Phytoestrogen Suppresses Efflux of the Diagnostic Marker Protoporphyrin IX in Lung Carcinoma

Home , Porphobilinogen

Published OnlineFirst February 2, 2016; DOI: 10.1158/0008-5472.CAN-15-1484 Cancer Molecular and Cellular Pathobiology Research

Phytoestrogen Suppresses Efﬂux of the Diagnostic Marker Protoporphyrin IX in Lung Carcinoma Hirofumi Fujita1, Keisuke Nagakawa1, Hirotsugu Kobuchi2, Tetsuya Ogino3, Yoichi Kondo1, Keiji Inoue4, Taro Shuin4, Toshihiko Utsumi5, Kozo Utsumi1,†, Junzo Sasaki1, and Hideyo Ohuchi1

Abstract

One promising method to visualize cancer cells is based on the treatment, a significant increase in PpIX accumulation was detection of the fluorescent photosensitizer protoporphyrin IX observed in A549 cells (3.7-fold) and in other cell lines. Systemic (PpIX) synthesized from 5-aminolevulinic acid (ALA), but this preconditioning with genistein in a mouse xenograft model of method cannot be used in cancers that exhibit poor PpIX lung carcinoma resulted in a 1.8-fold increase in accumulated accumulation. PpIX appears to be pumped out of cancer cells by PpIX. Long-term genistein treatment stimulated the expression the ABC transporter G2 (ABCG2), which is associated with of genes encoding enzymes involved in PpIX synthesis, such as multidrug resistance. Genistein is a phytoestrogen that appears porphobilinogen deaminase, uroporphyrinogen decarboxyl- to competitively inhibit ABCG2 activity. Therefore, we investigat- ase, and protoporphyrinogen oxidase. Accordingly, the rate of ed whether genistein can promote PpIX accumulation in human PpIX synthesis was also accelerated by genistein pretreatment. lung carcinoma cells. Here we report that treatment of A549 lung Thus, our results suggest that genistein treatment effectively carcinoma cells with genistein or a specific ABCG2 inhibitor enhances ALA-induced PpIX accumulation by preventing promoted ALA-mediated accumulation of PpIX by approximately the ABCG2-mediated efflux of PpIX from lung cancer cells 2-fold. ABCG2 depletion and overexpression studies further and may represent a promising strategy to improve ALA-based revealed that genistein promoted PpIX accumulation via func- diagnostic approaches in a broader set of malignancies. tional repression of ABCG2. After an extended period of genistein Cancer Res; 76(7); 1–10. 2016 AACR.

Introduction (3), and brain tumors (4), to identify precise tumor margins and prevent overlooking small lesions that are otherwise invisible. Protoporphyrin IX (PpIX) functions as a fluorescent photosen- However, ALA-based photodynamic diagnosis remains unsatis- sitizer, which is synthesized from 5-aminolevulinic acid (ALA). As factory in diagnosing some tumors that accumulate insufficient PpIX preferentially accumulates in malignant tissues (1), the amounts of PpIX (5–9). exogenous administration of ALA enables us to detect tumors Successful ALA-induced PpIX accumulation may rely on the exhibiting enhanced PpIX fluorescence. This technology, referred activity of enzymes that synthesize and metabolize PpIX and on to as photodynamic diagnosis, has been widely used clinically, the proteins that transport PpIX (10, 11). Exogenously added ALA especially during surgery for bladder cancer (2), prostate cancer is taken up by target cells and metabolized to coproporphyrinogen III in the cytosol by several enzymes, which include porpho- 1Department of Cytology and Histology, Okayama University Gradu- bilinogen deaminase (PBGD), uroporphyrinogen III synthase ate School of Medicine, Dentistry and Pharmaceutical Sciences, (UROS), which is the rate-limiting enzyme of porphyrin metab- Okayama, Japan. 2Department of Cell Chemistry, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, olism (12), and uroporphyrinogen decarboxylase (UROD). Okayama, Japan. 3Department of Nursing Science, Faculty of Health Coproporphyrinogen III is then translocated into mitochondria and Welfare Science, Okayama Prefectural University, Soja, Japan. through ATP-binding cassette (ABC) transporter B6 and metab- 4Department of Urology, Kochi University Medical School, Nankoku, Kochi, Japan. 5Department of Biological Chemistry, Faculty of Agri- olized to PpIX by coproporphyrinogen oxidase (CPOX) and culture, Yamaguchi University, Yamaguchi, Japan. protoporphyrinogen oxidase (PPOX; refs. 5, 7). PpIX is metab- Note: Supplementary data for this article are available at Cancer Research olized further to heme by ferrochelatase. Furthermore, accumu- Online (http://cancerres.aacrjournals.org/). lating evidence indicates that the elimination of PpIX from cells is H. Fujita and K. Nagakawa contributed equally to this article. carried out by ABC transporter G2 (ABCG2), which is a multidrug resistance-associated protein (11). Thus, heme synthesis enzymes †Deceased. and ABCG2 play important roles in regulating the cellular accu- Corresponding Author: Hirofumi Fujita, Okayama University Graduate School, mulation of PpIX in cancer. Our current research goal is to develop 2-5-1 shilatacho, Kitaku, Okayama 700-8558, Japan. Phone: 818-6235-7081; Fax: new combination regimens with compounds that enhance the 818-6235-7079; E-mail: [email protected] accumulation of PpIX to improve ALA-induced PpIX accumula- doi: 10.1158/0008-5472.CAN-15-1484 tion. A recent study reported that the systemic administration 2016 American Association for Cancer Research. of vitamin D3 for preconditioning significantly increased the

www.aacrjournals.org OF1

Fujita et al.

accumulation of PpIX in squamous tumor cell lines both in vitro cell lines were authenticated using short tandem repeat analysis and in vivo (13). The underlying mechanism involves increases in with the GenePrint 10 System (Promega) and Cell Line Authen- the expression of CPOX and decreases in that of ferrochelatase. We tication Database of JCRB and ATCC in 2015. ST-HEK cells were previously reported that the iron chelator deferoxamine promot- prepared by the stable transfection of ABCG2 and cells expressing ed the accumulation of PpIX in urothelial carcinoma in vitro and in high levels of the functional ABCG2 protein (18). These cells were vivo as well as in prostate cancer, oral squamous cell carcinoma, maintained in complete medium: RPMI1640 supplemented with and histiocytic lymphoma in vitro (14–17). Furthermore, the 10% heat-inactivated FBS, 100 U/mL penicillin, and 100 mg/mL inhibition of ABCG2 by specific inhibitors or knockdown using streptomycin in a humidified atmosphere with 5% CO2/air at RNAi facilitated the ALA-mediated accumulation of PpIX (8, 18, 37C. Typically, 1 105 cells were seeded in 1.5 mL of complete 19). Therefore, the use of compounds that stimulate the synthesis medium on 3.5-cm dishes and cultured for 24 hours before of PpIX or block the efflux of PpIX may become a good strategy to each experiment. Unless otherwise indicated, chemicals were improve ALA-induced PpIX accumulation. added at the following final concentrations: 0.5 mmol/L of ALA, Estrogens are known to induce porphyria cutanea tarda, which 5to50mmol/L of genistein, 1 mmol/L of Ko143, and 300 mmol/L is characterized clinically by cutaneous photosensitivity and the of Noc18. excessive excretion of porphyrins (20). This effect of estrogen is fi supported by the ndings of another study using cancer-bearing Flow cytometry of cellular PpIX female rats in which estrogen depletion by ovariectomy caused a After being incubated, cells were rinsed three times with PBS fi signi cant reduction in ALA-induced PpIX levels and PBGD and harvested by trypsinization. After centrifugation at 800 g for activity in tumors (21). On the other hand, a phytoestrogen 5 minutes, the cells were resuspended in 0.4 mL of PBS. Cellular genistein known as a tyrosine kinase inhibitor was found to PpIX contents were measured using the flow cytometer FACScan exhibit estrogen-like activity by interacting with estrogen recep- (BD Biosciences) and quantified with CellQuest software (BD – tors (ER) in mammals (22 24). Furthermore, a previous study Biosciences). A total of 10,000 cells were analyzed in each sample reported that genistein reversed ABCG2-mediated multidrug (excitation 488 nm, emission 650 nm). resistance and genistein was likely to competitively inhibit the efflux of anticancer agents such as SN-38 and mitoxantrone by ABCG2 (25). Fluorescence microscopy fi After being treated with ALA, cells were stained with 1 mmol/L These ndings prompted us to hypothesize that genistein promotes the accumulation of PpIX by increasing the synthesis MitoTracker Green for 20 minutes at 37 C and then observed by fl of PpIX and/or reducing the efflux of PpIX. However, the effects of uorescence microscopy (Axiovert 200, Carl Zeiss Inc.). Mito- fl genistein on the accumulation of PpIX have not yet been Tracker Green is a uorescent dye compound that is used for the elucidated. detection of mitochondria. Fluorescence images were taken using Therefore, we herein determined whether genistein increased a highly light-sensitive thermoelectrically cooled charge-coupled fi the accumulation of PpIX in vitro and in a xenograft model using device camera (Axio-Cam CCD camera, Zeiss). The lter combi- fi the human lung carcinoma A549 cell line, which expresses high nations were a 450-nm excitation lter, 510-nm beam splitter, fi levels of the endogenous ABCG2 protein. and 515 to 565-nm emission lter for MitoTracker green; a 400-nm excitation filter, 580-nm beam splitter, and 590-nm long-pass emission filter for PpIX. Materials and Methods Chemicals Western blot analysis ALA was purchased from COSMO OIL. The iron chelator Cells were solubilized in ice-cold lysis buffer (20 mmol/L deferoxamine, genistein, Ko143, FBS, and G418 were obtained Tris-HCl, pH 7.4, 0.15 mol/L NaCl, 1% NP-40, 0.1% SDS, 0.1% from Sigma-Aldrich. RPMI1640 medium was obtained from sodium deoxycholate, 5 mmol/L ethylenediaminetetraacetic Wako. (Z)-1-[N-(2-Aminoethyl)-N-(2-ammonioethyl)-amino]- acid (EDTA), 5 mmol/L ethylene glycol tetraacetic acid (EGTA), diazen-1-ium-1,2- diolate (NOC18) was obtained from Dojindo. 1 mmol/L phenylmethylsulfonylfluoride, and 1 mg/mL each of The antibody to ABCG2 was obtained from Cell Signaling Tech- leupeptin and pepstatin A). After centrifugation of the homogenate nology. The monoclonal anti-actin antibody (clone C4) was at 15,000 g for 15 minutes to remove cell nuclei, the supernatant from Millipore. The anti-ferrochelatase antibody was a gift from was collected and the protein content was determined using a BCA Dr. S. Taketani (Kyoto Institute of Technology, Kyoto, Japan). Protein Assay Kit. Samples were prepared by mixing with 2 SDS MitoTracker Green was obtained from Invitrogen. The BCA Pro- sample buffer and boiling for 5 minutes, and were then stored at tein Assay Kit was from Thermo Scientific. All other chemicals 80C until use. Protein content was determined using a BCA were of analytic grade and obtained from Nacalai Tesque. Defer- protein assay kit. The samples were subjected to SDS-PAGE and oxamine was dissolved in saline as a stock solution. Genistein, proteins in the gel were transferred electrophoretically onto Ko143, and MitoTracker Green were dissolved in DMSO and an Immobilon membrane (Millipore). The membrane was blocked stored in aliquots at 20 C until use. ALA was diluted in ultrapure by 5% skim milk in Tris–buffered saline tween 20 (TBST; 0.15 mol/ water to make a stock solution of 0.5 mol/L. L NaCl, 0.05% Tween 20, 10 mmol/L Tris-HCl, pH 7.4) and then incubated overnight with primary antibodies (1:1,000 for the Cell lines and culture conditions mouse anti-human actin antibody, 1:300 for the rabbit anti-human The following cell lines were purchased from the Japanese ABCG2 antibody, and 1:1,000 for the rabbit anti-bovine ferroche- Collection of Research Bioresources: A549, HEK293T, T98G, latase antibody) diluted in TBST containing 5% skim milk at 4C. T24, MDA-MB-231, MeWo, DLD-1, and H1299. U937 and After washing three times with TBST, the membrane was incubated HL-60 cell lines were obtained from the RIKEN Cell Bank. These for 1.5 hours with the horseradish peroxidase -conjugated

OF2 Cancer Res; 76(7) April 1, 2016 Cancer Research

Genistein Promotes PpIX Accumulation

secondary antibody diluted at 1:5,000 in TBST containing 5% skim reverse primer: CTGGGTGTGCAACTGTCTGATAC; human milk at room temperature. Immunoreactive bands ware visualized UROD forward primer CGGGAGTGTGTGGGAA, UROD reverse with Immunostar LD and exposed to Polaroid films. primer: AAGCAGACGTGAGTGTTTATGCA; human CPOX forward primer GGCGGAGATGTTGCCTAAGAC, CPOX reverse qRT-PCR primer: AATGCTCACCCCAGCCTTTT; human b-actin forward Total RNA was isolated from A549 cells treated with or without primer TGGCACCCAGCACAATGAA, b-actin reverse primer: 50 mmol/L genistein for 28 hours using RNeasy Mini Kit (Qiagen) CTAAGTCATAGTCCGCCTAGAAGCA; The PCR protocol was following the manufacturer's instructions. The RNA was treated 95C for 10 minutes followed by 40 cycles of denaturation at with a TURBO DNA-free kit (ThermoFisher Scientific) to remove 95C for 15 seconds and annealing/extension at 60C for genomic DNA. cDNA was prepared from 0.5 mg of total RNA using 1 minute. Revertra Ace qPCR RT master Mix (TOYOBO). One-twentieth of each of the obtained cDNA specimens was used for each PCR. Generation of an in vivo tumor xenograft model qRT-PCR was performed in LightCycler 8-Tube Strips using All animal procedures were approved by the Okayama Uni- the LightCycler Nano Instrument and FastStart Essential DNA versity Institutional Animal Care and Use Committee (approval Green Master (Roche). The primers used to amplify heme syn- number: OKU-2012634 and OKU-2015443). Male athymic nude thesis genes and an internal standard b-actin were: human mice (BALB/c nu/nu) were obtained from Charles River Japan, ALAD forward primer CCTCGGTTCCAACCAACTGAT, ALAD Inc. and they received intradermal injections of 2 106 A549 cells reverse primer: GATAGGVTGTATGTCATCAGGAACA; human in each flank. After 2 weeks, visible nodules approximately 5 mm PBGD forward primer CAAGGACCAGGACATCTTGGAT, PBGD in diameter had formed. Mice (n ¼ 5 per group) received vehicle reverse primer: CCAGACTCCTCCAGTCAGGTACA; human (50% ethanol in PBS) or genistein [10 mg/kg/day, intraperitone- UROS forward primer TCAGCACTGCCTCTTCTATTTCC, UROS ally (i.p.) daily for 3 days] as a preconditioning treatment. Zero, 1,

Figure 1. Effects of genistein on ALA-mediated accumulation of PpIX in ABCG2-expressing cells. A, protein expression of ABCG2 and ferrochelatase (FECH) in U937 and A549 cells. Cell lysates were analyzed by Western blotting using antibodies against ABCG2, ferrochelatase, and actin. B, the ALA-mediated accumulation of PpIX in U937 and A549 cells. Cells were treated with the indicated concentration of ALA for 3 hours. PpIX fluorescence was measured by a flow cytometer and expressed as a percentage of the control (0 mmol/L ALA). C, genistein (left graph) and Ko143 (an ABCG2 inhibitor, right graph) treatments stimulated ALA-mediated accumulation of PpIX in A549 cells in a dose-dependent manner. D, genistein and Ko143 stimulated ALA-mediated accumulation of PpIX in A549 cells in a time- dependent manner. Data in C are percent increases relative to the PpIX fluorescence intensity induced by 0.5 mmol/L ALA only. E and F, effects of genistein, Ko143, and Noc18 (ferrochelatase inhibitor) on the ALA-induced accumulation of PpIX in U937 (E) and A549 (F) cells. Cells were incubated with 0.5 mmol/L ALA for 3 hours in the presence or absence of 1 mmol/L Ko143, 25 mmol/L genistein, or 300 mmol/L Noc18 and intracellular PpIX fluorescence was measured. The mean SD for three independent experiments. Asterisks, a significant difference from ALA only.

www.aacrjournals.org Cancer Res; 76(7) April 1, 2016 OF3

Fujita et al.

4, 12, and 24 hours after the administration of ALA [75 mg/kg, diluted 1:750 for 1.5 hours at room temperature. Cell nuclei were intramuscularly (i.m.)] to both groups, the mice were euthanized stained with 40-6-diamidino-2-phenylindole (DAPI). The slides and the tumors and neighboring normal-appearing tissues were were mounted with the antifade mountant, SlowFade Gold harvested, embedded in optimal cutting temperature com- (Life Technologies). The sections were analyzed by confocal pound (Sakura Finetek), and frozen in liquid nitrogen. Sec- microscopy (ZEISS Confocal Laser Scanning Microscope Model tions were cut (15-mm thickness) and five consecutive sections LSM780) or BZ-X700. were mounted on separate Superfrost Plus glass slides (Thermo Fisher Scientific). PpIX was detected without fixation with a 400-nm excitation filter, a 565-nm beam splitter, and a Statistical analysis 605/70 nm bandpass emission filter on a fluorescence micro- Statistical analyses were performed using the Student t test. The fi scope (BZ-X700, KEYENCE). Sections containing tumor tissue mean of three distributions was considered signi cantly different < and neighboring normal-appearing tissue were identified and if P 0.05. analyzed with BZ-X Analyzer software (KEYENCE). Detection of E-cadherin was used to distinguish the tumor tissues and Results nontumor tissues, as described below. PpIX fluorescence was Effects of genistein on ALA-mediated accumulation of PpIX in measured as reported previously (13) with the following A549 cells modifications. The PpIX signal intensity per unit area was To determine whether genistein promoted the ALA-mediated measured separately in the tumor and nontumor tissues by accumulation of PpIX by inhibiting ABCG2, we examined the calculating the mean intensity of the red fluorescence in each relationship between the abundance of ABCG2 and effects of pixel of each digital image. The ratios of tumor PpIX signal genistein on the accumulation of PpIX. A Western blot analysis per unit area and nontumor PpIX signal per unit area were confirmed that A549 cells had a high content of ABCG2, whereas compared. human histiocytic lymphoma U937 cells had a low content of ABCG2 (Fig. 1A; ref. 18). The ALA-mediated accumulation of PpIX E-cadherin IHC was significantly lower in A549 cells than in U937 cells (Fig. 1B). Frozen sections adjacent to the sections used for PpIX detection Genistein increased the ALA-mediated accumulation of PpIX in were fixed in 4% paraformaldehyde for 10 minutes and then were A549 cells in dose- and time-dependent manners, which was incubated with blocking solution (5% goat serum and 0.25% similar to the effects of the specific ABCG2 inhibitor Ko143 Triton X-100 in PBS) for 30 minutes at room temperature. (Fig. 1C and D). In contrast, neither Ko143 nor genistein pro- The sections were incubated with the anti-E-cadherin antibody moted the ALA-mediated accumulation of PpIX in U937 cells diluted 1:200 in IMMUNO SHOT (Cosmobio) for 16 hours at (Fig. 1E). When heme synthesis was inhibited by the FEHC 4C. After washing, Alexa488-labeled secondary antibodies were inhibitor Noc18 (NO donor), cellular PpIX levels increased

Figure 2. Effects of genistein and Ko143 on the amount and distribution of PpIX in ALA-treated A549 cells in vitro. Cells were incubated with 0.5 mmol/L ALA for 3 hours in the presence or absence of 1 mmol/L Ko143 or 25 mmol/L genistein and then stained with MitoTracker Green, a mitochondrial-selective probe. Images were taken using ﬂuorescent microscopy. Magniﬁcation, 200.

OF4 Cancer Res; 76(7) April 1, 2016 Cancer Research

Genistein Promotes PpIX Accumulation

significantly in U937 cells and A549 cells (Fig. 1E and F). Fluo- generation of NO, but not with Ko143 or genistein than that in rescence microscopy revealed that accumulated PpIX mainly control cells. These results suggested that the efflux of PpIX by localized in mitochondria (Fig. 2). These results indicated that ABCG2 and heme synthesis were major factors regulating the the ALA-mediated accumulation of PpIX was facilitated by genis- intracellular accumulation of PpIX in A549 cells and also that tein in A549 cells in vitro. genistein suppressed the efflux of intracellular PpIX by ABCG2. We then investigated the effects of genistein on the ALA-medi- Genistein stimulated PpIX accumulation by the functional ated accumulation of PpIX in ABCG2-overexpressing HEK cells repression of ABCG2 (ST-HEK cells, Fig. 3C). The ALA treatment failed to increase the As genistein stimulated the ALA-mediated accumulation of accumulation of PpIX in ST-HEK cells, which was in contrast to PpIX in a similar manner to that of the ABCG2 inhibitor, we that in parental HEK cells. However, these two cells showed the expected genistein to suppress the efflux of PpIX through ABCG2. higher or similar accumulation of PpIX when cells were treated Thus, we examined the effects of genistein on the accumulation of with Ko143 or genistein, respectively (Fig. 3D). These results fl PpIX in ABCG2-knockdown cells. Figure 3A shows that the siRNA suggested that genistein suppressed the PpIX ef ux function of designed to silence the ABCG2 gene obliterated the protein ABCG2 in HEK cells. expression of ABCG2 in A549 cells. After the accumulation of PpIX was induced with ALA, washing the control cells with A longer genistein treatment markedly stimulated ALA-free medium significantly decreased the accumulated levels ALA-mediated accumulation of PpIX of PpIX in 3 hours. On the other hand, large amounts of PpIX still We also examined the effects of a longer genistein treatment for remained in ABCG2-knockdown cells after washing (Fig. 3B). up to 48 hours on the ALA-mediated accumulation of PpIX in When control cells were washed in the presence of Ko143 or A549 cells. The accumulation of PpIX was markedly higher by ALA genistein, larger amounts of PpIX were retained in the cells. The þ genistein than by ALA alone or ALA þ Ko143 (Fig. 4A). The accumulation of PpIX in ALA-treated ABCG2-knockdown cells ABCG2-inhibitory activity of Ko143 was not affected by the was greater with Noc18, which inhibits ferrochelatase through the incubation in culture medium for 48 hours (data not shown).

Figure 3. Genistein (Geni) stimulated PpIX accumulation by functionally repressing ABCG2. A and B, ABCG2 protein expression in A549 cells was knocked down by RNAi, and the effects of genistein on PpIX excretion/consumption were determined by removing the substrate. Cells were transfected with ABCG2-specific siRNA for 5 days, treated with ALA for 1.5 hours, rinsed with ALA-free medium, and then incubated in the medium for 3 hours in the presence of 1 mmol/L Ko143, 25 mmol/L genistein, or 300 mmol/L Noc18. A, Western blot for the ABCG2 protein from ABCG2-knockdown A549 cells. B, PpIX levels were measured by flow cytometry. Asterisks, a significant difference from the corresponding "control cells." #, a significant difference from the corresponding "Wash" samples. C, ABCG2 protein expression in parental HEK cells and ST-HEK cells, in which an ABCG2 expression vector was stably transfected. D, effects of genistein on the accumulation of PpIX in HEK cells and ST-HEK cells. HEK cells and ST-HEK cells were incubated with 0.5 mmol/L ALA for 3 hours in the presence of 1 mmol/L Ko143 or 25 mmol/L genistein. The mean SD for three independent experiments. Asterisks, a significant difference from the corresponding "ALA" samples.

www.aacrjournals.org Cancer Res; 76(7) April 1, 2016 OF5

Fujita et al.

Figure 4. A longer genistein treatment markedly stimulated ALA-mediated accumulation of PpIX. A, cells were treated with 0.5 mmol/L ALA for 48 hours in the presence of 1 mmol/L Ko143 or 50 mmol/L genistein and PpIX fluorescence was analyzed by flow cytometry. B, the amount and distribution of PpIX at 48 hours were analyzed by fluorescence microscopy. Magnification, 200. MitoTracker Green is a mitochondrial-sensitive probe.

Fluorescence microscopy revealed that the ALA þ genistein treat- cancer cells (26–28), the effect of genistein on the cell growth of ment for 48 hours increased PpIX fluorescence in A549 cells and A549 cells was examined. As shown in Fig. 5C, 25 and 50 mmol/L that PpIX was localized in the cytosol and mitochondria, which genistein inhibited the cell growth of A549 cells for 48 hours. was similar to that with ALA alone (Fig. 4B). Furthermore, the pretreatment with genistein significantly increased the ALA-medi- Systemic preconditioning with genistein enhanced the ated accumulation of PpIX in a manner that depended on the dose accumulation of PpIX in a tumor model established in vivo of genistein and time of the pretreatment (Fig. 5A and 5B). In To determine whether genistein can enhance the accumulation addition, genistein enhanced ALA-mediated PpIX accumulation of PpIX in tumor cells in vivo, nude mice were implanted with in another lung carcinoma cell line, H1299, and also in various A549 adenocarcinoma cells. After 2 weeks, the tumors were visible tumor cell lines that express ABCG2, including glioblastoma and the animals received systemic preconditioning with vehicle or T98G cells, breast cancer MDA-MB-231 cells, melanoma MeWo genistein over a 3-day period. On the third day, ALA was injected cells, and colorectal adenocarcinoma DLD-1 cells (Supplemen- intramuscularly and both tumor-bearing and nontumor tissues tary Fig. S1A and S1B). Importantly, genistein pretreatment also were harvested 0, 1, 4, 12, and 24 hours later and were analyzed enhanced the cell death induced by ALA-mediated photodynamic for PpIX fluorescence. In the vehicle control, ALA was found to treatment in A549 cells (Supplementary Fig. S1C). As estrogen induce a transient accumulation of PpIX in the xenografted tumor and estrogenic compounds often stimulate the proliferation of cells (Supplementary Fig. S2A), which was confirmed with the

Figure 5. The genistein pretreatment stimulated ALA-mediated accumulation of PpIX. A, A549 cells were treated with the indicated concentration of genistein for 48 hours and then with 0.5 mmol/L ALA þ genistein for 3 hours. PpIX fluorescence was analyzed by flow cytometry. B, preincubation time-dependent accumulation of ALA-mediated PpIX by genistein. C, effects of genistein on A549 cell growth at 48 hours. Cell growth was analyzed with a hemocytometer and the Trypan blue exclusion test. Asterisks, genistein significantly increased the accumulation of PpIX. Data are the mean SD derived from three independent experiments.

OF6 Cancer Res; 76(7) April 1, 2016 Cancer Research

Genistein Promotes PpIX Accumulation

detection of E-cadherin immunoreactivity in the tumor tissues. not suppress, but rather increased the protein expression of PpIX accumulation peaked 4 hours after the administration of ABCG2. We investigated the effects of genistein on the gene ALA (Supplementary Fig. S2A), although, the fluorescence inten- expression of heme synthesis enzymes. The gene expression of sity ratio for PpIX between the tumor and nontumor tissues was ALAD, PBGD, UROS, UROD, CPOX, and PPOX was stimulated by relatively low at the same time point (1.49 0.31). The greatest the genistein treatment (Fig. 6E). Accordingly, the rate of PpIX contrast in PpIX fluorescence was observed 1 hour after ALA synthesis was also accelerated by the genistein pretreatment for administration, and at the same time point the genistein pretreat- 48 hours when the efflux of PpIX by ABCG2 and its consumption ment significantly increased PpIX accumulation in the xeno- by ferrochelatase were inhibited by Ko143 and Noc18 (Fig. 6F). grafted tumor (Fig. 6A–C). Moreover, this increase in PpIX accu- These results suggested that the long-term treatment with genis- mulation did not affect the tumor:nontumor PpIX fluorescence tein increased the accumulation of PpIX by upregulating gene ratio (Fig. 6C). Treatment with genistein also increased the accu- expression involved in the synthesis of PpIX. mulation of PpIX in the normal epidermis and hair follicles (Supplementary Fig. S2B and S2C), suggesting that genistein promotes PpIX accumulation in proliferating tissues such as skin. Discussion We also attempted to elucidate the mechanism by which the The enhanced accumulation of PpIX in tumor cells represents long-term treatment with genistein increased the accumulation of a critical issue for successful ALA-mediated photodynamic PpIX in vitro. We examined whether genistein affected the expres- diagnosis. In the current study, we showed that genistein sion of ABCG2 in A549 cells. Figure 6D showed that genistein did stimulated the accumulation of PpIX in A549 human lung

Figure 6. Preconditioning with genistein stimulated the accumulation of PpIX in a xenograft tumor model and promoted the gene expression of heme metabolism enzymes in A549 cells. A, histologic detection of PpIX levels in A549 tumor cells established in nude mice as a xenograft model (see Materials and Methods). These mice received genistein (10 mg/kg, daily, i.p.) for 3 days and ALA (75 mg/kg, i.m.) for 1 hour to stimulate the synthesis of PpIX before the tumors were analyzed. Wide-field and confocal fluorescence photomicrographs of frozen sections showed PpIX fluorescence and E-cadherin immunoreactivity, respectively. Asterisk, calcified tissue. Scale bar, 300 mm. B, digital quantification of PpIX fluorescence in A549 tumors subjected to genistein as shown in A. PpIX signal per unit area in the E-cadherin positive region was calculated and expressed as a percentage of vehicle PpIX fluorescence. Data are the mean SD calculated from five samples. Asterisks, a significant difference from the vehicle control. C, the ratio of PpIX fluorescence in the tumor to that in adjacent nontumor tissue is shown. D, Western blot analysis of protein levels. A549 cells were treated with 50 mmol/L genistein for the indicated times in vitro. Proteins were analyzed by Western blotting. E, changes in the mRNA expression of heme metabolism enzymes in A549 cells treated with genistein in vitro. These mRNAs were analyzed by quantitative real-time PCR. Asterisks, a significant difference from the 0-h control of each gene of interest. FECH, ALAD, PBGD, UROS, UROD, CPOX, and PPOX are ferrochelatase (FECH), ALA dehydrogenase, PBGD, UROS, UROD, CPOX, and PPOX, respectively. F, the rate of PpIX synthesis in genistein-treated A549 cells. A549 cells were treated with or without 50 mmol/L genistein for 48 hours and then treated with 0.5 mmol/L ALA for 1 and 2 hours in the presence of Noc18 and Ko143.

www.aacrjournals.org Cancer Res; 76(7) April 1, 2016 OF7

Fujita et al.

non–small cell lung cancer cells (33), genistein did not stimulate the proliferation of A549 cells expressing ERb (Fig. 1 and 5C; refs. 34–36). Furthermore, estrogen depletion by ovariectomy in rats reduced ALA-induced PpIX levels in the tumors of these animals (21). These findings suggest that estrogen and the phytoestrogen genistein may accelerate the accumulation of PpIX by stimulating ERb signaling. Porphyria cutanea tarda is a metabolic disease of the heme synthesis pathway and is characterized by vesicles, bullae, and the fragility of skin and sclerodermoid changes that occur predomi- nantly in sun-exposed areas (37,38). This disorder is biochemically characterized by the accumulation of uroporphyrin and hepta- carboxyl porphyrin in the liver (39). Uroporphyrin and hepta- carboxyl porphyrin circulate in the plasma and mediate cutaneous photosensitivity. It is widely accepted that estrogen is one of the Figure 7. environmental factors of porphyria cutanea tarda and the use of Schematic representation of potential mechanisms by which phytoestrogen estrogen also plays a role its clinical expression (37, 39). In this genistein stimulates ALA-mediated accumulation of PpIX in malignant cells. context, itisimportant tonote thatgenistein pretreatment increased FECH, ALAD, PBGD, UROS, UROD, CPOX, PPOX, and PEPT are ferrochelatase ALA-induced PpIX accumulation in the mouse normal epidermis (FECH), ALA dehydrogenase, PBGD, UROS, UROD, CPOX, PPOX, and and hair follicles as well as in xenografted tumor in the current oligopeptide transporter, respectively. study. Thus, estrogens may be good candidates of potentiating agents for ALA-mediated photodynamic diagnosis in humans. adenocarcinoma cell line both in vitro and in vivo by preventing Genistein is known to enhance apoptosis induced by the ABCG2-mediated PpIX efflux and upregulating the gene expres- anticancer drug trichostatin A in A549 cells, but not in normal sion of heme synthesis enzymes (Fig. 7). Furthermore, the human lung fibroblasts by increasing the expression of ERb- enhancement of PpIX accumulation by genistein was not lim- mediated TNF receptor-1 (40, 41). Therefore, genistein may be ited to the A549, lung adenocarcinoma cell line, and similar beneficial not only for improving the quality of ALA-mediated accumulation patterns were observed in other tumor cell lines. photodynamic diagnosis, but also for better outcomes of subse- Thus, genistein appears to effectively improved ALA-induced quent chemotherapy in patients with lung cancer. PpIX accumulation in ABCG2 expressing tumors in vitro and in In the current study, ALA-induced PpIX accumulation increased vivo and these findings suggest that the phytoestrogen, genis- 3.7-fold in vitro and 1.8-fold in vivo following pretreatment with tein, may represent a potentiating agent for ALA-mediated genistein (Figs. 5B and 6B). This discrepancy between the in vitro photodynamic diagnosis in humans. and in vivo results may be multifactorial. One possible explana- Various approaches have been attempted to promote the tion is that genistein can be metabolized to 30-OH-genistein by accumulation of PpIX in tumors by regulating key factors in the cytochrome P450 1A2 (CYP1A2), a major enzyme in the phase I synthesis and metabolism of PpIX. The inhibition of heme hydroxylation process in mouse liver (42, 43). Another consid- synthesis by deferoxamine has been shown to improve the eration is that passive targeting systems [e.g., tumor targeting ALA-induced accumulation of PpIX in urothelial carcinoma, nanoparticles (44)] may improve the effects of genistein in vivo by leukemia, and gastric cancer (15–17, 29–31). However, the ane- achieving a better distribution of genistein among cancer cells. mic conditions commonly observed in patients with cancer may When ALA-induced PpIX is activated by visible light in tumor limit the use of such an iron chelator to locally. A treatment with cells, cytotoxic singlet oxygens are generated and these can kill the an ABCG2 inhibitor alone or in combination with ferrochelatase tumor cells. In the current study, cell death due to ALA-based inhibitors was shown to significantly increase the ALA-medi- photodynamic treatment increased 1.8-fold and PpIX accumula- ated accumulation of PpIX in human urothelial and oral tion increased 3.7-fold following 50 mmol/L genistein pretreat- squamous cell carcinoma in vitro (8, 19). Vitamin D3 has also ment. The mild increase in cytotoxicity despite a marked PpIX been reported to promote the ALA-mediated accumulation of increase suggests that genistein partially acts as an antioxidant PpIX by increasing the protein expression of the heme synthesis (45). In addition, genistein has been shown to upregulate the enzyme CPOX (13). In this context, our results demonstrated expression of antioxidant genes via ERs and ERK1/2 activation for the first time that genistein promoted the accumulation of (46). Thus, it is possible that some of the cytotoxic effects that are PpIX by regulating multiple key factors in the accumulation of mediated by ALA-based photodynamic treatment in the presence PpIX such as ABCG2, ALAD, PBGD, UROS, UROD, and PPOX of genistein may be cancelled by genistein itself. in A549. Furthermore, the combination of these approaches In the current study, the 48-hour genistein treatment was such as the genistein treatment with vitamin D3 may improve associated with an increase, rather than a decrease in ABCG2 ALA-induced PpIX accumulation more efficiently, which is the protein expression, despite the treatment markedly promoting the next subject to be explored. ALA-mediated accumulation of PpIX in A549 cells. Although the Genistein has a structural similarity to 17b-estradiol, which reason for this apparent discrepancy currently remains unknown, explains its estrogenic activity (32). Genistein binds to ERa and it may reflect the multifaceted functions of genistein. Namely, the ERb, and its affinity to ERb is similar to that of 17b-estradiol (22). synthesis of PpIX by genistein-induced heme synthesis enzymes Although the stimulation of ERb by its selective agonist 2,3-bis may override the efflux of PpIX by genistein-induced ABCG2 in (4-hydroxyphenyl) propionitrile was previously shown to pro- A549 cells and/or genistein may suppress preexisting and genis- mote the activation of ERK1/2 and cell growth in 201T human tein-induced ABCG2 protein levels.

OF8 Cancer Res; 76(7) April 1, 2016 Cancer Research

Genistein Promotes PpIX Accumulation

Disclosure of Potential Conﬂicts of Interest Administrative, technical, or material support (i.e., reporting or organizing No potential conﬂicts of interest were disclosed. data, constructing databases): H. Fujita, H. Kobuchi, Y. Kondo, T. Shuin, T. Utsumi, K. Utsumi, H. Ohuchi Authors' Contributions Study supervision: H. Fujita, K. Inoue, T. Shuin, J. Sasaki, H. Ohuchi Other (senior Author and I made an equal contribution): K. Nagakawa Conception and design: H. Fujita, T. Ogino, T. Shuin Development of methodology: K. Nagakawa Acquisition of data (provided animals, acquired and managed patients, The costs of publication of this article were defrayed in part by the provided facilities, etc.): H. Fujita, K. Nagakawa, H. Kobuchi, T. Shuin, payment of page charges. This article must therefore be hereby marked T. Utsumi, K. Utsumi advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate Analysis and interpretation of data (e.g., statistical analysis, biostatistics, this fact. computational analysis): H. Fujita, K. Nagakawa, T. Ogino, T. Shuin, H. Ohuchi Writing, review, and/or revision of the manuscript: H. Fujita, K. Nagakawa, Received June 3, 2015; revised October 22, 2015; accepted November 18, T. Ogino, Y. Kondo, T. Shuin, K. Utsumi, J. Sasaki, H. Ohuchi 2015; published OnlineFirst February 2, 2016.

References 1. Peng Q, Warloe T, Berg K, Moan J, Kongshaug M, Giercksky KE, et al. 5- 16. Inoue K, Fukuhara H, Kurabayashi A, Furihata M, Tsuda M, Nagakawa K, Aminolevulinic acid-based photodynamic therapy. Clinical research and et al. Photodynamic therapy involves an antiangiogenic mechanism and is future challenges. Cancer 1997;79:2282–308. enhanced by ferrochelatase inhibitor in urothelial carcinoma. Cancer Sci 2. Inoue K, Fukuhara H, Shimamoto T, Kamada M, Iiyama T, Miyamura M, 2013;104:765–72. et al. Comparison between intravesical and oral administration of 5- 17. Inoue K, Karashima T, Kamada M, Shuin T, Kurabayashi A, Furihata M, aminolevulinic acid in the clinical benefit of photodynamic diagnosis for et al. Regulation of 5-aminolevulinic acid-mediated protoporphyrin IX nonmuscle invasive bladder cancer. Cancer 2012;118:1062–74. accumulation in human urothelial carcinomas. Pathobiology 2009; 3. Fukuhara H, Inoue K, Satake H, Tamura K, Karashima T, Yamasaki I, et al. 76:303–14. Photodynamic diagnosis of positive margin during radical prostatectomy: 18. Kobuchi H, Moriya K, Ogino T, Fujita H, Inoue K, Shuin T, et al. Mito- preliminary experience with 5-aminolevulinic acid. Int J Urol 2011;18: chondrial localization of ABC transporter ABCG2 and its function in 5- 585–91. aminolevulinic acid-mediated protoporphyrin IX accumulation. PloS One 4. Friesen SA, Hjortland GO, Madsen SJ, Hirschberg H, Engebraten O, 2012;7:e50082. Nesland JM, et al. 5-Aminolevulinic acid-based photodynamic detec- 19. Ogino T, Kobuchi H, Munetomo K, Fujita H, Yamamoto M, Utsumi T, et al. tion and therapy of brain tumors (review). Int J Oncol 2002;21: Serum-dependent export of protoporphyrin IX by ATP-binding cassette 577–82. transporter G2 in T24 cells. Mol Cell Biochem 2011;358:297–307. 5. Messmann H, Mlkvy P, Buonaccorsi G, Davies CL, MacRobert AJ, Bown SG. 20. Vail JTJr. Porphyria cutanea tarda and estrogens. JAMA 1967;201: Enhancement of photodynamic therapy with 5-aminolaevulinic acid- 671–4. induced porphyrin photosensitisation in normal rat colon by threshold 21. Gibson SL, Anderson LT, Havens JJ, Hilf R. Effect of estrogenic perturba- and light fractionation studies. Br J Cancer 1995;72:589–94. tions on delta-aminolevulinic acid-induced porphobilinogen deaminase 6. Teng L, Nakada M, Zhao SG, Endo Y, Furuyama N, Nambu E, et al. Silencing and protoporphyrin IX levels in rat Harderian glands, liver, and R3230AC of ferrochelatase enhances 5-aminolevulinic acid-based fluorescence and tumors. Biochem Pharmacol 1999;58:1821–9. photodynamic therapy efficacy. Br J Cancer 2011;104:798–807. 22. Kuiper GG, Lemmen JG, Carlsson B, Corton JC, Safe SH, van der Saag PT, 7. Togsverd-Bo K, Lerche CM, Philipsen PA, Poulsen T, Wulf HC, Haedersdal et al. Interaction of estrogenic chemicals and phytoestrogens with estrogen M. Porphyrin biodistribution in UV-exposed murine skin after methyl- and receptor beta. Endocrinology 1998;139:4252–63. hexyl-aminolevulinate incubation. Exp Dermatol 2012;21:260–4. 23. Ming LG, Chen KM, Xian CJ. Functions and action mechanisms of 8. Yamamoto M, Fujita H, Katase N, Inoue K, Nagatsuka H, Utsumi K, et al. flavonoids genistein and icariin in regulating bone remodeling. J Cell Improvement of the efficacy of 5-aminolevulinic acid-mediated photody- Physiol 2013;228:513–21. namic treatment in human oral squamous cell carcinoma HSC-4. Acta Med 24. Prossnitz ER, Barton M. The G-protein-coupled estrogen receptor GPER in Okayama 2013;67:153–64. health and disease. Nat Rev Endocrinol 2011;7:715–26. 9. Zhao SG, Chen XF, Wang LG, Yang G, Han DY, Teng L, et al. Increased 25. Imai Y, Tsukahara S, Asada S, Sugimoto Y. Phytoestrogens/flavonoids expression of ABCB6 enhances protoporphyrin IX accumulation and reverse breast cancer resistance protein/ABCG2-mediated multidrug resis- photodynamic effect in human glioma. Ann Surg Oncol 2013;20:4379–88. tance. Cancer Res 2004;64:4346–52. 10. Ishizuka M, Abe F, Sano Y, Takahashi K, Inoue K, Nakajima M, et al. Novel 26. Beral V. Breast cancer and hormone-replacement therapy in the Million development of 5-aminolevurinic acid (ALA) in cancer diagnoses and Women Study. Lancet 2003;362:419–27. therapy. Int Immunopharmacol 2011;11:358–65. 27. Beral V, Bull D, Green J, Reeves G. Ovarian cancer and hormone replace- 11. Kuo MT. Redox regulation of multidrug resistance in cancer chemotherapy: ment therapy in the Million Women Study. Lancet 2007;369:1703–10. molecular mechanisms and therapeutic opportunities. Antioxid Redox 28. Lucki NC, Sewer MB. Genistein stimulates MCF-7 breast cancer cell growth Signal 2009;11:99–133. by inducing acid ceramidase (ASAH1) gene expression. J Biol Chem 12. Doss M, Sixel-Dietrich F, Verspohl F. "Glucose effect" and rate limiting 2011;286:19399–409. function of uroporphyrinogen synthase on porphyrin metabolism in 29. Berg K, Anholt H, Bech O, Moan J. The influence of iron chelators on the hepatocyte culture: relationship with human acute hepatic porphyrias. accumulation of protoporphyrin IX in 5-aminolaevulinic acid-treated cells. J Clin Chem Clin Biochem 1985;23:505–13. Br J Cancer 1996;74:688–97.

13. Anand S, Wilson C, Hasan T, Maytin EV. Vitamin D3 enhances the 30. Lin F, Geiger PG, Korytowski W, Girotti AW. Protoporphyrin IX-sensitized apoptotic response of epithelial tumors to aminolevulinate-based photo- photoinactivation of 5-aminolevulinate-treated leukemia cells: effects of dynamic therapy. Cancer Res 2011;71:6040–50. exogenous iron. Photochem Photobiol 1999;69:375–81. 14. Amo T, Kawanishi N, Uchida M, Fujita H, Oyanagi E, Utsumi T, et al. 31. Tan WC, Krasner N, O'Toole P, Lombard M. Enhancement of photody- Mechanism of cell death by 5-aminolevulinic acid-based photodynamic namic therapy in gastric cancer cells by removal of iron. Gut 1997;41:14–8. action and its enhancement by ferrochelatase inhibitors in human histio- 32. Mahmoud AM, Yang W, Bosland MC. Soy isoﬂavones and prostate cancer: cytic lymphoma cell line U937. Cell Biochem Funct 2009;27:503–15. a review of molecular mechanisms. J Steroid Biochem Mol Biol 15. Fukuhara H, Inoue K, Kurabayashi A, Furihata M, Fujita H, Utsumi K, et al. 2014;140:116–32. The inhibition of ferrochelatase enhances 5-aminolevulinic acid-based 33. Hershberger PA, Stabile LP, Kanterewicz B, Rothstein ME, Gubish CT, Land photodynamic action for prostate cancer. Photodiagnosis Photodyn Ther S, et al. Estrogen receptor beta (ERbeta) subtype-speciﬁc ligands increase 2013;10:399–409. transcription, p44/p42 mitogen activated protein kinase (MAPK)

www.aacrjournals.org Cancer Res; 76(7) April 1, 2016 OF9

Fujita et al.

activation and growth in human non-small cell lung cancer cells. J Steroid 41. Wu TC, Yang YC, Huang PR, Wen YD, Yeh SL. Genistein enhances Biochem Mol Biol 2009;116:102–9. the effect of trichostatin A on inhibition of A549 cell growth by 34. Verma MK, Miki Y, Abe K, Niikawa H, Sasano H. Cytoplasmic estrogen increasing expression of TNF receptor-1. Toxicol Appl Pharmacol receptor beta as a potential marker in human non-small cell lung carci- 2012;262:247–54. noma. Expert Opin Ther Targets 2012;16:S91–102. 42. Hu M, Krausz K, Chen J, Ge X, Li J, Gelboin HL, et al.et al. Identification of 35. Shen L, Li Z, Shen S, Niu X, Yu Y, Li Z, et al. The synergistic effect of EGFR CYP1A2 as the main isoform for the phase I hydroxylated metabolism of tyrosine kinase inhibitor gefitinib in combination with aromatase inhib- genistein and a prodrug converting enzyme of methylated isoflavones. itor anastrozole in non-small cell lung cancer cell lines. Lung Cancer Drug Metab Dispos 2003;31:924–31. 2012;78:193–200. 43. Kulling SE, Honig DM, Metzler M. Oxidative metabolism of the soy 36. Tang H, Liao Y, Chen G, Xu L, Zhang C, Ju S, et al. Estrogen upregulates the isoflavones daidzein and genistein in humans in vitro and in vivo. J Agr Food IGF-1 signaling pathway in lung cancer through estrogen receptor-beta. Chem 2001;49:3024–33. Med Oncol 2012;29:2640–8. 44. Cho K, Wang X, Nie S, Chen ZG, Shin DM. Therapeutic nano- 37. Elder GH. Porphyria cutanea tarda. SemLiver Dis 1998;18:67–75. particles for drug delivery in cancer. Clin Cancer Res 2008;14: 38. Urbanek RW, Cohen DJ. Porphyria cutanea tarda: pregnancy versus 1310–6. estrogen effect. J Am Acad Dermatol 1994;31:390–2. 45. Wei H, Wei L, Frenkel K, Bowen R, Barnes S. Inhibition of tumor promoter- 39. Phillips JD, Jackson LK, Bunting M, Franklin MR, Thomas KR, Levy JE, et al. induced hydrogen peroxide formation in vitro and in vivo by genistein. Nutr A mouse model of familial porphyria cutanea tarda. Proc Natl Acad Sci USA Cancer 1993;20:1–12. 2001;98:259–64. 46. Borras C, Gambini J, Gomez-Cabrera MC, Sastre J, Pallardo FV, Mann GE, 40. Shiau RJ, Chen KY, Wen YD, Chuang CH, Yeh SL. Genistein and beta- et al.et al. Genistein, a soy isoflavone, up-regulates expression of antiox- carotene enhance the growth-inhibitory effect of trichostatin A in A549 idant genes: involvement of estrogen receptors, ERK1/2, and NFkappaB. cells. Eur J Nutr 2010;49:19–25. FASEB J 2006;20:2136–8.

OF10 Cancer Res; 76(7) April 1, 2016 Cancer Research

Phytoestrogen Suppresses Efflux of the Diagnostic Marker Protoporphyrin IX in Lung Carcinoma

Hirofumi Fujita, Keisuke Nagakawa, Hirotsugu Kobuchi, et al.

Cancer Res Published OnlineFirst February 2, 2016.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-15-1484

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2021/03/04/0008-5472.CAN-15-1484.DC1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/early/2016/03/22/0008-5472.CAN-15-1484. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.