOX) (50 nm) in normoxia (white bars) or hypoxia (black bars) for 72 h, and cell viabilities were determined. Data are expressed as mean6SD, and differences were considered statistically signifi cant at ap value of ,0.05. In all experiments, the result is representative of three independent experiments.

to doxorubicin. Resveratrol at a concentration of 1 mm of CBR1, but not the expressions of b-actin or a-tubu- had no infl uence on cell viability in the presence or lin. Knockdown of CBR1 restored hypoxia-induced absence of doxorubicin in normoxia, but attenuated resistance to cytotoxicity of doxorubicin (Fig. 2B). On hypoxia-induced resistance to doxorubicin (Fig. 1B). On the other hand, siCBR1 had no infl uence on cell viability the other hand, resveratrol at a concentration of 10 mm in the absence of doxorubicin. completely canceled hypoxia-induced resistance to cyto- To investigate whether HIF-1a is involved in hypoxia- toxicity of doxorubicin. Next, we determined the effects increased expression of CBR1, HIF-1a was knocked of resveratrol and its derivatives on the hypoxia-induced down using siRNA for HIF-1a in MCF-7 cells, and then resistance to cytotoxicity of doxorubicin (Fig. 1C). Res- the expression level of CBR1 was assessed. Knock- veratrol and 3,5-dihydroxy-4′-methoxy-trans-stilbene down of HIF-1a canceled the hypoxic induction of the (4′-OMe) at a concentration of 10 mm suppressed the expressions of HIF-1a and CBR1 (Fig. 3A). Next, we viability of MCF-7 cells in hypoxia (Fig. 1D) and inhib- determined the effects of resveratrol on the expression ited the hypoxia-induced resistance to doxorubicin of CBR1 and HIF-1a in hypoxia. Resveratrol at concen- (Fig. 1E). In contrast, 3,5-dimethoxy-4′-hydroxy-trans- trations above 10 mm decreased the expression level of stilbene (3,5-OMe) had no infl uence on the cell viability HIF-1a protein, but not of HIF-1a mRNA (Fig. 3B). In or the resistance to doxorubicin in hypoxia. contrast, resveratrol at concentrations above 10 mm To determine whether CBR1 is involved in hypoxia- reduced the expression of CBR1 at both mRNA and induced resistance to doxorubicin, MCF-7 cells were protein levels. Resveratrol reduced hypoxia-activated transiently transfected with siRNA for CBR1, followed HIF-1 activity in a dose-dependent manner (Fig. 3C). by incubation with doxorubicin in normoxia or hypoxia. Furthermore, hypoxia-induced expression levels of HIF- Hypoxia increased the expression of CBR1 at both the 1a and CBR1 protein were reduced by resveratrol and mRNA and protein levels (Fig. 2A). In both normoxia 4′-OMe, but not by 3,5-OMe (Fig. 3D). and hypoxia, siRNA for CBR1 suppressed the expression We assessed whether proteasome is involved in res- Resveratrol and Doxorubicin 125

(A) (A) (B) Normoxia Hypoxia N H Res ( M) 0 1 10 25 0 1 10 25 N H siControl + + HIF-1α siControl + + W . siCBR1 + + siHIF-1α + + CBR1 B W CBR1 α . HIF-1 α-Tubulin B .

α-Tubulin CBR1 HIF-1α R T R - P T α-Tubulin CBR1 C CBR1 - P R

C β-Actin β-Actin R (C) (D) ia ia (B) rmox HRE-Luc No Hypox a a a a 8 a a Res +

) 100 l 6 4´-OMe +

b U 3,5-OMe +

75 L R on tr o 4 b ab ilit y c i HIF-1α v f c 50 2 c c c c o ll c c CBR1 e c % 0 C ( 25 Res ( M) 0 1 10 25 α-Tubulin 0 DOX + + Fig. 3. Effect of resveratrol on hypoxia-increased CBR1 siControl siCBR1 expression. (A) MCF-7 cells were transiently transfected with siControl or siHIF-1a, and incubated in normoxia Fig. 2. Involvement of CBR1 in hypoxia-induced resis- (N) or hypoxia (H) for 24 h. The cell lysates were ana- tance to doxorubicin. (A) MCF-7 cells were transiently lyzed by Western blotting. (B) MCF-7 cells were incu- transfected with siControl or siCBR1 for 24 h, followed bated with various concentrations of resveratrol by incubation in normoxia (N) or hypoxia (H) for 24 h. (0–25 mm) in normoxia or hypoxia for 24 h. The cell The cell lysates were analyzed by Western blotting lysates were analyzed by Western blotting (W.B.), and (W.B.), and the mRNA levels were semi-quantitatively the mRNA levels were semi-quantitatively measured measured by RT-PCR. (B) After siRNA transfection, the by RT-PCR. (C) MCF-7 cells were transiently trans- cells were incubated in the presence or absence of doxo- fected with pEpo-HRE-Luc and pRL-SV40 for 24 h. rubicin (50 nm) in normoxia (white bars) or hypoxia The cells were incubated with various concentrations (black bars) for 72 h. Cell viabilities were determined. of resveratrol (0–25 mm) in normoxia or hypoxia for Data are expressed as mean6SD, and differences 24 h. Data are expressed as mean6SD, and differences were considered statistically signifi cant at a p value of were considered statistically signifi cant at a p value of ,0.05. In all experiments, the result is representative ,0.05. (D) MCF-7 cells were incubated with resver- of three independent experiments. atrol or its derivatives (10 mm) in the presence of 50 nm doxorubicin in normoxia or hypoxia for 24 h. The cell lysates were analyzed by Western blotting. In all experi- ments, the result is representative of three independent a veratrol-decreased expression of HIF-1 in hypoxia. A experiments. proteasome inhibitor, MG132, increased the expression level of HIF-1a protein (fi rst and third lanes in Fig. 4) and the expression level of resveratrol-decreased HIF- Hypoxia 1a protein (second and forth lanes in Fig. 4). However, MG132 + + resveratrol decreased the expression level of HIF-1a Res + + protein even in the presence of MG132 (third and forth HIF-1α lanes in Fig. 4). In contrast, although the expression of HSP90 is induced by MG132 (25), resveratrol had no HSP90 infl uence on MG132-induced HSP90 expression level. α-Tubulin

DISCUSSION Fig. 4. Involvement of proteasome in resveratrol- decreased HIF-1a protein. MCF-7 cells were incubated Recently, combination treatment with a chemothera- with or without 10 mm resveratrol in the presence or peutic agent and functional foods has been increasing absence of MG132 (10 mm) in hypoxia for 24 h. The among cancer patients (5). However, combined usage of cell lysates were analyzed by Western blotting. In all functional foods with a chemotherapeutic agent might experiments, the result is representative of three inde- cause side effects or might attenuate the effect of the pendent experiments. therapeutic agent. CBR1 could be a promising target molecule for resistance to cytotoxicity of doxorubicin, because CBR1 converts doxorubicin to doxorubicinol, potential to be used in combination with doxorubicin in which is less effective against cancer and which induces the treatment of cancer. In this study, we demonstrated cardiotoxicity. We previously demonstrated that resver- that resveratrol suppresses hypoxia-induced resistance atrol directly interacts with CBR1 and inhibits its enzy- to cytotoxicity of doxorubicin in MCF-7 cells. matic activity (14). Therefore, resveratrol may have the Resveratrol inhibited hypoxia-induced resistance to 126 Mitani T et al. doxorubicin in a concentration-dependent manner. 4′-hydroxy-trans-stilbene. Thus, at least the hydroxyl Tumors increase or acquire resistance to doxorubicin group of the 3 or 5 position of resveratrol plays an in several ways, including (i) increased expression of important role in the mechanism by which resveratrol drug transporters (26), (ii) enhancement of doxorubi- decreases HIF-1a protein level. The 3,5-dihydroxyl cin metabolism (27), and (iii) alteration of intracellular groups of resveratrol are required for direct interaction localization of doxorubicin (28). Doxorubicin is deacti- with CBR1 (14). Furthermore, the transcriptional activ- vated by several carbonyl reducing , including ity of androgen receptor is inhibited by resveratrol and carbonyl reductases (e.g., CBR1 and CBR3) (29, 30) 3,5-dihydroxy-4′-methoxy-trans-stilbene, but not by and aldo-keto reductases (e.g., AKR1A1, AKR1B10, 3,5-dimethoxy-4′-hydroxy-trans-stilbene (40). In con- AKR1C3, and AKR1C4) (31–33). CBR1 activities are trast, the 4′-hydroxyl group of resveratrol is required increased in human lung and breast cancer tissues (21). for enhancing the transcriptional activity of estrogen AKR1C3 is predominantly expressed in human pros- receptor a (41). Thus, the number and the position of tate and mammary gland, compared to AKR1C4, and hydroxyl groups of resveratrol seem to play an important AKR1C3 and AKR1C4, which are expressed at similar role in interaction with resveratrol-binding proteins. levels in the (33). Thus, the activities and expres- Doxorubicin-based chemotherapy has improved the sions of these carbonyl reducing enzymes are different survival of many patients with breast cancer (18). between normal cells and cancer cells and among tis- However, cardiotoxicity is the unexpected side effect of sues. In the present study, knockdown of CBR1 can- doxorubicinol and limits the clinical application of doxo- celed hypoxia-induced resistance to doxorubicin in rubicin (42). Because mice with a null allele of CBR1 MCF-7 cells, and resveratrol repressed the expression of are protected from doxorubicin-induced cardiotoxicity CBR1. These results indicate that CBR1 is a key factor (43), CBR1 is a major enzyme to produce doxorubicinol. in hypoxia-induced resistance to cytotoxicity of doxo- Therefore, resveratrol is expected to down-regulate the rubicin in MCF-7 cells and that resveratrol suppress CBR1 expression, and the combined use of doxorubicin hypoxia-induced resistance to doxorubicin by decreas- and resveratrol is expected to decrease the effective con- ing the expression of CBR1. centration of doxorubicin (i.e., less doxorubicin will be Resveratrol decreased the expression of HIF-1a and needed if resveratrol is given also) in the treatment of CBR1 at concentrations above 10 mm. HIF-1a increases breast cancer. Furthermore, resveratrol suppresses doxo- the expression of CBR1 at the transcriptional level in rubicin-induced cardiotoxicity by increasing heme oxy- hypoxic hepatocellular carcinoma cells (22), consis- genase-1 expression (44). Therefore, resveratrol appears tent with the present results that knockdown of HIF-1a to be a promising candidate not only for enhancing the down-regulated CBR1 expression in hypoxic MCF-7 cells. anticancer activity of doxorubicin, but also for reducing Some flavonoids such as quercetin and luteolin, which the doxorubicin-induced cardiotoxicity. are structurally similar to resveratrol, suppress hypoxia- REFERENCES mediated resistance to doxorubicin. Luteolin restores hypoxia-induced tumor resistance to doxorubicin in 1) Glass AG, Lacey JV Jr, Carreon JD, Hoover RN. 2007. breast cancer 4T1 and MCF-7 cells (34) and decreases Breast cancer incidence, 1980–2006: combined roles of HIF-1a expression in hypoxic human brain astrocytes menopausal hormone therapy, screening mammogra- phy, and estrogen receptor status. J Natl Cancer Inst 99: (35). Quercetin improves the therapeutic index of doxo- 1152–1161. rubicin by degrading HIF-1a protein in hypoxic breast 2) Pouysségur J, Dayan F, Mazure NM. 2006. Hypoxia sig- cancer 4T1 cells (36). Thus, quercetin and luteolin nalling in cancer and approaches to enforce tumour may decrease the CBR1 protein level by suppressing regression. Nature 441: 437–443. the expression of HIF-1a. Quercetin-reduced HIF-1a 3) Trédan O, Galmarini CM, Patel K, Tannock IF. 2007. protein is not restored by the proteasome inhibitor in Drug resistance and the solid tumor microenvironment. prostate cancer LNCaP cells (37). Likewise, in hypoxic J Natl Cancer Inst 99: 1441–1454. MCF-7 cells, resveratrol decreased HIF-1a protein level 4) Kaur M, Agarwal C, Agarwal R. 2009. Anticancer even in the presence of MG132 (Fig. 4). Thus, resver- and cancer chemopreventive potential of grape seed atrol seems to decrease HIF-1a protein through a pro- extract and other grape-based products. J Nutr 139: teasome-independent pathway. In contrast, in A2780/ 1806S–1812S. 5) Li Y, Tollefsbol TO. 2010. Impact on DNA methylation in CP70 and HepG2 cells, resveratrol degrades HIF-1a cancer prevention and therapy by bioactive dietary com- protein through the ubiquitin-proteasome pathway (38, ponents. Curr Med Chem 17: 2141–2151. 39). Therefore, in hypoxia, the degradation pathway 6) Olaku O, White JD. 2011. Herbal therapy use by cancer of HIF-1a protein by resveratrol is different depending patients: a literature review on case reports. Eur J Cancer on the cell types. Taken together, these results indicate 47: 508–514. that resveratrol suppresses CBR1-mediated resistance to 7) Rimando AM, Kalt W, Magee JB, Dewey J, Ballington doxorubicin by decreasing the HIF-1a protein, perhaps JR. 2004. Resveratrol, pterostilbene, and piceatannol in through a proteasome-independent pathway, in hypoxic vaccinium berries. J Agric Food Chem 52: 4713–4719. MCF-7 cells. 8) Tsai SH, Lin-Shiau SY, Lin JK. 1999. Suppression of The expression level of HIF-1a protein was decreased nitric oxide synthase and the down-regulation of the by resveratrol and its derivative 3,5-dihydroxy-4′- activation of NFkappaB in macrophages by resveratrol. Br J Pharmacol 126: 673–680. methoxy-trans-stilbene, but not by 3,5-dimethoxy- Resveratrol and Doxorubicin 127

9) Frémont L, Belguendouz L, Delpal S. 1999. Antioxidant of androgen receptor through hypoxia-inducible factor- activity of resveratrol and alcohol-free wine polyphe- 1a in a low androgen environment. J Steroid Biochem nols related to LDL oxidation and polyunsaturated fatty Mol Biol 123: 58–64. acids. Life Sci 64: 2511–2521. 25) Mathew A, Mathur SK, Morimoto RI. 1998. Heat shock 10) Leifert WR, Abeywardena MY. 2008. Cardioprotective response and protein degradation: regulation of HSF2 actions of grape polyphenols. Nutr Res 28: 729–737. by the ubiquitin-proteasome pathway. Mol Cell Biol 18: 11) Um JH, Park SJ, Kang H, Yang S, Foretz M, McBurney 5091–5098. MW, Kim MK, Viollet B, Chung JH. 2010. AMP-acti- 26) Kartner N, Riordan JR, Ling V. 1983. Cell surface P-gly- vated protein kinase-deficient mice are resistant to the coprotein associated with multidrug resistance in mam- metabolic effects of resveratrol. Diabetes 59: 554–563. malian cell lines. Science 221: 1285–1288. 12) Filomeni G, Graziani I, Rotilio G, Ciriolo MR. 2007. 27) Gavelová M, Hladíková J, Vildová L, Novotná R, Von- trans-Resveratrol induces in human breast drácek J, Krcmár P, Machala M, Skálová L. 2008. cancer cells MCF-7 by the activation of MAP kinases Reduction of doxorubicin and oracin and induction of pathways. Nutr 2: 295–305. carbonyl reductase in human breast carcinoma MCF-7 13) Osman AM, Bayoumi HM, Al-Harthi SE, Damanhouri cells. Chem Biol Interact 176: 9–18. ZA, Elshal MF. 2012. Modulation of doxorubicin cyto- 28) Heibein AD, Guo B, Sprowl JA, Maclean DA, Parissenti toxicity by resveratrol in a human breast cancer cell line. AM. 2012. Role of aldo-keto reductases and other doxo- Cancer Cell Int 12: 47. rubicin pharmacokinetic genes in doxorubicin resis- 14) Ito Y, Mitani T, Harada N, Isayama A, Tanimori S, Yake- tance, DNA binding, and subcellular localization. BMC naka S, Nakano Y, Inui H, Yamaji R. 2013. Identification Cancer 12: 381. of carbonyl reductase 1 as a resveratrol-binding protein 29) Kassner N, Huse K, Martin HJ, Gödtel-Armbrust U, by affinity chromatography using ′4 -amino-3,5-dihy- Metzger A, Meineke I, Brockmöller J, Klein K, Zanger droxy-trans-stilbene. J Nutr Sci Vitaminol 59: 358–364. UM, Maser E, Wojnowski L. 2008. Carbonyl reductase 15) Gonzalez-Covarrubias V, Ghosh D, Lakhman SS, Pen- 1 is a predominant doxorubicin reductase in the human dyala L, Blanco JG. 2007. A functional genetic polymor- liver. Drug Metab Dispos 36: 2113–2120. phism on human carbonyl reductase 1 (CBR1 V88I) 30) Bains OS, Karkling MJ, Lubieniecka JM, Grigliatti TA, impacts on catalytic activity and NADPH binding affin- Reid RE, Riggs KW. 2009. Naturally occurring variants ity. Drug Metab Dispos 35: 973–980. of human CBR3 alter anthracycline in vitro metabolism. 16) Bateman RL, Rauh D, Tavshanjian B, Shokat KM. 2008. J Pharmacol Exp Ther 332: 755–763. Human carbonyl reductase 1 is an S-nitrosoglutathione 31) Martin HJ, Breyer-Pfaff U, Wsol V, Venz S, Block S, Maser reductase. J Biol Chem 283: 35756–35762. E. 2006. Purification and characterization of akr1b10 17) Bains OS, Karkling MJ, Grigliatti TA, Reid RE, Riggs KW. from human liver: role in carbonyl reduction of xenobi- 2009. Two nonsynonymous single nucleotide polymor- otics. Drug Metab Dispos 34: 464–470. phisms of human carbonyl reductase 1 demonstrate 32) Bains OS, Grigliatti TA, Reid RE, Riggs KW. 2010. Natu- reduced in vitro metabolism of daunorubicin and doxo- rally occurring variants of human aldo-keto reductases rubicin. Drug Metab Dispos 37: 1107–1114. with reduced in vitro metabolism of daunorubicin and 18) Lopez M. 2006. Anthracyclines in the adjuvant treat- doxorubicin. J Pharmacol Exp Ther 335: 533–545. ment of breast carcinoma: thirty years later. Clin Ter 33) Penning TM, Burczynski ME, Jez JM, Lin HK, Ma H, 157: 165–177. Moore M, Ratnam K, Palackal N. 2001. Structure- 19) Varela M, Real MI, Burrel M, Forner A, Sala M, Brunet function aspects and inhibitor design of type 5 17beta- M, Ayuso C, Castells L, Montañá X, Llovet JM, Bruix J. hydroxysteroid dehydrogenase (AKR1C3). Mol Cell 2007. Chemoembolization of hepatocellular carcinoma Endocrinol 171: 137–149. with drug eluting beads: efficacy and doxorubicin phar- 34) Du GJ, Song ZH, Lin HH, Han XF, Zhang S, Yang YM. macokinetics. J Hepatol 46: 474–481. 2008. Luteolin as a glycolysis inhibitor offers superior 20) Kassner N, Huse K, Martin HJ, Gödtel-Armbrust U, efficacy and lesser toxicity of doxorubicin in breast can- Metzger A, Meineke I, Brockmöller J, Klein K, Zanger cer cells. Biochem Biophys Res Commun 372: 497–502. UM, Maser E, Wojnowski L. 2008. Carbonyl reductase 35) Park SW, Cho CS, Jun HO, Ryu NH, Kim JH, Yu YS, Kim 1 is a predominant doxorubicin reductase in the human JS, Kim JH. 2012. Anti-angiogenic effect of luteolin on liver. Drug Metab Dispos 36: 2113–2120. retinal neovascularization via blockade of reactive oxy- 21) López de Cerain A, Marín A, Idoate MA, Tuñón MT, Bello gen species production. Invest Ophthalmol Vis Sci 53: J. 1999. Carbonyl reductase and NADPH cytochrome 7718–7726. P450 reductase activities in human tumoral versus nor- 36) Du G, Lin H, Wang M, Zhang S, Wu X, Lu L, Ji L, Yu L. mal tissues. Eur J Cancer 35: 320–324. 2010. Quercetin greatly improved therapeutic index of 22) Tak E, Lee S, Lee J, Rashid MA, Kim YW, Park JH, Park doxorubicin against 4T1 breast cancer by its opposing WS, Shokat KM, Ha J, Kim SS. 2011. Human carbonyl effects on HIF-1a in tumor and normal cells. Cancer Che- reductase 1 upregulated by hypoxia renders resistance mother Pharmacol 65: 277–287. to apoptosis in hepatocellular carcinoma cells. J Hepatol 37) Lee DH, Lee YJ. 2008. Quercetin suppresses hypoxia- 54: 328–339. induced accumulation of hypoxia-inducible factor- 23) Paul S, Mizuno CS, Lee HJ, Zheng X, Chajkowisk S, 1alpha (HIF-1alpha) through inhibiting protein synthe- Rimoldi JM, Conney A, Suh N, Rimando AM. 2010. In sis. J Cell Biochem 105: 546–553. vitro and in vivo studies on stilbene analogs as potential 38) Cao Z, Fang J, Xia C, Shi X, Jiang BH. 2004. trans-3,4,5′- treatment agents for colon cancer. Eur J Med Chem 45: Trihydroxystibene inhibits hypoxia-inducible factor 3702–3708. 1alpha and vascular endothelial growth factor expres- 24) Mitani T, Yamaji R, Higashimura Y, Harada N, Nakano Y, sion in human cells. Clin Cancer Res 10: Inui H. 2011. Hypoxia enhances transcriptional activity 5253–5263. 128 Mitani T et al.

39) Zhang Q, Tang X, Lu QY, Zhang ZF, Brown J, Le AD. cancer cells. Mol Nutr Food Res 53: 845–858. 2005. Resveratrol inhibits hypoxia-induced accumu- 42) Minotti G, Menna P, Salvatorelli E, Cairo G, Gianni L. lation of hypoxia-inducible factor-1alpha and VEGF 2004. Anthracyclines: molecular advances and phar- expression in human tongue squamous cell carcinoma macologic developments in antitumor activity and car- and hepatoma cells. Mol Cancer Ther 4: 1465–1474. diotoxicity. Pharmacol Rev 56: 185–229. 40) Iguchi K, Toyama T, Ito T, Shakui T, Usui S, Oyama M, 43) Olson LE, Bedja D, Alvey SJ, Cardounel AJ, Gabrielson Iinuma M, Hirano K. 2012. Antiandrogenic activity of KL, Reeves RH. 2003. Protection from doxorubicin- resveratrol analogs in prostate cancer LNCaP cells. J induced cardiac toxicity in mice with a null allele of car- Androl 33: 1208–1215. bonyl reductase 1. Cancer Res 63: 6602–6606. 41) Lappano R, Rosano C, Madeo A, Albanito L, Plastina 44) Gu J, Song ZP, Gui DM, Hu W, Chen YG, Zhang DD. P, Gabriele B, Forti L, Stivala LA, Iacopetta D, Dolce V, 2012. Resveratrol attenuates doxorubicin-induced car- Andò S, Pezzi V, Maggilini M. 2009. Structure-activity diomyocyte apoptosis in lymphoma nude mice by heme relationships of resveratrol and derivatives in breast oxygenase-1 induction. Cardiovasc Toxicol 4: 341–349.