novel thiosemicarbazones regulate the signal...

1521-0111/87/3/543–560$25.00 http://dx.doi.org/10.1124/mol.114.096529MOLECULAR PHARMACOLOGY Mol Pharmacol 87:543–560, March 2015Copyright ª 2015 by The American Society for Pharmacology and Experimental Therapeutics

Novel Thiosemicarbazones Regulate the Signal Transducer andActivator of Transcription 3 (STAT3) Pathway: Inhibition ofConstitutive and Interleukin 6–Induced Activation byIron Depletion

Goldie Y. L. Lui, Zaklina Kovacevic, Sharleen V. Menezes, Danuta S. Kalinowski,Angelica M. Merlot, Sumit Sahni, and Des R. RichardsonDepartment of Pathology and Bosch Institute, School of Medical Sciences, Sydney Medical School, University of Sydney,Sydney, New South Wales, Australia

Received October 22, 2014; accepted January 5, 2015

ABSTRACTPharmacologic manipulation of metal pools in tumor cells is apromising strategy for cancer treatment. Here, we reveal how theiron-binding ligands desferrioxamine (DFO), di-2-pyridylketone-4,4-dimethyl-3-thiosemicarbazone (Dp44mT), and di-2-pyridylketone4-cyclohexyl-4-methyl-3-thiosemicarbazone (DpC) inhibit con-stitutive and interleukin 6–induced activation of signal transducerand activator of transcription 3 (STAT3) signaling, which promotesproliferation, survival, and metastasis of cancer cells. Wedemonstrate that DFO, Dp44mT, and DpC significantly decreaseconstitutive phosphorylation of the STAT3 transcription factor atTyr705 in the pancreatic cancer cell lines PANC-1 and MIAPaCa-2 as well as the prostate cancer cell line DU145. Thesecompounds also significantly decrease the dimerized STAT3levels, the binding of nuclear STAT3 to its target DNA, and theexpression of downstream targets of STAT3, including cyclin D1,

c-myc, and Bcl-2. Examination of upstream mediators of STAT3in response to these ligands has revealed that Dp44mT and DpCcould significantly decrease activation of the nonreceptor tyrosinekinase Src and activation of cAbl in DU145 and MIAPaCa-2 cells.In contrast to the effects of Dp44mT, DpC, or DFO on inhibitingSTAT3 activation, the negative control compound di-2-pyridylketone2-methyl-3-thiosemicarbazone, or the DFO:Fe complex, whichcannot bind cellular iron, had no effect. This demonstrates the roleof iron-binding in the activity observed. Immunohistochemicalstaining of PANC-1 tumor xenografts showed a marked decreasein STAT3 in the tumors of mice treated with Dp44mT or DpCcompared with the vehicle. Collectively, these studies demon-strate suppression of STAT3 activity by iron depletion in vitro andin vivo, and reveal insights into regulation of the critical oncogenicSTAT3 pathway.

IntroductionIdentification of novel chemotherapeutics that target deregu-

lated signaling pathways responsible for tumorigenesis andmetastasis is urgently required, especially considering thatcancer remains a major cause of mortality (Goldstein et al.,1989; Lee et al., 2008; Vincent et al., 2011). In this regard, noveldi-2-pyridyl thiosemicarbazones (DpT) are well characterizedas promising chemotherapeutic agents (Yuan et al., 2004;Whitnall et al., 2006; Kovacevic et al., 2011; Chen et al., 2012;Liu et al., 2012; Merlot et al., 2013a; Sun et al., 2013).Numerous studies have demonstrated their potent and selec-tive antitumor activity and antimetastatic effects across

a broad range of cancer cell-types in vitro and in vivo (Yuan et al.,2004; Whitnall et al., 2006; Kovacevic et al., 2011; Chen et al.,2012; Liu et al., 2012; Merlot et al., 2013a; Sun et al., 2013).Notably, the second-generation DpT analog di-2-pyridylketone4-cyclohexyl-4-methyl-3-thiosemicarbazone (DpC; Fig. 1A) dis-plays marked antitumor activity and high tolerability in vivothat surpasses the former first-generation, lead compound di-2-pyridylketone 4,4-dimethyl-3-thiosemicarbazone (Dp44mT; Fig.1A) (Kovacevic et al., 2011; Lovejoy et al., 2012).Unlike desferrioxamine (DFO; Fig. 1A), the classic iron

chelator used to treat iron overload disease, DpT thiosemicarba-zones form redox active metal complexes that generate reactiveoxygen species (ROS), which plays a key role in their cytotoxicactivity (Yuan et al., 2004; Richardson et al., 2006; Lovejoy et al.,2011). The mechanism of action of these ligands also involvescellular iron depletion, which is known to affect cellular targetscritical for cell proliferation, survival, and metastasis, includingribonucleotide reductase (Yu et al., 2011), cyclin D1 (Kulp et al.,

This work was supported by a National Health and Medical ResearchCouncil of Australia (NHMRC) Project Grant [APP1060482], NHMRC SeniorPrincipal Research Fellowship [APP1062607], and a NHMRC Peter DohertyBiomedical Post-Doctoral Fellowship [APP1037323] and a Cancer InstituteNew South Wales Early Career Development Fellowship [12/ECF/2-17].

dx.doi.org/10.1124/mol.114.096529.

ABBREVIATIONS: BSA, bovine serum albumin; DAPI, 49,6-diamidino-2-phenylindole; DFO, desferrioxamine; DMSO, dimethylsulfoxide; Dp2mT,di-2-pyridylketone 2-methyl-3-thiosemicarbazone; Dp44mT, di-2-pyridylketone 4,4-dimethyl-3-thiosemicarbazone; DpC, di-2-pyridylketone4-cyclohexyl-4-methyl-3-thiosemicarbazone; DpT, di-2-pyridyl thiosemicarbazone; GSH, glutathione; IHC, immunohistochemical; JAK, Januskinase; JAK2, Janus tyrosine kinase; NAC, N-acetylcysteine; PaEC, normal pancreatic epithelial cells; PBS, phosphate-buffered saline; ROS,reactive oxygen species; TBS, Tris buffered saline.

543

at ASPE

T Journals on Septem

ber 9, 2018m

olpharm.aspetjournals.org

Dow

nloaded from

http://dx.doi.org/10.1124/mol.114.096529

http://dx.doi.org/10.1124/mol.114.096529

http://molpharm.aspetjournals.org/

1996; Gao and Richardson, 2001; Nurtjahja-Tjendraputra et al.,2007), p21 (Darnell and Richardson, 1999; Fu and Richardson,2007), and the metastasis suppressor N-myc downstreamregulated gene-1 (NDRG1) (Le and Richardson, 2004; Kovacevicet al., 2011; Kovacevic et al., 2013).

This group of thiosemicarbazones has also been demon-strated to induce apoptosis (Yuan et al., 2004; Noulsri et al.,2009) and autophagy (Sahni et al., 2014), as well as to inhibitthe epithelial-to-mesenchymal transition (Chen et al., 2012)in cancer cells. Dp44mT could target these critical molecules

Fig. 1. The iron chelators DFO,Dp44mT, andDpC decrease expression of phosphorylated STAT3 at Tyr705 in pancreatic and prostate cancer cells. (A) Linedrawings of the chemical structures of the iron chelators DFO, Dp44mT, and DpC, and the negative control compound Dp2mT. (B–E) The levels of p-STAT3and STAT3 protein in (B) PANC-1, (C)MIAPaCa-2, (D) DU145, and (E)mortal PaEC cells after a 24-hour/37°C incubationwith controlmedium, DFO (250 mM),Dp44mT (5 mM), DpC (5 mM), DFO:Fe (1:1; 250 mM), Dp44mT:Fe (2:1; 5 mM), DpC:Fe (2:1; 5 mM), FeCl3 (250 mM), or Dp2mT (5 mM). The Western blots aretypical of three independent experiments, with densitometric analysis representing mean 6 S.D. Relative to control: *P , 0.05; **P , 0.01; ***P , 0.001.

544 Lui et al.

at ASPE


ber 9, 2018m


Dow

nloaded from


and cellular processes through its demonstrated modulationof oncogenic signaling pathways, including the transforminggrowth factor-b (Kovacevic et al., 2013), c-Jun NH2-terminalkinase (JNK) and p38mitogen-activated protein kinase (MAPK)(Yu and Richardson, 2011), Ras/ERK (Kovacevic et al., 2013),protein kinase B (AKT)/phosphatidylinositol-3-kinase (Dixonet al., 2013; Kovacevic et al., 2013), and ROCK1/pMLC2pathways (Sun et al., 2013). Despite considerable advances inour understanding of the molecular mechanisms of action ofthiosemicarbazones, they remain incompletely understood.Detailed characterization is lacking in particular for thesecond-generation thiosemicarbazone DpC, which is impor-tant to elucidate for clinical development.The signal transducer and activator of transcription (STAT)

family consists of seven proteins that transduce extracellularsignals to regulate genes involved in cell growth, survival, anddifferentiation (Kamran et al., 2013). Of these, STAT3 is themost well characterized and has been linked to tumor pro-gression (Kamran et al., 2013). STAT3 signaling can be activatedby upstream receptor interactions (e.g., IL-6/Janus kinase[JAK]) and nonreceptor kinases (e.g., Src, c-Abl) that phosphor-ylate STAT3 at Tyr705 (p-STAT3) (Turkson and Jove, 2000).This leads to dimer formation, nuclear translocation, binding toSTAT3-specific DNA-binding elements, and transcription oftarget genes (e.g., Bcl-2, cyclin D1, and c-myc) that promote cellproliferation, survival, migration, apoptosis inhibition, andimmune evasion (Darnell et al., 1994; Ihle, 1995; Sasse et al.,1997; Zushi et al., 1998; Bromberg et al., 1999;Wang et al., 2004;Xie et al., 2004). Constitutive activation of STAT3 is a hallmarkof many human malignancies, including those of the prostate,pancreas, and breast, as well as leukemias and melanomas(Bowman et al., 2000; Scholz et al., 2003). Indeed, several studieshave demonstrated that STAT3 inhibition can induce apoptosisin a range of cancer types, including prostate and pancreaticcancers (Haura et al., 2005; Lewis et al., 2008). Understandingthe role of STAT3 signaling in tumor progression and how thisaffects responsiveness to therapies is an active area of clinicalinvestigation and is critical for the successful development ofpersonalized cancer treatments.Herein, for the first time, we demonstrate that inducing

cellular iron depletion using the iron chelators DFO, Dp44mT,and DpC can inhibit STAT3 signaling in vitro in prostate andpancreatic cancer cells, as well as in vivo using pancreatictumor xenografts grown in nude mice. These findings revealkey insights into the role of iron in tumor progression throughSTAT3 pathway inhibition and highlight its potential applica-tion in cancer therapy through the use of potent and novelthiosemicarbazones.

Materials and MethodsCell Culture. The human pancreatic cancer cell lines, PANC-1

andMIAPaCa-2, and the human prostate cancer cell line DU145 werepurchased from the American Type Culture Collection (Manassas,VA). PANC-1 andMIAPaCa-2 cells were grown in Dulbecco’s modifiedEagle medium (Life Technologies, Melbourne, Australia), and DU145cells were grown in RPMI 1640medium (Life Technologies). All mediawere supplemented with 10% fetal bovine serum, penicillin (100 IU/ml),streptomycin (100 mg/ml), glutamine (2 mM), nonessential amino acids(100 mM), and sodium pyruvate (100 mM; all supplements from LifeTechnologies).

Primary human pancreatic epithelial cells were purchased from theApplied Cell Biology Research Institute (Kirkland, WA) and grown and

maintained in Cell Systems Corporation (Kirkland, WA) CompleteMedium containing 10% serum, as well as 2% CultureBoost-R (AppliedCell Biology Research Institute) to support cell growth. All cells wereincubated at 37°C in a humidified atmosphere containing 5% CO2.

Cell Treatments. The thiosemicarbazones Dp44mT, DpC, andDp2mT were synthesized and characterized using standard methods(Richardson et al., 2006; Lovejoy et al., 2012). DFOwas purchased fromNovartis (Basel, Switzerland). Dp44mT, DpC, and di-2-pyridylketone2-methyl-3-thiosemicarbazone (Dp2mT) were dissolved in dimethyl-sulfoxide (DMSO) at 10 mM and then diluted in media containing 10%(v/v) fetal bovine serum (Sigma-Aldrich, Castle Hill, Australia) so thatthe final [DMSO] was #0.1% (v/v). Control studies have previouslyshown that these low DMSO levels do not affect cellular proliferation oriron metabolism (Richardson et al., 1995). IL-6 was purchased fromR&D Systems (Minneapolis, MN), reconstituted in sterile phosphate-buffered saline (PBS) containing 0.1% bovine serum albumin (BSA),and used at a final concentration of 10 ng/ml. This IL-6 concentrationwas used as it has been previously shown to effectively activate STAT3(Hahm and Singh, 2010).

Western Blotting. Western blot analysis was performed usingestablished procedures (Chen et al., 2012). The following primaryantibodies used in this study were from Cell Signaling Technology(Beverly, MA): phospho-STAT3 (Tyr705), STAT3, phospho-Janustyrosine kinase [p-JAK2] (Tyr1007/8), JAK2, phospho-Src (Tyr416),phospho-Src (Tyr527), Src, phospho-cAbl (Tyr245), cAbl, and Bcl-2.Antibodies against c-myc and cyclin D1 were from Santa CruzBiotechnology (Santa Cruz, CA), and antibody against b-actin waspurchased from Sigma-Aldrich. All primary antibodies were used ata 1:1000 dilution except for b-actin (1:10,000). All secondary antibodieswere from Sigma-Aldrich and used at a 1:10,000 dilution.

Membranes were probed for b-actin as a loading control, and allsample data values were normalized to the corresponding b-actinvalues. Densitometric analysis was performed using Image Labora-tory 4.0.1 software (Bio-Rad Laboratories, Hercules, CA).

Native PAGE. Native PAGE was performed according to themethod of Shin et al. (2009). Briefly, native cell extracts were preparedusing ice-cold isotonic buffer containing 20 mM Tris (pH 7.0), 150 mMNaCl, 6 mM MgCl2, 0.8 mM phenylmethylsulfonyl fluoride, and 20%glycerol, and homogenized with a 27-gauge syringe. Lysates werecleared by centrifugation at 13,000 rpm for 30minutes at 4°C. Samples(10 mg) were separated on 6% SDS-free PAGE gels, transferred topolyvinylidene difluoride membranes, and immunoblotted as describedfor Western blotting.

Chemiluminescent Electrophoretic Mobility Shift Assay. Toexamine DNA-STAT3 interactions, nuclear protein extracts wereprepared using NE-PERNuclear andCytoplasmic Extraction Reagents(Thermo Scientific, Scoresby, Australia) according to the manufac-turer’s instructions, and protein concentrations were determined usingthe BCA assay (Thermo Scientific). The double-stranded biotinylatedoligonucleotides used to detect functional STAT3 DNA-binding werebased on the STAT3 consensus sequence originally identified to bindthe acute-phase response element in a variety of target gene promoters(Yu et al., 1995). These oligonucleotides were purchased from LifeTechnologies, with the following sequences: STAT3 specific probe,59-GATCCTTCTGGGAATTCCTAGATC-39; STAT3 mutant probe,59-GATCCTTCTGGGCCGTCCTAGATC-39; nonspecific probe, 59-AGTC-TAGAGTCACAGTGAGTCGGCAAAATTTGAGCC-39. Equal amountsof nuclear extract (10 mg) were incubated with 20 fmol biotin-end-labeled DNA in a final volume of 20 ml with binding buffer [10 mMTrispH 7.5, 50 mM KCl, 1 mM dithiothreitol, 10% glycerol, 0.1 mg/ml poly(dI-dC), and 0.5 mg/ml BSA] and allowed to bind for 30 minutes at 4°C.The protein-DNA complexes were separated on a 6% native PAGE gelin 0.5� Tris borate-EDTA, then transferred onto a positively chargednylonmembrane (GEHealthcare, Rydalmere, Australia) for 30minutesin 0.5� Tris borate-EDTA. The DNA was crosslinked to the membraneusing a UV Stratalinker 1800 (Stratagene, Santa Clara, CA) anddetected using the Chemiluminescent Nucleic Acid Detection Module(Thermo Scientific).

Antitumor Thiosemicarbazones Inhibit STAT3 Signaling 545

at ASPE


ber 9, 2018m


Dow

nloaded from


Immunofluorescence. Cells were visualized by immunofluores-cence as described previously elsewhere (Chen et al., 2012). Briefly,cells were fixed with 4% (w/v) paraformaldehyde (Sigma-Aldrich) for10 minutes at 20°C then permeabilized using 0.1% (v/v) Triton X-100for 5 minutes at 20°C. Blocking was performed using 5% BSA in PBSfor 1 hour at 20°C (room temperature), and then the cells wereincubated with primary antibodies overnight at 4°C. The cells weresubsequently incubated with fluorescent secondary antibody for1 hour at 20°C, washed with PBS, and set onto slides using mountingsolution containing 49,6-diamidino-2-phenylindole (DAPI; Invitrogen).Imaging was performed using an Olympus Zeiss AxioObserver Z1fluorescence microscope (Carl Zeiss GmbH, Jena, Germany) witha 63� oil objective. Mean fluorescence intensities were measuredusing ImageJ (http://imagej.nih.gov/ij/; Schneider et al., 2012), withaverage values taken from three random images per sample.

Preparation of Tumor Xenografts in Nude Mice. Our in vivostudy was approved by the Sydney University Animal EthicsCommittee. Tumor xenografts were established in 8-week-old femalenude mice (BALBc nu/nu), which were routinely fed basal rodent chow,watered ad libitum, and housed under a 12-hour light/dark cycle.Briefly, 2 � 106 PANC-1 cells grown in culture were harvested andsuspended in Matrigel (BD Biosciences, San Jose, CA) and injectedsubcutaneously into the right flank of each mouse. Tumor size andvolume were monitored daily, and treatments began when tumorsreached an average size of 90 mm3. The mice were separated into threegroups (n 5 6), with each group receiving either Dp44mT (0.4 mg/kg),DpC (5 mg/kg), or the vehicle control intravenously via tail veininjection 5 days/week. This treatment regimen was established basedon previous studies performed in our laboratory using these agents thathad demonstrated that this administration schedule was well toleratedand showed high antitumor efficacy (Whitnall et al., 2006; Kovacevicet al., 2011; Lovejoy et al., 2012). The Dp44mT and DpC treatmentswere prepared by dissolving each of the compounds in 30% propyleneglycol/0.9% saline, and the vehicle control treatment consisted of 30%propylene glycol/0.9% saline only (Whitnall et al., 2006; Kovacevic et al.,2011; Lovejoy et al., 2012). After 22 days of treatment, the mice wereeuthanized, and the tumors were harvested, fixed in 10% neutralbuffered formalin for 24 hours, then processed in paraffin wax.

Immunohistochemistry. Tumor tissue sections were sectioned at5 mm onto Superfrost Plus slides (Menzel-Gläser Braunschweig,Germany), deparaffinized through xylene rinses, and rehydratedthrough graded ethanols. Antigen retrieval was performed by boilingslides in a water bath for 30 minutes using Dako Target RetrievalSolution (Dako, Carpinteria, CA) at pH 9.0 for p-STAT3 staining, andpH 6.0 for STAT3 staining. Endogenous peroxidase activity was blockedby 5 minutes of incubation with 3% hydrogen peroxide, followed bya wash with 1� Tris buffered saline (TBS)/Tween (0.05%). The sampleswere incubated with primary antibody for 30 minutes at 20°C, followedby a TBS/Tween wash. Bound antibody was detected using DakoEnvision1 anti-rabbit horseradish peroxidase–labeled polymer second-ary antibody (Dako) for 30 minutes at 20°C. Slides were rinsed ina series of TBS/Tween washes, then incubated with DAB1 chromogen(Dako) for 10 minutes. Finally, the slides were counterstained withHarris hematoxylin, dehydrated in graded ethanols, cleared throughxylene rinses, and mounted. IHC staining was performed using thefollowing primary antibodies from Cell Signaling Technology: p-STAT3and STAT3.

Evaluation of Immunohistochemical Staining. As the IHCp-STAT3 staining demonstrated nuclear localization in this study, wequantified the nuclear p-STAT3 staining in ImageJ (Schneider et al., 2012)using the ImmunoRatio plug-in (Tuominen et al., 2010), as previouslydescribed elsewhere (Jahangiri et al., 2013). Theaverage valueswere takenfrom three random images per sample, and background correction wasperformed for each slide using a blank field image.

As STAT3 staining was not exclusively localized in the nucleus, theabove method of quantification was not appropriate. Instead, STAT3staining was scored for each sample by considering both the percentageand intensity of cells displaying positive staining. Scores for the

percentage of positive cells were assigned based on standard procedures(Krajewskaet al., 1996;Gonget al., 2005) as follows: 0–10%50; 11–25%51;26–50%5 2; 51–75%5 3; and 75–100%5 4. Scores for staining intensitywere assigned as follows: light brown, weak staining 5 1; brown, mod-erate staining5 2; and dark brown, intense staining5 3. The individualscores for percentage and intensity were multiplied to obtain an IHCscore, as previously described elsewhere (Krajewska et al., 1996; Gonget al., 2005). Two researchers independently scored three random fieldsfor each sample in a blinded manner, with the average IHC score beingpresented.

Statistical Analysis. Data are expressed as mean 6 S.D. Statis-tical analyses were performed using GraphPad Prism (GraphPadSoftware, La Jolla, CA). Data were compared against the respectivecontrol in each experiment by one-way analysis of variance followed byDunnett’s post-test. P , 0.05 was considered statistically significant.

ResultsPrevious studies have revealed that the p38 mitogen-

activated protein kinase (MAPK), AKT, and transforminggrowth factor b (TGF-b) pathways are molecular targets thatplay a role in the antiproliferative activity of DFO and Dp44mT(Yu and Richardson, 2011; Dixon et al., 2013; Kovacevic et al.,2013). To further elucidate the molecular mechanisms ofthiosemicarbazones, we considered evidence that suggested cross-talk between these and other major signaling pathways, in-cluding the STAT3 pathway (Jain et al., 1998; Zhao et al.,2008; Assinder et al., 2009), and we examined the potential roleof the latter pathway in mediating the antiproliferative ac-tivities of the potentDpT thiosemicarbazonesDpCandDp44mT.In the current investigation, we used the pancreatic cancer celllines PANC-1 andMIAPaCa-2 as well as the prostate cancer cellline DU145 because the potent antitumor efficacy of Dp44mTand DpC have previously been demonstrated in these cell types(Whitnall et al., 2006; Kovacevic et al., 2011; Dixon et al., 2013;Sun et al., 2013). Furthermore, these cell lines display con-stitutive phosphorylation of STAT3 that leads to its activation(Mora et al., 2002; Lian et al., 2004).The cells were incubated with DFO (250 mM), Dp44mT

(5 mM), or DpC (5 mM) for 24 hours at 37°C (Fig. 1A). Theseconditions have been previously demonstrated to efficientlyinduce iron depletion and increase the expression of iron-regulatedmolecules (Richardson et al., 1994; Yuan et al., 2004; Kovacevicet al., 2011). Furthermore, we have shown that Dp44mT up-take by cells saturates at 5–10 mM (Merlot et al., 2013b). Atthis concentration, the level of agent is pharmacologically rel-evant in humans, as the structurally related thiosemicarbazoneTriapine has been observed in the serum at similar levels(Wadler et al., 2004; Chao et al., 2012). The much higher DFOconcentration was used because of its limited membranepermeability compared with Dp44mT and DpC, which displayhigh permeability and marked iron chelation efficacy (Yuanet al., 2004; Kovacevic et al., 2011; Lovejoy et al., 2012; Sun et al.,2013). As a negative control, under the same conditions, cellswere also incubated with Dp2mT (5 mM) (Fig. 1A), a structuralanalog of Dp44mT and DpC that cannot bind metal ions and dis-plays negligible antitumor activity (Yuan et al., 2004; Richardsonet al., 2006; Chen et al., 2012; Sun et al., 2013).Dp44mT and DpC Decrease Constitutive STAT3 Phos-

phorylation at Tyr705. Initial studies revealed that incuba-tion of PANC-1,MIAPaCa-2, andDU145 cellswithDFO,Dp44mT,or DpC significantly (P, 0.001–0.01) decreased levels of phos-phorylated (Tyr705) STAT3 (i.e., p-STAT3; Fig. 1, B–D). Under

546 Lui et al.

at ASPE


ber 9, 2018m


Dow

nloaded from

http://imagej.nih.gov/ij/


the same conditions, total STAT3 expression was also signif-icantly (P , 0.05) decreased after incubation with Dp44mTor DpC, but not with DFO. In contrast, the negative controlcompound for Dp44mT—namely, Dp2mT, which cannot bindmetal ions (Yuan et al., 2004; Chen et al., 2012)—did notsignificantly (P. 0.05) alter p-STAT3 or STAT3 levels in any ofthe cell types examined (Fig. 1, B–D). Herein, levels of STAT3activation were determined by densitometric analysis of theWestern blots and calculation of the p-STAT3/STAT3 ratio.To understand the role of iron depletion in the decrease of the

p-STAT3/STAT3 ratio induced by these compounds, the cellswere also incubatedwith the iron complexes of these ligands thatwere prepared by precomplexing the chelators with FeCl3 usingstandard methods (Yu and Richardson, 2011). Because DFO isa hexadentate ligand (Merlot et al., 2013a), the iron complex ofDFO (DFO:Fe) was prepared in a 1:1 ligand/iron molar ratio tosaturate the coordination sphere at a concentration of 250mM. Incontrast, the iron complexes of the tridentate chelators Dp44mTand DpC (Richardson et al., 2006; Lovejoy et al., 2012; Merlotet al., 2013a) (i.e., Dp44mT:Fe and DpC:Fe, respectively) wereprepared in 2:1 ligand/iron molar ratios to saturate the coordi-nation sphere at a final concentration of 5 mM. As another rel-evant control, cellswere also incubatedwithFeCl3 alone (250mM),which resulted in no significant (P . 0.05) effects on p-STAT3 orSTAT3 levels versus the control (Fig. 1, B–D).For all three cell-types, incubation with the nonredox-active

DFO:Fe complex (Kalinowski and Richardson, 2005) resultedin p-STAT3/STAT3 levels that were significantly (P , 0.001–0.01) greater than those found for DFO alone, and comparableto its respective control (Fig. 1, B–D). This finding indicatesthat the ability of DFO to inhibit constitutively activated STAT3depends on its ability to bind intracellular iron. In contrast,incubation of PANC-1, MIAPaCa-2, or DU145 cells with theDp44mT:Fe or DpC:Fe complex significantly (P , 0.001–0.05)decreased the p-STAT3/STAT3 ratio compared with the control.The extent of the decrease in the p-STAT3/STAT3 ratio by thecomplexes was similar (PANC-1) or slightly less (MIAPaCa-2,DU145) than that observed for Dp44mT or DpC alone. Theseresults suggest that, in contrast to the nonredox active DFO:Fecomplex, the well characterized ability of Dp44mT:Fe and DpC:Fe complex to generate ROS (Yuan et al., 2004; Richardson et al.,2006; Lovejoy et al., 2011) may also play a role in the decreasedp-STAT3/STAT3 ratio.To compare the effects of these compounds on mortal,

nontransformed cells, we also examined the effects of DFO,Dp44mT, or DpC on primary cultures of normal pancreaticepithelial cells (PaEC; Fig. 1E). Incubation of PaEC withDFO, Dp44mT, and DpC for 24 hours at 37°C did not result inany significant (P . 0.05) alteration in p-STAT3 or STAT3levels relative to the control. In fact, p-STAT3 expression wasbarely detectable in these normal cells (Fig. 1E) comparedwith their malignant counterparts (Fig. 1, B–D). This observa-tion is consistent with 1) previous reports that STAT3 is over-activated in pancreatic carcinoma specimens, but not in normalpancreatic tissues (Lian et al., 2004); and 2) that these com-pounds can selectively target cancer cells over normal cells invitro and in vivo (Yuan et al., 2004; Kovacevic et al., 2011;Lovejoy et al., 2012; Dixon et al., 2013).Collectively, these studies demonstrate that iron chelation

decreases the p-STAT3/STAT3 ratio and that the redox activityof the Dp44mT:Fe and DpC:Fe complexes may be important ininducing this effect.

The Antioxidant and Glutathione PrecursorN-acetylcysteine Prevents the Ability of the Dp44mT- orDpC-Fe Complexes to Decrease the p-STAT3/STAT3 Ratioin PANC-1 and MIAPaCa-2 Cells. Considering that incuba-tion of PANC-1, MIAPaCa-2, and DU145 cells with Dp44mT:Feor DpC:Fe also significantly decreased the p-STAT3/STAT3ratio (Fig. 1, B–D), we further examined whether the redoxactivity of these complexes (Yuan et al., 2004; Richardsonet al., 2006; Lovejoy et al., 2011) was at least partly respon-sible for this effect (Fig. 2). This hypothesis was investigatedby using the antioxidant N-acetylcysteine (NAC), which hasbeen well characterized to increase intracellular glutathione(GSH) and quench ROS (Schafer and Buettner, 2001; Wattsand Richardson, 2001). Previous studies have demonstratedthat NAC markedly reduces the antiproliferative activity ofDp44mT and prevents its ability to induce lysosomal mem-brane permeabilization due to its ability to act as a GSH donor(Lovejoy et al., 2011).In this investigation, cells were initially preincubated for

1 hour at 37°C with control medium alone or NAC (10 mM)dissolved in medium. Then, the cells were incubated withDp44mT:Fe (1 or 5 mM) or DpC:Fe (1 or 5 mM) in controlmedium alone, or in medium containing NAC (10 mM) fora further 24 hours at 37°C.Consistent with the results in Fig. 1B, incubation of PANC-1

cells with Dp44mT:Fe or DpC:Fe decreased the p-STAT3/STAT3ratio, particularly at 5 mM relative to 1 mM (Fig. 2A). How-ever, incubation of cells with Dp44mT:Fe (1 mM) or DpC:Fe(1 mM) in the presence of NAC prevented the decrease in thep-STAT3/STAT3 ratio in PANC-1 cells (Fig. 2A). This findingsuggests the ability of Dp44mT:Fe andDpC:Fe to inhibit STAT3activation could depend on their ability to generate ROS (Yuanet al., 2004). When the concentration of the Dp44mT:Fe andDpC:Fe complexes was increased to 5 mM, NAC no longer sig-nificantly prevented their ability to decrease the p-STAT3/STAT3ratio (Fig. 2A). This observation suggested ROS generation by thecomplexes at this higher concentration was markedly greaterthan that induced at 1 mM and could not be sufficiently rescuedby NAC.A similar result was observed with theMIAPaCa-2 cells (Fig.

2B), with bothDp44mT:Fe andDpC:Fe significantly (P, 0.01–0.05) decreasing the p-STAT3/STAT3 ratio relative to thecontrol, as shown in Fig. 1C. The presence of NAC inhibited thedecrease in the p-STAT3/STAT3 ratio for both Dp44mT:Fe(1 and 5 mM) and DpC:Fe (1 and 5 mM) (Fig. 2B). In examiningthe DU145 cells, we found that both Dp44mT:Fe and DpC:Fecomplexes also markedly (P , 0.01) reduced the p-STAT3/STAT3 ratio at the 5 mM concentration (Fig. 2C), which wasconsistent with the results in Fig. 1D. However, in contrast tothe PANC-1 and MIAPaCa-2 cells, incubation of the DU145cells with the Dp44mT:Fe or DpC:Fe complexes in the presenceof NAC did not significantly (P. 0.05) rescue the effect of theseagents on reducing the p-STAT3/STAT3 ratio (Fig. 2C).Considering these results, it could be suggested that DU145

cells have an altered redox state, which may modulate theirresponse to NAC and the complexes. Indeed, a previous studyshowed that DU145 cells have higher basal GSH levels and areducing redox environment when compared with other prostatecancer cells, enhancing their tolerability to ROS (Jayakumaret al., 2014). Alternatively, the Dp44mT:Fe and DpC:Fe com-plexes may have additional molecular effects in this cell linethat are not accounted for by ROS production alone.


at ASPE


ber 9, 2018m


Dow

nloaded from


DFO, Dp44mT, and DpC Inhibit STAT3 Dimerization.Activation of STAT3 by phosphorylation at Tyr705 induces di-merization, which allows subsequent nuclear translocation (Ihle,1995; Bromberg et al., 1999). Thus, native PAGE was used toassess whether the decrease in p-STAT3/STAT3 levels inducedby the ligands (Fig. 1, B–D) resulted in a reduction in STAT3dimers, as this is vital for STAT3 activity (Haura et al., 2005).Incubation of PANC-1, MIAPaCa-2 and DU145 cells with

DFO, Dp44mT, or DpC for 24 hours significantly (P , 0.001–0.01) decreased dimerized STAT3 levels compared with theuntreated controls (Fig. 3). Notably, Dp44mT and particularlyDpC were significantly (P , 0.05) more effective than DFO at

decreasing STAT3 dimer formation. Further, incubation ofthese cells with Dp44mT:Fe or DpC:Fe complexes also resultedin significantly (P , 0.001–0.01) decreased levels of dimerizedSTAT3, whereas incubation with either DFO:Fe, FeCl3, orDp2mT did not significantly (P . 0.05) alter dimerized STAT3levels compared with the control (Fig. 3).These findings agreewith the decrease in the p-STAT3/STAT3

ratio observed in Fig. 1, indicating that the ability of DFO toinhibit STAT3 dimer formation depends solely on its iron-binding activity. In contrast, the ability of Dp44mT and DpCto inhibit STAT3 dimer formation could also be dependent onROS generation in addition to their known ability to chelate

Fig. 2. The antioxidant NAC prevents the ability of Dp44mT- or DpC-Fe complexes to suppress STAT3 phosphorylation in pancreatic cancer cells. Thelevels of p-STAT3 and STAT3 protein are shown as determined by Western blotting in (A) PANC-1, (B) MIAPaCa-2, and (C) DU145 cells aftera preincubation with control mediumalone or NAC (10mM) for 1 hour at 37°C, followed by a second incubation with control medium alone, Dp44mT:Fe (2:1;1 or 5 mM), or DpC:Fe (2:1; 1 or 5 mM) in control medium or in the continued presence of NAC (10 mM) for 24 hours at 37°C. TheWestern blots are typical ofthree independent experiments, with densitometric analysis representing mean 6 S.D. Relative to control: *P , 0.05; **P , 0.01; ***P , 0.001.

548 Lui et al.

at ASPE


ber 9, 2018m


Dow

nloaded from


intracellular iron (Yuan et al., 2004; Whitnall et al., 2006;Kovacevic et al., 2011; Lovejoy et al., 2012). It is notable thatthe effect observed after treating cells with the ligands couldnot be reproduced by adding the agents directly to control celllysates, indicating a biologic response that required cellularand metabolic integrity.DFO, Dp44mT, and DpC Inhibit STAT3 DNA-Binding

Activity. To examinewhether the chelator-induced decrease inp-STAT3/STAT3 and STAT3 dimer levels resulted in decreasedSTAT3 DNA-binding activity, we performed an electrophoreticmobility shift assay using a biotin-labeled, double-strandedDNAprobe containing the STAT3 consensus-binding motif (Fig. 4). Inthese studies, incubation of nuclear extracts with this proberesulted in a pronounced band (Fig. 4A, lane 2) that was notpresent with the probe alone (Fig. 4A, lane 1).

To assess whether the STAT3 DNA-binding activity ob-served in Fig. 4A (lane 2) was a specific interaction, a series ofcontrols was used. First, addition of a 50-fold molar excess ofunlabeled specific competitor oligonucleotide (i.e., Spec. comp.;Fig. 4A) markedly and significantly (P , 0.001) reduced theintensity of this band (Fig. 4A, lane 3), whereas addition of a50-fold molar excess of a nonspecific competitor oligonucleotide(i.e., Nonspec. comp.; Fig. 4A) did not significantly (P . 0.05)alter the intensity of the STAT3 band compared with the con-trol in lane 2 (Fig. 4A, lane 4). Second, incubation with a biotin-labeled probe containing a mutated STAT3-binding motif didnot result in the appearance of any clear bands (Fig. 4A, lane 5).Third, addition of an antibody against STAT3 (i.e., Spec. Ab;Fig. 4A) to the binding reaction resulted in a significant (P ,0.001) reduction of the STAT3 band, but also the apparent

Fig. 3. DFO,Dp44mT, andDpC inhibit STAT3dimerization in pancreatic and prostate cancercells.We used native PAGE to determine STAT3dimer levels in (A) PANC-1, (B)MIAPaCa-2, and(C) DU145 cells after a 24-hour/37°C incubationwith control medium, DFO (250 mM), Dp44mT(5 mM), DpC (5 mM), DFO:Fe (1:1; 250 mM),Dp44mT:Fe (2:1; 5 mM), DpC:Fe (2:1; 5 mM),FeCl3 (250 mM), or Dp2mT (5 mM). TheWesternblots are typical of three independent experi-ments, with densitometric analysis representingmean 6 S.D. Relative to control: **P , 0.01;***P , 0.001.


at ASPE


ber 9, 2018m


Dow

nloaded from


presence of a smear that ran higher in the gel (see asterisk),suggesting an antibody-induced supershift of STAT3 (Fig. 4A,lane 6). Fourth, as a further control, incubation of an isotypecontrol antibody (i.e., control Ab; Fig. 4A) to the binding reactionunder the same conditions (Fig. 4A, lane 7) did not significantly(P . 0.05) alter the appearance of the band compared with thecontrol in lane 2 (Fig. 4A). Collectively, these controls demon-strate the identified STAT3 band was representative of STAT3DNA-binding activity.To assess the effect of the ligands on STAT3 DNA-binding

activity, PANC-1 cells were incubated with DFO (250 mM),Dp44mT (5 mM), or DpC (5 mM) for 24 hours at 37°C, nuclearlysates were prepared, and electrophoretic mobility shift as-say was performed (Fig. 4B). Incubation with these ligandssignificantly (P , 0.001) reduced the intensity of the STAT3band compared with control cells (Fig. 4B; compare lanes 3–5with lane 2).Moreover, this effectwas also observed inMIAPaCa-2(Fig. 4C) andDU145 (Fig. 4D) cells, indicating thatDFO,Dp44mT,and DpC could significantly (P , 0.001–0.01) reduce bindingof STAT3 to its target DNA in each cell line examined. Thus,these data agree with the effects of the ligands on STAT3 ac-tivation (p-STAT3/STAT3 ratio; Fig. 1) and STAT3 dimerization(Fig. 3), and demonstrate that they inhibit STAT3 DNA-bindingactivity, which may compromise its downstream oncogeniceffects.DFO, Dp44mT, and DpC Downregulate the Expres-

sion of the Downstream STAT3 Targets, c-myc, CyclinD1, and Bcl-2. Considering the results indicating these li-gands can decrease STAT3 activation, dimerization, and DNAbinding (Figs. 1–4), we examined whether this subsequentlyaffected the expression of multiple downstream STAT3 targetsthat regulate cell cycle progression and apoptosis, namely,c-myc, cyclin D1, and Bcl-2 (Darnell, 1997; Sasse et al., 1997;Zushi et al., 1998; Bromberg et al., 1999; Zhong et al., 1999). Inthese studies, incubation of cells with DFO (250 mM), Dp44mT(5 mM), or DpC (5 mM) for 24 hours at 37°C significantly (P ,0.001–0.05) reduced the protein levels of c-myc, and cyclinD1 in PANC-1 (Fig. 5A), MIAPaCa-2 (Fig. 5B), and DU145(Fig. 5C) cells. Notably, Bcl-2 was also significantly (P , 0.01)reduced by Dp44mT and DpC in each of the three cell types,whereas DFO only significantly (P, 0.01) reduced Bcl-2 levelsin DU145 cells (Fig. 5C). These findings are in agreement withprevious reports that cellular iron depletion using chelatorscan regulate cyclin D1 and Bcl-2 levels, which could mediatethe potent antiproliferative activity of these compounds (Gaoand Richardson, 2001; Yuan et al., 2004; Kovacevic et al., 2011;Dixon et al., 2013).The iron complex DFO:Fe did not significantly alter the

expression of these downstream targets compared with thecontrol, indicating that the ability of DFO to sequester cellulariron is important for its ability to inhibit STAT3 signaling andits downstream effectors. In contrast, incubation of cells with theDp44mT:Fe andDpC:Fe complexes significantly (P, 0.001–0.05)reduced expression of the downstream targets c-myc, cyclin D1,and Bcl-2 in all three cell-types (Fig. 5, A–C). Again, thissuggests that the redox activity of Dp44mT:Fe and DpC:Fe com-plexes could play a role in Dp44mT- and DpC-mediated suppres-sion of downstream STAT3 signaling.Use of FeCl3 as a control for the Dp44mT:Fe and DpC:Fe

complexes indicated that it did not significantly (P . 0.05)affect c-myc expression in PANC-1 or MIAPaCa-2 cells (Fig. 5,A and B), nor did it significantly (P , 0.05) affect cyclin D1 or

Bcl-2 expression in any cell types examined. Interestingly, forDU145 cells only, FeCl3 was found to significantly (P, 0.001)decrease c-myc expression (Fig. 5C). This observation may re-flect cell-type heterogeneity that results in enhanced sensitivityof c-myc to high intracellular iron levels in DU145 cells. For allcell types, incubationwith the negative control compoundDp2mT,which cannot bind metals (Yuan et al., 2004; Chen et al., 2012),resulted in no significant (P . 0.05) alteration in the levels ofc-myc, cyclin D1, or Bcl-2, demonstrating the key role of ironchelation by Dp44mT and DpC in decreasing the expression ofthese proteins (Fig. 5).To compare the activity of these agents in mortal, nontrans-

formed cells, we also examined the effects of DFO, Dp44mT,and DpC on these downstream target molecules in primaryPaEC cultures (Fig. 5D). In agreement with the results indicat-ing that iron chelation did not significantly alter STAT3 ac-tivation in these cells (Fig. 1E), c-myc and Bcl-2 expressionlevels were not significantly (P . 0.05) altered compared withthe control in response to any of the ligands examined (Fig. 5D).Interestingly, cyclin D1 expression was significantly (P, 0.001)increased in response to Dp44mT, DpC, Dp44mT:Fe, and DpC:Fe in PaEC (Fig. 5D), which may possibly reflect a prosurvivalresponse of these normal cells to the stress conditions inducedby these compounds.Pretreatment of PANC-1 and DU145 cells with Dp44mT

or DpC Inhibits IL-6–Induced Activation of STAT3. IL-6 isimplicated in the hyperactivation of STAT3 signaling in tumorprogression by stimulating this pathway through a receptor-mediated mechanism (Turkson and Jove, 2000). To investigatewhether cellular iron depletion could also inhibit IL-6–inducedactivation of STAT3, we preincubated PANC-1 and DU145cells with Dp44mT, DpC, or Dp2mT in serum-free medium orwith the serum-free medium alone for 24 hours at 37°C toremove cytokines/growth factors (Hahmand Singh, 2010), thenstimulated the cells with IL-6 (10 ng/ml) in the serum-freemediumor the serum-freemedium alone for 30minutes at 37°C(Fig. 6).Preincubation with the control medium alone followed by

incubation with IL-6 for 30 minutes markedly and signifi-cantly (P , 0.001) increased the p-STAT3/STAT3 ratio inPANC-1 (Fig. 6A; lane 2) and DU145 (Fig. 6B; lane 2) cellscompared with the same cells incubated with medium alone(Fig. 6, A and B; lane 1). Notably, the low p-STAT3 levels inFig. 6 relative to Fig. 1, are consistent with serum starvation(Hahm and Singh, 2010). Preincubation of cells with Dp44mTor DpC significantly (P , 0.001) inhibited IL-6–inducedactivation of STAT3 in PANC-1 and DU145 cells (Fig. 6, Aand B; lanes 4 and 6). In contrast to Dp44mT and DpC,Dp2mT, which does not bind metal ions (Yuan et al., 2004;Richardson et al., 2006; Chen et al., 2012; Sun et al., 2013),had no significant (P . 0.05) effect on STAT3 activationinduced by IL-6 relative to the control (Fig. 6, A and B; lane 8).In contrast to the PANC-1 andDU145 cell types, MIAPaCa-2

cells showed a different response to the ligands and incubationwith IL-6 (Fig. 6C). In fact, IL-6 stimulation of cells preincubatedwith Dp44mT or DpC resulted in a significant (P , 0.001–0.01)increase in p-STAT3/STAT3 levels compared with cells pre-incubated with control media alone (Fig. 6C; compare lane 2with lanes 4 and 6). This observation was unexpected giventhe previous results demonstrating that Dp44mT and DpCcould inhibit constitutive STAT3 activation and its down-stream targets in this particular cell line (Figs. 1C, 2B, 3B,

550 Lui et al.

at ASPE


ber 9, 2018m


Dow

nloaded from


4C, 5B) and may reflect tumor cell heterogeneity. Alterna-tively, this result may reflect a compensatory mechanism inthe MIAPaCa-2 cells in response to Dp44mT or DpC that

leads to a more marked IL-6 response, which is then subse-quently inhibited by these agents downstream, as demonstratedin Figs. 1C, 2B, 3B, 4C, and 5B.

Fig. 4. DFO, Dp44mT, and DpC inhibit STAT3 DNA-binding activity in pancreatic and prostate cancer cells. (A) Electrophoretic mobility shift assay(EMSA) examining STAT3 DNA-binding activity of PANC-1 cells in the presence of relevant controls to confirm specific DNA binding. Asterisk indicatesantibody-induced supershift of STAT3-DNA band. Control Ab, isotype control antibody; Non-spec. comp., nonspecific competitor oligonucleotide; Spec.Ab, antibody against STAT3; Spec. comp., specific competitor oligonucleotide. (B–D) EMSA examining STAT3 DNA-binding activity after incubation of(B) PANC-1, (C) MIAPaCa-2, and (D) DU145 cells for 24 hours at 37°C with control medium, DFO (250 mM), Dp44mT (5 mM), or DpC (5 mM). Nuclearlysates were then prepared and EMSA performed. The results are typical of three independent experiments, with densitometric analysis representingmean 6 S.D. Relative to control: **P , 0.01; ***P , 0.001.


at ASPE


ber 9, 2018m


Dow

nloaded from


The positive findings in PANC-1 and DU145 cells usingIL-6 (Fig. 6, A and B) were further confirmed by immunoflu-orescence staining that demonstrated a marked induction ofp-STAT3 expression (green fluorescence) upon incubation of

control cells with IL-6, which was predominantly nuclear asdemonstrated by the merged stain showing overlap with DAPIstaining (blue fluorescence; Fig. 7, A and B). Compared withtheir respective control cells, the mean p-STAT3 staining

Fig. 5. DFO, Dp44mT, and DpC downregulate the expression of downstream STAT3 targets. The expression of c-myc, cyclin D1, and Bcl-2 proteins in(A) PANC-1, (B) MIAPaCa-2, (C) DU145, or (D) PaEC cells after a 24-hour/37°C incubation with control medium, DFO (250 mM), Dp44mT (5 mM), DpC(5 mM), DFO:Fe (1:1; 250 mM), Dp44mT:Fe (2:1; 5 mM), DpC:Fe (2:1; 5 mM), FeCl3 (250 mM), or Dp2mT (5 mM). The Western blots are typical of threeindependent experiments, with densitometric analysis representing mean 6 S.D. Relative to control: *P , 0.05; **P , 0.01; ***P , 0.001.

552 Lui et al.

at ASPE


ber 9, 2018m


Dow

nloaded from


intensities in cells treated with IL-6 were significantly (P ,0.01) increased by 3.9-fold in PANC-1 cells and 3.5-fold inDU145 cells. However, stimulation with IL-6 of PANC-1 orDU145 cells preincubated with Dp44mT or DpC did not resultin marked p-STAT3 staining, with mean p-STAT3 stainingintensities that were not significantly different compared withthe control cells (Fig. 7, A and B).Using MIAPaCa-2 cells, the mean p-STAT3 staining was also

significantly (P , 0.05) increased 2.1-fold upon treatment withIL-6 when compared with their respective control cells (Fig. 7C).However, as demonstrated in Fig. 6C, preincubation of thesecells with either Dp44mT or DpC resulted in a significantly

(P, 0.01–0.05) greater 2.4-fold and 2.5-fold increase in p-STAT3levels in response to IL-6 treatment, respectively, when com-pared with the IL-6–treated control cells (Fig. 7C). Together, theresults from immunoblotting and fluorescencemicroscopy demon-strate that pretreatment of PANC-1 and DU145 cells withDp44mT or DpC inhibits IL-6–induced STAT3 activation, andthis process is stimulated by these ligands in MIAPaCa-2 cells.DFO, Dp44mT, andDpCDecrease Activation of Kinases

Upstream of STAT3, Namely, Src and cAbl. It is possiblethat the observed decrease in STAT3 activation induced byDFO, Dp44mT, and DpC may be mediated by their effect onupstream kinases known to promote STAT3 signaling (Yu and

Fig. 6. Pretreatment of PANC-1 and DU145 cells with Dp44mT or DpC inhibits IL-6–induced STAT3 phosphorylation. The expression of p-STAT3 andSTAT3 protein levels as determined by Western blotting of (A) PANC-1, (B) DU145, and (C) MIAPaCa-2 cells after a 24-hour/37°C incubation withserum-free medium alone, Dp44mT (5 mM), or DpC (5 mM) in serum-free medium, followed by a 30-minute/37°C incubation with serum-free mediumalone or IL-6 (10 ng/ml) in serum-free medium. The Western blots are typical of three independent experiments, with densitometric analysisrepresenting mean 6 S.D. Relative to control: **P , 0.01; ***P , 0.001.


at ASPE


ber 9, 2018m


Dow

nloaded from


Richardson, 2011). To investigate this hypothesis, we examinedthe effect of these compounds on key protein tyrosine kinasesthat phosphorylate and activate STAT3, namely, the nonrecep-tor tyrosine kinases Src and cAbl, and the receptor-associatedJanus tyrosine kinase (JAK2) (Turkson et al., 1998; Haura et al.,

2005; Srinivasan et al., 2008; Hedvat et al., 2009). Specifically,activation of Src and cAbl were examined through assessinglevels of an activating phosphorylation of Src (p-SrcY416), aninactivating phosphorylation of Src (p-SrcY527) (Hunter, 1987),and an activating phosphorylation of cAbl (p-cAblY245) (Brasher

Fig. 7. Pretreatment of PANC-1 and DU145 cells with Dp44mT or DpC inhibits IL-6–induced nuclear translocation of p-STAT3. Immunofluorescencemicroscopy examining p-STAT3 staining was performed in (A) PANC-1 and (B) DU145 cells and (C) MIAPaCa-2 cells under the same conditions asdescribed in Fig. 6. Individual and merged images were taken to show staining of p-STAT3 (green) with the cell nucleus (blue) as stained by DAPI. Thescale bar in the bottom left corner of the first image represents 20 mm and is the same across all images. The results are typical of three independentexperiments.

554 Lui et al.

at ASPE


ber 9, 2018m


Dow

nloaded from


and Van Etten, 2000). Activation of JAK2 was also assessed byinvestigating the levels of phospho-JAK2Y1007/8 (p-JAK2), anactivating phosphorylation that occurs in response to receptor-mediated cytokine or growth factor stimulation (Feng et al.,1997; Hedvat et al., 2009). It is important to note that JAK2,rather than JAK1, was investigated because numerous studieshave identified that the JAK2/STAT3 cascade has beenspecifically implicated in pancreatic (Thoennissen et al., 2009;Corcoran et al., 2011) and other cancers (Xiong et al., 2008;Zhang et al., 2009; Colomiere et al., 2009; Marotta et al., 2011)as promoting tumor progression.Incubation with Dp44mT, DpC, Dp44mT:Fe, or DpC:Fe

significantly (P , 0.001–0.01) decreased the p-SrcY416/Srcratio relative to the control, with no significant (P . 0.05)alteration being observed with total Src levels in PANC-1(Fig. 8A), MIAPaCa-2 (Fig. 8B), and DU145 (Fig. 8C) cells. Incontrast, incubation of these cells with DFO, DFO:Fe, FeCl3,or Dp2mT did not significantly (P . 0.05) alter levels ofp-SrcY416 or total Src compared with the untreated control.Further, none of the treatments were found to significantly(P . 0.05) alter p-SrcY527 levels, indicating that these com-pounds did not affect Src activity by regulating the phosphor-ylation of this site. Examination of cAbl activity revealed thatincubation with Dp44mT, DpC, Dp44mT:Fe, and DpC:Fecould significantly (P , 0.001–0.05) reduce the p-cAbl/cAblratio with no significant (P . 0.05) alterations to total cAbllevels in MIAPaCa-2 (Fig. 8B) and DU145 (Fig. 8C) cells, butnot in PANC-1 cells (Fig. 8A). Interestingly, incubation ofDU145 cells with FeCl3 alone resulted in significantly (P ,0.05) decreased p-cAbl and cAbl levels (Fig. 8C). This resultappeared analogous to that seen with c-myc (Fig. 5C), andagain, may reflect an enhanced sensitivity of these moleculesto high cellular iron levels in DU145 cells (Li et al., 2011).In all three cell types, incubation with any of the chelators,

their corresponding iron complexes, FeCl3, or the negativecontrol compoundDp2mT did not significantly (P. 0.05) alterp-JAK2 or JAK2 levels (Fig. 8). This indicates that the abilityof iron chelators to inhibit STAT3 activation in these cellslikely does not occur through inhibition of constitutive JAK2activation.Collectively, these findings indicate the ability of Dp44mT,

DpC, Dp44mT:Fe, and DpC:Fe to inhibit STAT3 activationcould potentially be mediated by the upstream inhibition of Srcactivation in PANC-1, MIAPaCa-2, and DU145 cells as well asinhibition of cAbl activation in MIAPaCa-2 and DU145 cells.Dp44mT and DpC Inhibit STAT3 Signaling In Vivo in

a PANC-1 Tumor Xenograft Model. To further validatethe data observed in vitro, we used a previously establishedPANC-1 tumor xenograft model in nude mice that assessedthe antitumor activity of Dp44mT and DpC (Kovacevic et al.,2011). In this study, PANC-1 tumor xenografts were estab-lished in nude mice and allowed to grow to 90 mm3, whentreatment was initiated. Each group of mice (n 5 6) receivedeither the vehicle alone, Dp44mT (0.4 mg/kg, 5 days/week), orDpC (5 mg/kg, 5 days/week) intravenously (via tail vein) over22 days. At the end of the experiment, the mice weresacrificed, and tumors were harvested.In these studies, Dp44mT and DpC reduced PANC-1 tumor

growth to 29.9 6 8.8% (n 5 6) and 19.3 6 1.4% (n 5 6) of thecontrol, respectively, after 22 days. Tumor tissues were ex-amined by immunohistochemistry for p-STAT3 and STAT3levels (Fig. 9). The staining for p-STAT3 was predominantly

located in nuclei, and thus staining was quantified in eachsample as the percentage of DAB-positive nuclei to the total nu-clear area in each sample, as per established methods (Tuominenet al., 2010; Schneider et al., 2012; Jahangiri et al., 2013).Staining of p-STAT3 in control-treated tumor samples dis-played 10.09 61.73% (n 5 18, six tumor samples/treatmentgroup, with an average taken from three random images/sample)DAB-positive nuclei/total nuclear area. The Dp44mT- andDpC-treated tumor samples displayed significantly (P ,0.01–0.05) lower values of 5.60 6 2.07% (n 5 18) and 5.20 63.24% (n 5 18) DAB-positive nuclei/total nuclear area, respec-tively (Fig. 9A).IHC staining of total STAT3 in the control-treated tumor

samples was noticeably marked and intense in both thecytoplasm and nuclei, with an average IHC score of 10.89 61.77 (n 5 18). However, STAT3 staining was found to besignificantly (P , 0.001–0.01) lower in tumor samples frommice treated with Dp44mT or DpC, with IHC scores fromthese groups of 6.38 6 2.70 (n 5 18) and 5.50 6 2.06 (n 5 18),respectively (Fig. 9B). Collectively, these results demonstratethat Dp44mT and DpC can also inhibit STAT3 signaling invivo, which corroborates well with our findings in vitro.

DiscussionIn this study, we have further elucidated the antitumor mech-

anism of the novel di-2-pyridyl thiosemicarbazones Dp44mT andDpC, which are emerging as a promising class of anticanceragents that demonstrate marked and selective antitumor andantimetastatic efficacy in vitro and in vivo in a range of cancercell types (Whitnall et al., 2006; Kovacevic et al., 2011; Chenet al., 2012; Liu et al., 2012; Lovejoy et al., 2012; Sun et al.,2013). Here, for the first time, we have identified that theseagents can inhibit the STAT3 pathway, which could be a crucialmechanism by which they exhibit their potent antitumor activity(Fig. 10A).We have demonstrated that DFO,Dp44mT, andDpC, inhibit

the STAT3 pathway by markedly decreasing constitutive phos-phorylation and STAT3 activation in three cancer cell lines—PANC-1, MIAPaCa-2, and DU145 (Fig. 1, B–D)—but not inprimary cultures of mortal PaEC (Fig. 1E). We also showedthese compounds inhibit STAT3dimerization (Fig. 3), the bindingof nuclear STAT3 to its target DNA (Fig. 4, B–D), and subse-quently attenuate the expression of the downstream STAT3targets c-myc, cyclin D1, and Bcl-2 (Fig. 5, A–C). Theseeffects were dependent on the ability of the ligands to bindiron, because incubation of cells with the inert DFO:Fe com-plex, which is not redox active (Kalinowski and Richardson,2005), or the control compound Dp2mT, which does not bindmetal ions (Yuan et al., 2004; Chen et al., 2012), did not lead toinhibition.Further studies herein demonstrate that the redox-active

Dp44mT:Fe and DpC:Fe complexes inhibit STAT3 phosphor-ylation, dimerization, and expression of downstream STAT3targets. These observations suggest the ability of these com-plexes to form redox-active species within cells also plays arole in their STAT3-inhibitory properties. In fact, the de-creased STAT3 activation induced by Dp44mT:Fe and DpC:Fe could be preventedby theantioxidantNAC in thePANC-1andMIAPaCa-2 cells (Fig. 2), indicating that ROS generation is in-volved in mediating this effect. Interestingly, ROS can inhibitSTAT3 signaling by oxidation of conserved cysteines on JAK


at ASPE


ber 9, 2018m


Dow

nloaded from


and STAT proteins that decrease their activation, implicatingSTAT3 as a mediator of redox homeostasis (Mamoon et al.,2007; Li et al., 2010). More recently, STAT3 transcriptional

activity has been shown to be directly regulated by the redoxfunction of the APE1/Ref-1 endonuclease (Cardoso et al.,2012).

Fig. 8. DFO, Dp44mT, andDpC decrease activation of kinases upstream of STAT3. The levels of p-SrcY416, p-SrcY527, Src, p-cAbl, cAbl, p-JAK2, and JAK2weremeasured by Western blot analysis in (A) PANC-1, (B) MIAPaCa-2, and (C) DU145 cells after a 24-hour/37°C incubation with control medium, DFO (250 mM),Dp44mT (5 mM), DpC (5 mM), DFO:Fe (1:1; 250 mM), Dp44mT:Fe (2:1; 5 mM), DpC:Fe (2:1; 5 mM), FeCl3 (250 mM), or Dp2mT (5 mM). The Western blots aretypical of three independent experiments, with densitometric analysis representing mean 6 S.D. Relative to control: *P , 0.05; **P , 0.01; ***P , 0.001.

556 Lui et al.

at ASPE


ber 9, 2018m


Dow

nloaded from


Our studies demonstrating the effect of the DpT agents onthe STAT3 pathway through their ability to bind iron andredox cycle to generate ROS illustrate a unique mechanism ofaction that is very different compared with other drugs tar-geting thismolecule. In fact, previous attempts to target STAT3in cancer therapy have included direct chemical inhibitors ofSTAT3, JAK and Src, IL-6 receptor antagonists, antisensestrategies, decoy phosphopeptides, decoy duplex oligonucleotides,dominant negative proteins, RNA interference, chemopreventiveagents, andG-quartet oligodeoxynucleotides (Jing and Tweardy,2005; Sansone and Bromberg, 2012; Wang et al., 2012; Siveenet al., 2014). Recently, resistance to standard anticancer ther-apies has been correlated to constitutive or unabated acti-vation of STAT3, indicating that combination therapy withSTAT3 inhibitors may be important for patient treatment(Tan et al., 2014). Hence, this finding underlines the im-portance of developing new STAT3 inhibitors such as Dp44mTand DpC that demonstrate potent and selective antitumoractivity.

To investigate whether the decreased STAT3 activation ob-served with DFO, Dp44mT, and DpC was due to the reducedactivity of well known upstream activators of this protein, ourstudies examined the effect of these agents on JAK2, Src, andcAbl activation. These agents did not significantly alter consti-tutive phosphorylation of JAK2 in any of the three cell types(Fig. 8). However, Dp44mT and DpC were found to decreaseactivation of the nonreceptor kinase Src in all cell lines as wellas decrease activation of cAbl in MIAPaCa-2 and DU145 cells(Fig. 8). This observation suggests the ability of Dp44mT, DpC,Dp44mT:Fe, and DpC:Fe to inhibit STAT3 activation in thesecells could potentially be mediated by the upstream inhibitionof Src activation. However, considering that cAbl activationwas not significantly altered in response to Dp44mT or DpC inPANC-1 cells, and that DFO did not significantly alter Src andcAbl activation, these latter proteins may not be the only con-tributors to inhibiting the STAT3 pathway.In addition to inhibiting constitutive activation of STAT3 in

tumor cells, Dp44mT and DpC could inhibit IL-6–inducedactivation of STAT3 in PANC-1 andDU145 cells (Figs. 6 and 7).This finding is critical given that several studies have indi-cated that IL-6–mediated activation of STAT3 is a principalpathway implicated in promoting tumorigenesis (Sansone andBromberg, 2012). Interestingly, this effect was not observedwith MIAPaCa-2 cells, where Dp44mT and DpC enhanced sub-sequent IL-6–induced STAT3 phosphorylation compared withthe control (Figs. 6C and 7C). This latter observation was un-expected given that these compounds inhibit constitutive STAT3activation in MIAPaCa-2 cells, and may reflect cell-type geneticheterogeneity resulting in a differential response to the com-bined stimulus of Dp44mT or DpC with IL-6. Importantly, thiseffect is not pro-oncogenic because these cells are sensitive tothe antitumor activity of these compounds in vitro and in vivo(Kovacevic et al., 2011). In fact, Dp44mT and DpC markedlyinhibit STAT3 activation (Figs. 1C, 2B, 3B, 4C) as well as ex-pression of downstream tumorigenic effectors c-myc, cyclin D1,and Bcl-2 (Fig. 5B) in MIAPaCa-2 cells. Moreover, Dp44mTand DpC increase the expression of the cyclin-dependentkinase inhibitor p21, the apoptotic markers cleaved PARPand Bax, and the growth and metastasis suppressor NDRG1,which mediates the pronounced inhibition of proliferationby these compounds in MIAPaCa-2 cells (Kovacevic et al.,2011).In light of a recent study demonstrating the potential of

Dp44mT andDpC to be effective for pancreatic cancer treatmentin vivo (Kovacevic et al., 2011), we used tumors from micetreated with Dp44mT, DpC, or the vehicle alone and performedIHC staining to examine p-STAT3 and STAT3 levels. Bothp-STAT3 and STAT3 stainingwere significantly lower in tumorsfrom mice treated with Dp44mT and DpC compared with thevehicle (Fig. 9). These studies suggest Dp44mT andDpC preventtumor growth, in part, by inhibition of STAT3.Apart from the effect of these agents on the STAT3 pathway,

it has been demonstrated that these compounds have multiplemolecular targets in tumor cells, which leads to their markedantiproliferative efficacy (Fig. 10B). These include: 1) enhance-ment of ironmobilization from cells (Yuan et al., 2004; Richardsonet al., 2006; Lovejoy et al., 2011); 2) inhibition of iron uptakefrom transferrin [both (1) and (2)], leading to cellular iron de-pletion,which is essential for growth (Yuanet al., 2004;Richardsonet al., 2006; Lovejoy et al., 2011); 3) upregulation of the potentgrowth andmetastasis suppressor NDRG1 (Le and Richardson,

Fig. 9. Dp44mT andDpC inhibit STAT3 signaling in vivo in a PANC-1 tumorxenograft model. PANC-1 tumor xenografts were grown subcutaneously innude mice that were separated into three groups, with each group receivingDp44mT (0.4 mg/kg), DpC (5 mg/kg), or the vehicle control intravenously asdescribed in detail in Materials and Methods. Immunohistochemistry wasperformed on formalin-fixed tumor tissue sections to examine (A) p-STAT3 or(B) STAT3 protein expression in tumors obtained from mice treated with thevehicle control, Dp44mT, or DpC. Scale bar: 1 mm. Quantitation of p-STAT3staining and scoring of STAT3 staining was performed as described. Theanalysis of p-STAT3 or STAT3 staining represents themean6S.D. (n= 18; sixtumor samples/treatment group, with an average taken from three randomimages/sample). Relative to control: *P , 0.05; **P , 0.01; ***P , 0.001.


at ASPE


ber 9, 2018m


Dow

nloaded from


2004; Kovacevic et al., 2011; Kovacevic et al., 2013), leading toinhibition of the epithelial-mesenchymal transition (Chen et al.,2012); 4) lysosomal membrane permeabilization mediated bythe generation of redox active copper complexes (Yuan et al.,2004; Richardson et al., 2006; Lovejoy et al., 2011), resulting inimpaired autophagy (Gutierrez et al., 2014) and the induction ofapoptosis (Gutierrez et al., 2014; Sahni et al., 2014); 5) in-hibition of multiple oncogenic cellular signaling pathways suchas phosphoinositide 3-kinase/protein kinase B (PI3K/AKT) andtransforming growth factor-b (TGF-b) pathways and upregu-lation of tumor-suppressive phosphatase and tensin homologdeleted on chromosome 10 (PTEN) and SMAD4 (Dixon et al.,2013; Kovacevic et al., 2013); 6) inhibition of cyclin D1 expression(Kulp et al., 1996; Gao and Richardson, 2001; Nurtjahja-Tjendraputra et al., 2007); and 7) inhibition of the rate-limitingstep of DNA synthesis catalyzed by the iron-containing enzyme

ribonucleotide reductase (Yu et al., 2011). The relative con-tribution of each of these targets to the inhibition of tumor cellgrowth is yet to be deciphered, but may depend on the type andstage of the neoplasm. In fact, the multiple molecular targets ofthese agents probably explains their marked effect on a broadrange of aggressive tumor types (Whitnall et al., 2006; Kovacevicet al., 2011; Lovejoy et al., 2012) and their ability to overcomedrug resistance (Whitnall et al., 2006; Kovacevic et al., 2011;Lovejoy et al., 2012).Considering other molecular targets of thiosemicarbazones, it

is notable that the loss of STAT1 expression, which is closelyrelated to STAT3, corresponds to advanced-stage pancreaticcancer with lymph node metastasis and poor prognosis (Sunet al., 2014). Conversely, STAT1 signaling has also been de-scribed as a tumor promoter (Kovacic et al., 2006), and thusthe importance of STAT1 in pancreatic and prostate tumor

Fig. 10. (A) Schematic summary of the effect of thechelators on the STAT3 signaling pathway. The ligandsDFO, Dp44mT, and DpC, decrease (1) phosphorylatedand activated STAT3; (2) STAT3 dimerization; (3) STAT3-DNA binding; and (4) expression of the downstreamSTAT3 targets c-myc, cyclin D1, and Bcl-2. BothDp44mTand DpC can also inhibit activation of the upstream non-receptor tyrosine kinases Src and cAbl, suggesting thatthese molecules could potentially mediate, at least in part,the observed effects on STAT3 activation. (B) Schematic ofthe multiple effects of the DpT class of thiosemicarba-zones (in addition to the effects on the STAT3 signalingpathway) that lead to the inhibition of tumor cell pro-liferation. The effects of these agents include 1) enhance-ment of iron mobilization from cells; 2) inhibition of ironuptake from transferrin; 3) upregulation of the growth andmetastasis suppressor N-myc downstream regulated gene1 (NDRG1); 4) lysosomalmembrane permeabilizationmedi-ated by the generation of redox active copper complexesleading to impaired autophagy and the induction of apoptosis;5) inhibition ofmultiple oncogenic cellular signaling path-ways such as phosphoinositide 3-kinase/protein kinase B(PI3K/AKT) and transforming growth factor-b (TGF-b)pathways, and upregulation of tumor-suppressive phos-phatase and tensin homolog deleted on chromosome 10(PTEN) and SMAD4; (6) inhibition of cyclin D1 expression;and (7) inhibition of the rate-limiting step of DNA synthesiscatalyzed by the iron-containing enzyme ribonucleotidereductase (RR).

558 Lui et al.

at ASPE


ber 9, 2018m


Dow

nloaded from


progression remains to be more extensively characterized.However, it may be of interest to examine the potential roleof STAT1 in mediating the activities of thiosemicarbazonesin future studies.Our investigation is the first to examine in detail the regulation

of the STAT3 pathway by cellular iron in pancreatic and prostatecancer (Fig. 10A). Importantly, our findings provide furtherrationale for the use of novel thiosemicarbazones in chemother-apy.Moreover, because STAT3 inhibition increases the responseto conventional therapies and delays pancreatic cancer progression(Venkatasubbarao et al., 2013), Dp44mT and DpC could be aneffective chemotherapeutic strategy in combination with exist-ing anticancer agents.

Acknowledgments

The authors wish to thank Sanaz Maleki for her valuable assistancewith immunohistochemical staining.

Authorship Contributions

Participated in research design: Lui, Kovacevic, Richardson.Conducted experiments: Lui, Kovacevic, Menezes.Contributed new reagents or analytic tools: Richardson.Performed data analysis: Lui, Kovacevic, Richardson.Wrote or contributed to the writing of the manuscript: Lui, Kovacevic,

Menezes, Kalinowski, Merlot, Sahni, Richardson.

References

Assinder SJ, Dong Q, Kovacevic Z, and Richardson DR (2009) The TGF-beta, PI3K/Akt and PTEN pathways: established and proposed biochemical integration inprostate cancer. Biochem J 417:411–421.

Bowman T, Garcia R, Turkson J, and Jove R (2000) STATs in oncogenesis. Oncogene19:2474–2488.

Brasher BB and Van Etten RA (2000) c-Abl has high intrinsic tyrosine kinase activitythat is stimulated by mutation of the Src homology 3 domain and by autophos-phorylation at two distinct regulatory tyrosines. J Biol Chem 275:35631–35637.

Bromberg JF, Wrzeszczynska MH, Devgan G, Zhao Y, Pestell RG, Albanese C,and Darnell JE, Jr (1999) Stat3 as an oncogene. Cell 98:295–303.

Cardoso AA, Jiang Y, Luo M, Reed AM, Shahda S, He Y, Maitra A, Kelley MR,and Fishel ML (2012) APE1/Ref-1 regulates STAT3 transcriptional activity andAPE1/Ref-1-STAT3 dual-targeting effectively inhibits pancreatic cancer cell sur-vival. PLoS ONE 7:e47462.

Chao J, Synold TW, Morgan RJ, Jr, Kunos C, Longmate J, Lenz HJ, Lim D, ShibataS, Chung V, and Stoller RG et al. (2012) A phase I and pharmacokinetic study oforal 3-aminopyridine-2-carboxaldehyde thiosemicarbazone (3-AP, NSC #663249) inthe treatment of advanced-stage solid cancers: a California Cancer ConsortiumStudy. Cancer Chemother Pharmacol 69:835–843.

Chen Z, Zhang D, Yue F, Zheng M, Kovacevic Z, and Richardson DR (2012) The ironchelators Dp44mT and DFO inhibit TGF-b-induced epithelial-mesenchymal tran-sition via up-regulation of N-Myc downstream-regulated gene 1 (NDRG1). J BiolChem 287:17016–17028.

Colomiere M, Ward AC, Riley C, Trenerry MK, Cameron-Smith D, Findlay J, Ack-land L, and Ahmed N (2009) Cross talk of signals between EGFR and IL-6Rthrough JAK2/STAT3 mediate epithelial-mesenchymal transition in ovarian car-cinomas. Br J Cancer 100:134–144.

Corcoran RB, Contino G, Deshpande V, Tzatsos A, Conrad C, Benes CH, Levy DE,Settleman J, Engelman JA, and Bardeesy N (2011) STAT3 plays a critical role inKRAS-induced pancreatic tumorigenesis. Cancer Res 71:5020–5029.

Darnell G and Richardson DR (1999) The potential of iron chelators of the pyridoxalisonicotinoyl hydrazone class as effective antiproliferative agents III: the effect ofthe ligands on molecular targets involved in proliferation. Blood 94:781–792.

Darnell JE, Jr (1997) STATs and gene regulation. Science 277:1630–1635.Darnell JE, Jr, Kerr IM, and Stark GR (1994) Jak-STAT pathways and transcrip-tional activation in response to IFNs and other extracellular signaling proteins.Science 264:1415–1421.

Dixon KM, Lui GY, Kovacevic Z, Zhang D, Yao M, Chen Z, Dong Q, Assinder SJ,and Richardson DR (2013) Dp44mT targets the AKT, TGF-b and ERK pathwaysvia the metastasis suppressor NDRG1 in normal prostate epithelial cells andprostate cancer cells. Br J Cancer 108:409–419.

Feng J, Witthuhn BA, Matsuda T, Kohlhuber F, Kerr IM, and Ihle JN (1997) Acti-vation of Jak2 catalytic activity requires phosphorylation of Y1007 in the kinaseactivation loop. Mol Cell Biol 17:2497–2501.

Fu D and Richardson DR (2007) Iron chelation and regulation of the cell cycle: 2mechanisms of posttranscriptional regulation of the universal cyclin-dependentkinase inhibitor p21CIP1/WAF1 by iron depletion. Blood 110:752–761.

Gao J and Richardson DR (2001) The potential of iron chelators of the pyridoxalisonicotinoyl hydrazone class as effective antiproliferative agents, IV: The mech-anisms involved in inhibiting cell-cycle progression. Blood 98:842–850.

Goldstein LJ, Galski H, Fojo A, Willingham M, Lai SL, Gazdar A, Pirker R, Green A,Crist W, and Brodeur GM et al. (1989) Expression of a multidrug resistance gene inhuman cancers. J Natl Cancer Inst 81:116–124.

Gong W, Wang L, Yao JC, Ajani JA, Wei D, Aldape KD, Xie K, Sawaya R, and HuangS (2005) Expression of activated signal transducer and activator of transcription3 predicts expression of vascular endothelial growth factor in and angiogenic phe-notype of human gastric cancer. Clin Cancer Res 11:1386–1393.

Gutierrez E, Richardson DR, and Jansson PJ (2014) The anticancer agent di-2-pyridylketone 4,4-dimethyl-3-thiosemicarbazone (Dp44mT) overcomes prosurvivalautophagy by two mechanisms: persistent induction of autophagosome synthesisand impairment of lysosomal integrity. J Biol Chem 289:33568–33589.

Hahm ER and Singh SV (2010) Sulforaphane inhibits constitutive and interleukin-6-induced activation of signal transducer and activator of transcription 3 in prostatecancer cells. Cancer Prev Res (Phila) 3:484–494.

Haura EB, Turkson J, and Jove R (2005) Mechanisms of disease: Insights into theemerging role of signal transducers and activators of transcription in cancer. NatClin Pract Oncol 2:315–324.

Hedvat M, Huszar D, Herrmann A, Gozgit JM, Schroeder A, Sheehy A, Buettner R,Proia D, Kowolik CM, and Xin H et al. (2009) The JAK2 inhibitor AZD1480 potentlyblocks Stat3 signaling and oncogenesis in solid tumors. Cancer Cell 16:487–497.

Hunter T (1987) A tail of two src’s: mutatis mutandis. Cell 49:1–4.Ihle JN (1995) Cytokine receptor signalling. Nature 377:591–594.Jahangiri A, De Lay M, Miller LM, Carbonell WS, Hu YL, Lu K, Tom MW, PaquetteJ, Tokuyasu TA, and Tsao S et al. (2013) Gene expression profile identifies tyrosinekinase c-Met as a targetable mediator of antiangiogenic therapy resistance. ClinCancer Res 19:1773–1783.

Jain N, Zhang T, Fong SL, Lim CP, and Cao X (1998) Repression of Stat3 activity byactivation of mitogen-activated protein kinase (MAPK). Oncogene 17:3157–3167.

Jayakumar S, Kunwar A, Sandur SK, Pandey BN, and Chaubey RC (2014) Differ-ential response of DU145 and PC3 prostate cancer cells to ionizing radiation: roleof reactive oxygen species, GSH and Nrf2 in radiosensitivity. Biochim Biophys Acta1840:485–494.

Jing N and Tweardy DJ (2005) Targeting Stat3 in cancer therapy. Anticancer Drugs16:601–607.

Kalinowski DS and Richardson DR (2005) The evolution of iron chelators for thetreatment of iron overload disease and cancer. Pharmacol Rev 57:547–583.

Kamran MZ, Patil P, and Gude RP (2013) Role of STAT3 in cancer metastasis andtranslational advances. Biomed Res Int 2013:421821.

Kovacevic Z, Chikhani S, Lovejoy DB, and Richardson DR (2011) Novel thio-semicarbazone iron chelators induce up-regulation and phosphorylation of themetastasis suppressor N-myc down-stream regulated gene 1: a new strategy for thetreatment of pancreatic cancer. Mol Pharmacol 80:598–609.

Kovacevic Z, Chikhani S, Lui GY, Sivagurunathan S, and Richardson DR (2013) Theiron-regulated metastasis suppressor NDRG1 targets NEDD4L, PTEN, andSMAD4 and inhibits the PI3K and Ras signaling pathways. Antioxid Redox Signal18:874–887.

Kovacic B, Stoiber D, Moriggl R, Weisz E, Ott RG, Kreibich R, Levy DE, Beug H,Freissmuth M, and Sexl V (2006) STAT1 acts as a tumor promoter for leukemiadevelopment. Cancer Cell 10:77–87.

Krajewska M, Krajewski S, Epstein JI, Shabaik A, Sauvageot J, Song K, Kitada S,and Reed JC (1996) Immunohistochemical analysis of bcl-2, bax, bcl-X, and mcl-1expression in prostate cancers. Am J Pathol 148:1567–1576.

Kulp KS, Green SL, and Vulliet PR (1996) Iron deprivation inhibits cyclin-dependentkinase activity and decreases cyclin D/CDK4 protein levels in asynchronous MDA-MB-453 human breast cancer cells. Exp Cell Res 229:60–68.

Le NT and Richardson DR (2004) Iron chelators with high antiproliferative activityup-regulate the expression of a growth inhibitory and metastasis suppressor gene:a link between iron metabolism and proliferation. Blood 104:2967–2975.

Lee JT, Lehmann BD, Terrian DM, Chappell WH, Stivala F, Libra M, Martelli AM,Steelman LS, and McCubrey JA (2008) Targeting prostate cancer based on signaltransduction and cell cycle pathways. Cell Cycle 7:1745–1762.

Lewis HD, Winter A, Murphy TF, Tripathi S, Pandey VN, and Barton BE (2008)STAT3 inhibition in prostate and pancreatic cancer lines by STAT3 binding se-quence oligonucleotides: differential activity between 59 and 39 ends. Mol CancerTher 7:1543–1550.

Li L, Cheung SH, Evans EL, and Shaw PE (2010) Modulation of gene expression andtumor cell growth by redox modification of STAT3. Cancer Res 70:8222–8232.

Li Z, Tanaka H, Galiano F, and Glass J (2011) Anticancer activity of the iron facil-itator LS081. J Exp Clin Cancer Res 30:34.

Lian JP, Word B, Taylor S, Hammons GJ, and Lyn-Cook BD (2004) Modulation of theconstitutive activated STAT3 transcription factor in pancreatic cancer prevention:effects of indole-3-carbinol (I3C) and genistein. Anticancer Res 24:133–137.

Liu W, Xing F, Iiizumi-Gairani M, Okuda H, Watabe M, Pai SK, Pandey PR, HirotaS, Kobayashi A, and Mo YY et al. (2012) N-myc downstream regulated gene 1modulates Wnt-b-catenin signalling and pleiotropically suppresses metastasis.EMBO Mol Med 4:93–108.

Lovejoy DB, Jansson PJ, Brunk UT, Wong J, Ponka P, and Richardson DR (2011)Antitumor activity of metal-chelating compound Dp44mT is mediated by formationof a redox-active copper complex that accumulates in lysosomes. Cancer Res 71:5871–5880.

Lovejoy DB, Sharp DM, Seebacher N, Obeidy P, Prichard T, Stefani C, Basha MT,Sharpe PC, Jansson PJ, and Kalinowski DS et al. (2012) Novel second-generationdi-2-pyridylketone thiosemicarbazones show synergism with standard chemo-therapeutics and demonstrate potent activity against lung cancer xenografts afteroral and intravenous administration in vivo. J Med Chem 55:7230–7244.

Mamoon NM, Smith JK, Chatti K, Lee S, Kundrapu K, and Duhé RJ (2007) Multiplecysteine residues are implicated in Janus kinase 2-mediated catalysis. Bio-chemistry 46:14810–14818.

Marotta LL, Almendro V, Marusyk A, Shipitsin M, Schemme J, Walker SR,Bloushtain-Qimron N, Kim JJ, Choudhury SA, and Maruyama R et al. (2011) TheJAK2/STAT3 signaling pathway is required for growth of CD441CD242 stem cell-like breast cancer cells in human tumors. J Clin Invest 121:2723–2735.

Merlot AM, Kalinowski DS, and Richardson DR (2013a) Novel chelators for cancertreatment: where are we now? Antioxid Redox Signal 18:973–1006.


at ASPE


ber 9, 2018m


Dow

nloaded from


Merlot AM, Pantarat N, Menezes SV, Sahni S, Richardson DR, and Kalinowski DS(2013b) Cellular uptake of the antitumor agent Dp44mT occurs via a carrier/receptor-mediated mechanism. Mol Pharmacol 84:911–24.

Mora LB, Buettner R, Seigne J, Diaz J, Ahmad N, Garcia R, Bowman T, Falcone R,Fairclough R, and Cantor A et al. (2002) Constitutive activation of Stat3 in humanprostate tumors and cell lines: direct inhibition of Stat3 signaling induces apo-ptosis of prostate cancer cells. Cancer Res 62:6659–6666.

Noulsri E, Richardson DR, Lerdwana S, Fucharoen S, Yamagishi T, Kalinowski DS,and Pattanapanyasat K (2009) Antitumor activity and mechanism of action of theiron chelator, Dp44mT, against leukemic cells. Am J Hematol 84:170–176.

Nurtjahja-Tjendraputra E, Fu D, Phang JM, and Richardson DR (2007) Iron chela-tion regulates cyclin D1 expression via the proteasome: a link to iron deficiency-mediated growth suppression. Blood 109:4045–4054.

Richardson D, Ponka P, and Baker E (1994) The effect of the iron(III) chelator,desferrioxamine, on iron and transferrin uptake by the human malignant mela-noma cell. Cancer Res 54:685–689.

Richardson DR, Sharpe PC, Lovejoy DB, Senaratne D, Kalinowski DS, Islam M,and Bernhardt PV (2006) Dipyridyl thiosemicarbazone chelators with potent andselective antitumor activity form iron complexes with redox activity. J Med Chem49:6510–6521.

Richardson DR, Tran EH, and Ponka P (1995) The potential of iron chelators of thepyridoxal isonicotinoyl hydrazone class as effective antiproliferative agents. Blood86:4295–4306.

Sahni S, Bae DH, Lane DJ, Kovacevic Z, Kalinowski DS, Jansson PJ, and RichardsonDR (2014) The metastasis suppressor, N-myc downstream-regulated gene 1(NDRG1), inhibits stress-induced autophagy in cancer cells. J Biol Chem 289:9692–9709.

Sansone P and Bromberg J (2012) Targeting the interleukin-6/Jak/stat pathway inhuman malignancies. J Clin Oncol 30:1005–1014.

Sasse J, Hemmann U, Schwartz C, Schniertshauer U, Heesel B, Landgraf C,Schneider-Mergener J, Heinrich PC, and Horn F (1997) Mutational analysis ofacute-phase response factor/Stat3 activation and dimerization. Mol Cell Biol 17:4677–4686.

Schafer FQ and Buettner GR (2001) Redox environment of the cell as viewed throughthe redox state of the glutathione disulfide/glutathione couple. Free Radic Biol Med30:1191–1212.

Schneider CA, Rasband WS, and Eliceiri KW (2012) NIH Image to ImageJ: 25 yearsof image analysis. Nat Methods 9:671–675.

Scholz A, Heinze S, Detjen KM, Peters M, Welzel M, Hauff P, Schirner M, WiedenmannB, and Rosewicz S (2003) Activated signal transducer and activator of transcription3 (STAT3) supports the malignant phenotype of human pancreatic cancer. Gastro-enterology 125:891–905.

Shin DS, Kim HN, Shin KD, Yoon YJ, Kim SJ, Han DC, and Kwon BM (2009)Cryptotanshinone inhibits constitutive signal transducer and activator of tran-scription 3 function through blocking the dimerization in DU145 prostate cancercells. Cancer Res 69:193–202.

Siveen KS, Sikka S, Surana R, Dai X, Zhang J, Kumar AP, Tan BK, Sethi G,and Bishayee A (2014) Targeting the STAT3 signaling pathway in cancer: role ofsynthetic and natural inhibitors. Biochim Biophys Acta 1845:136–154.

Srinivasan D, Sims JT, and Plattner R (2008) Aggressive breast cancer cells aredependent on activated Abl kinases for proliferation, anchorage-independentgrowth and survival. Oncogene 27:1095–1105.

Sun J, Zhang D, Zheng Y, Zhao Q, Zheng M, Kovacevic Z, and Richardson DR (2013)Targeting the metastasis suppressor, NDRG1, using novel iron chelators: regula-tion of stress fiber-mediated tumor cell migration via modulation of the ROCK1/pMLC2 signaling pathway. Mol Pharmacol 83:454–469.

Sun Y, Yang S, Sun N, and Chen J (2014) Differential expression of STAT1 and p21proteins predicts pancreatic cancer progression and prognosis. Pancreas 43:619–623.

Tan FH, Putoczki TL, Stylli SS, and Luwor RB (2014) The role of STAT3 signaling inmediating tumor resistance to cancer therapy. Curr Drug Targets 15:1341–1353.

Thoennissen NH, Iwanski GB, Doan NB, Okamoto R, Lin P, Abbassi S, Song JH, YinD, Toh M, and Xie WD et al. (2009) Cucurbitacin B induces apoptosis by inhibitionof the JAK/STAT pathway and potentiates antiproliferative effects of gemcitabineon pancreatic cancer cells. Cancer Res 69:5876–5884.

Tuominen VJ, Ruotoistenmäki S, Viitanen A, Jumppanen M, and Isola J (2010)ImmunoRatio: a publicly available web application for quantitative image analysis

of estrogen receptor (ER), progesterone receptor (PR), and Ki-67. Breast Cancer Res12:R56.

Turkson J, Bowman T, Garcia R, Caldenhoven E, De Groot RP, and Jove R (1998)Stat3 activation by Src induces specific gene regulation and is required for celltransformation. Mol Cell Biol 18:2545–2552.

Turkson J and Jove R (2000) STAT proteins: novel molecular targets for cancer drugdiscovery. Oncogene 19:6613–6626.

Venkatasubbarao K, Peterson L, Zhao S, Hill P, Cao L, Zhou Q, Nawrocki ST,and Freeman JW (2013) Inhibiting signal transducer and activator of transcription-3increases response to gemcitabine and delays progression of pancreatic cancer.Mol Cancer 12:104.

Vincent A, Herman J, Schulick R, Hruban RH, and Goggins M (2011) Pancreaticcancer. Lancet 378:607–620.

Wadler S, Makower D, Clairmont C, Lambert P, Fehn K, and Sznol M (2004) Phase Iand pharmacokinetic study of the ribonucleotide reductase inhibitor, 3-aminopyridine-2-carboxaldehyde thiosemicarbazone, administered by 96-hour intravenous continuousinfusion. J Clin Oncol 22:1553–1563.

Wang T, Niu G, Kortylewski M, Burdelya L, Shain K, Zhang S, Bhattacharya R,Gabrilovich D, Heller R, and Coppola D et al. (2004) Regulation of the innate andadaptive immune responses by Stat-3 signaling in tumor cells. Nat Med 10:48–54.

Wang X, Crowe PJ, Goldstein D, and Yang JL (2012) STAT3 inhibition, a novel approachto enhancing targeted therapy in human cancers (review). Int J Oncol 41:1181–1191.

Watts RN and Richardson DR (2001) Nitrogen monoxide (NO) and glucose: un-expected links between energy metabolism and no-mediated iron mobilization fromcells. J Biol Chem 276:4724–4732.

Whitnall M, Howard J, Ponka P, and Richardson DR (2006) A class of iron chelatorswith a wide spectrum of potent antitumor activity that overcomes resistance tochemotherapeutics. Proc Natl Acad Sci USA 103:14901–14906.

Xie TX, Wei D, Liu M, Gao AC, Ali-Osman F, Sawaya R, and Huang S (2004) Stat3activation regulates the expression of matrix metalloproteinase-2 and tumor in-vasion and metastasis. Oncogene 23:3550–3560.

Xiong H, Zhang ZG, Tian XQ, Sun DF, Liang QC, Zhang YJ, Lu R, Chen YX,and Fang JY (2008) Inhibition of JAK1, 2/STAT3 signaling induces apoptosis, cellcycle arrest, and reduces tumor cell invasion in colorectal cancer cells. Neoplasia10:287–297.

Yu CL, Meyer DJ, Campbell GS, Larner AC, Carter-Su C, Schwartz J, and Jove R(1995) Enhanced DNA-binding activity of a Stat3-related protein in cells trans-formed by the Src oncoprotein. Science 269:81–83.

Yu Y and Richardson DR (2011) Cellular iron depletion stimulates the JNK and p38MAPK signaling transduction pathways, dissociation of ASK1-thioredoxin, andactivation of ASK1. J Biol Chem 286:15413–15427.

Yu Y, Suryo Rahmanto Y, Hawkins CL, and Richardson DR (2011) The potent andnovel thiosemicarbazone chelators di-2-pyridylketone-4,4-dimethyl-3-thiosemicarbazoneand 2-benzoylpyridine-4,4-dimethyl-3-thiosemicarbazone affect crucial thiol systemsrequired for ribonucleotide reductase activity. Mol Pharmacol 79:921–931.

Yuan J, Lovejoy DB, and Richardson DR (2004) Novel di-2-pyridyl-derived ironchelators with marked and selective antitumor activity: in vitro and in vivo as-sessment. Blood 104:1450–1458.

Zhang X, Yin P, Di D, Luo G, Zheng L, Wei J, Zhang J, Shi Y, Zhang J, and Xu N(2009) IL-6 regulates MMP-10 expression via JAK2/STAT3 signaling pathway ina human lung adenocarcinoma cell line. Anticancer Res 29:4497–4501.

Zhao S, Venkatasubbarao K, Lazor JW, Sperry J, Jin C, Cao L, and Freeman JW(2008) Inhibition of STAT3 Tyr705 phosphorylation by Smad4 suppresses trans-forming growth factor beta-mediated invasion and metastasis in pancreatic cancercells. Cancer Res 68:4221–4228.

Zhong H, De Marzo AM, Laughner E, Lim M, Hilton DA, Zagzag D, Buechler P,Isaacs WB, Semenza GL, and Simons JW (1999) Overexpression of hypoxia-inducible factor 1alpha in common human cancers and their metastases. CancerRes 59:5830–5835.

Zushi S, Shinomura Y, Kiyohara T, Miyazaki Y, Kondo S, Sugimachi M, HigashimotoY, Kanayama S, and Matsuzawa Y (1998) STAT3 mediates the survival signal inoncogenic ras-transfected intestinal epithelial cells. Int J Cancer 78:326–330.

Address correspondence to: Dr. Des R. Richardson, Department of Pathologyand Bosch Institute, University of Sydney, Sydney, New South Wales, 2006Australia. E-mail: [email protected]

560 Lui et al.

at ASPE


ber 9, 2018m


Dow

nloaded from

mailto:[email protected]


novel thiosemicarbazones regulate the signal...

Documents