Lefranc et al. Neurosurgery 2008.pdf

0 downloads 0 Views 583KB Size Report
Isabelle Roland, Ms.C. Unibioscreen SA,. Brussels, Belgium. Olivier Debeir, Ph.D. Department of Logical and. Numerical Systems,. Faculty of Applied Sciences,.

Florence Lefranc, M.D., Ph.D.


Laboratory of Toxicology, Institute of Pharmacy, and Department of Neurosurgery, Erasmus University Hospital, Free University of Brussels, Brussels, Belgium

Tatjana Mijatovic, Ph.D. Unibioscreen SA, Brussels, Belgium

Yasuko Kondo, M.D., Ph.D. Department of Neurosurgery, University of Texas M. D. Anderson Cancer Center, Houston, Texas

Sébastien Sauvage, Ms.C. Unibioscreen SA, Brussels, Belgium

Isabelle Roland, Ms.C. Unibioscreen SA, Brussels, Belgium

Olivier Debeir, Ph.D. Department of Logical and Numerical Systems, Faculty of Applied Sciences, Free University of Brussels, Brussels, Belgium

Danijela Krstic, Ph.D. Vincˇa Institute of Nuclear Sciences, Department of Physical Chemistry, Belgrade, Serbia

Vesna Vasic, Ph.D. Vincˇa Institute of Nuclear Sciences, Department of Physical Chemistry, Belgrade, Serbia

Philippe Gailly, M.D., Ph.D. Department of Physiology and Pharmacology, Université Catholique de Louvain, Brussels, Belgium

Seiji Kondo, M.D., Ph.D. Department of Neurosurgery, University of Texas M.D. Anderson Cancer Center, Houston, Texas

Gustavo Blanco, Ph.D. Department of Molecular and Integrative Physiology, University of Kansas Medical Center, Kansas City, Kansas

Robert Kiss, Ph.D. Laboratory of Toxicology, Institute of Pharmacy, Free University of Brussels, Brussels, Belgium Reprint requests: Florence Lefranc, M.D., Ph.D., Department of Neurosurgery, Erasme Hospital, Free University of Brussels, Route de Lennik, 808, 1070 Brussels, Belgium. Email: fl[email protected] Received, March 13, 2007. Accepted, July 13, 2007.



TARGETING THE α1 SUBUNIT OF THE SODIUM PUMP TO COMBAT GLIOBLASTOMA CELLS OBJECTIVE: Ion transporters play pivotal roles in cancer cell migration in general and in glioblastomas (GBMs) in particular. However, the specific role of Na+/K+-ATPase (the sodium pump) and, in particular, its α1 subunit, has remained unexplored in GBMs. MATERIALS AND METHODS: The expression of Na+/K+-ATPase α1 in GBM clinical samples, normal brain tissue, and a human GBM cell line has been investigated. Using the novel cardenolide UNBS1450 (Unibioscreen, Brussels, Belgium), which is a ligand of the sodium pump, we have characterized the effects of inhibiting Na+/K+-ATPase α1 in human GBM cells with respect to cell proliferation; morphology; impact on intracellular Na+, Ca2+, and adenosine triphosphate; and changes in the actin cytoskeleton. We have investigated the mechanism by which UNBS1450 overcomes the apoptosis resistance of GBMs and determined its anti-tumor effects in comparative studies in vitro in GBM cell viability assays and in vivo using an orthotopic human GBM xenograft model. RESULTS: Overall, the α1 subunit of Na+/K+-ATPase is highly expressed in a majority of glioblastomas compared with normal brain tissues, and by binding to this subunit in human U373-MG GBM cells, UNBS1450 impairs cell proliferation and migration via an intracellular adenosine triphosphate decrease-mediated disorganization of the actin cytoskeleton and cytotoxic proautophagic effects. UNBS1450 also significantly increases the in vivo survival of mice orthotopically grafted with U373-MG GBM cells. CONCLUSION: Inhibition of the Na+/K+-ATPase α1 subunit in human GBM cells impairs both cell migration and cell proliferation. KEY WORDS: Autophagy, Glioblastoma, Migration, Orthotopic xenografts, Sodium pump, Temozolomide Neurosurgery 62:211–222, 2008


DOI: 10.1227/01.NEU.0000296985.39041.52

lioblastomas (GBMs) are characterized by dismal prognoses. Active migration of GBM cells through the narrow extracellular spaces in the brain makes them elusive targets for surgical management (19, 37). Glioma cells are “self-propelled” (37) and are able to adjust their shape and volume rapidly as they invade the brain parenchyma. Essential to this process is the activity of Cl⫺ channels and anion transport mechanisms (35). Na+/K+ATPase, which is also called the sodium pump, is one such ion transporter that is present in the plasma membrane of most cells of higher eukaryotic organisms. Na+/K+-ATPase itself is a transmembrane heterodimer composed of a catalytic α subunit (exchanges intracellular Na+ for extracellular K+) and a glycosylated β subunit (controls α/β heterodimer assembly and insertion


into the plasma membrane [4, 29, 43, 44]). To date, four α and three β isoforms have been characterized in mammalian cells with tissuedependent distribution (4, 44). Numerous studies have reported changes in Na+/K+ATPase activity in the course of malignant transformation (4, 10, 29), including gliomas (36), with evidence that these occur at the very early stages of tumorigenesis (29). There are in fact two pools of Na + /K + -ATPase within the plasma membrane with two distinct functions. One constitutes the energytransducing pool of the enzyme and is broadly distributed in the plasma membrane (44). The other pool, which is restricted to caveolae, is independent of changes in intracellular Na + and K + concentrations (44), requires the initial association of Na + /K + ATPase with tyrosine kinase Src (39), and

VOLUME 62 | NUMBER 1 | JANUARY 2008 | 211


functions as a signal-transducing receptor for cardiotonic steroids (29, 43, 44). In addition to Src (19) and epidermal growth factor receptor (19), Ras (13, 19, 30), phosphatidyl inositol 3-kinase (3, 19) and caveolin-1 (1, 41) also closely interact with Na+/K+-ATPase within the restricted space of caveolae. Moreover, caveolae functions rely on caveolin-1, their major protein, which directly interacts with epidermal growth factor receptor (1, 7), drives the formation of plasma membrane caveolae, and anchors caveolae to the actin cytoskeleton (33). Importantly, caveolin-1 depletion results in the loss of focal adhesion sites and overall cell adhesion (33). Migrating GBM cells are resistant to apoptosis (19) and thus to proapoptotic drugs, which are presently the main chemotherapeutic agents in medical oncology. However, the possibility of overcoming apoptosis resistance has previously been reported for tumor astrocytes by decreasing their migration (19) or by inducing autophagy (20), for example, with temozolomide (15, 19, 40). The sodium pump is characteristically inhibited by cardiotonic steroids (44), a family of compounds that includes cardenolide (28, 42) and primarily bind to extracellular domains of α subunits (44). In this study, we investigated whether Na+/K+ATPase α1 subunits, which are a selective ligand for the cardenolide UNBS1450 (Unibioscreen, Brussels, Belgium) (29), are highly expressed in GBM samples compared with normal brain tissues. We then determined the ability of UNBS1450 to: 1) inhibit the α1 subunit of the sodium pump, 2) limit the in vitro proliferation and migration of GBM cells with and without prior small interfering ribonucleic acid (siRNA) α1 downregulation, and 3) improve survival in vivo in a preclinical model of human GBM.

MATERIALS AND METHODS Compounds Drugs were purchased as follows: digoxin (Lanoxin; GlaxoSmithKline, Genval, Belgium), ouabain (Acros Organics, Geel, Belgium), vincristine (Yick Vic Chemicals and Pharmaceuticals, Hong Kong, China), temozolomide (Temodal; Schering-Plough, Brussels, Belgium), tamoxifen and hydroxy-tamoxifen (Sigma, Bornem, Belgium), lomustine (Medac, Hamburg, Germany), procarbazine (Sigma-Tau Ethifarma, Asten, The Netherlands), and carmustine (Bristol Myers Squibb, Braine l’Alleud, Belgium). UNBS1450 was obtained as previously described (42) by hemisynthesis from 2ⴖ-oxovoruscharin (itself extracted from the African plant Calotropis procera) in the laboratories of Unibioscreen.

Surgical Specimens and Cell Cultures The clinical specimens evaluated in this study (four normal brain samples from autopsy and 10 GBMs) were histologically diagnosed according to World Health Organization classification. These samples were kindly provided by Dr. Isabelle Salmon (head of the Department of Anatomopathology at Erasme Hospital, Free University of Brussels, Brussels, Belgium). Cell cultures were prepared as previously described (6, 22). All cell lines under study were obtained from the American Type Culture Collection (ATCC, Manassas, VA) and included four human GBM models: Hs683 (ATCC HTB-138), U373-MG (ATCC HTB-17), U87-MG

212 | VOLUME 62 | NUMBER 1 | JANUARY 2008

(ATCC HTB-14), and T98G (ATCC CRL-1690); the rat glioma model C6 (ATCC CCL-107); and two normal human cell lines: WI-38 lung fibroblasts (ATCC CCL-75) and WS1 skin fibroblasts (ATCC CRL-1502).

Animal Xenograft Model In vivo orthotopic xenografts of human U373-MG GBM were obtained as previously described (6, 21) by injecting two million cells into the left temporal lobes of 8-week-old female immunodeficient nu/nu mice (weight, 21–23 g; Iffa Credo; Charles River, Arbresle, France). These mice were then randomly divided into three groups of nine on Day 10 post tumor graft. On Day 14 posttumor graft, the mice were treated as follows: Group 1 (controls) received 12 intravenous injections (three times a week on Mondays, Wednesdays, and Fridays for 4 consecutive weeks) of 50 µL saline. The first injection was administered on Day 14. Group 2 received 12 intravenous injections of temozolomide at a dose of 40 mg/kg under an identical regimen to Group 1. Group 3 received 12 intravenous injections of UNBS1450 at a dose of 7.5 mg/kg under an identical regimen to Group 1. Animal survival with and without drug treatment was then followed. To ensure a humane endpoint, animals were sacrificed before being rendered moribund. All the in vivo experiments described in the present study were carried out on the basis of authorization number LA1230509 of the Animal Ethics Committee of the Federal Department of Health, Nutritional Safety, and the Environment (Belgium).

Cytology and Histology Immunofluorescence staining of cells was performed using selective antibodies against Na+/K+-ATPase α1 (AbCam, Cambridge, UK), α2 (Upstate, Huissen, The Netherlands), and α3 (Sigma, St. Louis, MO), and against the human forms of caveolin-1 (BD Biosciences Pharmigen, Erembodegem, Belgium). Cell viability with and without UNBS1450 treatment was assessed using the MitoFluor Red 589 test (Molecular Probes, Invitrogen, Merelbeke, Belgium) that allows the in vitro detection of living and functional mitochondria. Fluorescent phallacidin conjugated with Alexa Fluor 488 fluorochrome (Molecular Probes, Invitrogen) was used to label fibrillar actin, whereas Fluor 594conjugated DNAseI (Molecular Probes, Invitrogen) was used to stain globular actin as previously described (21). Various specific and scrambled siRNAs were designed against the sodium pump α1 subunit whose knockdown in U373-MG cells (verified by means of immunofluorescence and Western blotting) was near total with the sequences of sense, 5ⴕ-GGGCAGUGUUUCAGGCUAA3ⴕ and antisense, 5ⴕ-UUAGCCUGAAACACUGCCC-3ⴕ (Eurogentec, Liège, Belgium). The scrambled sequence of sense, 5ⴕ-UCUACGAGGCACGAGACUU-3ⴕ and antisense, 5ⴕ-AACUCUCGUGCC UCGUAGA-3ⴕ (Eurogentec), did not modify the levels of expression of the sodium pump α1 subunit.

Quantitative Reverse Transcriptase-Polymerase Chain Reaction The determination of α1 subunit messenger ribonucleic acid (mRNA) expression in 1) U373-MG GBM cells and 2) clinical GBM and normal brain samples by quantitative reverse transcriptase (RT)polymerase chain reaction (PCR) was undertaken against a standard curve established through serial dilutions (108 to 102 copies/µL) of the PCR products generated with specific primers. The sequence of the forward primer used was ACTTAGCCTTGATGA ACTTCA, and that of the reverse primer was GCTTGGATGCTATAAGCCAA. The quantitative PCR reactions were carried out with 20 ng of purified complimentary deoxyribonucleic acid in the LightCycler thermocycler (Roche Diagnostics, Vilvoorde, Belgium) using LC-Fastart DNA Master SYBR



Green 1 (Roche Diagnostics). After amplification, data analysis was carried out using the “Fit points” algorithm of the LightCycler quantification software.

Human Erythrocyte Membrane Na+/K+-ATPase Inhibition Assays The effects of UNBS1450 and digoxin on Na+/K+-ATPase activity were investigated in the concentration range 10⫺10 to 10⫺3 M in a reaction mixture containing 50 mmol/L Tris-HCl (pH 7.4), 100 mM NaCl, 20 mM KCl, 5 mM MgCl2, 2 mM adenosine triphosphate, and 50 µg of red blood cell membrane proteins in a final volume of 200 µL. Human erythrocyte membranes were prepared as previously detailed (17).

Na+/K+-ATPase Inhibition Assays in Insect Cells A heterologous protein expression system was used, given its advantage over eukaryotic systems in being able to produce large amounts of active recombinant Na+/K+-ATPase in a background with very little or no endogenous Na+/K+-ATPase (5). Na+/K+-ATPase activity was therefore assayed in homogenates of baculovirus-transfected Sf-9 cells expressing rat α1β1, α2β1, and α3β1 isozymes in the presence or absence of UNBS1450 or ouabain (dose-response curves were performed for each compound) as previously described (5, 29).

In Vitro Cell Proliferation Measurements Cell proliferation was assessed with the colorimetric MTT (3-[4, 5dimethylthiazol-2yl]-diphenyl tetrazolium bromide; Sigma, Bornem, Belgium) assay as described previously (28, 42). The cells were incubated for 72 hours in the presence and absence of the indicated drugs. Drug concentrations ranged between 10⫺9 and 10⫺5 M with semilog concentration increases. The experiments were carried out in sextuplicate.

Computer-assisted Phase-contrast Videomicroscopy Human U373-MG GBM cell migration, cell death, and cell proliferation with and without UNBS1450 treatment were characterized in vitro by the use of computer-assisted phase-contrast videomicroscopy as described previously (21). To minimize the file size of the generated video clips, they have been compressed using the DivX codec. The codec and a movie player are available at http://divx.com.

Intracellular Ca2+ Concentration Measurement Cells were cultured on glass coverslips to approximately 30 to 50% confluence to permit the measurement of intracellular Ca2+ concentration ([Ca2+]i) for individual cells, as we have detailed elsewhere (26, 27). The cells were loaded with 1 µM fura-2 acetoxymethyl ester (Molecular Probes, Invitrogen) for 60 minutes at room temperature. A standard intracellular calibration procedure effected after cell permeabilization with 5 µM ionomycin (Sigma) was then performed to calculate the [Ca2+]i from the ratio of the fluorescence intensities emitted at the two wavelengths. The Grynkiewicz equation (12) was used, providing a dissociation constant of 223 nM of fura-2 for Ca2+. The Rmin value was obtained by perfusing permeabilized cells with Krebs buffer in which CaCl2 was replaced by 2 mM ethyleneglycoltetraacetic acid.

Intracellular Na+ Concentration Measurements Cells were cultured on glass coverslips to approximately 30 to 50% confluence to permit measurement of intracellular Na+ concentration ([Na+]i) for individual cells. The tetraammonium salt fluorescence indicator SBFI-AM (Molecular Probes, Invitrogen) was mixed in a 1 to 1 ratio with a 20% (volume/volume) solution of Pluronic F-127 (Molecular Probes, Invitrogen) in dimethyl sulfoxide. This loading solution


was diluted in serum-free minimal essential medium to reach a final dye concentration of 10 µM. The cells were loaded for 90 minutes at 37⬚C. The coverslips were then rinsed for 30 minutes in Krebs buffer and mounted in a 1-mL cuvette placed in the light path of an inverted Nikon microscope (Nikon, Melville, NY) equipped with epifluorescence illumination. The SBFI-AM-loaded cells were excited alternatively at 340 and 380 nm. The fluorescence emission was monitored at 510 nm by means of a Deltascan spectrofluorimeter (Photon Technology International, Monmouth Junction, NJ) coupled to an inverted microscope (Nikon Diaphot, oil-immersion objective ⫻40; numerical objective, 1.3). The emission was corrected for background using Felix software (Felix NMR, San Diego, CA). The [Na+]i was estimated from the 340-to-380-nm ratio according to the Grynkiewicz equation (12) with 11.3 mM as the dissociation constant of SBFI for [Na+]i. The cells were continuously perfused with washing buffer at 1.5 mL/minute. The extra- and intracellular [Na+] values were equilibrated using 3 µM gramicidin, 10 µmol/L monensin, and 1 mM ouabain. To reach Rmin, this buffer was replaced by a solution in which NaCl was replaced by 130 mM KCl to maintain cell osmolarity.

Intracellular Adenosine Triphosphate Measurements Cellular adenosine triphosphate levels were measured as detailed elsewhere (29).

In Vitro Autophagy Evaluation

Detection and Quantification of Acidic Vesicular Organelles with Acridine Orange Staining Acidic vesicular organelles were stained as previously described (15, 26). Although the cytoplasm and nucleus fluoresce green in acridine orange-stained cells, the acidic compartments fluoresce red. The intensity of the red fluorescence is proportional to the degree of acidity and the volume of acidic vesicular organelles, including autophagic vacuoles.

Green Fluorescent Protein Light Chain-3 Plasmid Transfection The green fluorescent protein (GFP) and the microtubule-associated protein-1 light chain-3 (LC3) fusion vector (GFP-LC3) were provided by Dr. Noboru Mizushima (Tokyo Metropolitan Institute of Medical Science, Tokyo, Japan). LC3, which is the homolog of the yeast Apg8/Aut7p gene, is located on the autophagosomal membrane during autophagy. The GFP-LC3-fused protein was used to pinpoint the site of exogenous LC3. Cells cultured on a chamber slide dish were transfected with 1 µg of the GFP-LC3 expression vector. FuGENE6 transfection reagent (Roche, Indianapolis, IN) was used according to the manufacturer’s instructions. After 24 hours, the cells were treated with various concentrations of UNBS1450 for 3 days and were then fixed in 1% paraformaldehyde and observed under fluorescence microscopy.

Light Chain-3 Western Blotting LC3 Western blotting was performed as detailed elsewhere (26). The bound antibody complex was detected by enhanced chemiluminescence. The antibodies used were the rabbit anti-LC3 antibody raised in our laboratory and the mouse anti-β-actin antibody (Sigma).

Beclin-1 Western Blotting Cell extracts were prepared after lysis of U373-MG cells directly in the boiling lysis buffer (10 mmol/L Tris-HCl, pH 7.4, 1 mmol/L Na3O4V, and 1% sodium dodecyl sulfate at pH 7.4). Western blotting

VOLUME 62 | NUMBER 1 | JANUARY 2008 | 213


analyses were performed as previously described (16). The primary anti-beclin-1 antibody was obtained from BD Transduction Laboratories (Erembodegem, Belgium) and used at a 1-to-250 dilution.

Data Analyses Data are expressed as a means plus/minus the standard error of the mean. Data obtained from independent groups were compared by the nonparametric Kruskal-Wallis (more than two groups) or MannWhitney U tests (two groups). The standard survival time analyses were carried out using the Kaplan-Meier curves and the Gehan generalized Wilcoxon test. Statistical analysis was performed using Statistica software (Statsoft, Tulsa, OK).

RESULTS Na+/K+-ATPase α1 Subunits Are Highly Expressed in Human GBMs and Collocate with Caveolin-1 in the Lamellipodia of U373-MG GBM Cells Quantitative RT-PCR analysis revealed that the levels of α1 mRNA are much higher in most of the GBMs analyzed than in normal brain tissue (Fig. 1A). Such measurements also revealed high levels of α1 mRNA in human U373-MG GBM cells (Fig. 1A). Additionally, using immunofluorescent staining, it was shown that the α1 subunit of the sodium pump is more appreciably expressed in the lamellipodia of U373-MG cells than are the α2 and α3 subunits (Fig. 1, B and C). As further revealed in Figure 1D, the α1 subunits located in the lamellipodia of U373-MG cells colocalize with caveolin-1. This observation is consistent with 1) the high expression of α1 subunits in U373-MG cells; 2) the Na + /K + -ATPase pool involved in the migration of cancer cells being restricted to caveolae; and 3) caveolin-1, the major protein of caveolaemodulating cell interaction with the extracellular matrix, regulating the interaction of different signaling molecules and having a significant role in cell movement by anchoring the plasma membrane to the actin cytoskeleton (33).

UNBS1450 Displays Significant Inhibitory Effects on Na+/K+-ATPase α1 Subunits and Has Potent Antiproliferative Effects In Vitro against GBM Cell Lines UNBS1450 and ouabain both inhibited rat Na+/K+-ATPase isozymes produced in Sf-9 insect cells in an isozyme-specific manner with a relative potency that followed the sequence α3β1 ⬎ α2β1 ⬎ α1β1. UNBS1450 inhibited α3β1, α2β1, and α1β1 Na+/K+-ATPases with inhibition constant (Ki) values of 1.4 ⫻ 10 ⫺9 , 1.9 ⫻ 10 ⫺8 , and 1.7 ⫻ 10 ⫺7 M, respectively. UNBS1450 was the more potent inhibitor of all three isozymes and strikingly inhibited α1β1 with a potency more than 200 times greater than that of ouabain (Fig. 2A). The Na + /K + -ATPase α1 subunit could be a target for UNBS1450 as preliminary data from our group recently suggested with respect to non-small cell lung cancer cells (29). To evaluate this with respect to GBM cells, the in vitro antiproliferative effects of UNBS1450 and ouabain against U373-MG GBM and rat C6 glioma cells were first evaluated. Rodent α1

214 | VOLUME 62 | NUMBER 1 | JANUARY 2008

subunit is between 100 and 1000 times less sensitive to cardiotonic steroids than the human one, because rodent α1 displays two mutations (4). The data in Figure 2B clearly indicate that human U373 GBM cells are indeed much more sensitive to the anti-tumor effects of both UNBS1450 and ouabain than are rat C6 GBM cells. Ouabain was ineffective against rat C6 GBM cells, whereas UNBS1450 markedly inhibited the proliferation of these C6 cells at high concentration (Fig. 2B). The fact that UNBS1450 was able to decrease the proliferation of rat C6 GBM cells at high dose, whereas ouabain failed to do it (Fig. 2B), can be related, at least partly, to the higher binding affinity of UNBS1450 to the sodium pump α1 subunit as compared with ouabain (Fig. 2A). Additionally, decreasing the expression of α1 in human U373-MG GBM cells using specific siRNA (Fig. 2D) led to an increase in the magnitude of UNBS1450-induced antiproliferative effects in these cells compared with control cells and cells treated with scrambled siRNA (Fig. 2C). These data suggest that α1 is important for U373-MG GBM cell viability. Decreasing α1 subunit expression using specific siRNA was also noted to markedly alter U373-MG GBM cell morphology with transformation of the large lamellipodia into long cytoplasmic cell processes (Fig. 2G). This also had an impact on cell migration (similar to that shown in Fig. 3, F–I). The in vitro antiproliferative effects of UNBS1450 were further confirmed in several human GBM cell lines (Hs683, U373-MG, and T98G GBM) and found to be similar to vincristine but greater than those of temozolomide, lomustine, procarbazine, and carmustine, which are routinely used to treat patients with GBM (Fig. 4A) and greater than those of ouabain, digoxin, and digitoxin (data not shown). UNBS1450 was also found to be a more potent inhibitor of GBM cell proliferation (U373-MG and T98G cells) than normal fibroblasts (WI-38 and WS1 fibroblasts; Fig. 2E). This may be explained, at least in part, by the fact that the intracellular ATP concentration ([ATP]i) is higher in normal cells (WI-38) than in GBM cells (U373-MG) and UNBS1450 decreases [ATP]i more markedly in GBM than in normal cells (Fig. 2F). A decrease in [ATP]i is known to lead to hypoxia in GBM cells, and hypoxia has been shown to decrease Na+/K+-ATPase activity in alveolar epithelial cells by triggering endocytosis of the pump through the mitochondrial reactive oxygen species and protein kinase C-ζ-mediated phosphorylation of the Na+/K+-ATPase α1 subunit (8).

UNBS1450 Impairs U373-MG GBM Cell Proliferation and Migration Through Disorganization of the Actin Cytoskeleton

Computer-assisted Phase-contrast Videomicroscopy (see video at web site) Double-click on the image to activate the video once the downloading procedure is complete (as detailed in Materials and Methods) to reveal round-shaped cells as the morphological endpoint of 72 hours of treatment of both proliferating (Fig. 3, B–E) and migrating (Fig. 3, F–I) U373-MG GBM cells with 10






D FIGURE 1. Tissue expression and cellular location of Na+/K+-ATPase α1. A, RT-PCR quantitative determination of Na+/K+-ATPase α1 mRNA in normal human brain tissue (open bars), human GBMs (black bars), and human U373-MG GBM cells (blue bar). B (G ⫻ 100) and C, (G ⫻ 400), conventional immunofluorescence localization of the α1 (green fluorescence), α2 (red fluorescence), and α3 (blue fluorescence) subunits of the sodium pump in U373-MG GBM cells. Merged images are also shown (B

nM UNBS1450. The single white arrow in Figure 3A indicates a U373-MG cell (Fig. 3B) that enters mitosis (Fig. 3C) but fails to complete cytokinesis and therefore does not divide into two daughter cells (Fig. 3, D and E). The double white arrows in Figure 3A point to a migrating U373-MG cell (Fig. 3F) in which the lamellipodia (yellow and green arrows in Fig. 3, F and H) become progressively long cytoplasmic cell processes over time by 10 nM UNBS1450 (Fig. 3, G and H) resulting in a rounded cell with adjacent broken-off cytoplasmic cell processes (Fig. 3I). These marked UNBS1450-induced morphological changes were not paralleled by compound-induced increases in [Ca2+]i


and C, bottom, right). The data show that the Na+/K+-ATPase α1 subunit is localized at the edge of U373 GBM cell lamellipodia, in contrast with the α2 and α3 subunits. D, immunofluorescent colocalization of sodium pump α1 subunit (red fluorescence) and caveolin-1 (green fluorescence) in U373MG GBM cells with the merged image. The data show that the Na+/K+ATPase α1 subunit colocalizes with caveolin-1 in U373 GBM cells.

(Fig. 3J) or [Na+]i (Fig. 3K) even at higher concentrations (data not shown). Although damaged by 10 nM UNBS1450 treatment (Fig. 3, E and I), the round-shaped U373-MG cells remained alive for several days as revealed by the MitoFluor red staining of their living mitochondria (Fig. 3, L and M). Using the mitochondrial membrane potential (δψm)-specific cationic, lipophilic dye JC-1, it was also confirmed that UNBS1450 at 10 and 100 nM for 72 hours did not decrease δψm values (data not shown). The UNBS1450-induced changes in the morphology of U373-MG cells (and other types of tumor cells) appear to correspond to profound and irreversible com-

VOLUME 62 | NUMBER 1 | JANUARY 2008 | 215









Ga Dd C De


Df FIGURE 2. Effects of UNBS1450 and anti-Na+/K+-ATPase α1 siRNA. A, inhibitory effect of ouabain and UNBS1450 on rat Na+/K+-ATPase α1β1, α2β1, and α3β1 isozymes. The inhibition constants (Ki values, in moles) for each isozyme are expressed as the means ⫾ the standard errors of the means. B, antiproliferative activity of ouabain (open symbols) and UNBS1450 (solid symbols) on rat C6 (red) and human U373-MG GBM (blue) cells. Data are presented as means ⫾ the standard errors of the mean values with the standard error of mean values included within the symbol sizes not exceeding 3% of the mean values. C, UNBS1450-induced antiproliferative activity on human U373-MG GBM cells in which expression of the Na+/K+-ATPase α1 subunit was decreased using siRNA. U373-MG cells treated with the transfection agent only (*) were used as controls. U373-MG cells were treated with scrambled siRNA (䊊) or with the anti-α1 siRNA (䊉). Cell proliferation was measured in vitro using the MTT colorimetric assay. D, expression of the Na+/K+-ATPase α1 subunit in U373-MG GBM cells. Bright-field and fluorescence microscopy of cells treated with the transfection agent only (Da

216 | VOLUME 62 | NUMBER 1 | JANUARY 2008

and Db, respectively), scrambled siRNA (Dc and Dd, respectively), and anti-α1 siRNA (De and Df, respectively). E, dose-response curve of the antiproliferative activity of UNBS1450 on normal human cell lines (WI-38 lung fibroblasts and WS1 skin fibroblasts) and human GBM cell lines (T98G and U373-MG). Cell proliferation was measured in vitro using the MTT colorimetric assay. Data are presented as means ⫾ standard errors of mean values with the standard error of mean values included within the symbols. F, intracellular adenosine triphosphate concentration ([ATP]i) was determined in human WI-38 normal lung fibroblasts (open bars) and human U373MG GBM cells (black bars) using a bioluminescence assay. Values are expressed as means ⫾ standard errors of the means. G, influence of the transfection agent (Ga) and of the anti-α1 siRNA (Gb) on U373-MG GBM cell morphology. The data show that anti-α1 siRNA markedly modifies U373 cell morphology. The scrambled siRNA induced no modifications in U373MG cell morphology as compared with control U373-MG cells (data not shown).



B: 0hr


F: 0hr L

C: 1hr

G: 10hr

D: 3hr

E: 5hr

H: 25hr

I: 50hr





FIGURE 3. Effects of UNBS1450 on cell proliferation, morphology, and mobility, and [Ca2+]i and [Na+]i in GBM cells. A, computer-assisted phase-contrast microscopy of the effects of UNBS1450 (10 nM for 72 hours) on human U373MG GBM cells (see the video on the web site). B–E, effects of 10 nM UNBS1450 over 5 hours on U373-MG GBM cells is indicated (A, single white arrow); this cell (B) enters the mitosis stage (C) and fails to divide (D and E). F–I, effects of UNBS1450 (10 nM for 50 hours) on U373-MG GBM cells is shown (A, double white arrows). This cell (F) is a migratory cell (see the yellow and the green arrows pointing to lamellipodia) whose lamellipodia are progressively destroyed by UNBS1450 (G and H) with a morphological endpoint similar (I) to the cells, which failed to divide (E). J, quantitative determination of [Ca2+]i in

pound-induced modifications to the organization of the actin cytoskeleton (Fig. 3, N and O).

UNBS1450 Has Potent Proautophagic Effects on Human U87-MG and U373-MG Glioblastoma Cells The potent antiproliferative effects of UNBS1450 on Hs683, T98G, and U373-MG GBM cells (Fig. 4A) and also on U87-MG, H4, SW1783, and SW1088 GBM cells (data not shown) do not appear to occur through the induction of apoptosis, certainly at concentrations between 10 and 1000 nM (data not shown).


O U373-MG GBM cells treated first with 10 nM UNBS1450 and then with 1 µM thapsigargin. K, quantitative determination of [Na+]i in U373-MG tumor cells treated first with 100 nM UNBS1450 and then with a mixture of 3 µM gramicidine, 10 µM monensin, and 1 mM ouabain. L and M, fluorescence microscopic detection of living mitochondria identified with MitoTracker dye (Molecular Probes) both in U373-MG GBM cells (M) and in U373-MG cells treated with 10 nM UNBS1450 for 50 hours (L). N and O, fluorescence microscopy of the organization of the actin cytoskeleton. Red fluorescence indicates globular (nonpolymerized) actin; green fluorescence reveals fibrillary (polymerized) actin in control U373-MG GBM cells (N) and in U373-MG cells treated for 40 hours with 10 nM UNBS1450 (O).

In contrast, UNBS1450-induced antiproliferative activity seems to occur as the result of proautophagic effects. To confirm this, acidic vesicular organelles, including autophagic vacuoles, were first quantified in U373-MG and U87-MG GBM cell lines with and without UNBS1450 treatment using acridine orange staining (Fig. 5A). The percentage of cells with positive red staining increased from approximately 3 to 42% in the U373-MG cells and from approximately 4 to 65% in the U87-MG cells 72 hours after treatment with 100 and 50 nM UNBS1450, respectively.

VOLUME 62 | NUMBER 1 | JANUARY 2008 | 217




72 hours of treatment with UNBS1450 at 50 nM, the percentage of cells with GFP–LC3 speckling (Fig. 5E), which is so typical of autophagy (15, 16), had increased significantly from 7 to 20% in U373-MG cells (P ⫽ 0.001) and from 7 to 13% in U87-MG cells (P ⫽ 0.014; Fig. 5F). Taken together, these results clearly indicate that UNBS1450 induces autophagy in GBM cells.

UNBS1450 is Active In Vivo in the Diffusely Invasive Orthotopic Human Glioblastoma Model Figure 4C illustrates the growth pattern of the U373MG orthotopic GBM xenograft in vivo. Figure 4E is a high-magnification image of the center of Figure 4D, which further indicates that the process of diffuse invasion of U373-MG tumor cells into the C E brain parenchyma (see the area delineated by the white FIGURE 4. Comparative in vitro and in vivo activity of UNBS1450 in GBMs. A, comparison of the antiproliferahatched line) can occur far tive activity (y-axis: IC50 values in logarithmic scale) of UNBS1450 and a series of drugs in human GBM cell lines from the tumor bulk (Fig. 4C). Hs683 (light-gray bars), U373-MG (dark-gray bars), and T98G (black bars). Cell proliferation was determined The U373-MG GBM model using the MTT colorimetric assay. B, the survival of immunodeficient mice (n ⫽ 9 mice per group) with orthotopic was used because it displays brain grafts of U373-MG tumor cells, which received 12 intravenous injections (into the tail vein three times per week marked astrocytic differentiafor 4 consecutive weeks) of 7.5 mg/kg UNBS1450 (blue line) or 40 mg/kg temozolomide (red line). Drug treatment was initiated on Day 14 posttumor graft. C, typical morphological illustration (hematoxylin-eosin staining at G ⫻ tion (6), is nondeleted in 40) of an orthotopic U373-MG GBM xenograft invading the normal brain parenchyma (NBP) of an immunocom1p19q, and behaves rather like promised mouse. Area delineated in C (hatched white line) is magnified in D (G ⫻ 100), the center is further maga chemoresistant high-grade nified in E (G ⫻ 200). malignant glioma (6), that diffusely invades the brain parenchyma (Fig. 4, C and E) Additionally, increased expression of LC3 (Fig. 5B), a spefor which surgery has been shown to be of limited benefit as cific autophagy marker (16, 17), was detected on treatment of shown in immunodeficient rats bearing brain xenografts (22). GBM cell lines with increasing concentrations of UNBS1450 We chose temozolomide to challenge UNBS1450 in this U373 (0–100 nM). In untreated control cells, LC3 Type I (LC3-I) orthotopic model, because 1) temozolomide contributes higher and Type II (LC3-II) were expressed at similar levels in U373therapeutic benefits than do proapoptotic drugs in this model MG cells, whereas a higher level of LC3-I than of LC3-II was (6, 19); and 2) temozolomide partly bypasses the resistance to expressed in the U87-MG cells. After 72 hours treatment with apoptosis in GBM cells (19) by inducing autophagy (15), as 50 nM UNBS1450, LC3-II levels had increased to a greater does UNBS1450 (Fig. 5), and temozolomide is the drug extent than the LC3-I levels, and the ratio of LC3-II to LC3-I presently associated with the most encouraging therapeutic had increased in both cell types. The expression of beclin-1, benefits (20, 40). a second specific autophagy marker (34), also increased in In vivo in this orthotopic U373-MG GBM model, UNBS1450 U373-MG GBM cells incubated with 10 or 100 nM UNBS1450 displays anti-tumor effects that are similar to temozolomide for up to 40 hours (Fig. 5C). (Fig. 4B), the reference proautophagic compound used to treat Finally, GMB cells transfected with the GFP–LC3 plasmid to patients with GBM (15, 19, 40). We did not observe toxic side highlight the exogenous LC3 sites revealed them to be distribeffects at the doses used in the present experiment. uted homogeneously in the cytoplasm (Fig. 5D). However, after Experiments are ongoing to analyze the therapeutic benefits


218 | VOLUME 62 | NUMBER 1 | JANUARY 2008









FIGURE 5. UNBS1450 induces autophagy in GBM cells. A, detection and quantification of acidic vesicular organelles by means of acridine orange staining in human U373-MG and U87-MG GBM cells treated with increasing concentrations of UNBS1450 for 72 hours. B, Western blotting analyses of microtubule-associated protein 1 LC3 (an autophagosomal ortholog of yeast Atg8/Aup7) Type I (LC3-I) and Type II (LC3-II) in human U373-MG and U87-MG GBM cells treated with increasing concentrations of UNBS1450 for 72 hours. C, Western blotting analyses of the beclin-1 (upper gel, with tubulin chosen as the control for the bottom gel) in human U373-MG GBM cells treated with 10 and 100 nM UNBS1450 for 10 to 40 hours. D, transfection of a GFP-LC3 plasmid into human U373-MG and U87-MG GBM cells. LC3 is located on the autophagosomal membrane during autophagy and the GFP-LC3-fused protein enables exogenous LC3 to be detected. In the control condition (open bars), exogenous LC3 was distributed homogeneously in the cytoplasm of the U87-MG cells (D), whereas in the U87-MG cells treated for 72 hours with 50 nM UNBS1450 (black bars), the percentage of cells with GFP-LC3 dots (E) increased significantly, a feature also observed in U373-MG tumor cells (F).

contributed by combining temozolomide with UNBS1450 under various regimens in the U373 orthotopic model.

DISCUSSION We have previously shown that both non-small cell lung cancer (28, 29) and prostate cancer (manuscript in prepara-


tion) overexpress Na+/K+-ATPase α1 compared with normal tissues. The present study also suggests overall higher levels of α1 mRNA in a large proportion of clinical GBM samples compared with normal brain tissue. Additional measurements also revealed high levels of α1 mRNA in human U373MG GBM cells and appreciable expression of α1 in U373-MG cell lamellipodia, which was found to be coexpressed with caveolin-1. Additionally, specific siRNA knockdown of α1 in U373-MG cells markedly modified their morphology and reduced their proliferation in the presence of the novel cardenolide UNBS1450, which is known to bind to the sodium pump and display potent anti-tumor activity both in vitro and in vivo in experimental models of human cancer (28, 42). Collectively, these data support the view that Na + /K + ATPase α1 could potentially be a target in cancer therapy and notably in GBM. The question remains: If this is the same pump as exists in normal brain, how can it be a target for therapy? At present, we do not know whether it is the same pump as in the normal brain. However, as is mentioned at the beginning of this article, it seems that two pools of sodium pumps exist, i.e., the one directly involved in ion exchange, and another (the one located in caveolae) implicated in the signal transduction in partnerships with Src and epidermal growth factor receptor. The possibility thus remains that anti-GBM therapy would consist of targeting this caveolae-related pool of sodium pumps. In addition, only those patients whose GBMs overexpress by far over the control values for the α1 subunit of the Na+/K+-ATPase should be involved in an anti-Na+-K+ therapy based on the use of UNBS1450. Cardiotonic steroids, the pharmacological class to which cardenolide belongs, are modulators of Na+/K+-ATPase expression and activity. They remain important in the treatment of congestive heart failure, and several epidemiological studies have suggested their possible use in oncology (38). A number of cardiotonic steroids have demonstrated in vitro anti-tumor activity against a range of cancer types (14, 23, 25, 38) including glioma (23, 38) with oleandrin, for example, demonstrating suppression of nuclear factor-κB activation, induction of apoptosis, and potentiation of the proapoptotic effects of cytotoxic drugs. However, a narrow therapeutic index related to their inotropic effects (11, 45) has hindered their development as anticancer agents. The novel cardenolide UNBS1450, which differs structurally significantly from the classic cardiotonic steroids ouabain, digitoxin, and digoxin (28, 29, 42), demonstrates much higher rates of inhibition of all Na+/K+-ATPase isozymes (notably, α1β1) than classic cardenolides (Fig. 2A) (29). UNBS1450 has more marked antiproliferative effects in human GBM cell lines than ouabain and a range of marketed drugs routinely used to treat patients with GBM (notably temozolomide). Unlike ouabain, UNBS1450 also induces antiproliferative effects in the more resistant rat C6 GBM cell line. In contrast with ouabain, digoxin, and digitoxin (14, 25), UNBS1450 at an antiproliferative (Figs. 2–4) and antimigratory (Fig. 3) concentration (10 nM) does not induce apoptosis or [Ca2+]i increases involved in car-

VOLUME 62 | NUMBER 1 | JANUARY 2008 | 219


denolide-mediated cardiotoxicity (11, 45). The antiproliferative effects of UNBS1450 on human GBM cells appear to be mediated by proautophagic effects (Fig. 5), a feature that enables the compound to circumvent the resistance of GBM cells to apoptosis (19). The data also indicate that the antiproliferative and antimigratory effects of UNBS1450 in human GBM cells (and other cell types; data not shown) result from disorganization of the actin cytoskeleton (Fig. 3). The actin cytoskeleton is involved in many cellular processes that are essential for growth, differentiation, division, membrane organization, and motility (2, 19). Moreover, the association of actin filaments with the plasma membrane provides mechanical stability, maintains cell shape and adhesion, and regulates dynamic surface protrusions such as lamellipodia and filopodia, which are fundamental determinants of the motility and migratory potential of cells (9). UNBS1450 is known to markedly inhibit sodium pump activity (42) and, with Na+/K+-ATPase directly linked to the actin membrane cytoskeleton through its α subunit, it is evident how the molecule might elicit its activity (32). As previously reported by Molitoris et al. (31) and confirmed in the present study, cellular adenosine triphosphate depletion (Fig. 2F) results in a rapid-duration-dependent dissociation of the sodium pump from the actin cytoskeleton, during which total cell sodium pump activity remains unaltered (31). This is a factor that at least partly explains why cells remain alive while profoundly modified in shape (Fig. 3L). Adenosine triphosphate depletion results in the redistribution of F-actin from a primarily cortical concentration to a perinuclear location (31) as also observed when treating U373-MG tumor cells with UNBS1450 (Fig. 3O). UNBS1450-induced adenosine triphosphate depletion (Fig. 3F) could also explain, at least in part, the induction by the compound of autophagic processes (16, 18) (Fig. 5). Indeed, adenosine triphosphate depletion is known to trigger autophagy through the mammalian target of rapamycin pathway (24). In conclusion, impairment of GBM cell proliferation and migration through targeted cardenolide-mediated inhibition of the signaling function of Na+/K+-ATPase occurs through intracellular adenosine triphosphate depletion that markedly disorganizes the actin cytoskeleton and induces proautophagic cytotoxic effects. We are the first to propose that the sodium pump, and more specifically, its α1 subunit, is a new target in the context of malignant GBM treatment. The structural uniqueness of the novel cardenolide UNBS1450, which displays increased affinity for α1, proautophagic characteristics, and improved anti-tumor activity, supports the possibility of its development as an anticancer agent targeting overexpressed α1 in GBMs.

REFERENCES 1. Abulrob A, Giuseppin S, Andrade MF, McDermid A, Moreno M, Stanimirovic D: Interactions of EGFR and caveolin-1 in human glioblastoma cells: Evidence that tyrosine phosphorylation regulates EGFR association with caveolae. Oncogene 23:6967–6979, 2004. 2. Bamburg JR, Wiggan ON: ADF/cofilin and actin dynamics in disease. Trends Cell Biol 12:598–605, 2002.

220 | VOLUME 62 | NUMBER 1 | JANUARY 2008

3. Barwe SP, Anilkumar G, Moon SY, Zheng Y, Whitelegge JP, Rajasekaran SA, Rajasekaran AK: Novel role for Na,K-ATPase in phosphatidylinositol 3-kinase signaling and suppression of cell motility. Mol Biol Cell 16:1082–1094, 2005. 4. Blanco G: Na,K-ATPase subunit heterogeneity as a mechanism for tissuespecific ion regulation. Semin Nephrol 25:292–303, 2005. 5. Blanco G: The Na/K-ATPase and its isozymes: What we have learned using the baculovirus expression system. Front Biosci 10:2397–2411, 2005. 6. Branle F, Lefranc F, Camby I, Jeuken J, Geurts-Moespot A, Sprenger S, Sweep F, Kiss R, Salmon I: Evaluation of the efficiency of chemotherapy in in vivo orthotopic models of human glioma cells with and without 1p19q deletions and in C6 rat orthotopic allografts serving for the evaluation of surgery combined with chemotherapy. Cancer 95:641–655, 2002. 7. Couet J, Sargiacomo M, Lisanti MP: Interaction of a receptor tyrosine kinase, EGF-R, with caveolins. Caveolin binding negatively regulates tyrosine and serine/threonine kinase activities. J Biol Chem 272:30429–30438, 1997. 8. Dada LA, Chandel NS, Ridge KM, Pedemonte C, Bertorello AM, Sznajder JI: Hypoxia-induced endocytosis of Na,K-ATPase in alveolar epithelial cells is mediated by mitochondrial reactive oxygen species and PKC-zeta. J Clin Invest 111:1057–1064, 2003. 9. Denker SP, Barber DL: Ion transport proteins anchor and regulate the cytoskeleton. Curr Opin Cell Biol 14:214–220, 2002. 10. Espineda CE, Chang JH, Twiss J, Rajasekaran SA, Rajasekaran AK: Repression of Na,K-ATPase beta1-subunit by the transcription factor snail in carcinoma. Mol Biol Cell 15:1364–1373, 2004. 11. Gheorghiade M, Adams KF, Colucci WS: Digoxin in the management of cardiovascular disorders. Circulation 109:2959–2964, 2004. 12. Grynkiewicz G, Poenie M, Tsien RY: A new generation of Ca2+ indicators with greatly improved fluorescence properties. J Biol Chem 260:3440–3450, 1985. 13. Holmen SL, Williams BO: Essential role for Ras signaling in glioblastoma maintenance. Cancer Res 65:8250–8255, 2005. 14. Huang YT, Chueh SC, Teng CM, Guh JH: Investigation of ouabain-induced anticancer effect in human androgen-independent prostate cancer PC-3 cells. Biochem Pharmacol 67:727–733, 2004. 15. Kanzawa T, Germano IM, Komata T, Ito H, Kondo Y, Kondo S: Role of autophagy in temozolomide-induced cytotoxicity for malignant glioma cells. Cell Death Differ 11:448–457, 2004. 16. Kondo Y, Kanzawa T, Sawaya R, Kondo S: The role of autophagy in cancer development and response to therapy. Nat Rev Cancer 5:726–734, 2005. 17. Krstic D, Krinulovic K, Spasojevic-Tisma V, Joksic G, Momic T, Vasic V: Effects of digoxin and gitoxin on the enzymatic activity and kinetic parameters of Na+/K+-ATPase. J Enzyme Inhib Med Chem 19:409–415, 2004. 18. Lang-Rollin IC, Rideout HJ, Noticewala M, Stefanis L: Mechanisms of caspase-independent neuronal death: Energy depletion and free radical generation. J Neurosci 23:11015–11025, 2003. 19. Lefranc F, Brotchi J, Kiss R: Possible future issues in the treatment of glioblastomas: Special emphasis on cell migration and the resistance of migrating glioblastoma cells to apoptosis. J Clin Oncol 23:2411–2422, 2005. 20 Lefranc F, Kiss R: Autophagy, the Trojan horse to combat glioblastomas. Neurosurg Focus 20:E7, 2006. 21. Lefranc F, Mijatovic T, Mathieu V, Rorive S, Decaestecker C, Debeir O, Brotchi J, Van Ham P, Salmon I, Kiss R: Characterization of gastrin-induced proangiogenic effects in vivo in orthotopic U373 experimental human glioblastomas and in vitro in human umbilical vein endothelial cells. Clin Cancer Res 10:8250–8265, 2004. 22. Lefranc F, Sadeghi N, Metens T, Brotchi J, Salmon I, Kiss R: Characterization of gastrin-induced cytostatic effect on cell proliferation in experimental malignant gliomas. Neurosurgery 52:881–891, 2003. 23. López-Lázaro M, Pastor N, Azrak SS, Ayuso MJ, Austin CA, Cortès F: Digitoxin inhibits the growth of cancer cell lines at concentrations commonly found in cardiac patients. J Nat Prod 68:1642–1645, 2005. 24. Lum JJ, DeBerardinis RJ, Thompson CB: Autophagy in metazoans: Cell survival in the land of plenty. Nat Rev Mol Cell Biol 6:439–448, 2005. 25. McConkey DJ, Lin Y, Nutt LK, Ozel HZ, Newman RA: Cardiac glycosides stimulates Ca2+ increases and apoptosis in androgen-independent, metastatic human prostate adenocarcinomas cells. Cancer Res 60:3807–3812, 2000. 26. Mégalizzi V, Mathieu V, Mijatovic T, Gailly P, Debeir O, De Neve N, Van Damme M, Bontempi G, Haibe-Kains B, Decaestecker C, Kondo Y, Kiss R, Lefranc F: 4-IBP: A sigma-1 receptor agonist decreases the migration of






30. 31.


33. 34. 35.


37. 38. 39.




43. 44. 45.

human cancer cells including glioblastoma cells, in vitro and sensitizes them in vitro and in vivo to the cytotoxic insults of pro-apoptotic and proautophagic drugs. Neoplasia 9:358–369, 2007. Mijatovic T, Gailly P, Mathieu V, De Néve N, Yeaton P, Kiss R, Decaestecker C: Neurotensin is a versatile modulator of in vitro human pancreatic ductal adenocarcinomas cell (PDAC) migration. Cell Oncol 29:315–326, 2007. Mijatovic T, Op De Beeck A, Van Quaquebeke E, Dewelle J, Darro F, de Launoit Y, Kiss R: The cardenolide UNBS1450 is able to deactivate nuclear factor kappaB-mediated cytoprotective effects in human non-small cell lung cancer cells. Mol Cancer Ther 5:391–399, 2006. Mijatovic T, Roland I, Van Quaquebeke E, Nilsson B, Mathieu A, Van Vynckt F, Darro F, Blanco G, Facchini V, Kiss R: The alpha1 subunit of the sodium pump could represent a novel target to combat non-small cell lung cancers. J Pathol 212:170–179, 2007. Mischel PS, Cloughesy TF: Targeted molecular therapy of GBM. Brain Pathol 13:52–61, 2003. Molitoris BA, Geerdes A, McIntosh JR: Dissociation and redistribution of Na+,K+-ATPase from its surface membrane actin cytoskeleton complex during cellular ATP depletion. J Clin Invest 88:462–469, 1991. Morrow JS, Cianci CD, Ardito T, Mann AS, Kashgarian M: Ankyrin links fodrin to the alpha subunit of Na,K-ATPase in Madin-Darby canine kidney cells and in intact renal tubule cells. J Cell Biol 108:455–465, 1989. Navarro A, Anand-Apte B, Parat MO: A role for caveolae in cell migration. FASEB J 18:1801–1811, 2004. Pattingre S, Levine B: Bcl-2 inhibition of autophagy: A new route to cancer? Cancer Res 66:2885–2888, 2006. Ransom CB, O’Neal JT, Sontheimer H: Volume-activated chloride currents contribute to the resting conductance and invasive migration of human glioma cells. J Neurosci 21:7674–7683, 2001. Senner V, Schmidtpeter S, Braune S, Puttmann S, Thanos S, Bartsch U, Schachner M, Paulus W: AMOG/beta2 and glioma invasion: Does loss of AMOG make tumour cells run amok? Neuropathol Appl Neurobiol 29:370–377, 2003. Sontheimer H: Malignant gliomas: Perverting glutamate and ion homeostasis for selective advantage. Trends Neurosci 26:543–549, 2003. Stenkwist B: Cardenolides and cancer. Anticancer Drugs 12:635–638, 2001. Stettner MR, Wang W, Nabors LB, Bharara S, Flynn DC, Grammer JR, Gillespie GY, Gladson CL: Lyn kinase activity is the predominant cellular SRC kinase activity in glioblastoma tumor cells. Cancer Res 65:5535–5543, 2005. Stupp R, Mason WP, van den Bent MJ, Weller M, Fisher B, Taphoorn MJ, Belanger K, Brandes AA, Marosi C, Bogdahn U, Curschmann J, Janzer RC, Ludwin SK, Gorlia T, Allgeier A, Lacombe D, Cairncross JG, Eisenhauer E, Mirimanoff RO; European Organisation for Research and Treatment of Cancer Brain Tumor and Radiotherapy Groups; National Cancer Institute of Canada Clinical Trials Group: Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med 352:987–996, 2005. Tian J, Cai T, Yuan Z, Wang H, Liu L, Haas M, Maksimova E, Huang XY, Xie ZJ: Binding of Src to Na+/K+-ATPase forms a functional signaling complex. Mol Biol Cell 17:317–326, 2006. Van Quaquebeke E, Simon G, André A, Dewelle J, El Yazidi M, Bruyneel F, Tuti J, Nacoulma O, Guissou P, Decaestecker C, Braekman JC, Kiss R, Darro F: Identification of a novel cardenolide (2ⴖ-oxovoruscharin) from Calotropis procera and the hemisynthesis of novel derivatives displaying potent in vitro antitumor activities and high in vivo tolerance: Structure–activity relationship analyses. J Med Chem 48:849–856, 2005. Xie Z, Askari A: Na+/K+-ATPase as a signal transducer. Eur J Biochem 269:2434–2439, 2002. Xie Z, Cai T: Na+-K+-ATPase-mediated signal transduction: From protein interaction to cellular function. Mol Interv 3:157–168, 2003. Yuan Z, Cai T, Tian J, Ivanov AV, Giovannucci DR, Xie Z: Na/K-ATPase ethers phospholipase C and IP3 receptor into a calcium-regulatory complex. Mol Biol Cell 16:4034–4045, 2005.

Acknowledgments This work was supported by grants awarded by the Fonds Yvonne Boël (Brussels, Belgium) and the Région de Bruxelles-Capitale (Brussels, Belgium).



n this report, Lefranc et al. have continued their research on overcoming apoptosis resistance in glioblastoma (GBM) cells by decreasing migration and inducing autophagy. They propose these mechanisms as possible avenues for glioma management. This current article contains a series of elegant in vitro and in vivo studies on a novel cardenolide (UNBS1450) that supposedly impairs GBM growth by inhibiting migration and promoting autophagy. The article convincingly demonstrates that UNBS1450 displays significant inhibitory effects on Na⫹/K⫹ATPase α1 subunits, has potent antiproliferative effects on several human GBM cell lines (Hs683, U373-MG, and T98G) in vitro, and prolongs survival of tumor-bearing mice in vivo. The in vitro studies are elegant, mechanistic, and appropriately controlled. Their results demonstrate that UNBS1450 impairs GBM cell migration through disorganization of the actin cytoskeleton and that UNBS1450 has potent proautophagic effects on GBM cell lines. The authors also found that genetic inhibition of the Na⫹/K⫹-ATPase α1 subunit via small interfering ribonucleic acid increased the magnitude of the antiproliferative effect of UNBS1450, suggesting that this α1 subunit is important for UNBS1450 sensitivity. Despite the potential issues associated with using highly passaged cell lines, as opposed to primary human tumor explants, the authors found their results to be consistent in at least three different GBM cell lines. This suggests that their findings are not merely an artifact of a single cell line and allows for adequate interpretation of this experimental work. Overall, this study clearly suggests that UNBS1450 may be a novel antitumor agent against primary brain tumors. I look forward to future studies aimed at further assessing the in vivo efficacy of this agent in relevant glioma models and the design of possible clinical trials based on these findings. Linda M. Liau Los Angeles, California efranc et al. present intriguing evidence for involvement of the α1 subunit of the ubiquitous Na⫹/K⫹-ATPase in GBM cell migration by its enriched colocalization with caveolin and at the subcellular lamellipodia, likely needed for adhesion and migration into brain parenchyma. In addition, specific targeting of the α1 subunit with a novel cardionolide compound, UNBS1450, showed inhibition of GBM cell proliferation in vitro via actin cytoskeletal disorganization and induced autophagy. There is evidence that this strategy holds promise for treating GBM xenografts in a preclinical model. The authors are to be commended for a careful study establishing a role for a previously overlooked membrane protein in GBM tumor biology and for a possible therapeutic strategy to augment and improve on the current armamentarium against a pernicious disease. I look forward to further work extending these results in an in vivo model and possibly for translation to clinical trial in the near future.


John S. Kuo Madison, Wisconsin

Disclosures Robert Kiss, Ph.D., is a Director of Research and Florence Lefranc, M.D., Ph.D., is a Clinical Research Fellow with the Belgian National Fund for Scientific Research (FNRS, Belgium). Tatjana Mijatovic, Ph.D., Sébastien Sauvage, Ms.C., and Isabelle Roland, Ms.C., are employees of Unibioscreen SA, Brussels, Belgium.



he authors have described a unique target not previously noted in high-grade glial tumors that has potential as a therapeutic strategy. The sodium pump and its α1 subunit appear to be overexpressed in GBMs. Blocking the expression of the pump’s α1 subunit leads to

VOLUME 62 | NUMBER 1 | JANUARY 2008 | 221


decreased proliferation and cellular migration along with promoting an increased lifespan in animals. Although one of the major concerns for this type of strategy will be cardiotoxicity, the authors have very nicely shown that the therapeutic ratio is greater than digoxin without showing any increased cardiotoxicity. Therefore, this peculiar side effect should not limit the antitumor treatment potential of blocking this specific sodium pump. It will also be interesting to know whether or not there is potential synergy between agents that are commonly used, such as other signaling inhibitors or temozolomide, along with this specific pump inhibitor. This could increase the cytotoxicity of the more traditional agents and prove to be a useful combination strategy. The authors have reported on innovative work that has a real therapeutic potential. Mitchel S. Berger San Francisco, California


n this fascinating laboratory study, the authors present compelling evidence for an antiproliferative and antimigratory effect on glioma cell lines by the drug UNBS1450, an inhibitor of the sodium pump. These studies are performed both in vitro and in a transplanted human glioma model in mice. In the latter, animal survival was significantly prolonged upon treatment with the drug and, in fact, the anticancer effect appears superior to that of temozolomide. Although

the sodium pump is essential for cell survival, the background relevant finding is that expression of the pump’s messenger ribonucleic acid seems much reduced in normal brain compared with GBM samples. There are some questions not addressed in the report that readers should be aware of. The current experiments do not tell us if the sodium pump messenger ribonucleic acid differential between GBM and brain is also reflected at the level of protein expression or if inhibition of the pump in normal brain cells, at the same concentration of drug used to provide the observed antiglioma effects, is tolerated. In addition, the authors assume that downstream effects of inhibition are related to disruptions of the actin cytoskeleton primarily based on morphological colocalization analyses, but this may require more detailed biochemical analyses to prove. Furthermore, they show effects on migration at the single cell level, but more detailed analyses of invasion are not provided. Overall, the study brings forth proof of concept experiments that the sodium pump may be another useful target for drug development against gliomas. Additional toxicological studies to determine therapeutic indices and studies against freshly excised human glioma cells in vitro and in vivo may be required before clinical trials in humans will validate the hypothesis that inhibition of the pump aids in glioma therapy. E. Antonio Chiocca Columbus, Ohio

View of a Cometary Impact Into Aerogel. Closeup view of a cometary impact (upper right) into aerogel was inspected by scientists at a laboratory at the Johnson Space Center, Houston, Texas, hours after the STARDUST Sample Return Canister was delivered to the Johnson Space Center from the spacecraft’s landing site in Utah. STARDUST successfully completed a seven-year, 2.8 billion mile journey to fly by the comet Wild 2 and returned samples to Earth on January 15, 2006. Image credit: National Aeronautics and Space Administration. From Jet Propulsion Laboratory California Institute of Technology, (www.nasa.gov/mission_pages/stardust/multimedia/jsc2006e01007.html). See Elder et al., pp 1–20.