Calpain-Mediated Integrin Deregulation as a Novel Mode of ......mediated mechanism as a novel mode...

Cancer Biology and Signal Transduction

Calpain-Mediated Integrin Deregulation as a Novel Mode ofAction for the Anticancer Gallium Compound KP46

Ute Jungwirth1, Johannes Gojo1, Theresa Tuder1, Gernot Walko2, Martin Holcmann1, Thomas Sch€ofl1,KarinNowikovsky3,NastasiaWilfinger3, SushillaSchoonhoven1,ChristianR.Kowol4,RosaLemmens-Gruber5,Petra Heffeter1,4, Bernhard K. Keppler4, and Walter Berger1,4

AbstractOn thebasis of enhanced tumor accumulation andbone affinity, galliumcompounds areunderdevelopment

as anticancer and antimetastatic agents. In this study, we analyzed molecular targets of one of the lead

anticancer gallium complexes [KP46, Tris(8-quinolinolato)gallium(III)] focusing on colon and lung cancer.

Within a fewhours,KP46 treatment at lowmicromolar concentrations induced cell body contraction and loss of

adhesion followed by prompt cell decomposition. This rapid KP46-induced cell death lacked classic apoptotic

features and was insensitive toward a pan–caspase inhibitor. Surprisingly, however, it was accompanied by

upregulation of proapoptotic Bcl-2 family members. Furthermore, a Bax- but not a p53-knockout HCT-116

subline exhibited significant KP46 resistance. Rapid KP46-induced detachment was accompanied by down-

regulation of focal adhesion proteins, including several integrin subunits. Loss of integrin-b1 and talin plasma

membrane localization corresponded to reduced binding of RGD (Arg–Gly–Asp) peptides to KP46-treated

cells. Accordingly, KP46-induced cell death and destabilization of integrins were enhanced by culture on

collagen type I, a major integrin ligand. In contrast, KP46-mediated adhesion defects were partially rescued by

Mg2þ ions, promoting integrin-mediated cell adhesion. Focal adhesion dynamics are regulated by calpains via

cleavage of multiple cell adhesionmolecules. Cotreatment with the cell-permeable calpain inhibitor PD150606

diminished KP46-mediated integrin destabilization and rapid cell death induction. KP46 treatment distinctly

inhibited HCT-116 colon cancer xenograft in vivo by causing reduced integrin plasma membrane localization,

tissue disintegration, and intense tumor necrosis. This study identifies integrin deregulation via a calpain-

mediated mechanism as a novel mode of action for the anticancer gallium compound KP46. Mol Cancer

Ther; 13(10); 2436–49. �2014 AACR.

IntroductionMetal compounds have a long history as therapeutic

agents, especially in oncology. However, since the suc-cessful clinical approval of cisplatin, platinum com-pounds have dominated the field of anticancermetal drugresearch. More recently, other metals have also movedinto focus for drug development due to their tumor-targeting properties, redox-based activationmechanisms,and unexpected molecular targeting characteristics (1, 2).Major aims of these attempts are to optimize the thera-peutic activity, reduce unwanted adverse effects, andprevent resistance development (3).

The development of gallium compounds for anticancertherapy is especially interesting due to the fact that gal-lium shares several characteristics with iron. On the basisof the enhanced needs of the rapidly proliferating malig-nant tissue for iron, gallium compounds harbor intrinsictumor-targeting properties by preferential uptake via thetransferrin/transferrin receptor system (4). Moreover,gallium drugs interfere with several iron-containingenzymes, including ribonucleotide reductase, essentialfor the maintenance of the malignant phenotype. In addi-tion, gallium nitrate has been approved for the treatment

1Department of Medicine I, Institute of Cancer Research and Comprehen-sive Cancer Center, Medical University of Vienna, Vienna, Austria. 2Depart-ment of Biochemistry and Cell Biology, Max F. Perutz Laboratories, CentreforMolecular Biology, University of Vienna, Vienna, Austria. 3Department ofInternalMedicine I, AnnaSpiegel Centre of Translational Research,MedicalUniversity Vienna, Vienna, Austria. 4Institute of Inorganic Chemistry andResearch Platform "Translational Cancer Therapy Research," University ofVienna, Vienna, Austria. 5Department of Pharmacology and Toxicology,University of Vienna, Vienna, Austria.

Note: Supplementary data for this article are available at Molecular CancerTherapeutics Online (http://mct.aacrjournals.org/).

Current address for U. Jungwirth: Breakthrough Breast Cancer ResearchCentre, The Institute of Cancer Research, 237 Fulham Road, London, SW36JB, United Kingdom; current address for J. Gojo, Department of Pediat-rics, Medical University of Vienna, W€ahringer G€urtel 18-20, 1090 Vienna,Austria; and current address for G. Walko, Centre for Stem Cells andRegenerative Medicine, King's College London School of Medicine, Guy'sHospital, Great Maze Pond, London SE1 9RT, United Kingdom.

U. Jungwirth and J. Gojo contributed equally to the main findings of thisarticle.

Corresponding Author:Walter Berger, Department of Medicine I, InstituteofCancerResearch,MedicalUniversity of Vienna,Borschkegasse8a, 1090Vienna, Austria. Phone: 43-1-40160-57555; Fax: 43-1-40160-957555;E-mail: [email protected]

doi: 10.1158/1535-7163.MCT-14-0087

�2014 American Association for Cancer Research.

MolecularCancer

Therapeutics

Mol Cancer Ther; 13(10) October 20142436

on April 22, 2021. © 2014 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst July 31, 2014; DOI: 10.1158/1535-7163.MCT-14-0087

http://mct.aacrjournals.org/

of cancer-related hypercalcemia and was proved to exertpositive effects onbone turnover andosteolysis inpatientswith multiple myeloma (5) and bone metastases (6).Currently, the orally bioavailable galliummaltolate and

tris(8-quinolinolato)gallium(III) (KP46) are in early clini-cal development as anticancer agents (7–9). KP46 wasdesigned to optimize hydrolytic stability and membranepenetration abilities. A recent X-ray absorption studyproved high stability under physiologic conditions in cellculture media and even in tissue samples of treated mice(10). Antihypercalcemic as well as antitumor effects weredemonstrated in theWalker carcinosarcoma 256model inrats (11). A first-in-human phase I study revealed favor-able tolerability and preliminary evidence of activityagainst renal cancer (9). Moreover, gallium compounds,including KP46 and gallium nitrate, preferentially accu-mulate in the bone (12), reflected by the use of radioactivegallium compounds as PET tracers (13). This enhancedbone accumulation also suggests direct activity againstprimary bone tumors and metastases.Although tumor-targeting mechanisms of KP46 are

well characterized, the molecular mechanisms underly-ing cancer cell death induction by KP46 are widely unex-plored. In this study,we elucidated for thefirst time loss ofintegrin-mediated cell adhesion followed by rapid celldeath induction by a caspase-independent but a calpain-promoted mechanism as a novel mode of action for thegallium compound KP46.

Materials and MethodsReagentsKP46 [tris(8-quinolinolato)gallium(III)] was synthe-

sized according to previously described methods (12) atthe Institute of Inorganic Chemistry, University of Vienna(Vienna, Austria). For in vitro studies, compounds weredissolved in dimethyl sulfoxide (DMSO) and diluted intoculture media. Maximum DMSO content was 0.3%. Allother substanceswerepurchased fromSigma-Aldrich.Allsolutions were freshly prepared before use.

Cell cultureThe following cancer cell lines were used in this study:

the human non–small cell lung cancer (NSCLC) cell linesA549, A427, SW1537, and Calu-6; the human colon carci-noma cell lines SW480 and Caco2 and the murine coloncancer cell model CT-26; the human leukemic cell modelHL60; the human osteosarcoma cell lines HOS and MG63(all obtained between 1995 to 2009 from American TypeCulture Collection); the human small cell lung carcinomacell line GLC-4 (generously provided in 1995 by Dr. DeVries, University Groningen, Groningen, The Nether-lands; ref. 14); the colon carcinoma cell model HCT-116;and respective sublines with deleted p53 or Bax genes(generously provided in 2005 by Dr. Vogelstein, JohnHopkinsUniversity, Baltimore,MD). HCT-116was growninMcCoy’s, SW480 inminimumessentialmedium(MEM),A427 cells in MEM supplemented with pyruvate andnonessential amino acids, CT-26 in DMEM/F12, and all

other cell lines in RPMI-1640. Culture media were supple-mented with 10% fetal calf serum (FBS, PAA). All cellswere cultured at 37�C in humidified atmosphere and 5%CO2. The cell lines were authenticated in all cases by arraycomparative genomic hybridization (Agilent; 44k humanwhole genome DNA arrays) as published previously (15)and/or STR fingerprinting before the start of this study.

Cytotoxicity and clonogenic assaysImpact of drug exposure on cell viability was deter-

mined using a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide (MTT)-based vitality assay (EZ4U;Biomedica; ref. 16). Experiments were carried out intriplicate in growth media with 10% FBS and repeatedthree times. For viability assays with magnesium (Mg2þ),mediumwas replaced with serum-free medium (contain-ing standard 0.4 mmol/L Mg2þ) or supplemented withMg2þ to 3.2 mmol/L and cells were exposed to KP46for 24 hours. For vitality assay with calpain inhibitors,cells were preexposed to PD150606 for 1 hour before KP46treatment. For clonogenicity assays, 6-well plates werecoated with calf skin collagen type I (Sigma) for 24 hoursand washed twice with phosphate-buffered saline (PBS).Then, 1,000 cells per well were seeded in uncoated andcoated wells, treated for 24 hours with increasing KP46concentrations and, after drug removal, cultured for 7 to12 days until clones had reached the optimal size formicroscopical evaluation after crystal violet staining aspublished previously (17). Experiments were performedtwice in duplicate.

Annexin V/PI stainingAfter 24 and 48 hours of KP46 treatment, cells were

stained with Annexin V (Annexin V–FITC; BD Bios-ciences) and propidium iodide (200 ng/mL PI; Sigma-Aldrich). Cells were analyzed according to the manufac-turer’s protocol using fluorescence-activated cell sorting(FACSCalibur; BectonDickinson). CellQuest Pro software(Becton Dickinson and Co.) was used to analyze the data.Experiments were repeated three times.

Western blot analysisCells were exposed to KP46, and total protein lysates

were prepared, resolved by SDS/PAGE, and transferredonto a polyvinylidene difluoride membrane for Westernblotting as published previously (16, 18). The p53 anti-body was from Thermo Scientific (clone DO-1). The fol-lowing antibodies were purchased from Cell SignalingTechnology: Bid, Bax, Bim, and Bak from the Pro-Apo-ptosis Bcl-2 Family Sampler Kit (#9942), P-myosin lightchain (Ser19) and myosin light chain from the MyosinLight Chain 2 Sampler Kit (#9776), all integrin subunitsexcept b1 from the Integrin Sampler Kit (#4749), all Rho/Rac antibodies from the Rho-GTPase Sampler Kit (#9968),p21Waf1/Cip1 from the Cell Cycle Regulation SamplerKit (#9932), focal adhesion kinase (FAK; #13009). Theantibodies against talin and integrin-b1 were from Sigma(#T3287) and Becton Dickinson (#610467), respectively.

Integrin Deregulation by the Anticancer Gallium Compound KP46

www.aacrjournals.org Mol Cancer Ther; 13(10) October 2014 2437




Secondary antibodies labeled with horseradish peroxi-dase (Santa Cruz Biotechnology) were used at workingdilutions of 1:10,000. Western blot bands were quantifiedwith QuantiScan software (Biosoft).

Live cell imagingTime-lapse video microscopy was performed using an

AxioObserver Z1 microscope coupled to AxioCam MRm(Carl Zeiss MicroImaging) equipped with phase contrastoptics. Cells were plated in 6-well dishes at a density of1� 105 cells per well and kept in the appropriate mediumduring thewhole period of observation. Control andKP46(10 mmol/L)-treated cells were monitored in parallel in aPM S1 incubator (Carl Zeiss MicroImaging) at 37�C and5% CO2 using the "mark and find" module of AxioVision(version 4.8.1) image analysis software (19). Recordingsstarted 20 hours after plating and 1 hour after addition ofthe drug. Frames were taken with an EC Plan-Neofluar10�/0.3NA objective lens in 10-minute intervals over aperiod of up to 24 hours. Images were processed withZeiss AxioVision 4.8.1 image analysis software and fur-ther analyzed with ImageJ (NIH, Bethesda, MD).

Immunofluorescence staining and confocalmicroscopy

For immunofluorescence staining, cells were treated asindicated. After washing with PBS, cells were fixed with3% (w/v) paraformaldehyde in PBS followed by permea-bilization for 7 minutes with 0.5% Triton-X100 (v/v) inPBS. For mitochondrial staining, MitoTracker Red (Invi-trogen)was used according to themanufacturer’s instruc-tion. For membrane staining, rhodamine-labeled wheatgerm agglutinin (WGA; Vector Labs) was used. Briefly,cells were incubated with 5 mg/mL rhodamine-WGA for10minutes at 37�C.Mitochondria- andmembrane-stainedcells were washedwith PBS and fixed as described above.FITC-phalloidin was used at a dilution of 1:250 (Sigma).Primary antibodies were diluted in 1% BSA/PBS (w/v) atthe following dilutions: a-tubulin (1:200; Sigma), talin(1:100; Sigma), and integrin-b1 (1:200; BD; #610467).Anti-mouse IgG-FITC (1:500; Sigma; #F2012) and anti-rabbit-Dylight549 (1:500; Santa Cruz Biotechnology) wereused as secondary antibodies. In all cases, controls with-out or with isotype-specific control antibodies instead ofthe first antibody were used. After mounting with Vec-torshield (containing DAPI), samples were examinedunder a confocal laser scanning microscope (Zeiss InvertAxio Observer.Z1, Two-channel LSM 700 URGB). Digitalimages were acquired and processed with the Zeiss ZEN(2011) software. Stacks (scan zoom 1.0, z distance 1.5 mm)were recorded at the lowest focal plane set close to thebasal cell membrane.

Adhesion assayCells were detached with Accutase (Sigma) and

allowed to adhere for 1 hour on plates coated with colla-gen-1 (Sigma; C8919; 0.5 mg/well) in medium withoutserum but supplemented with 1% BSA (w/v). In short-

term assays, cells were treated with KP46 at the indicatedconcentration during the period of adhesion. Otherwise,cells were treated with KP46 and PD150606 for 14 hoursbefore adhesion assay. After cell adhesion, wells werewashed twice with PBS, fixed with methanol, and stainedwith crystal violet. Differences in cell adhesion werequantified using ImageJ software.

RGD peptide bindingHL60 cells were treated with indicated KP46 concen-

trations for 6 hours. Afterward, cells were incubated withFITC-labeled RGD (Arg–Gly–Asp) peptide (Anaspec) inbinding buffer (10 mmol/L HEPES, 150 mmol/L NaCl,and 1%w/vBSA) for 20minutes andwashedwith PBS. PIwas used to exclude dead cells. Binding of the FITC-labeled RGD peptide to the cell surface HL60 leukemiccells was measured by fluorescence-activated cell sorting(FORTESSA; BD Biosciences).

Intracellular calcium levelsHCT-116 cells were loaded with FURA 2AM (Sigma-

Aldrich) and an equivalent of pluronic 20% (MolecularProbes) at room temperature for 30 to 45 minutes. Afterwashout, cells were allowed to equilibrate for 30 minutes.KP46wasdiluted to thebathing solution in anappropriateconcentration and experiments were performed as previ-ously described (20). Ratiometric measurements wererealized following background subtractionwith the AxonImaging Workbench 2.2 software (Axon Instruments)averaging two frames. Results are presented as the rela-tive change (�DF/Fo) of the F340 nm/F380 nm signal.

In vivo experimentsSix- to 8-week-old female SCID/BALB/c mice were

purchased from Harlan. For xenograft experiments, ani-malswere kept in a pathogen-free environment and everyprocedure was done in a laminar airflow cabinet. Theexperiments were done according to the Federation ofLaboratory Animal Science Association (FELASA) guide-lines for the use of experimental animals and approved bythe Ethics Committee for the Care and Use of LaboratoryAnimals at the Medical University Vienna and the Min-istry of Science and Research, Austria.

Solid tumor xenograft modelFor local tumor growth experiments, 5 � 105 HCT-116

cells were injected subcutaneously into the right flank ofmice. Animals were randomly assigned to treatmentgroups and therapy started when tumors were palpable.Mice were treated with KP46 (15 mg/kg dissolved in 5%DMSO and diluted inmedium) for 2weeks, five times perweek, intraperitoneal (i.p.). The control group received100 mL solvent. Animals were controlled for distressdevelopment daily and tumor sizewas assessed regularlyby caliper measurement. Tumor volume was calculatedas [(length � width2) � 0.5; ref. 21]. Body weight wasdetermined before drug administration and recordedregularly during the experiment. Tissue sections were

Jungwirth et al.

Mol Cancer Ther; 13(10) October 2014 Molecular Cancer Therapeutics2438




paraffin-embedded and hematoxylin and eosin (H&E)stained by routine procedures. Immunohistochemicalstaining for integrin-b1 and detection of DNA breaksby terminal deoxynucleotidyltransferase dUTP nick endlabeling (TUNEL) assay were performed as publishedpreviously (22). The percentage of mitotic figures anddead cells (TUNEL-positive þ TUNEL-negative, nonmi-totic cells with condensed chromatin) was determinedmicroscopically in nonnecrotic tumor regions (at least6 optical fields from 4 different tumors/animals) usingDAPI/TUNEL staining.

Statistical analysisWhere not mentioned otherwise, data are expressed as

mean � SEM. Results were analyzed and illustrated withGraphPad Prism (version 5; GraphPad Software). Statis-tical analyses were performed using one- and two-wayANOVAwith drug treatment, time, concentration, or celltype as independent variables and conducted with Bon-ferroni posttests to examine the differences between thedifferent drug treatment regimens and the diverseresponses. A P value of 0.05 was considered statisticallysignificant.

ResultsAnticancer activity of KP46 in vitro: minor impact ofthe p53 statusIn an initial in vitro screen, KP46 proved to be highly

active against 22 cell lines from different tumor entitieswith IC50 values for 72-hour drug exposure generally inthe low mmol/L range (Supplementary Fig. S1A). Besidescell lines from melanoma (23) and osteosarcoma (manu-script in preparation), also A427 (NSCLC) and HCT-116p53/wt (colon cancer) cellswere among themost sensitive

cell models (IC50 values of 0.5 and 1.2 mmol/L, respec-tively). In contrast, theNSCLCcell lineA549 and the coloncancer cell line SW480were relatively KP46 resistant (IC50

values above 3.0 mmol/L). Differences of the KP46 IC50

values were highly significant for A427 versus A549 andfor HCT-116 versus SW480 cells (both Student t test, P <0.001) and not a consequence of different proliferativecapacities (data not shown). Thus, we decided to focus onthesemajor tumor entities (IC50 values for a cell line panelfrom these tumor entities are shown in Table 1) and tocompare A549 and SW480 as "KP46-resistant" cells incontrast to HCT-116 and A427 as "KP46-sensitive"representatives.

The tumor-suppressor p53 was recently suggested tobe an important factor concerning KP46-induced celldeath (24). However, we did not find any significantdifferences in KP46 activity exerted against multiplep53 wild-type (wt) and p53-mutated/knockout cancercell lines (Supplementary Fig. S1A). In addition, theKP46 IC50 values of an isogenic HCT-116 cell line pairwith p53 wt and knockout status were comparable(Table 1), excluding the p53 status as a general deter-minant of KP46 responsiveness.

Caspase-mediatedapoptosis isnot essential for rapidcell death execution induced by KP46

Time-dependency experiments revealed that the num-ber of viable cells was reduced by up to 75% in thesensitive cell lines already after 24 hours of exposure to2.5 mmol/L KP46, and even the viability of the resistantcell models was inhibited by >50% at 10 mmol/L (Fig. 1A).Surprisingly, however, Annexin V staining revealed anunexpectedly low percentage of apoptotic cells (Fig. 1B).Thus, solely in the highly sensitive cell lines a rathermoderate increase in the percentage of Annexin Vþ cells

Table 1. IC50 values of KP46 tested against diverse lung and colon cancer cell models

Cell line Histology p53 status IC50 KP46 (mmol/L � SD)

LungA549 NSCLC (AC) wt 3.89 � 0.87A427 NSCLC (AC) wt 0.40 � 0.23SW1573 NSCLC (AC) wt 1.74 � 1.15Calu6 NSCLC (AC) mut 1.97 � 0.46VL-8 NSCLC (SCC) mut 2.30 � 0.06VL-10 NSCLC (SCC) unknown 1.89 � 0.66GLC-4 SCLC mut 2.73 � 0.09ColonHCT-116 p53/wt Colorectal wt 1.06 � 0.35HCT-116 p53/ko Colorectal ko 1.14 � 0.29HCT-116 bax/ko Colorectal wt 1.77 � 0.44SW480 Colorectal (AC) mut 3.02 � 0.93CaCo2 Colorectal (AC) mut 2.30 � 0.13CT-26a Colon carcinoma mut 2.17 � 0.31

aOriginating from mouse, all others human.






Figure 1. KP46 induces rapid cell detachment and loss of viability without massive apoptosis induction. A, relative viability of KP46-sensitive and -resistant lungand colon cancer cell lines untreated (100%) or following treatment with the indicated KP46 concentrations for 24 hours was determined by microscopicalcounting inTrypanbluesolution.B, inductionof apoptosisbyKP46wasanalyzedbyFACSofAnnexinVþcells treatedas inA.Meanpercentages�SDofAnnexinVþ (PI� and PIþ) cells were determined from two independent experiments. C, impact of the pan–caspase inhibitor Z-VAD-FMK on the anticancer activityof KP46 exerted by 24 hours of exposure was determined by MTT assay in the indicated KP46-sensitive cell models. (Continued on the following page.)

Jungwirth et al.





(both PIþ and PI�) was detectable after 24 hours of KP46treatment, while the effects in the resistant cell models didnot even reach significance. Accordingly, mitochondrialmembrane depolarization as an earlymarker of apoptosisinduction was not enhanced in response to KP46 (Sup-plementary Fig. S1B). In contrast to other anticancer drugslike doxorubicin (data not shown), the pan–caspase inhib-itor Z-VAD-FMK did not significantly reduce the cyto-toxic activity of KP46 (Fig. 1C). Surprisingly, however, wedetected in KP46-treated cells an early upregulation ofproapoptotic proteins, such as Bim, Bax, and Bak (Sup-plementary Fig. S2A), loss of antiapoptotic Bcl-2, andtranslocation of Bax into the mitochondria (Supplemen-tary Fig. S2B). In case of Bim EL, additionally a KP46-induced shift of the respective protein band could bedetected.Accordingly, but in contrast to thep53knockout,a Bax-deleted HCT-116 subline displayedmoderate resis-tance against KP46-induced cell death (SupplementaryFig. S2C and Table 1). Together, our data suggest thatKP46 induces a form of cell death different from classiccaspase-mediated apoptosis that is supported by p53-independent activation of proapoptotic Bcl-2 familymembers.

Rapid KP46-induced cell death involves cell-detachment by loss of integrin-mediated focaladhesion contactsTo further characterize this unusual form of cancer

cell death, we monitored cell morphology duringKP46 treatment for 24 hours using time-lapse micros-copy. Pronounced morphologic changes were alreadydetected after short-term exposure to KP46 (Fig. 1D andE). A427 and HCT-116 cells were found to retract cellbodies and round-up rapidly, whereas A549 and SW480cells detached slowly (Fig. 1F and G). The retractionphase was accompanied by highly dynamic formationand loss of branching filopodia (Fig. 1H). Although thedetached cells at this stage were still viable (proved bythe lack of PI accumulation; Fig. 1I), the process endedin rapid and massive cell disintegration.Next, confocal immunofluorescence microscopy was

performed to follow dynamics of the cytoskeletal cha-nges during KP46-induced cell death (Fig. 2). Phalloidin-staining of the actin filament network revealed thatKP46-treated A427 cells had lost most of the centralstress fibers, whereas peripheral actin bundles wereunaffected (Fig. 2A and B). Orthographic projectionsgenerated from optical sections through short-termKP46-treated cells immunostained for a-tubulin revea-led the perikaryal region bulging out of the extremely

flattened cell body (Fig. 2C and D). Similar observationswere made using WGA as an unspecific membrane stain(Fig. 2E and F). The contracting cells left back severalplasma membrane remedies ripped-off from retractionfibers (Fig. 2F).

Rapid cell detachment by KP46 was accompanied byaltered expression of cytoskeletal proteins involved inadhesion and contraction. Thus,KP46 treatmentdistinctlyreduced the levels of several integrin subunits with thestrongest effect on integrin-b1, especially in KP46-sensi-tive cell models (Fig. 2G, top). Time-course analysesrevealed that downregulation of integrin-b1 and its focaladhesion–specific binding partner talin occurred alreadyat 6-hour KP46 treatment (Fig. 2G, bottom and Supple-mentary Fig. S3A), paralleling the loss of cell adhesion.Expression levels of integrin-b1 were variable in thedifferent cell lines but KP46-mediated loss was less pro-nounced in KP46-resistant A549 and SW480 cells (Sup-plementary Fig. S3A) but massive in KP46-sensitive oste-osarcoma cell lines (Supplementary Fig. S3B). In all cases,phosphorylation of the myosin light chain was upregu-lated by KP46 (HCT-116 cells representatively in Fig. 2G),indicative of increased actin–myosin–mediated cell con-traction (25).

Immunofluorescence microscopy proved that KP46-treated A427 cells were devoid of peripheral focal adhe-sion–associated integrin-b1 (Fig. 2H and I) and talin (Fig.2J and K) localization. Costaining with F-actin (Supple-mentary Fig. S3C) again depicted the complete loss ofintegrin and talin staining on the cell surface. Loss of cellsurface integrin exposurewas alsoprovedbydeterminingbinding of FITC-labeled RGD peptide, the major bindingsite for several integrins in various extracellular matrix(ECM) proteins (26). KP46 treatment for 6 hours signifi-cantly reduced peptide binding to approximately 60%(Fig. 2L). HL60 leukemic cells were used in these experi-ments to avoid proteolytic cleavage of integrins duringtrypsinization.

Culture on collagen sensitizes while Mg2þ ionsprotect against KP46

Integrins are the cellular receptors for ECM compo-nents resulting in activation of outside-in signaling,promoting survival and proliferation (26, 27). Hence,we investigated whether cultivation of KP46-sensitivecancer cells on collagen type I, a major ligand forintegrins, alters sensitivity against KP46. Indeed, plat-ing on collagen resulted in significantly enhanced sen-sitivity against KP46 in both A427 and HCT-116 cells(Fig. 3A). In addition, we tested expression of a panel of

(Continued.) D and E, morphologic changes in cell shape as a consequence of KP46 exposure (10 mmol/L) were monitored over 24 hours by time-lapsemicroscopy in the lung cancer cell models A427 and A549 (D) and in the colon cancer cells HCT-116 and SW480 (E). Different time points are shownin selected pictures. Arrowheads, cell rounding events; stars, cell divisions. F and G, distribution analysis of the time needed for full cell rounding (without celldivision). KP46-sensitive and -resistant lung (F) andcoloncancer (G) cellmodels are compared.H, imagesofA427cells treatedwithKP46 (10mmol/L) for 6 (left)and 8 hours (right; �80 objectives). Arrows indicate highly dynamic branching filopodia and arrowheads membranous bubble formation immediatelybefore cell disintegration. I, percentage of dead cells (PIþ) induced by the indicated KP46 concentrations after 16 hours of drug exposure. Statistical analysiswas performed by two-way ANOVA and Bonferroni posttest (�, P < 0.05; ��, P < 0.01; ��, P < 0.001); bars, 20 mm.






integrin-a and -b subunits in cells grown on cultureplastic as compared with collagen type I and the impactof 5 hours of KP46 treatment (Fig. 3B, top). Collagentended to result in upregulated expression of integrin-b5 but distinctly reduced levels of a5 and b1 subunits.Integrin-a5 and several b subunit (b1, b4, b5) levels were

markedly reduced by KP46 treatment. This integrindestabilization was, at least in HCT-116 cells, distinctlystronger in cultures on collagen. Also, reduced expres-sion levels of integrin outside-in signaling molecules(28) were detectable after KP46 treatment, includingFAK and members of the Rho/Rac GTPase family. Rho

F-Actin

F-Actin

α-Tubulin

α-Tubulin

Membrane

Membrane

Control

Control Talin TalinKP46

Integrin β1 Integrin β1KP46HCT-116

HCT-116

Integrin β1

A C E

B

G

L

D

H I

J K

F

Integrin β1

p-mlc (S19)

β-Actin

β-Actin

KP46 (h)

KP46 (h)

Talin

A427

1.0

1.0 1.0 1.1 0.5

1.0 1.4 1.6 0.6

1.0 1.3 1.3 1.5

0 1 3

1.4

1.2

1.0

0.8

0.6

0.4

0.20.0

0.0 2.5KP46 (mmol/L)

Fo

ld c

han

ge

inR

GD

-pep

tid

e b

ind

ing

10.0

6

0 024 24

0.4 1.0 0.6

Figure 2. Cytoskeletal changes andloss of integrin-b1–mediated focaladhesion contacts caused byKP46 treatment. Representativeconfocal microscopy pictures ofuntreated (A, C, E, H, and J) andKP46-treated (10 mmol/L, 4 hours)A427 cells (B, D, F, I, and K). A andB, F-actin staining in control andtreated cells was performed usingFITC-phalloidin. Note the reducedstress fibers and cortical actin inKP46-treated cells. A single opticalsection is shown. C and D, theorthographic projection of z-stacked images of cells stained fora-tubulin (1.5 mmZ-stack distance)shows nucleus protrusion in thetreated cells (arrow in D) in contrastto flattened control cells (C). E andF, cell membranes were stainedwithWGA-rhodamine. Arrowheadscell membrane remnants after cellcontraction. G, Western blotanalysis of protein extractsprepared from HCT-116 and A427cell lines treated with 10 mmol/LKP46 for 24 hours (top) or HCT-116for the indicated time points(bottom). P-mlc, phosphorylationof the myosin light chain at residueserine 19. H and I, integrin-b1- and(J and K) talin-staining in A427 cellstreated with KP46 for 4 hours.Arrows, peripheral localization ofintegrin-b1and talin in control cells.Costaining of actin and nuclei aredepicted in Supplementary Fig. S2.Bars, 20 mm. L, impact (given asfold change) of KP46 treatment(6 hours) on binding of the FITC-labeled RGD peptide to the surfaceof HL60 cells was determined byFACS analysis. Statistical analysiswas performed by ANOVA withBonferroni posttest (��, P < 0.01).

Jungwirth et al.





A was downmodulated in both cell lines but FAK onlyin A427 cells in a collagen-dependent manner. The effectwas even more pronounced for Rho B in the HCT-116cell line. In contrast, the negative cell cycle–regulatedp21 was slightly enhanced by KP46 (Fig. 3B, bottom).These data suggest that KP46 treatment leads to awidespread loss of ECM adhesion molecules. Conse-quently, we decided to determine the impact of KP46 oncell adhesion dynamics. KP46 treatment for 1 hoursignificantly inhibited the re-adhesion of A427 cells (Fig.

3C). Again, this effect was much stronger in the KP46-sensitive than in the -resistant cell models (A549 in Fig.3D). Because divalent metal cations are essential forintegrin function and facilitate adhesion (29), we testedwhether enhanced levels of Mg2þ could antagonizeadhesion destabilization by KP46. Indeed, addition ofMg2þ inhibited the effect of KP46 (Fig. 3C and D, whitecolumns). Supplementation of the serum-free mediumwith Mg2þ also significantly reduced the cytotoxicityinduced by short-term KP46 exposure (Fig. 3E and F).

Figure 3. Impact of collagen type Iand Mg2þ ions on KP46-mediatedloss of viability and cell adhesion.A, cells (103/well, 6-well plates)were seeded either on uncoatedcell culture plastic (-coll) or onwellscoated with collagen type I (þcoll),treated for 24 hours with theindicated KP46 concentrationsand, after drug removal, culturedfor another 7 to 10 days beforemethanol fixation and crystal violetstaining. Clones with >10 cellswere counted microscopically.A representative experiment withHCT-116 cells (left) and thequantitative evaluation for A427and HCT-116 cells (right) areshown. B, the impact of a 5-hourKP46 treatment (10 mmol/L) ofA427 and HCT-116 cells onexpression of the indicated integrinsubunits (left) and cell adhesionmolecules as well as the cell-cycleinhibitor p21 was analyzed.Cultures on culture plastic arecompared with those on collagentype I–coated wells (coll). Theb-actin blot shown in this figure is arepresentative of all the individualb-actin blots with similar results. Cand D, impact of KP46 exposure (1hour) at the indicatedconcentrations on re-adhesionwas tested in A427 and A549 cells.E and F, the impact of Mg2þ

supplementation on the reductionof A427 and HCT-116 cell viabilitywas analyzed after 24 hourstreatment with KP46 at theindicated concentrations in serum-free medium. Statistical analysiswas performed by ANOVA withBonferroni posttest (�, P < 0.05;��, P < 0.01; ��, P < 0.001).






Calpain activity is a driver of KP46-induced celldeath

Migratory and invasive functions of cells are stronglylinked to turnover of focal adhesion complexes. Thisprocess is partially mediated and regulated by calpainsand interestingly several cell adhesion and migration-regulating proteins deregulated by KP46 (integrin-b sub-units, talin, FA, RhoA) represent calpain substrates (30,31). Furthermore, calpain activation is connected toincreased intracellular calcium levels. In accordance witha previous study (24), also in our experiments KP46 led toa slow but distinct increase of intracellular calcium inHCT-116 cells, in contrast to the rapid increase induced bythe Ca2þ ionophore ionomycin used as a positive control(Supplementary Fig. S4). Therefore, we further analyzedwhether inhibition of calpain influences the anticanceractivity of KP46. Indeed, KP46-induced cytotoxicity wasdistinctly inhibitedby the cell-permeable calpain inhibitorPD150606 (50 mmol/L; Fig. 4A and B). In addition, time-lapse microscopy proved that cells cotreated withPD150606 detached later from the cell culture plate sur-face than cells solely treated with KP46 (Fig. 4C andSupplementary Video S1). In contrast, a weakly mem-brane-permeable calpain inhibitor (E-64, 20 mmol/L)exerted distinctly minor effects (Supplementary Fig.S5), indicating that cell-associated calpain activity drivesKP46-induced detachment. Accordingly, coincubationwith the calpain inhibitor PD150606 completely rescuedbinding of the labeledRGDpeptide to the surface ofKP46-treated HL60 cells, indicating restoration of functionalintegrin complexes at the cell surface (Fig. 4D). To showthat calpain is also involved in the KP46-induced celldetachment of adherent cells, re-adhesion assays after14-hour drug treatment with or without PD150606 wereperformed. Similar to short time experiments, KP46strongly inhibited re-adhesion, which was again par-tially rescued by PD150606 (Fig. 4E). As mentionedabove, KP46 treatment weakly enhanced markers ofclassic apoptosis in the hypersensitive cell models, only.Nevertheless, even the low levels of KP46-induced earlyand late apoptosis in KP46-hypersensitive cell modelscould be significantly inhibited by cotreatment withPD150606 (Fig. 4F). Likewise, upregulation of Bimexpression and the band shift of Bim EL in Westernblot analysis was reduced in the presence of the calpaininhibitor (Fig. 4G).

KP46 is active in vivo: loss of tissue integrity andintegrin membrane localization

The in vivo anticancer activity of KP46 was analyzed inthe solid human colon cancer xenograftmodelHCT-116 inSCID mice. Subcutaneous tumor growth was distinctlyretarded by KP46 therapy (Fig. 5A), resulting in signifi-cantly reduced tumor weight as compared with controls(day 18; Fig. 5B). Histologic sections stained with H&E aswell as TUNEL assay for apoptotic cell death induction(Fig. 5C, top and middle, respectively) revealed massiveareas of cancer cell necrosis induced by KP46 treatment

that stained strongly positive in TUNEL assays. A com-parison of viable tumor parts at highermagnification (Fig.5C, bottom) indicated distinct loss of tissue integrity withreduced cell density in response to KP46. Althoughmitot-ic figures were significantly reduced, TUNEL-positiveand -negative dead cells (with condensed chromatin inDAPI staining) were dispersed throughout the nonnecro-tic parts of KP46-treated tumors (Fig. 5C and D). Inaccordance with the in vitro data, membrane localizationof integrin-b1 immunostaining was distinctly reduced inthe remaining viable parts of the KP46-treated HCT-116xenografts (Fig. 5E), indicating that downregulation ofintegrin-b1–mediated focal adhesion complexes mightalso play an essential role in the anticancer activity ofKP46 in vivo.

DiscussionAlthough the (pre)clinical development of the oral

gallium compound KP46 has already advanced to a suc-cessful phase I study (9), the precise molecular targets incancer cells are widely unexplored. Here, we describerapid cancer cell death induction by calpain-mediatedfocal adhesion deregulation as a novel molecular mode-of-action for KP46. Although promoted by activation ofproapoptotic Bcl-2 family proteins, KP46-induced celldeathwas lacking classic features of apoptosis. Significantin vivo anticancer activity of KP46 against a human coloncancer xenograft was also characterized by reduced integ-rin-b1 plasma membrane localization, tissue disintegra-tion, and enhanced cell death induction.

One recent study proposed that KP46-mediated cyto-toxic effects are triggered via Ca2þ signaling–mediatedp53 activation or a p53-independent activation of FAS-mediated extrinsic apoptosis (24). However, despite con-firmation of the slow Ca2þ increase by KP46, we foundneither an impact of the p53 status on the overall antican-cer activity in a larger panel of cell lines from differentcancer types nor a significant impact of targeted p53 genedisruption in HCT-116 colon cancer cells. This does notgenerally preclude a regulatory role of p53 in KP46-medi-ated cell death, but excludes a key role of the p53mutationstatus in determining the overall level of KP46 respon-siveness of cancer cells.

In contrast, wewere puzzled by amassive loss of viablecells within the first 24 hours of KP46 exposure withoutappearance of an equivalent amount of apoptotic (ornecrotic) cells as would especially be expected from ap53-driven cell death (32). Consequently—at least in theKP46-hypersensitive cell models-–another mode of celldeath obviously had eradicated the majority of the malig-nant cells before execution of classic apoptosis. Indeed,time-lapsemicroscopy revealed an early onset of dramaticmorphologic changes uponKP46 treatmentwith cell bodycontraction followed by cell detachment. The cell contrac-tion and rounding-up of KP46-treated cells shared simi-larities with mitotic cell rounding (33). This was corrob-orated by increased phosphorylation at Ser-19 and, thus,

Jungwirth et al.





activation of the myosin light chain (34). However, incontrast to mitosis, no spindle formation, no M-phasearrest (data not shown), or chromatin condensation couldbe detected. Instead, after a convulsive period with rapid

appearance and loss of branching filopodia, cellscompletely disintegrated, explaining the massive loss ofviable cells without appearance of the respective amountof Annexin Vþ and/or PIþ cells.

A427A

C

D E

F G

B

A427KP46 KP46 + KP46

PD150606KP46 +PD150606

KP46 KP46 + KP46PD150606

KP46 +PD150606

7 h

21 h

0 h

14 h

KP46+ PD150606

1.25

1.00

0.75

0.50

0.25

0.00

1.25

1.00

0.75

Fo

ld s

urv

ival

Fo

ld s

urv

ival

0.50

0.25

0.00

HCT-116

0 1 2.5

HL60

Fo

ld c

han

ge

inR

GD

-pep

tid

e b

ind

ing

Fo

ld c

han

ge

Fo

ld c

han

ge

in a

dh

esio

n1.41.21.0

0.80.60.40.2

0.0

1.2

1.0

0.8

0.6

0.4

0.2

0.00 2.5

– PD150606

A427

4

3

2

1

0

Control

KP46

PD1506

06

KP46 +

PD15

0606

Control

KP46

PD1506

06

KP46 +

PD15

0606

Control

KP46

PD1506

06

KP46 +

PD15

0606

Control

KP46

PD1506

06

KP46 +

PD15

0606

HCT-116

Early apoptotic –– – + +

–+ + KP46PD150606

Bim EL

Bim LBim S

Late apoptotic

+ PD150606KP46 (mmol/L)

KP46 (mmol/L) KP46 (mmol/L)

KP46

KP46 +

PD15

0606

PD1506

0610.0

A427

5 10 0 1 2.5 5 10

10

8

6

4

2

0

Figure 4. Role of calpain activity inKP46-mediated loss of celladhesion and cell death induction.A and B, the impact of the cellpermeable calpain inhibitorPD150606 (50 mmol/L) on theviability of A427 and HCT-116 cellswas analyzed after 24 hoursKP46 treatment. C, A427 cellstreated with KP46 and cotreatedwith PD150606 were monitored bytime-lapse microscopy. Selectedphotomicrographs taken at theindicated time points are shown.Bars, 20 mmol/L. D, impact (givenas fold change) of KP46 on bindingof the FITC-labeled RGDpeptide tothe surface of HL60 cells in thepresence or absence of PD150606for 6 hours was determined byFACS analysis. Impact ofPD150606 (1 hour) on KP46 (14hours)-mediated inhibition of A427cell re-adhesion (E) and on low-level apoptosis induction by KP46(F) as determined by detectingearly-stage (AVþ/PI�) andlate-stage (AVþ/PIþ) apoptoticcells by FACS analysis. G, impactof KP46 (10 mmol/L, 8 hours)without and with PD150606 onexpression and gel migrationpattern of the indicated Bimisoforms was analyzed by Westernblot analysis in HCT-116 cells.Statistical analysis was performedbyANOVAwithBonferroni posttest(�, P < 0.05; ��, P < 0.01;��, P < 0.001). Bars, 20 mm.






On the basis of this profound impact on cell adhesionand survival, we hypothesized that KP46 might interferewith adhesion-dependent survival signal complexes.Major contributors to cell adhesion, especially to the ECM,are integrins. These families of a and b heterodimericreceptors regulate adhesion predominantly to the ECMand, in turn, cell viability by enabling cells to sense andrespond to their chemical and physical environment (26,27). Integrin deregulation is involved in many pathologicprocesses such as cardiovascular disease and cancer inva-sion (26, 27). Consequently, integrins are attractive targetsfor anticancer therapeutic interventions (35). Several of

our observations clearly demonstrate that KP46-mediatedcell detachment is critically involving loss of integrin-mediated focal adhesions. First, expression levels of integ-rin-a5 and several integrin-b subunits, together with theintracellular focal adhesion–specific bindingpartner talin,were significantly reduced, especially in KP46-hypersen-sitive cells in a time frame corresponding to cell detach-ment. Second, downregulation was accompanied by amassive loss of the appropriated integrin-b1 and talinlocalization to focal adhesion complexes. Third, KP46treatment led to a reduced FITC-labeled RGD peptidebinding of cells mimicking the integrin binding motif

Figure 5. In vivo anticancer activityof KP46 and impact on integrin-b1expression. A, impact of KP46 onxenograft growth of HCT-116p53/wt cells was determined aftersubcutaneous tumor cell injectionin SCIDmice. After the tumors werepalpable, mice were treated for2 weeks on 5 consecutive days/week with 15 mg/kg KP46 (N ¼ 5).Data are means � SEM. Statisticalanalysis was performed by two-way ANOVA with Bonferroniposttest (��,P <0.01; ��,P <0.001).B, tumor weights of control andtreatment groups on day 18 areshown. Box blot, inner quartilerange and median; whiskers,minimum to maximum; þ, mean.Statistical analysis was performedwith column statistics and t test(�, P < 0.05). C, H&E staining of arepresentative tumor of the solventand the treatment group (top) areopposed to photomicrographs ofTUNEL staining (middle, TUNEL inred, DAPI counterstain in bluefluorescence). Massive necroticareas of the treated tumor werestrongly reactive of the TUNELassay (red, middle). Bottom, onerepresentative optical field withinthe nonnecrotic parts of the tumorsdepicting mitotic figures in DAPIstaining (blue) and apoptotic cellsas TUNEL-positive (red).D, mitotic and apoptotic cells insolvent- and KP46-treated tumorswere counted in at least eightoptical fields from three tumorseach (�40 objective; Leica DMX).E, immunhistochemical staining ofintegrin-b1 in nonnecrotic tumorareas is shown for a control and aKP46-treated tumorrepresentatively.

Jungwirth et al.





(Arg–Gly–Asp) in ECM ligands such as collagen andfibronectin (26). Fourth, culture of cells on the majorintegrin ligand collagen I sensitized cancer cells againstKP46-induced cell death. Fifth, high-dose Mg2þ substitu-tion, a cation essential for integrin–ligand binding (29, 36),reduced the KP46-mediated re-adhesion block and alsoreduced cytotoxicity of the drug. Sixth, also expression ofproteins involved in focal adhesion dynamics down-stream of integrin outside-in signaling, such as FAK andmembers of the small GTPase Rho/Rac family (28), wasreduced by short-term KP46 treatment.Loss of integrin engagement and cell–ECM interac-

tions result in activation of a particular form of apo-ptotic cell death termed anoikis. This form of pro-grammed cell death is executed either via the intrinsicor extrinsic apoptosis pathways (37, 38). Furthermore,the promotion of BH3-only activators and sensitizers,such as Bim activation, as well as inhibition of Bcl-2 aredistinctive for this process (38, 39). Accordingly, wefound the levels of Bax, Bim, and Bak to be upregulatedafter short-term treatment with KP46. In addition, Baxwas distinctly translocated to mitochondria paralleledby a decrease in Bcl-2 levels. This corresponds to a mildbut significant KP46 resistance in a Bax- but surprising-ly not in a p53-deleted HCT-116 subline. However,hallmarks of classic apoptosis were only detected afterlonger drug exposure (48 hours; data not shown), whilethe massive viability loss within 24 hours was insensi-tive to a pan–caspase inhibitor. These data might indi-cate that KP46 targets an additional cellular factorinterfering with a downstream mechanism of KP46-induced anoikis and thus blocking or preceding caspaseactivation. Interestingly, also for cilengitide, an integrininhibitor already in advanced stage of clinical develop-ment, a comparable rapid cell detachment, massive cellloss within 24 hours, but lack of caspase activation orapoptosis induction has been described recently inhuman glioma cell lines (40). The rapid cilegitide-induced cell death was connected to autophagy. Asenhanced autophagic features were also found afterKP46 treatment (data not shown), we are currentlyaiming to dissect the contribution of this dual cell rescueor cell death–inducing mechanism.Nevertheless, the question remains how KP46 induces

integrin loss and degradation of focal adhesion com-plexes. The first observation coming into mind is that theligand 8-hydroxyquinoline is capable of chelating Mg2þ

ions.Deprival of this divalent cationwoulddefinitely leadto loss of the integrin–ligand interaction. Indeed, Mg2þ

overload reduced KP46-induced detachment and celldeath. On the contrary, KP46 has been demonstrated tobe extremely stable in biologic environments (10, 41),making the presence of substantial amounts of free ligandquestionable. In addition, even the formation of smallamounts of the 8-hydroxyquinolineMg2þ complexwouldnot distinctly reduce the extracellularMg2þ concentrationdue to the high excess of Mg2þ (400 mmol/L in the culturemedium) comparedwith the used KP46 concentrations in

the lowmmol/L range. These considerationsmake amajorrole for Mg2þ chelation in KP46-induced rapid cell deathunlikely.

KP46 treatment, in this study and in a previous one(24), induced a continuous Ca2þ release into the cytosolthat has been connected to activation of the Ca2þ chan-nel TRPC6 (42). This led our attention to calpains repre-senting a conserved family of Ca2þ-activated cysteineproteases that regulate cytoskeletal remodeling, cellularsignaling, apoptosis, and cell survival by controlledproteolysis. Also cell migration is a complex processincluding integrin-mediated (de)adhesion and actin-based membrane protrusion, which involve the activityof calpains (30). Calpain is associated with the controlof focal adhesion turnover by cleavage of integrinsand talin but also of integrin-recruited FAK and mem-bers of the Rho/Rac small GTPase family (30, 31).Moreover, calpain activity has been connected with thecleavage of Bcl-2 family members (43). Indeed, thepresence of the specific and cell-permeable calpaininhibitor (PD150606) reduced KP46-indcued cell round-ing and rapid cell death. In addition, a KP46-inducedband shift in the longer splice variant of Bim (Bim EL)was reduced by calpain inhibition. The longer isoformssuch as Bim EL are inactive by association with micro-tubules and the dynein light chain 1 (DLC-1) andreleased by certain stress stimuli to activate apoptosis(44). In another study, calcium-mediated calpain acti-vation led to focal adhesion disassembly and de-adhe-sion, which could be blocked by lanthanum (La3þ;ref. 45), a calcium channel blocker (46, 47). Supplemen-tation of the medium with lanthanum (La3þ) also in ourexperiments inhibited the anticancer activity of KP46(unpublished data). Together, these observations sug-gest a central role of Ca2þ release to the cytoplasm and,in turn, hyperactivation of calpain in KP46-induced celladhesion loss and rapid cell death induction.

In summary, this study elucidates disruption of integ-rin-mediated cell adhesion via a calpain-regulated pro-cess as a novel and targetedmode of action contributing tothe anticancer activity of the gallium compound KP46currently in early clinical development.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

Authors' ContributionsConception anddesign:U. Jungwirth, P.Heffeter, B.K.Keppler,W.BergerDevelopment of methodology: U. Jungwirth, G. Walko, W. BergerAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): U. Jungwirth, J. Gojo, T. Tuder, G. Walko,M. Holcmann, T. Sch€ofl, K. Nowikovsky, N. Wilfinger, S. Schoonhoven,R. Lemmens-Gruber, P. Heffeter, W. BergerAnalysis and interpretation of data (e.g., statistical analysis, biostatis-tics, computational analysis): U. Jungwirth, J. Gojo, T. Tuder, R. Lem-mens-Gruber, W. BergerWriting, review, and/or revision of themanuscript:U. Jungwirth, J. Gojo,T. Tuder, G. Walko, M. Holcmann, K. Nowikovsky, C.R. Kowol, R. Lem-mens-Gruber, P. Heffeter, B.K. Keppler, W. BergerAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): S. SchoonhovenStudy supervision: U. Jungwirth, P. Heffeter, W. Berger






AcknowledgmentsThe authors thank Christian Balcarek and Rosa-Maria Weiss for com-

petent technical assistance, Pakiza Rawnduzi for Ca2þ imaging, andGerhard Zeitler for animal care.

Grant SupportThe project was supported by the Austria Science Fund (FWF) grants

L212 and L568 and the Genome Austria GENAU program (PLACEBO; all

toW. Berger), theHerzfelder’sche Familienstiftung (to P.Heffeter) and theCOST action CM1105 (to W. Berger and B.K. Keppler).

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

Received February 10, 2014; revised July 3, 2014; accepted July 19, 2014;published OnlineFirst July 31, 2014.

References1. Sava G, Jaouen G, Hillard EA, Bergamo A. Targeted therapy vs. DNA-

adduct formation-guided design: thoughts about the future of metal-based anticancer drugs. Dalton Trans 2012;41:8226–34.

2. Jungwirth U, Kowol CR, Keppler BK, Hartinger CG, BergerW, HeffeterP. Anticancer activity of metal complexes: involvement of redoxprocesses. Antioxid Redox Signal 2011;15:1085–127.

3. Heffeter P, JungwirthU, JakupecM,Hartinger C,GalanskiM, Elbling L,et al. Resistance against novel anticancer metal compounds: differ-ences and similarities. Drug Resist Updat 2008;11:1–16.

4. Chitambar CR, Antholine WE. Iron-targeting antitumor activity ofgallium compounds and novel insights into triapine((R))-metal com-plexes. Antioxid Redox Signal 2013;18:956–72.

5. Warrell RPJr, Lovett D, Dilmanian FA, Schneider R, Heelan RT. Low-dose gallium nitrate for prevention of osteolysis in myeloma: results ofa pilot randomized study. J Clin Oncol 1993;11:2443–50.

6. Warrell RPJr, Alcock NW, Bockman RS. Gallium nitrate inhibits accel-erated bone turnover in patients with bone metastases. J Clin Oncol1987;5:292–8.

7. Bernstein LR, Tanner T, Godfrey C, Noll B. Chemistry and pharma-cokinetics of gallium maltolate, a compound with high oral galliumbioavailability. Metal Based Drugs 2000;7:33–47.

8. Collery P, Keppler B, Madoulet C, Desoize B. Gallium in cancertreatment. Crit Rev Oncol Hematol 2002;42:283–96.

9. Hofheinz RD, DittrichC, JakupecMA, Drescher A, JaehdeU, GneistM,et al. Early results from a phase I study on orally administered tris(8-quinolinolato)gallium(III) (FFC11, KP46) in patients with solid tumors–aCESAR study (Central European Society for Anticancer DrugResearch–EWIV). Int J Clin Pharmacol Ther 2005;43:590–1.

10. Hummer AA, Heffeter P, Berger W, Filipits M, Batchelor D, Buchel GE,et al. X-ray absorption near edge structure spectroscopy to resolve thein vivo chemistry of the redox-active indazolium trans-[Tetrachlorobis(1H-indazole)ruthenate(III)] (KP1019). J Med Chem 2013;56:1182–96.

11. Thiel M, Schilling T, Gey D, Ziegler R, Collery P, Keppler B. Tris(8-quinolinolato)gallium(III), a novel orally applied antitumor gallium com-pound. Relevance of TumorModels for Anticancer DrugDevelopment.Contrib Oncol Basel, Karger 1999;54:439–43.

12. Collery P, Domingo JL, Keppler BK. Preclinical toxicology and tissuegallium distribution of a novel antitumour gallium compound: tris (8-quinolinolato) gallium (III). Anticancer Res 1996;16:687–91.

13. Chitambar CR. Gallium-containing anticancer compounds. FutureMed Chem 2012;4:1257–72.

14. Zijlstra JG, de Vries EG,Mulder NH.Multifactorial drug resistance in anadriamycin-resistant human small cell lung carcinoma cell line. CancerRes 1987;47:1780–4.

15. Mathieu V, Pirker C, Schmidt WM, Spiegl-Kreinecker S, Lotsch D,Heffeter P, et al. Aggressiveness of human melanoma xenograftmodels is promoted by aneuploidy-driven gene expression deregu-lation. Oncotarget 2012;3:399–413.

16. Heffeter P, JakupecMA, Korner W, Chiba P, Pirker C, Dornetshuber R,et al.Multidrug-resistant cancer cells are preferential targets of the newantineoplastic lanthanum compound KP772 (FFC24). Biochem Phar-macol 2007;73:1873–86.

17. Lotsch D, Steiner E, Holzmann K, Spiegl-Kreinecker S, Pirker C,Hlavaty J, et al. Major vault protein supports glioblastoma survivaland migration by upregulating the EGFR/PI3K signalling axis. Onco-target 2013;4:1904–18.

18. Heffeter P, PongratzM,Steiner E,ChibaP, JakupecMA, Elbling L, et al.Intrinsic and acquired forms of resistance against the anticancer

ruthenium compound KP1019 [indazolium trans-[tetrachlorobis(1H-indazole)ruthenate (III)] (FFC14A). J Pharmacol Exp Ther 2005;312:281–9.

19. Walko G, Vukasinovic N, Gross K, Fischer I, Sibitz S, Fuchs P, et al.Targeted proteolysis of plectin isoform 1a accounts for hemidesmo-some dysfunction in mice mimicking the dominant skin blisteringdisease EBS-Ogna. PLoS Genet 2011;7:e1002396.

20. Kamyar MR, Kouri K, Rawnduzi P, Studenik C, Lemmens-Gruber R.Effects of moniliformin in presence of cyclohexadepsipeptides onisolated mammalian tissue and cells. Toxicol In Vitro 2006;20:1284–91.

21. Fischer H, Taylor N, Allerstorfer S, Grusch M, Sonvilla G, Holzmann K,et al. Fibroblast growth factor receptor-mediated signals contribute tothe malignant phenotype of non–small cell lung cancer cells: thera-peutic implications and synergismwith epidermal growth factor recep-tor inhibition. Mol Cancer Ther 2008;7:3408–19.

22. Hoda MA, Mohamed A, Ghanim B, Filipits M, Hegedus B, Tamura M,et al. Temsirolimus inhibits malignant pleural mesothelioma growth invitro and in vivo: synergism with chemotherapy. J Thorac Oncol2011;6:852–63.

23. Valiahdi SM, Heffeter P, Jakupec MA, Marculescu R, Berger W,Rappersberger K, et al. The gallium complex KP46 exerts strongactivity against primary explanted melanoma cells and induces apo-ptosis in melanoma cell lines. Melanoma Res 2009;19:283–93.

24. Gogna R, Madan E, Keppler B, Pati U. Gallium compound GaQ(3)-induced Ca(2þ) signalling triggers p53-dependent and -indepen-dent apoptosis in cancer cells. Br J Pharmacol 2012;166:617–36.

25. Cunningham KE, Turner JR. Myosin light chain kinase: pulling thestrings of epithelial tight junction function. Ann N Y Acad Sci 2012;1258:34–42.

26. Hynes RO. Integrins: bidirectional, allosteric signaling machines. Cell2002;110:673–87.

27. Zhong X, Rescorla FJ. Cell surface adhesionmolecules and adhesion-initiated signaling: understanding of anoikis resistance mechanismsand therapeutic opportunities. Cell Signal 2012;24:393–401.

28. Hu P, Luo BH. Integrin bi-directional signaling across the plasmamembrane. J Cell Physiol 2013;228:306–12.

29. Zhang K, Chen J. The regulation of integrin function by divalentcations. Cell Adhes Migr 2012;6:20–9.

30. Franco SJ, Huttenlocher A. Regulating cell migration: calpains makethe cut. J Cell Sci 2005;118:3829–38.

31. Storr SJ, Carragher NO, Frame MC, Parr T, Martin SG. The calpainsystem and cancer. Nat Rev Cancer 2011;11:364–74.

32. Vazquez A, Bond EE, Levine AJ, Bond GL. The genetics of the p53pathway, apoptosis and cancer therapy. Nat Rev Drug Discov2008;7:979–87.

33. Cramer LP, Mitchison TJ. Investigation of the mechanism of retractionof the cell margin and rearward flow of nodules during mitotic cellrounding. Mol Biol Cell 1997;8:109–19.

34. Maddox AS, Burridge K. RhoA is required for cortical retractionand rigidity during mitotic cell rounding. J Cell Biol 2003;160:255–65.

35. Desgrosellier JS, Cheresh DA. Integrins in cancer: biological impli-cations and therapeutic opportunities. Nat Rev Cancer 2010;10:9–22.

36. Xiong JP, Stehle T, Zhang R, Joachimiak A, Frech M, Goodman SL,et al. Crystal structure of the extracellular segment of integrin alpha


Jungwirth et al.




Vbeta3 in complex with an Arg–Gly–Asp ligand. Science 2002;296:151–5.

37. Galluzzi L, Vitale I, Abrams JM, Alnemri ES, Baehrecke EH, Blagosk-lonny MV, et al. Molecular definitions of cell death subroutines:recommendations of the Nomenclature Committee on Cell Death2012. Cell Death Differ 2012;19:107–20.

38. Taddei ML, Giannoni E, Fiaschi T, Chiarugi P. Anoikis: an emerginghallmark in health and diseases. J Pathol 2012;226:380–93.

39. Chiarugi P, Giannoni E. Anoikis: a necessary death program foranchorage-dependent cells. Biochem Pharmacol 2008;76:1352–64.

40. Lomonaco SL, Finniss S, Xiang C, Lee HK, Jiang W, Lemke N, et al.Cilengitide induces autophagy-mediated cell death in glioma cells.Neuro Oncol 2011;13:857–65.

41. Enyedy EA, Domotor O, Varga E, Kiss T, Trondl R, Hartinger CG, et al.Comparative solution equilibrium studies of anticancer gallium(III)complexes of 8-hydroxyquinoline and hydroxy(thio)pyrone ligands.J Inorg Biochem 2012;117:189–97.

42. Madan E, Gogna R, Keppler B, Pati U. p53 increases intra-cellularcalcium release by transcriptional regulation of calcium channelTRPC6 in GaQ3-treated cancer cells. PLoS ONE 2013;8:e71016.

43. Lopatniuk P, Witkowski JM. Conventional calpains and programmedcell death. Acta Biochim Pol 2011;58:287–96.

44. Pinon JD, Labi V, Egle A, Villunger A. Bim and Bmf in tissuehomeostasis and malignant disease. Oncogene 2008;27:S41–52.

45. SuLT, AgapitoMA, LiM, SimonsonWT,Huttenlocher A,HabasR, et al.TRPM7 regulates cell adhesion by controlling the calcium-dependentprotease calpain. J Biol Chem 2006;281:11260–70.

46. Kunzelmann-Marche C, Freyssinet JM, Martinez MC. Regulation ofphosphatidylserine transbilayer redistribution by store-operated Ca2þ

entry: role of actin cytoskeleton. J Biol Chem 2001;276:5134–9.47. Aussel C, Marhaba R, Pelassy C, Breittmayer JP. Submicromolar La3þ

concentrations block the calcium release-activated channel, andimpair CD69 and CD25 expression in CD3- or thapsigargin-activatedJurkat cells. Biochem J 1996;313:909–13.






2014;13:2436-2449. Published OnlineFirst July 31, 2014.Mol Cancer Ther Ute Jungwirth, Johannes Gojo, Theresa Tuder, et al. for the Anticancer Gallium Compound KP46Calpain-Mediated Integrin Deregulation as a Novel Mode of Action

Updated version

10.1158/1535-7163.MCT-14-0087doi:

Access the most recent version of this article at:

Material

Supplementary

http://mct.aacrjournals.org/content/suppl/2014/08/01/1535-7163.MCT-14-0087.DC1

Access the most recent supplemental material at:

Cited articles

http://mct.aacrjournals.org/content/13/10/2436.full#ref-list-1

This article cites 46 articles, 12 of which you can access for free at:

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

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://mct.aacrjournals.org/content/13/10/2436To request permission to re-use all or part of this article, use this link



http://mct.aacrjournals.org/lookup/doi/10.1158/1535-7163.MCT-14-0087

http://mct.aacrjournals.org/content/suppl/2014/08/01/1535-7163.MCT-14-0087.DC1

http://mct.aacrjournals.org/content/13/10/2436.full#ref-list-1

http://mct.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://mct.aacrjournals.org/content/13/10/2436


Calpain-Mediated Integrin Deregulation as a Novel Mode of ......mediated mechanism as a novel mode...

Documents

Transcript of Calpain-Mediated Integrin Deregulation as a Novel Mode of ......mediated mechanism as a novel mode...