Redox Signaling and Bioenergetics Influence Lung Cancer Cell …jpet.aspetjournals.org/content/jpet/358/2/199.full.pdf · ME-344 [(3R,4S)-3,4-bis(4-hydroxyphenyl)-8-methyl-3,4-dihydro-2H-chromen-7-ol]

1521-0103/358/2/199–208$25.00 http://dx.doi.org/10.1124/jpet.115.229344THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS J Pharmacol Exp Ther 358:199–208, August 2016Copyright ª 2016 by The American Society for Pharmacology and Experimental Therapeutics

Redox Signaling and Bioenergetics Influence Lung Cancer CellLine Sensitivity to the Isoflavone ME-344

Yefim Manevich, Leticia Reyes, Carolyn D. Britten, Danyelle M. Townsend,and Kenneth D. TewDepartments of Cell and Molecular Pharmacology and Experimental Therapeutics (Y.M., L.R., K.T.), Medicine (C.B.), and DrugDiscovery and Biomedical Sciences (D.T.) of the Medical University of South Carolina, Charleston, South Carolina

Received September 15, 2015; accepted May 26, 2016

ABSTRACTME-344 [(3R,4S)-3,4-bis(4-hydroxyphenyl)-8-methyl-3,4-dihydro-2H-chromen-7-ol] is a second-generation derivative naturalproduct isoflavone presently under clinical development.ME-344 effects were compared in lung cancer cell lines thatare either intrinsically sensitive or resistant to the drug and inprimary immortalized human lung embryonic fibroblasts (IHLEF).Cytotoxicity at low micromolar concentrations occurred only insensitive cell lines, causing redox stress, decreased mitochon-drial ATP production, and subsequent disruption of mitochon-drial function. In a dose-dependent manner the drug causedinstantaneous and pronounced inhibition of oxygen consump-tion rates (OCR) in drug-sensitive cells (quantitatively signifi-cantly less in drug-resistant cells). This was consistent withtargeting of mitochondria by ME-344, with specific effects on

the respiratory chain (resistance correlated with higher glyco-lytic indexes). OCR inhibition did not occur in primary IHLEF.ME-344 increased extracellular acidification rates in drug-resistant cells (significantly less in drug-sensitive cells), imply-ing that ME-344 targets mitochondrial proton pumps. Only indrug-sensitive cells did ME-344 dose-dependently increasethe intracellular generation of reactive oxygen species andcause oxidation of total (mainly glutathione) and protein thiolsand the concomitant immediate increases in NADPH levels.We conclude that ME-344 causes complex, redox-specific,and mitochondria-targeted effects in lung cancer cells, whichdiffer in extent from normal cells, correlate with drug sensitivity,and provide indications of a beneficial in vitro therapeuticindex.

IntroductionNatural product isoflavones have been studied as agents

useful in cancer prevention (de Souza et al., 2010; Leclercq andJacquot, 2014). However, one isoflavone, phenoxodiol, hascytotoxic antitumor activities more characteristic of a thera-peutic agent; consequently, consideration has been given as tohow this agent and some of its derivatives might induce celldeath (Sapi et al., 2004; Aguero et al., 2005; Silasi et al., 2009).Mechanistically, proapoptotic effects may result from inter-ference with mitochondrial functions (Kamsteeg et al., 2003)and interference with plasma membrane electron transportand/or cell surface thiol: disulfide exchange reactions (Morre

and Morre, 2003; Herst et al., 2007). Downstream of suchevents, phenoxodiol derivatives can activate caspase-dependentand independent apoptotic pathways, perhaps through proteinkinase B (Akt) signaling events (Wang et al., 2011; Pant et al.,2014). Related isoflavones alter mitochondrial membranepotential, deplete ATP, reduce steady-state levels of cyto-chrome c oxidase (complex IV of electron transport chain), andmight eventually inhibit mammalian target of rapamycin(mTOR) activity (Alvero et al., 2011). One commonality ofthese effects is interference with mitochondrial energy pro-duction. Related to phenoxodiol, ME-344 [(3R,4S)-3,4-bis(4-hydroxyphenyl)-8-methyl-3,4-dihydro-2H-chromen-7-ol] (Fig.1) is a second-generation derivative of triphendiol currentlyunder development as a clinical therapeutic drug candidatein both small-cell lung and ovarian cancers. ME-344 is thelead compound selected from a number of analogs, wherestructure–activity relationship analyses have revealed thatthe ring hydroxyl groups provides characteristics that areprerequisite for cytotoxicity (Silasi et al., 2009).Initial in vitro cytotoxicity screens demonstrated that only

20 of 240 human tumor cell lines (mostly solid tumor origins)were intrinsically (naturally) resistant to the effects ofME-344.Within the context of the present study, this provided

This work was supported by grants from the National Institutes of HealthNational Cancer Institute [Grants CA08660, CA117259] and National Centerfor Research Resources [Grant NCRR P20RR024485], COBRE in Oxidants,Redox Balance and Stress Signaling] and support from the South CarolinaCenters of Excellence program and was conducted in a facility constructed withthe support from the National Institutes of Health [Grant RR015455] from theExtramural Research Facilities Program of the National Center for ResearchResources. C.D.B. and K.D.T. hold endowed Chairs from the South CarolinaSmartState program. Supported in part by the Drug Metabolism and ClinicalPharmacology shared Resource, Hollings Cancer Center, Medical University ofSouth Carolina. Further financial support was through an unrestricted grantfrom MEI Pharma Inc.

dx.doi.org/10.1124/jpet.115.229344.

ABBREVIATIONS: AFC, 7-amino-4-trifluoromethyl coumarin; DMSO, dimethylsulfoxide; ECAR, extracellular acidification rates; GSSG, glutathionedisulfide; HPLC, high-performance liquid chromatography; IHLEF, immortalized human lung embryonic fibroblasts; IMM, inner mitochondrialmembrane; ME-344, (3R,4S)-3,4-bis(4-hydroxyphenyl)-8-methyl-3,4-dihydro-2H-chromen-7-ol; MPA, metaphosphoric acid; NNT, nicotinamidenucleotide transhydrogenase; OCR, oxygen consumption rates; PBS, phosphate-buffered saline; ROS, reactive oxygen species.

199

at ASPE

T Journals on M

arch 31, 2021jpet.aspetjournals.org

Dow

nloaded from

http://dx.doi.org/10.1124/jpet.115.229344http://dx.doi.org/10.1124/jpet.115.229344http://jpet.aspetjournals.org/

the opportunity to compare the effects of ME-344 in these celllines and compare with an immortalized cell line of lungfibroblast origin. Such an approach obviates the complexity ofusing drug selection conditions to derive acquired resistantclones and facilitates the identification of traits that might beimportant to themechanism of action ofME-344. In turn, suchinformation should have relevance to how the drug is clinicallyapplied. Our results indicate that ME-344 impacts redoxhomeostasis and may act in mitochondria through interfer-ence with pathways that regulate switching from oxidativephosphorylation to glycolysis. Our present data implicatecertain proteins and pathways that may be linked with thedrug’s mechanism of action.

Materials and MethodsTissue Culture. SHP-77–small cell lung carcinoma (epithelial

cells, sensitive) and H-460 large cell lung cancer (epithelial cells,sensitive) cells were purchased from the National Cancer Institute(Frederick, MD) and cultured according to supplier recommendations.Primary immortalized human lung embryonic fibroblasts (MRC-5),SW 900 grade IV squamous cell lung carcinoma (epithelial cells,resistant), and NCI-H596 adenosquamous lung carcinoma (epithelialcells, resistant) were purchased from American Type Culture Collec-tion (Manassas, VA) and cultured according to the supplier recom-mendations. A panel of 240 cell lines was initially screened by MEIPharma (San Diego, CA), and the present lines were chosen torepresent natural (intrinsic) sensitivity or resistance to ME-344.

Seahorse Bioscience Experiments. For standard and glycolyticmitochondrial stress test experiments, 96-well plates (XF96; SeahorseBioscience/Agilent, Santa Clara, CA) with attached confluent cellswere used in the Seahorse Bioscience apparatus in theMUSCCOBREMetabolomics Core. Standard protocols for OCR and extracellularacidification rates (ECAR) recording were used. Incubation time andME-344 concentration dependence of OCR and ECAR were analyzed.The results were normalized for cell number in each well andpresented as mean 6 S.E., for n 5 8.

Cell Viabilities and Fluorometric Detection of Caspase-3Protease. Weused a standardCaspase-3/CPP32 Fluorometric AssayKit from BioVision (Milpitas, CA) to study the effects of ME-344induction of apoptosis in lung cancer cells and immortalized lung

fibroblasts. Cells were grown in standard Petri dishes (60 mm)to ∼80% confluence, and the attached cells (control: dimethylsulfoxide[DMSO], or ME-344 treated) were washed with phosphate-bufferedsaline (PBS) 3 times, lyzed, combined in reaction buffer and fluores-cent substrate DEVD-AFC (caspase-3 substrate) and incubated at37°C for 1.5 hours. Finally, samples were placed in a quartz cuvette(5� 5� 40mm), and the intensity of the DEVD-AFC cleavage product(AFC: 7-amino-4-trifluoromethyl coumarin, Ex/Em 5 400/505 nm)was recorded using a standard kinetic mode (resolution, 0.1 second) ofthe QM-4 fluorometer for 200 seconds. The resultant emissions wereaveraged and plotted as mean 6 S.D., n 5 400. We recorded thekinetics of caspase-3 activation during incubation of cells withME-344for 0, 2, 4, and 6 hours. Each time point was measured in triplicate,and the final results are presented as mean 6 S.D.

Viability after incubation with ME-344 for 0, 2, 4, and 6 hours at37°C were assessed. After incubation, the cells were washed with PBSand detached from Petri dishes with trypsin, optimized for each celltype. After washing cells with fresh media, their counts/viabilitieswere determined using a Cellometer Auto T4 (Nexcelom Bioscience,Lawrence, MA) and trypan blue exclusion. All experiments were intriplicate and results presented as mean 6 S.D.

Real-Time Kinetic Detection of Intracellular ReactiveOxygen Species and NADPH. For reactive oxygen species (ROS)/NADPH fluorescent analyses, cells were grown as confluent mono-layers adherent to Aclar plastic slides (14� 25mm). These slides wereplaced in a quartz cuvette (10� 10� 40mm)with PBS (with 100mMofCaCl2, 37°C) at a 45° angle to excitation light and the dynamics ofNADPH emission (under constant stirring) were detected (Ex 340, Em460 nm) before and after ME-344 (or other effectors) in real-timekinetics with a resolution of 0.1 second using aQM-4 fluorometer (PTI,Edison, NJ) (Manevich et al., 2002). In other experiments, adheredcells were labeled with specific dyes (ROS measurements) andsimilarly analyzed in the QM-4 fluorometer. Intracellular O2

*2 wasmonitored with CellTracker Deep Red fluorescent dye (Ex/Em,630/650 nm; Life Technologies, Carlsbad, CA), intracellular H2O2generation with H2DCF-DA dye (Sigma-Aldrich, St. Louis, MO), and*OH radical with 3-CCA-DA dye (Manevich et al., 1997). Live cellsattached to Aclar slides were incubated in complete media with 5 mMof the appropriate dye for 45minutes at 37°C. Slides with labeled cellswere washed 3 times with PBS to eliminate excess dye before placingthem in a quartz cuvette for data recording.

High-Performance Liquid Chromatography Analyses ofATP, ADP, and AMP. Incubation medium from cell cultures insix-well plastic plates was replaced with 5% metaphosphoric acid(MPA) at 4°C. After 3 minutes, the plates were scraped, and theextracts centrifuged at 15,000g for 5minutes at 4°C. The supernatantswere then neutralized with 0.1 M Tris-acetate buffer. Adeninenucleotides in the extracts were separated using a Model 1525 BinaryBreeze high-performance liquid chromatography (HPLC) pump equip-ped with a Model 717 Plus Autosampler and detected using a Model2487 UV-Vis Detector (all from Waters, Milford, MA). Experimentswere performed using a C18, 5 mm, 4.6 � 150-mm reverse phasecolumn (SunFire; Waters), isocratic elution (1 ml/min) with 100 mMNa-phosphate buffer (pH 5.5) and UV detection at 260 nm.

The MPA lysates (25 ml) were injected into the column and ATP,ADP, and AMP were detected at approximate retention times of 4.1,4.9, and 11.5 minutes, respectively (Maldonado et al., 2013). Authen-ticity of detection was confirmed by spiking actual samples withknown amounts of ATP, ADP, and AMP as standards. Appropriatepeak areas were quantified using Empower 2 software (Waters), andquantification of individual compounds was achieved with the use ofparticular calibration curves with purified ATP, ADP, and AMPstandards.

HPLC Analysis of Glutathione Disulfide. Attached cells wereincubated with ice-cold 5% (final concentration) MPA. After 3 to5 minutes, the cells were scraped, and the extracts centrifuged at15,000g for 5 minutes at 4°C. The supernatants were collected andused for analysis. Glutathione disulfide (GSSG) in the extracts was

Fig. 1. Structure of ME-344.

200 Manevich et al.

at ASPE

T Journals on M


Dow

nloaded from

http://jpet.aspetjournals.org/

detected using an HPLC method similar to that described for ATP(ADP, AMP) analyses. The supernatants from cell lysates treatedwithMPA (25 ml) were injected into the column, and the GSSG peak wasdetected (absorbance at 260 nm) at a retention time of ∼2.85 minutes.Authenticity of detection was confirmed by spiking actual sampleswith known amounts of GSSG as standards. Peak areas werequantified using Empower 2 software (Waters), and quantification ofGSSG was achieved using calibration curves with purified GSSG.

Fluorescent Analysis of Low-Molecular-Weight and ProteinReduced Thiols. Detection and quantification of sulfhydryls is basedon their specific reaction with the maleimide probe TG-1 (Calbiochem,San Diego, CA), which becomes fluorescent (Bowers et al., 2012).Cell lysates were passed through a size-exclusion chromatographymicrocolumn (Bio-Spin 6, cutoff ∼6 kDa; Bio-Rad Laboratories,Hercules, CA) to separate proteins from lower molecular weightthiols. Both size-exclusion chromatography processed and intactsamples were diluted in a quartz cuvette 1:100 (by volume) with20 mM PB (pH 7.4) at 37°C under constant stirring in a sampleholder of the QM-4 fluorometer. After addition of TG-1 (DMSOsolution, final concentration 5 mM), the fluorescent detection(Ex 379 nm, Em 513 nm) of labeled sulfhydryls was performedusing a standard kinetic mode of the QM-4 to reach saturation.Where pertinent, background fluorescence’s of cell lysates, effectsof DMSO and background fluorescence of TG-1 in 20 mM PB (pH 7.4at 37°C) were subtracted from the final results. The final emissionvalues were averaged for 50 seconds (500 time points) using Sigma

Plot 10.0 software (Systat Software, San Jose, CA) and presented asthe mean 6 S.D.

Fluorescent Analysis of Cell Surface Thiols. Either lungcancer or immortalized primary (MRC-5) cells attached to Aclar plasticslides (∼85% confluent) were treated with ME-344 for 30 minutes incomplete media. After washing these slides in PBS (with 100 mM ofCaCl2) they were placed in a quartz cuvette with PBS (with 100 mMCaCl2) at 4°C (to prevent TG-1 endocytosis) similar to that for NADPHdetection. Both before and after addition of 5 mMTG-1, emissions weredetected in real-time kinetics until saturation. Background fluorescencefor the cells and TG-1 were each subtracted from the final results.Samples were also corrected for DMSO and the final TG-1 fluorescencewas normalized for cell number and presented asmean6 S.D. for threeindependent experiments. After those measurements, cells were de-tached from the slides and their counts/viabilities estimated usingCellometer Auto T4 (Nexcelom Bioscience LLC, Lawrence, MA) andtrypan blue exclusion. Cell viabilities under our experimental condi-tions were very close (98% to 99%) before and after treatment withME-344 (30 minutes) and TG-1.

Mitochondrial Membrane Depolarization Measurements inReal-Time Kinetics. Cells on the Aclar plastic slides (∼85% conflu-ent) were incubated with 5.0 mM of JC-9 fluorescent dye (LifeTechnology, CA) in complete media for 45 minutes at 37°C. Labeledcells were washed 3 times with PBS containing 100 mM CaCl2 andplaced in a quartz cuvette similar to the above measurements forNADPH dynamics. JC-9 fluorescence was measured as a ratio of

Fig. 2. Standard mitochondrial stress test traces for lung cancer cells. (A) OCR and (B) ECAR changes upon addition of: ME-344 (4.0 mg/ml),O (oligomycin, 0.5 mM), F (FCCP [p-triflouromethoxyphenylhydrazone], 1.0 mM), and A + R (antimycin A, 1.0 mM + rotenone, 1.0 mM) to lung cancer H460(sensitive) and H596 (resistant) cells in standard Seahorse 96-well microplates. The traces are representative of six independent experiments. (The resultswere essentially the same for SHP-77 [sensitive] and SW900 [resistant] lung cancer cells.) Data were normalized for cell numbers. (C) ME-344 effects on ATPsynthase inhibition by oligomycin. Presented as a difference (%) between OCR with or without ME-344 before and after oligomycin. Data are normalized forvalues before oligomycin addition, and they represent the mean6 S.D. for three independent experiments. &,«,#Statistical significance P# 0.05. (D) ME-344dose effects onOCR changes (% of initial OCRwithout effector). The datawere averaged for either sensitive or resistant cells and are presented asmean6S.E.for six independent experiments, P # 0.009.

Redox Active Isoflavone in Lung Cancer 201

at ASPE

T Journals on M


Dow

nloaded from


(485/529 nmmonomeric dye fluorescence) to (535/590 nm J-aggregatefluorescence) in real-time kinetics before and after addition of in-dicated amounts of those agents shown (ME-344 and glucose).

Statistical Analysis. We used SigmaStat 3.5 (Systat Software)for statistical analysis of data. One-way analysis of variance (ANOVA)was used for statistical relevance of data comparison.

ResultsRecent work has implicated inhibition of mitochondrial

complex I as a possible contributory mechanism for this classof drugs (Lim et al., 2015). We now provide further interroga-tion through SeaHorse technology and mitochondrial-stresstesting measurements of the effects of ME-344 (Fig. 1) onoxygen consumption and extracellular acidification rates(OCR and ECAR) in drug-sensitive (SHP-77 and H460) andresistant (H596 and SW900) lung cancer and IHLEF (MRC-5)cell lines.Mitochondrial Stress in Lung Cancer Cells. ME-344

caused a substantial and essentially instantaneous inhibition ofOCR (similar in extent to oligomycin) in the lung cancer cell lines,correlating with the natural sensitivity of cells to the drug.Application of an uncoupler p-triflouromethoxyphenylhydrazone(FCCP) caused a fast recovery in the drug-resistant lung cancercells, but was substantially retarded in the sensitive counterpart(Fig. 2A). Subsequent addition of antimycin A and rotenone

completely inhibited respiration, with the effectmore pronouncedin resistant cells. Mitochondrial stress testing of the same cellsrevealed a substantial increase of ECAR in resistant cells afterME-344 (Fig. 2B), an effect thatwas barely detectable in sensitivecells. The impact of other standard stressors on ECAR wassimilar for both sensitive and resistant cell lines (Fig. 2B).A further interesting difference was found between the

resistant and sensitive cells. While comparing the effects ofoligomycin on cells treated with either DMSO or ME-344, wemeasured OCR inhibition and normalized it for the OCR valuebefore oligomycin addition. Pretreatment of cells withME-344inhibited the effects of subsequent oligomycin, but only in thesensitive cell lines. In the resistant cells, ME-344 was shownto potentiate the effects of oligomycin. (Fig. 2C). The effects ofME-344 on OCR were concentration-dependent and reachedsaturation at ∼57.4 mM (ED50 ∼11.5 mM, Fig. 2D).Glycolytic Mitochondrial Stress Test. We next studied

the effect of ME-344 on glycolysis in the lung cancer cells.Pretreatment of sensitive lung cancer cells resulted in anincrease of ECAR (Fig. 3A) after addition of glucose. Interest-ingly the effects of oligomycin treatment were suppressed insensitive lung cancer cells following pretreatment withME-344 (Fig. 3B).Initial ECAR values under conditions of glucose deprivation

were lower in IHLEF (MRC-5) cells (Fig. 3B) compared withresistant (averaged H596 and SW900) lung cancer cell lines

Fig. 3. Effects of ME-344 on glycolysis in lung cancer and IHLEF cells. (A) Glycolysis stress tests were presented as ECAR changes upon addition of:ME-344 (11.48 mM), Glu (glucose, 10.0 mM), O (oligomycin, 1.0 mM), and 2-DG (2-deoxy-glucose, 10.0 mM) using SHP-77 lung cancer cells (sensitive). Thetrace is representative of six independent experiments. (B) Averaged differences between ME-344 (11.48 mM) and DMSO (same volume)-treatedsensitive (H460 and SHP-77), IHLEF MRC-5, and resistant (H596 and SW900) cells. The differences are statistically significant only where sensitivecells are compared with either resistant or IHLEF cells (P # 0.001). Data were normalized for cell numbers. (C) ME-344 dose effects on glycolysis(percentage of initial ECAR changes after glucose addition, averaged for sensitive [H460, SHP-77] cells). (D) ME-344 dose effects on glycolysis(percentage of initial ECAR changes after addition of glucose) averaged for IHLEF (MRC-5) and resistant (H596, SW900) cells. All data are presented asmean 6 S.D. for six independent experiments and are statistically different (P # 0.001).

202 Manevich et al.

at ASPE

T Journals on M


Dow

nloaded from


(Fig. 3B). While ME-344 decreased the glycolytic response inIHLEFMRC-5 cells (Fig. 3B), in resistant cells, therewasagreaterdecrease (Fig. 3B). Subsequent treatments with oligomycincaused a pronounced inhibitory response in both sensitive andthe primary MRC-5 cells and an increased response inresistant cells (Fig. 3B). The effects of ME-344 on glycolysiswere concentration dependent (Fig. 3C) in the sensitive cellsand reached saturation at ∼57.4 mM (ED50 ∼11.5 mM) a valuesimilar to that for OCR. Our data also suggest that ME-344had dose dependent effects upon glycolytic stress that weredifferent between the immortalized primary cells and drug-resistant cell lines (ED50 ∼11.5 mM, Fig. 3D).Bioenergetic and Redox Effects in Cancer and

Immortalized Primary Cells. Following ME-344, ATPlevels were decreased ∼12% in sensitive cells, with essen-tially no depletion in resistant cells (Fig. 4A). ME-344 alsomildly altered redox stress in sensitive compared withresistant cells (as detected by GSSG levels; Fig. 4B). InitialATP to ADP hydrolysis was 2-fold lower in primary cells(Fig. 4C) when compared with sensitive cells (Fig. 4D). ATPto ADP hydrolysis and redox stress (GSSG) were practicallyunaffected by ME-344 in primary cells (Fig. 4C) but theformer was dose-dependently facilitated in sensitive lungcancer cells (Fig. 4D). These data again imply a degree ofspecificity for ME-344 effects in sensitive lung cancer cells.Cytotoxicity and Apoptosis. Cytotoxicity of ME-344 was

assessed through its effects on apoptosis and viability. Thedrug rapidly (2 hours) induced apoptosis in the sensitive cell

lines (H460 and SHP-77; Fig. 5A) through an activation ofcaspase-3 in a time dependent manner. In contrast, there wasminimal (if any) induction of apoptosis in the IHLEF (MRC-5)and resistant cells. Figure 5B shows that progressive death ofsensitive cells occurred after ME-344, but in resistant cellsand IHLEF (MRC-5) these effects were minimal. Figure 5Cshows the dose effects of ME-344 on caspase-3 activation in allcell types. The ED50 of ME-344-mediated caspase-3 activationoccurred at 2.13 mM. Figure 5D shows viability of sensitive,resistant lung cancer and IHLEF after 24 hours incubationwith 5.74 and 11.48 mM of ME-344. Approximately 90% of thesensitive cells were killed, but only minimal cytotoxicity wasapparent in resistant and IHLEF cells.ME-344 Effects on Intracellular Oxidant Stress. ME-

344-mediated inhibition of OCR may contribute to oxidativestress through impairment of the mitochondrial electrontransfer chain. ME-344 induced pronounced, concentration-dependent generation of H2O2 and *OH in sensitive cells(SHP-77 and H460; ED50 ∼11.5 mM, Fig. 6, A–C). This effectwas not observed in resistant cells and wasminimal in IHLEF(Fig. 6, B and D), implying some degree of specificity ofME-344 toward the sensitive lung cancer cells. ME-344treatments did not lead to superoxide anion radical generationin either lung cancer cells or IHLEF (Fig. 7A).ME-344 Effects on General Redox Stress. NADPH is a

universal intracellular source of electrons for redox reactions.Both NADH (mainly mitochondrial) and NADPH (mainlycytosolic) have indistinguishable biofluorescences. Dynamic

Fig. 4. Effects of ME-344 on bioenergetics and redox stress in IHLEF and lung cancer cells. HPLC analyses of (A) ATP and (B) GSSG in intact (DMSO,black bars) andME-344 treated (11.48 mM,white bars) sensitive (SENS) or resistant (RES) cell lysates. Data were averaged for sensitive (H460, SHP-77)or resistant (H596, SW900) cells and presented as mean6 S.D. for three independent experiments. HPLC analyses of ME-344 dose effects on ATP/ADPratios (black bars) and GSSG (normalized for its value before treatment, gray bars) in (C) IHLEF (MRC-5) or (D) sensitive cells (H460) are presented asmean 6 S.E. for three independent experiments. ATP/ADP changes in sensitive (H460) cells were statistically different (P # 0.05).


at ASPE

T Journals on M


Dow

nloaded from


changes in NADPH fluorescence frequently accompany activeredox signaling and have been used previously for evalua-tion of drugs that impact intermediate metabolism (Biaglowet al., 1998). Our present data indicate that under conditionsof glucose deprivation, ME-344 treatment of sensitive cells(H460) caused a rapid increase and subsequent slow de-crease of NADPH to below normal resting levels (Fig. 7B,red trace). As detected by fluorescence measurements, sub-sequent addition of glucose to the same cells resulted in arapid increase in the generation and cycling of NADPH (Fig.7B, red trace). A similar dose of glucose given before ME-344did not alter NADPH, but further addition of ME-344induced a similar rapid increase and subsequent cycling ofNADPH fluorescence (Fig. 7B, green trace). Such effects didnot occur in resistant (H596, Fig. 7B, black trace) or IHLEF(not shown).Because proton fluxes control mitochondrial functions, we

measured the dynamics of inner mitochondrial membranepotential (IMM) changes (DCmit) in sensitive (H460), resistant(SW900) cells and IHLEF using a ratiometric fluorescent dyeJC-9 (Fig. 7, C and D). Our data showed rapid hyperpolariza-tion of IMM immediately after ME-344 addition (under glucosedeprivation conditions), which was significant (∼12%) andsustained in sensitive (H460) cells (Fig. 7C). In IHLEF andresistant cells the effects were smaller (∼5%), and hyperpolar-izationwas followed by rapid depolarization of IMM,whichwasslower and of smaller intensity in resistant cells (Fig. 7D).

In the cancer cells the addition of glucose after ME-344caused a fast, sustained depolarization of IMM almost up toresting levels (Fig. 7C). Similar treatments of IHLEF causedslower depolarization of IMM, exceeding resting levels by∼6%and further hyperpolarization of IMM in resistant cells (Fig.7D). These results again suggest that ME-344 has somedegree of specificity for effecting mitochondrial proton fluxesin sensitive cells.ME-344 Effects on Intra- and Extracellular Redox

States. The dynamics of ME-344-mediated induction of oxi-dant stress (H2O2 and *OH, Fig. 6, A and C) were furtherinvestigated using fluorescent detection of intracellular thioloxidation. The data showa progressive increase of redox stress(corresponding with decreases in reduced thiols) in lungcancer cells with saturation approximately 1 hour after drugtreatment (Fig. 8). Maximal decreases in reduced thiols wereseen in sensitive cells (Fig. 8A). Interestingly,ME-344 inducedredox stress was reversible in IHLEF (Fig. 8B), inferring aplausible therapeutic advantage. ME-344 triggered pro-nounced oxidation of cell surface thiols with similar kineticsand magnitude (∼40%–50% in 1 hour) in both sensitive cellsand IHLEF (Fig. 8, C and D).

DiscussionBecause of the hydroxyl constituents on its phenol rings

(Silasi et al., 2009), ME-344 has some properties that

Fig. 5. Me-344 effects on caspase-3 activation and cell viability. (A) The kinetics ofME-344 (11.4mM) inhibition of caspase-3 activity (normalized for DMSO)in H596 and SW-900 (resistant:u ands), H460 and SHP-77 (sensitive:j andd) andMRC-5 (IHLEF,,). (B) Temporal viability of sensitive (j andd) andresistant (u ands) cells and of IHLEF (MRC-5, ,) afterME-344 (11.48mM). (C) Dose effects ofME-344 on caspase-3 activity in all cell types. (D) Viabilities ofsensitive (H460) and resistant (H596) cells and IHLEF (MRC-5) after 24 hours’ incubation withME-344 (5.74 and 11.48mM). Data aremean6S.D. for threeindependent experiments.

204 Manevich et al.

at ASPE

T Journals on M


Dow

nloaded from


distinguish it from most of this family of chemicals. Ofparticular importance, our present results demonstrate thatat low micromolar concentrations, cytotoxic effects are notinduced in either immortalized normal lung fibroblasts orin lung cancer cell lines that have intrinsic resistance tothe drug. Because ME-344 is in early stage clinical trials,there would be benefit in defining the mechanism(s) ofaction. In the present study, we describe a series of drugeffects, which may be associated with a beneficial therapeu-tic index and can be linked with the unusual cytotoxicproperties of the drug. Previous work has identified in-hibition of energy production downstream of isoflavoneeffects, so we focused our attention on redox signaling andits consequent impact on bioenergetics in lung cancer celllines sensitive and resistant to the drug. The cell lines werechosen from a preliminary screen of over 200 where thoseidentified as resistant have not been previously exposed todrug selection.Energy production through the mitochondrial respiratory

chain requires an intensive flux of protons from the matrixinto the inner-membrane space. Inhibition of respirationthrough impairment of electron flow diminishes this protonflux and causes IMM depolarization (Alberts, 2002). ME-344can inhibit complex I of the respiratory chain and causedissipation of inner mitochondrial membrane potential (Limet al., 2015). Our OCR inhibition data indicate that sensitivelung cancer cells follow a similar response pattern. Insteadof depolarization, ME-344 causes hyperpolarization of the

inner mitochondrial membrane, serving to minimize thesubsequent impact of oligomycin on OCR and ECAR. In thiscontext, oligomycin is a known inhibitor of the F0 component(proton pump) of ATP synthase, which regulates protonefflux into the matrix from the intermembrane space(Shchepina et al., 2002). In the sensitive cells, ME-344 maybe targeting the oligomycin-sensitive F0 component of ATPsynthase, and some degree specificity is implied because ATPlevels and redox signaling were affected only in thesesensitive cells.On the basis of dose, the efficacy of ME-344 is less than

that of oligomycin (compare 11.5 mM to 1.0 mM), but thespecificity of this effect may be beneficial for the eventualtherapeutic index of ME-344. Also, in sensitive lung cancercells ME-344 dose dependently and effectively (∼50% at23.0 mM, 30 minutes’ incubation) facilitated ATP hydrolysis.This did not occur in IHLEF cells, perhaps again indicatingsome degree of selectivity for transformed cells. Takentogether, these data are consistent with one interpretation tosuggest that ME-344 targets the inner mitochondrial mem-brane proton channel (F0 component of ATP synthase), causesreduced levels of ATP production, alters proton fluxes, andinitiates inner membrane hyperpolarization. This would pro-vide a rational explanation for the unusually rapid andimmediate effects of ME-344 on bioenergetics and redoxsignaling in these cells. In this regard, there are existingexamples of other isoflavones that target ATP synthase(Zheng and Ramirez, 2000; Chinnam et al., 2010).

Fig. 6. ME-344 treatments linked with hydrogen peroxide and *OH generation. (A, C) ME-344 (11.4 mM) effects on intracellular H2O2 (*OH) generationin H460 sensitive (black traces) and H596 resistant (gray traces) cells in real-time kinetics. The traces are representative of three independentexperiments and similar for SHP-77 (sensitive) and SW900 (resistant) cells. (B, D) Averaged dose effects of ME-344 on intracellular generation ofaveraged H2O2 (*OH) in sensitive (H460 and SHP-77, d) and resistant (H596 and SW900, s) cells and IHLEF (MRC-5, ,) as mean 6 S.D. for threeindependent experiments, where the effects are statistically different (P # 0.001). The lines represent sigmoid fit (three parameters) using Sigma Plotsoftware. The arrows indicate points where effector was added.


at ASPE

T Journals on M


Dow

nloaded from


Under hypoxic conditions glycolysis is a complementarypathway for cancer cells to generate ATP. We showed thatresting ECAR in immortalized primary cells was∼2-fold lowerthan in cancer cells. With glucose deprivation, ME-344 hadlittle effect on ECAR in any of the cells. Glycolytic rates insensitive lung cancer cells were∼54.0mpH/min, but because oftheWarburg effect (Vander Heiden et al., 2009) the rates wereexpectedly lower in IHLEF cells (∼12.0 mpH/min with non-glycolytic acidification at ∼4.0 mpH/min). In resistant lungcancer cells ME-344 did not evince any glycolytic response.Although this could be interpreted to suggest a direct impactof ME-344 on glycolysis, at this stage, such a conclusion wouldbe premature. In addition, the results for estimating glycolysis(Fig. 3B) indicate that the drug may selectively inhibit ATPsynthase (following oligomycin addition) only in the sensitivecells and IHLEF.Two other effects of ME-344 in sensitive cancer cells are

described. Drug treatment was linked with increases inintracellular hydroxyl radicals (*OH) and NADPH. Genera-tion of *OH is surprising because traditionally isoflavones arenot viewed as pro-oxidants. Biologic formation of *OH radicalscan occur through Fenton chemistry, where transition metalsreact with H2O2 or other organic peroxides (Winterbourn,1995). In our present study we show that ME-344 increasesH2O2 generation in cytosol, and such a process would beconsistent with interference with the mitochondrial respiratory

chain and subsequent leakage of electrons (Halliwell et al.,1992) causing cytosolic redox stress. Again, this would beconsistent with our data showing progressive thiol oxidation.Moreover, this process has been previously shown to induceheme oxygenase (HO-1; Choi and Alam, 1996).In addition, Park et al. (2011) showed that certain isoflavones

can induce HO-1 in primary astrocytes and this can beaccompanied by its translocation from the ER to mitochondria(Bindu et al., 2011; Bansal et al., 2013). The first intermediatestep of HO-1 catalysis produces Fe21, an effective componentof the Fenton reaction. It is plausible that this sequence ofevents could be involved in the mechanism by which ME-344induces cytotoxicity. The precise chain of events by whichME-344 impacts HO-1 induction and/or translocation tomitochondria requires further investigation, but it is perhapspertinent that HO-1 has been suggested as a possible anti-cancer target (Choi and Alam, 1996).The rapid drug-induced increases in NADPH could also

result from direct or indirect effects on the F0 component ofATP synthase. In turn, this could be linked with IMMhyperpolarization and compensatory proton fluxes, whichcould influence nicotinamide nucleotide transhydrogenase(NNT), an enzyme localized in the inner mitochondrialmembrane catalyzing NADH-mediated reduction of NADP1

to NADPH (Mather et al., 2004; Gameiro et al., 2013). NNT-controlled catalysis is dynamic, efficient, and does seem to

Fig. 7. ME-344 effects on the dynamics of superoxide radical generation, NADPH redox, and inner mitochondrial membrane potentials (Cin). (A)ME-344 (11.4 mM) addition does not affect intracellular superoxide radical generation (detected with Deep Red fluorescent dye in real-time kinetics) inH460 (sensitive: black trace) and H596 (resistant: gray trace) cells. (B) ME-344 (11.48 mM) effects on NADPH emission either before (red trace) or after(green trace) glucose (Glu, 2.0 mM) as indicated by arrows in sensitive (H460) cells or in IHLEF (MRC-5, black trace). Traces are representative of threeindependent experiments and similar for SHP-77 (sensitive, not shown) cells. (C, D) Effects of ME-344 (11.48 mM) and glucose (2.0 mM, arrows) on theIMM potentials of sensitive (H460, solid line) cells, IHLEF (MRC-5, dotted and dashed line) and resistant (SW900, dashed line) cells; detected with JC-9ratiometric fluorescent dye in real-time kinetics. Traces are representative of three independent experiments and were similar for SHP-77 sensitive andH596 resistant cells.

206 Manevich et al.

at ASPE

T Journals on M


Dow

nloaded from


parallel the instantaneous effects of ME-344 (Gameiro et al.,2013). In concert with some of the other results, the immediacyof the effects of ME-344 on NADPH in the sensitive lungcancer cells could also imply that the drug has a direct orindirect impact on this enzyme, although further experimen-tation in this area will be required.Earlier genetic characterizations of the lung cancer cell

lines used here shed a little light on targeting redox pathwaysand their plausible influence drug response data. For exam-ple, SHP-77 cells express the cell surface markers that arereminiscent of HLA-DRmacrophages and hematopoietic cells,and these may influence cell attachment (Ruff et al., 1986).These are not found in H460 (NSCLC) cells and may help toexplain cell surface thiol differences (Davidson et al., 2004)and perhaps influence ME-344 sensitivities (likely only themagnitude of such responses). In the resistant cells, SW900(Wesley et al., 2004) and H596 cells (Nguyen et al., 2015) areeach characterized by the Notch1 transmembrane receptor,linked with the malignant phenotype and absent from thesensitive cell lines. In theory, this difference might alsoinfluence drug response. In general, however, the quantitativeaspects of the differences in response to ME-344 are unlikelyto be a function of simple monogenic differences between thecell lines.In summary, our study shows that ME-344 is an unusual

isoflavone with mitochondria-centered activities engender-ing pronounced and specific inhibition of bioenergetics andredox signaling pathways in lung cancer cells that are

sensitive to the drug. At equivalent concentrations theseeffects do not occur in naturally resistant or immortalizedprimary cells. Two major mitochondrial enzymes: ATP-synthase and NNT are potentially direct or indirect targetsfor ME-344. Hypothetically, translocation of HO-1 intomitochondria as well as its activity could also be influenced byME-344 and further studies in this area are warranted. Theclinical use of ME-344 either as a single agent or in drugcombinationsmay be contingent upon the cancer cell’s balancebetween oxidative phosphorylation and glycolysis. As a pre-diction, patient tumors more dependent upon glycolysis wouldbemore responsive to the drug. In addition, the drug has lowertoxicities in immortalized primary cells, indicating that atleast in vitro an advantageous therapeutic index may beattainable.

Acknowledgments

The authors thankDr. C. Beeson,Ms. G. Beeson, andMr. B. Hooverof the Metabolomics Core of the COBRE for their help and advice inconducting SeaHorse experiments.

Authorship Contributions

Participated in research design: Manevich, Britten, Townsend,Tew.

Conducted experiments: Manevich, Reyes, Townsend.Performed data analysis: Manevich, Townsend, Tew.Wrote or contributed to the writing of the manuscript: Manevich,

Tew.

Fig. 8. ME-344 effects on dynamics of intra- and extracellular thiol redox status in lung cancer and IHLEF cells. Averaged reduced thiol (sulfhydryls)dynamics in lysates of (A) sensitive (H460 and SHP-77), and (B) IHLEF (MRC-5) before (DMSO control [CTR], d and ,) and after incubation withME-344 (22.8 mM, s and ☆), analyzed with sulfhydryl-specific fluorescent dye TG-1. Emissions were normalized for protein concentrations andpresented as mean6 S.E. for three independent experiments. (C) Sensitive (H460 and SHP-77 cells) and (D) IHLEF (MRC-5) were treated with ME-344(22.8 mM) for the indicated times, and cell surface thiol (sulfhydryls) were detected with TG-1. Emissions were normalized for cell numbers, and areaveraged and presented as mean 6 S.E. for three independent experiments.


at ASPE

T Journals on M


Dow

nloaded from


References

Aguero MF, Facchinetti MM, Sheleg Z, and Senderowicz AM (2005) Phenoxodiol, anovel isoflavone, induces G1 arrest by specific loss in cyclin-dependent kinase2 activity by p53-independent induction of p21WAF1/CIP1. Cancer Res 65:3364–3373.

Alberts B (2002) Molecular Biology of the Cell, Garland Sciences, New York.Alvero AB, Montagna MK, Holmberg JC, Craveiro V, Brown D, and Mor G (2011)Targeting the mitochondria activates two independent cell death pathways inovarian cancer stem cells. Mol Cancer Ther 10:1385–1393.

Bansal S, Biswas G, and Avadhani NG (2013) Mitochondria-targeted hemeoxygenase-1 induces oxidative stress and mitochondrial dysfunction in macro-phages, kidney fibroblasts and in chronic alcohol hepatotoxicity. Redox Biol 2:273–283.

Biaglow JE, Manevich Y, Leeper D, Chance B, Dewhirst MW, Jenkins WT, TuttleSW, Wroblewski K, Glickson JD, and Stevens C, et al. (1998) MIBG inhibits res-piration: potential for radio- and hyperthermic sensitization. Int J Radiat OncolBiol Phys 42:871–876.

Bindu S, Pal C, Dey S, Goyal M, Alam A, Iqbal MS, Dutta S, Sarkar S, Kumar R,and Maity P, et al. (2011) Translocation of heme oxygenase-1 to mitochondria is anovel cytoprotective mechanism against non-steroidal anti-inflammatory drug-induced mitochondrial oxidative stress, apoptosis, and gastric mucosal injury. JBiol Chem 286:39387–39402.

Bowers RR, Manevich Y, Townsend DM, and Tew KD (2012) Sulfiredoxin redox-sensitive interaction with S100A4 and non-muscle myosin IIA regulates cancer cellmotility. Biochemistry 51:7740–7754.

Chinnam N, Dadi PK, Sabri SA, Ahmad M, Kabir MA, and Ahmad Z (2010) Dietarybioflavonoids inhibit Escherichia coli ATP synthase in a differential manner. Int JBiol Macromol 46:478–486.

Choi AM and Alam J (1996) Heme oxygenase-1: function, regulation, and implicationof a novel stress-inducible protein in oxidant-induced lung injury. Am J Respir CellMol Biol 15:9–19.

Davidson JD, Ma L, Flagella M, Geeganage S, Gelbert LM, and Slapak CA (2004) Anincrease in the expression of ribonucleotide reductase large subunit 1 is associatedwith gemcitabine resistance in non-small cell lung cancer cell lines. Cancer Res 64:3761–3766.

de Souza PL, Russell PJ, Kearsley JH, and Howes LG (2010) Clinical pharmacologyof isoflavones and its relevance for potential prevention of prostate cancer. NutrRev 68:542–555.

Gameiro PA, Laviolette LA, Kelleher JK, Iliopoulos O, and Stephanopoulos G (2013)Cofactor balance by nicotinamide nucleotide transhydrogenase (NNT) coordinatesreductive carboxylation and glucose catabolism in the tricarboxylic acid (TCA)cycle. J Biol Chem 288:12967–12977.

Halliwell B, Gutteridge JM, and Cross CE (1992) Free radicals, antioxidants, andhuman disease: where are we now? J Lab Clin Med 119:598–620.

Herst PM, Petersen T, Jerram P, Baty J, and Berridge MV (2007) The anti-proliferative effects of phenoxodiol are associated with inhibition of plasma mem-brane electron transport in tumour cell lines and primary immune cells. BiochemPharmacol 74:1587–1595.

Kamsteeg M, Rutherford T, Sapi E, Hanczaruk B, Shahabi S, Flick M, Brown D,and Mor G (2003) Phenoxodiol–an isoflavone analog—induces apoptosis inchemoresistant ovarian cancer cells. Oncogene 22:2611–2620.

Leclercq G and Jacquot Y (2014) Interactions of isoflavones and other plant derivedestrogens with estrogen receptors for prevention and treatment of breast cancer-considerations concerning related efficacy and safety. J Steroid Biochem Mol Biol139:237–244.

Lim SC, Carey KT, and McKenzie M (2015) Anti-cancer analogues ME-143 andME-344 exert toxicity by directly inhibiting mitochondrial NADH: ubiquinoneoxidoreductase (complex I). Am J Cancer Res 5:689–701.

Maldonado EN, Sheldon KL, DeHart DN, Patnaik J, Manevich Y, Townsend DM,Bezrukov SM, Rostovtseva TK, and Lemasters JJ (2013) Voltage-dependent anion

channels modulate mitochondrial metabolism in cancer cells: regulation by freetubulin and erastin. J Biol Chem 288:11920–11929.

Manevich Y, Held KD, and Biaglow JE (1997) Coumarin-3-carboxylic acid as a de-tector for hydroxyl radicals generated chemically and by gamma radiation. RadiatRes 148:580–591.

Manevich Y, Sweitzer T, Pak JH, Feinstein SI, Muzykantov V, and Fisher AB (2002)1-Cys peroxiredoxin overexpression protects cells against phospholipidperoxidation-mediated membrane damage. Proc Natl Acad Sci USA 99:11599–11604.

Mather OC, Singh A, van Boxel GI, White SA, and Jackson JB (2004) Active-siteconformational changes associated with hydride transfer in proton-translocatingtranshydrogenase. Biochemistry 43:10952–10964.

Morré DJ and Morré DM (2003) Cell surface NADH oxidases (ECTO-NOX proteins)with roles in cancer, cellular time-keeping, growth, aging and neurodegenerativediseases. Free Radic Res 37:795–808.

Nguyen D, Rubinstein L, Takebe N, Miele L, Tomaszewski JE, Ivy P, Doroshow JH,and Yang SX (2015) Notch1 phenotype and clinical stage progression in non-smallcell lung cancer. J Hematol Oncol 8:9.

Pant S, Burris HA, 3rd, Moore K, Bendell JC, Kurkjian C, Jones SF, Moreno O, KuhnJG, McMeekin S, and Infante JR (2014) A first-in-human dose-escalation study ofME-143, a second generation NADH oxidase inhibitor, in patients with advancedsolid tumors. Invest New Drugs 32:87–93.

Park JS, Jung JS, Jeong YH, Hyun JW, Le TK, Kim DH, Choi EC, and Kim HS (2011)Antioxidant mechanism of isoflavone metabolites in hydrogen peroxide-stimulatedrat primary astrocytes: critical role of hemeoxygenase-1 and NQO1 expression.J Neurochem 119:909–919.

Ruff MR, FarrarWL, and Pert CB (1986) Interferon gamma and granulocyte/macrophagecolony-stimulating factor inhibit growth and induce antigens characteristic ofmyeloid differentiation in small-cell lung cancer cell lines. Proc Natl Acad Sci USA83:6613–6617.

Sapi E, Alvero AB, Chen W, O’Malley D, Hao XY, Dwipoyono B, Garg M, KamsteegM, Rutherford T, and Mor G (2004) Resistance of ovarian carcinoma cells todocetaxel is XIAP dependent and reversible by phenoxodiol. Oncol Res 14:567–578.

Shchepina LA, Pletjushkina OY, Avetisyan AV, Bakeeva LE, Fetisova EK, IzyumovDS, Saprunova VB, Vyssokikh MY, Chernyak BV, and Skulachev VP (2002)Oligomycin, inhibitor of the F0 part of H1-ATP-synthase, suppresses the TNF-induced apoptosis. Oncogene 21:8149–8157.

Silasi DA, Alvero AB, Rutherford TJ, Brown D, and Mor G (2009) Phenoxodiol:pharmacology and clinical experience in cancer monotherapy and in combinationwith chemotherapeutic drugs. Expert Opin Pharmacother 10:1059–1067.

Vander Heiden MG, Cantley LC, and Thompson CB (2009) Understanding theWarburg effect: the metabolic requirements of cell proliferation. Science 324:1029–1033.

Wang X, McKernan R, Kim KH, Alvero AB, Whiting A, Thompson JA, Mor G, SaifMW, Husband AJ, and Brown DM, et al. (2011) Triphendiol (NV-196), developmentof a novel therapy for pancreatic cancer. Anticancer Drugs 22:719–731.

Wesley UV, Tiwari S, and Houghton AN (2004) Role for dipeptidyl peptidase IV intumor suppression of human non small cell lung carcinoma cells. Int J Cancer 109:855–866.

Winterbourn CC (1995) Toxicity of iron and hydrogen peroxide: the Fenton reaction.Toxicol Lett 82-83:969–974.

Zheng J and Ramirez VD (2000) Inhibition of mitochondrial proton F0F1-ATPase/ATPsynthase by polyphenolic phytochemicals. Br J Pharmacol 130:1115–1123.

Address correspondence to: Dr. Kenneth D. Tew, Department of Cell andMolecular Pharmacology and Experimental Therapeutics, 173 Ashley Ave.,BSB358, Charleston, SC 29425. E-mail: [email protected]

208 Manevich et al.

at ASPE

T Journals on M


Dow

nloaded from
mailto:[email protected]://jpet.aspetjournals.org/

Documents

Redox Signaling and Bioenergetics Influence Lung Cancer Cell …jpet.aspetjournals.org/content/jpet/358/2/199.full.pdf · ME-344 [(3R,4S)-3,4-bis(4-hydroxyphenyl)-8-methyl-3,4-dihydro-2H-chromen-7-ol]