Dicoumarol: a Unique Microtubule Stabilizing Natural Product That Is Synergistic with Taxol1

Home , Dicoumarol

[CANCER RESEARCH 63, 1214–1220, March 15, 2003] Dicoumarol: A Unique Microtubule Stabilizing Natural Product that Is Synergistic with Taxol1

Hamta Madari, Dulal Panda, Leslie Wilson, and Robert S. Jacobs2 Departments of Ecology, Evolution, and Marine Biology [H. M., R. S. J.] and Molecular, Cellular, and Developmental Biology [L. W.], University of California, Santa Barbara, California 93106, and Bhupat and Jyoti Mehta School of Biosciences and Bioengineering, Indian Institute of Technology, Bombay, Mumbai, India 400076 [D. P.]

ABSTRACT prostate cancer, malignant melanoma, and metastatic renal cell carcinoma (12–14). In studies on the antiproliferative actions of coumarin compounds, we In addition, the coumarin anticoagulants, dicoumarol (Dicumarol) -discovered that dicoumarol (a coumarin anticoagulant; 3,3؅-methyl and its synthetic derivative warfarin sodium (Coumadin), have been enebis[4-hydroxycoumarin]) inhibits the first cleavage of Strongylocentro- tus purpuratus (sea urchin) embryos in a concentration-dependent manner shown to decrease metastases in experimental animals (15). Warfarin with 50% inhibition occurring at a concentration of 10 ␮M. Because first sodium, largely replacing dicoumarol therapeutically as an anticoag- cleavage in sea urchin embryos is highly selective for microtubule- ulant, has been used for the treatment of a variety of cancers and targeted agents, we thought that the active compounds might inhibit cell shown to improve tumor response rates and survival in patients with division by interacting with tubulin or microtubules. We found that several types of cancer (16–22). However, despite numerous studies, ␮ dicoumarol binds to bovine brain tubulin with a Kd of 22 M and that 0.1 little information has been acquired on the cellular mechanism of ␮ M dicoumarol strongly stabilizes the growing and shortening dynamics action of coumarin compounds in the treatment of malignancies. at the plus ends of the microtubules in vitro. Dicoumarol reduces the rate Possibly for this reason, the coumarin compounds have not received and extent of shortening, it increases the percentage of time the microtubules spend in an attenuated (paused) state, and it reduces the overall much attention for the treatment of cancer. dynamicity of the microtubules. The antimitotic effects of the widely used On further investigation of the coumarin literature, we developed cancer chemotherapeutic agent Taxol (paclitaxel) are also mediated by the hypothesis that coumarin compounds might inhibit cell prolifer- suppressing microtubule dynamics. We demonstrate that exposure to ation by interfering with mitotic spindle microtubule function. Earlier combinations of Taxol and dicoumarol results in a synergistic inhibition of studies revealed that coumarin, 7-hydroxycoumarin, and 4-hydroxy- cell division of sea urchin embryos. The results suggest that the antipro- coumarin inhibit mitosis in Allium cepa root tips (23, 24). Interest- liferative mechanism of action of dicoumarol and possibly related phar- ingly, 7-hydroxycoumarin disorganized the mitotic spindle microtu- macophores may be mediated by tubulin binding and the stabilization of bules in A. cepa cells, leading to the random distribution of the spindle microtubule dynamics. Because of its low toxicity and simple chromosomes at metaphase (24). This is a form of cytotoxicity com- chemical structure, there is potential interest to explore combinations of antimitotic coumarins with other chemotherapeutic agents to improve mon to mitotic spindle poisons that inhibit mitosis by modifying efficacy and lower toxicity. microtubule dynamics (25). Microtubules are intrinsically dynamic polymers composed of the heterodimeric protein tubulin. Both in vitro and in cells, microtubule INTRODUCTION ends switch between growing and shortening states, a behavior termed dynamic instability (see Ref. 25 for a review). Microtubule dynamics Dicoumarol, a natural anticoagulant drug chemically designated as are essential for the proper movements of chromosomes during mi- 3,3Ј-methylenebis[4-hydroxycoumarin], is metabolized from couma- tosis and play a crucial role in passage through the metaphase/ rin in the sweet clover (Melilotus alba and Melilotus officinalis)by anaphase checkpoint. At the onset of mitosis, the interphase microtu- molds, such as Penicillium nigricans and Penicillium jensi. Coumarin bule network is replaced by a population of spindle microtubules that (1,2-benzopyrone), the parent molecule of dicoumarol, is the simplest are 10–100 times more dynamic than the interphase network (26). compound of a large class of naturally occurring phenolic substances These rapid dynamics are crucial for attachment at the kinetochores made of fused benzene and ␣-pyrone rings (1). Our investigation of and proper chromosome alignment for segregation at anaphase (25, coumarin compounds originated when translated Persian literature3 27). Cells with improper spindle assembly or chromosome alignment revealed that a wide spectrum of medicinal plant extracts that were in are prevented from transitioning from metaphase to anaphase by use as early as 1000 A.D. contains a high content of coumarins.4 mitotic checkpoint mechanisms, resulting in cell cycle arrest in mi- Subsequent analysis of scientific literature revealed numerous reports on the antiproliferative and antitumor activities of a variety of cou- tosis followed by apoptosis (28–30). marin compounds, e.g., both coumarin itself and 7-hydroxycoumarin In the present study, we investigated the antimitotic actions of have been reported to inhibit the proliferation of a number of human several coumarin compounds on the first division of sea urchin malignant cell lines in vitro (2–6) and have demonstrated activity embryos. The first few cell divisions of sea urchin embryos exhibit a against several types of animal tumors (7–11). These compounds have remarkable degree of pharmacological selectivity toward mitotic spin- also been reported in clinical trials to demonstrate activity against dle poisons (31). We found that several coumarin compounds inhibited the first cleavage of Strongylocentrotus purpuratus embryos in a concentration-dependent manner, the most potent of the coumarins Received 8/12/02; accepted 1/15/03. The costs of publication of this article were defrayed in part by the payment of page tested being dicoumarol. By video microscopy and fluorescence spec- charges. This article must therefore be hereby marked advertisement in accordance with troscopy, we determined that dicoumarol binds to tubulin and sup- 18 U.S.C. Section 1734 solely to indicate this fact. 1 Supported by Grants 14350100CA31063 and NSI3560 from the United States presses the dynamic instability of individual bovine brain microtu- Department of Interior, Mineral Management Service, and USPHS, respectively. bules in vitro. In addition, we evaluated the combined antiproliferative 2 To whom reprint requests should be addressed, at Department of Ecology, Evolution activity of Taxol (paclitaxel) and dicoumarol in the sea urchin cell and Marine Biology, University of California, Santa Barbara, CA 93106. Phone: (805) 893-4007; Fax: (805) 893-4724; E-mail: [email protected]. division model and found that the combination did produce synergis- 3 Received as a gift from the personal archives of ancient Persian literature of Seyedeh tic inhibition of cell division. In view of the relative simplicity of the Nasrin Sohrab (in memorandum) and translated from the original Farsi by Amin Madari. 4 H. Madari and R. S. Jacobs. Coumarin content of medicinal plant extracts used in coumarin compounds and their microtubule-targeted mechanism of ancient Persian medicines, manuscript in preparation. action, we believe the coumarin pharmacophore may serve as an 1214

Downloaded from cancerres.aacrjournals.org on September 28, 2021. © 2003 American Association for Cancer Research. DICOUMAROL IS SYNERGISTIC WITH TAXOL important model on which to develop new avenues of cancer chem- achieve three to six seeds per microscope field. After 35 min of incubation, otherapy. samples of microtubule suspensions (2 ␮l) were prepared for video microscopy, and the dynamic instability behavior of individual microtubules was recorded at 37°C and analyzed as described previously (34). Microtubules MATERIALS AND METHODS were observed for a maximum of 45 min after they had reached steady state. Materials. All drugs and reagents were obtained from Sigma Chemical Co. We considered a microtubule to be in a growth phase if it increased in length Ͼ ␮ Ͼ ␮ (St. Louis, MO) unless otherwise indicated. by 0.2 m at a rate 0.15 m/min and in a shortening phase, if it shortened Ͼ ␮ Ͼ ␮ Inhibition of Cell Cleavage. Sea urchins, either S. purpuratus or Strongy- by 0.2 m at a rate 0.3 m/min. Microtubules undergoing length Յ ␮ locentrotus franciscanus, were induced to spawn by injection of 0.5 M KCl into changes 0.2 m over the duration of six data points were considered to be the coelomic cavity and cultured in filtered seawater, as described previously in an attenuated state. The same tubulin preparation was used for all experi- (31). Eggs were used in fresh filtered seawater at a concentration of ϳ1% ments; 20–30 microtubules were analyzed for each experimental condition. (v/v).5 Sperm were collected directly from the aboral surface of male sea The catastrophe frequency [a catastrophe is a transition from the growing or urchins and diluted into seawater (one drop of concentrated sperm/1 ml of attenuated state to shortening (35)] was determined by dividing the number of seawater). catastrophes by the sum of the total time spent in the growing plus attenuated Eggs were fertilized by adding diluted sperm to the diluted egg suspensions states for all microtubules for a particular condition. The rescue frequency [a (1 ml of sperm suspension/100 ml of diluted egg suspension) and then added rescue is a transition from shortening to growing or attenuation, excluding new to various drug concentrations. The embryos were then incubated at 15°Cina growth from a seed (35)] was calculated by dividing the total number of rescue water bath for ϳ120 min after fertilization. Quantification of division was events by the total time spent shortening for all microtubules for a particular performed visually by light microscopy by counting the number of cleaved and condition. Dynamicity is the sum of all growing and shortening events divided noncleaved embryos after the control embryos had progressed to the end of by the total time measured, including time spent in the attenuated state (32). first cleavage. Fluorescence Spectroscopy. Fluorescence spectroscopy was performed The effect of dicoumarol at different stages of the cell cycle was determined using a Perkin-Elmer LS50B spectrofluorometer. Spectra were taken by mul- by adding dicoumarol at 10-min intervals after fertilization to aliquots of sea tiple scans, and buffer blanks were subtracted from all measurements. The urchin embryos and scoring the inhibition percentage of first cleavage. To inner filter effects were corrected as described by Sackett (36) and empirically examine the reversibility of the compound, dicoumarol was added to embryo by measuring the change of fluorescence intensity of a tryptophan solution suspensions within 1 min of fertilization. Exposure to the compound was equivalent to the tubulin concentration in the presence of dicoumarol. Dicou- terminated at 10-min intervals by sedimenting and resuspending the embryos marol did not quench the fluorescence of tryptophan in solution after inner three times in fresh seawater. filter effect correction. The excitation and emission wavelengths were 295 and ␤ Purification of MTP, Tubulin, and Flagellar Axonemal “Seeds.” Bo- 336 nm, respectively. The fraction of binding sites ( ) occupied by dicoumarol ␤ ϭ Ϫ Ϫ vine brain MTP (70% tubulin and 30% microtubule-associated proteins) was was determined using the following relationship: (Fo F)/(Fo Fm), purified by three cycles of warm polymerization and cold depolymerization in where Fo is the fluorescence intensity of tubulin in the absence of dicoumarol, vitro. Tubulin was purified from the MTP by phosphocellulose chromatogra- F is the corrected fluorescence intensity when the tubulin and dicoumarol are phy as described previously (32). The MTP and purified tubulin were quickly in equilibrium, and Fm is the calculated fluorescence intensity of the fully Ϫ frozen as drops in liquid nitrogen and stored at Ϫ70°C until used. Axoneme liganded tubulin. Fm was determined by plotting 1/(Fo F) versus 1/L ϭ ϭ seeds were prepared from sea urchin sperm (S. purpuratus; Ref. 32). All (L total ligand concentration) and extrapolating 1/L 0. The association ϭ ␤ Ϫ ␤ ϫ protein concentrations were determined by the method of Bradford (33) using constant, Ka, was determined using the relationship: Ka ( /1 ) 1/Lf, ϭ Ϫ ␤ BSA as the standard. where Lf L [C] and [C] is the molar concentration of ligand-binding Assembly of Microtubules and Determination of Steady-state Microtu- sites, assuming a single binding site per tubulin dimer. bule Polymer Mass. Bovine brain MTP and phosphocellulose-purified tubu- Analysis of Drug Synergism. The CI was used to determine whether the lin pellets were thawed and centrifuged at 4°C to remove any aggregated or drugs interacted synergistically, additively, or antagonistically (37). The CI is denatured tubulin. Several reaction mixtures were prepared to determine the defined by the following equation: effects of dicoumarol on microtubule assembly. Tubulin (2.5 mg/ml) was polymerized in the absence or presence of a range of dicoumarol concentra- (D)1 (D)2 CI ϭ ϩ , tions (35 min; 37°C) in PME buffer (100 mM piperazinediethanesulfonic acid, (Dx)1 (Dx)2 1mM MgCl2,1mM EGTA (pH 6.8), and 100 mM GTP). The same reaction mixture was used as described above, using S. purpuratus axoneme seeds or in which (D)1 is the concentration necessary for a particular effect in glycerol microtubule seeds for assembly initiation. To prepare glycerol micro- the combination, (Dx)1 is the dose of the same drug that will produce tubule seeds for nucleation, 30 ␮M tubulin was polymerized in PME buffer the identical level of effect by itself, (D)2 is the concentration of a containing 10% gylcerol (v/v) and 100 mM GTP for 30 min at 30°C. After 30 min of polymerization, microtubules were sheared by passage (Յ10 times) second drug that will produce a particular effect in the combination, through the needle of a 50-␮l Hamilton syringe to form seeds. Seeds were then and (Dx)2 is the concentration of the second drug that will produce the added in a ratio of 1:5 v/v to the polymerization mixture and warmed to room same level of effect by itself. For combinations, a CI Ͼ 1 implies temperature. The final glycerol concentration was 2%. For studies with MTP, antagonism, a CI Ͻ 1 indicates synergy, and a CI ϭ 1 indicates samples containing 2.5 mg of MTP/ml were incubated in PME buffer con- additivity. taining 100 mM GTP in the presence or absence of various dicoumarol concentrations without seeds. All of the above reaction mixtures were polymerized for 35–45 min at 37°C to achieve near steady state, and microtubule RESULTS polymerization was monitored by turbidimetry at 350 nm using a Gilford Response Spectrophotometer. To determine the microtubule mass, microtu- Inhibition of Cell Division by Dicoumarol. In preliminary studies bules assembled in the above reaction mixtures were pelleted by centrifugation on the antimitotic actions of coumarins, we found that several cou- at 150,000 ϫ g for1hat37°C. Microtubule pellets were solubilized in PME marin compounds inhibited the first cleavage of S. purpuratus em- buffer at 0°C for protein determination. bryos in a concentration-dependent manner with 50% inhibition Determination of Microtubule Dynamic Instability by Video Micros- occurring at 10 ␮M dicoumarol, 27 ␮M coumarin, 36 ␮M 7-hydroxy- copy. Tubulin (12 ␮M) was polymerized to steady state onto flagellar seeds in coumarin, and 25 ␮M warfarin (data not shown). Dicoumarol was the the absence or presence of dicoumarol. The seed concentration was adjusted to ϭ ␮ most potent of the coumarin compounds examined; IC50 10 M (Fig. 1) with a complete block of cell division at concentrations Ͼ 20 5 The abbreviations used are: v/v, volume or volume; MTP, microtubule protein; PME, ␮M 100 nm piperazinediethanesulfonic acid, 1 mM MgCl2, 1 mM EGTA (pH 6.8) and 100 nm . Dicoumarol-treated embryos showed no lysis or morphological GTP; PC, phosphocellulose chromotagraphy; CI, combination index. changes at any of the concentrations treated. Identical concentration- 1215

growth or shortening at microtubule ends was not detectable. Visual analysis of the traces revealed that 0.1 ␮M dicoumarol significantly suppressed the dynamics. The effects of dicoumarol (0.04–50 ␮M) on the individual dynamic instability parameters are shown quantitatively in Table 1. Dicouma- rol inhibited the rate and extent of microtubule shortening in a concentration-dependent manner (Fig. 5A). The addition of 1 ␮M dicoumarol significantly reduced the mean shortening rate by ϳ58%, from 18.5 to 7.8 ␮m/min, and it reduced the length of a shortening excursion by ϳ40%, from 2.3 to 1.4 ␮m. In contrast to the strong Fig. 1. Log concentration response curve for inhibition of S. purpuratus embryo action of dicoumarol on the rate and length of shortening, there was cleavage by dicoumarol. Pooled results from eight representative experiments are shown. no significant change in the average rate or length of growth at the Error bars, SE. various dicoumarol concentrations examined (Fig. 5B). Steady-state microtubules in vitro and in cells spend a portion of response curves were also obtained using Lytechinus pictus and S. time in an attenuated (or paused) state (32). At concentrations as low franciscanus embryos (data not shown). as 0.1 ␮M, dicoumarol significantly increased the overall percentage Cell Cycle Analysis and Reversibility of Dicoumarol. To inves- of time in the attenuated state. tigate where in the cell cycle dicoumarol exerts its activity, sea urchin The effects of dicoumarol on the frequency of transition among embryos were exposed to dicoumarol at progressively later times after fertilization. This allows rapid identification of phases of the cell cycle that are sensitive to the drug. Fig. 2 shows that maximal inhibition of embryo cleavage occurred when the dicoumarol was added as late as 75 min after fertilization, well after the time of DNA synthesis. Inhibition declined steeply when dicoumarol was added 90 min after fertilization, during prophase and pro-metaphase of mitosis. These results suggest that dicoumarol inhibits cell division by blocking cells at prometaphase or metaphase of mitosis. We examined whether inhibition of cell division by dicoumarol could be reversed. Embryos were incubated with 50 ␮M dicoumarol (a concentration which inhibits cell division by 100%); beginning im- mediately after fertilization, the embryos were washed by sedimentation and resuspended in fresh seawater to remove the compound at 10-min intervals until the time of first cleavage (2 h). As shown in Fig. 2B, the effects of dicoumarol appear to be ϳ30% (t ϭ 35–65 min) to 60% (t ϭ 15 min) reversible when the compound is removed before M phase of the cell cycle. Thus, dicoumarol is reversible until the onset of mitosis. Binding of Dicoumarol to Tubulin. Tubulin (2 ␮M) was incubated with a range of dicoumarol concentrations for 30 min at 34°Cin 50 mM PME buffer. Dicoumarol quenched the intrinsic fluorescence of tubulin in a concentration-dependent manner. The binding constant Fig. 2. A, the effect of dicoumarol at different stages of the cell cycle in S. purpuratus determined from the data was 22 ␮M (Fig. 3), and the dicoumarol embryos. Dicoumarol (50 ␮M) was added to embryo suspensions at the times indicated bound to the tubulin with a 1:1 stoichiometry. after fertilization. Incubation was continued until control embryos had completed first cleavage. Pro, Met, Ana, and Tel are prophase, metaphase, anaphase, and telophase of Effects of Dicoumarol on Microtubule Polymerization. As de- mitosis, respectively. B, reversibility of 50 ␮M dicoumarol during the first division cycle. termined by sedimentation and light scattering assays, a wide range of The drug was added 1-min postfertilization, then at 10-min intervals; a 1-ml sample of 1% dicoumarol concentrations (0–100 ␮M) incubated with PC-tubulin, (v/v) suspension of eggs in seawater was washed three times with 10 ml of filtered seawater. axoneme-seeded PC-tubulin, glycerol-seeded PC-tubulin, or MTP did not produce any significant alteration in the polymer mass or the rate and extent of polymerization of free tubulin into microtubules (data not shown). However, because recent data suggest that the suppression of microtubule dynamics rather than the stimulation or inhibition of microtubule polymerization is the primary anticancer mechanism of action of many antimitotic drugs (see Ref. 25 for review), we examined the effects of dicoumarol on microtubule dynamic instability in vitro. Effects of Dicoumarol on Microtubule Dynamic Instability. The effects of dicoumarol on the growth and shortening dynamics at the plus ends of individual microtubules in vitro were analyzed by video microscopy. Several life history traces of microtubule Fig. 3. Effects of dicoumarol on tubulin fluorescence. Tubulin (2 ␮M) was incubated length changes with time in the presence and absence of 0.1 ␮M in the absence or presence of various dicoumarol concentrations at 25°C for 30 min, and dicoumarol are shown in Fig. 4. Control microtubules alternated the fluorescence was determined as described in “Materials and Methods.” The above figure shows protein fluorescence quenching titration after binding of dicoumarol to between phases of growing and shortening but also spent a small tubulin. Fluorescence values at 336 nm were taken for the calculation. Data are repre- percentage of the time in an attenuated state, periods during which sentative of four different experiments. 1216

chemotherapeutic agent, Taxol, we investigated the combined effects of the two compounds on cleavage. In the presence of 5 ␮M dicou- ϳ marol (a concentration that inhibits cell division by 25%), the IC50 for Taxol was effectively lowered by Ͼ2-fold (Fig. 6A). Combina- tions of Taxol with dicoumarol produced a more pronounced inhibition of cell division than expected. As seen in Fig. 6B, the inhibition percentage of cell division in cells incubated with 0.3 ␮M Taxol increased by Ͼ70% in the presence of 5 ␮M dicoumarol. Calculation of the CI indicates that Taxol and dicoumarol act synergistically to inhibit cell proliferation (CI ϭ 0.31).

DISCUSSION Mechanism of Action of Dicoumarol. The fertilized sea urchin egg has shown utility as an experimental cell model for investigating mechanisms of drug action. Eggs fertilized at the same time undergo numerous synchronous divisions, allowing rapid identification of drug-induced inhibition of cell division. In addition, the first cell cycle exhibits a remarkable degree of pharmacological selectivity for mitotic spindle poisons while being relatively insensitive to agents acting by other common inhibitory mechanisms (31). In the present study we found that dicoumarol produced a concentration-dependent inhibition ϭ ␮ of the first cell division of the sea urchin embryo (IC50 10 M). We then examined the ability of the compound to block cells in Fig. 4. Length changes of individual microtubules at their plus ends at steady state in the absence (A) and presence (B)of0.1␮M dicoumarol. mitosis, as would be expected of a mitotic spindle poison. The first mitosis typically occurs between 70 and 90 min after fertilization, and cytokinesis is usually complete at 120 min. By adding dicoumarol to phases of growing, shortening, and attenuation were calculated as fertilized embryos at different times before the completion of the first events per unit of time and are considered important parameters in the cleavage, we found that the ability of dicoumarol to inhibit the first regulation of microtubule dynamics in cells (28). They are also cell division occurred when added as late as 75 min after fertilization. indicators of whether the drug is acting on the stabilizing cap. Dicou- These results suggested that dicoumarol selectively inhibits cell cycle marol decreased the catastrophe frequency at concentrations as low as progression at prometaphase/metaphase of mitosis, and actions on any 0.1 ␮M, by 65%, from 0.62 to 0.22 events/min. These results indicate events before the beginning of M phase are not crucial to the activity that dicoumarol may stabilize the cap at microtubule plus ends. of the drug. Dicoumarol had no apparent effect on the rescue frequency. Because dicoumarol inhibits cell division by acting on a cellular Dynamicity is the summed gain and loss (exchange) of tubulin process essential during mitosis, we reasoned that the cellular target subunits at the microtubule ends and is a measure of overall dynamic for these compounds might be tubulin or microtubules. We first instability behavior (32). Dynamicity was reduced by 57% at 0.1 ␮M. analyzed the effects of dicoumarol on microtubule assembly in vitro Thus, dicoumarol significantly decreases the dynamic instability be- and found that even at very high concentrations (100 ␮M), the drug did havior of microtubules, primarily by decreasing the catastrophe fre- not produce any appreciable changes in the rate or extent of polym- quency and inhibiting the rate and extent of microtubule shortening. erization. However, because the suppression of microtubule dynamics Synergistic Inhibition of Cell Division by Dicoumarol and by antimitotic drugs in the absence of appreciable change in polymer Taxol. To determine whether the effect of dicoumarol on division of mass occurs at the lowest effective concentrations with most of these sea urchin embryos could be additive or possibly synergistic with the compounds (25), we then examined the effects of dicoumarol on

Table 1 Effects of dicoumarol on the dynamic instability parameters at microtubule plus ends at steady state

Concentration, ␮M

0 0.04 0.1 1 10 50 Mean rate (␮m/min) Growing 0.79 Ϯ 0.1a 0.61 Ϯ 0.07 0.53 Ϯ 0.1 0.63 Ϯ 0.08 0.63 Ϯ 0.1 0.54 Ϯ 0.07 Shortening 18.5 Ϯ 7.2 18.9 Ϯ 8.1 12.1 Ϯ 4.6 7.8 Ϯ 4.8 7.4 Ϯ 2.4 1.1 Ϯ 0.2 Mean length (␮m/event) Growing 0.86 Ϯ 0.15 0.70 Ϯ 0.09 0.61 Ϯ 0.1 0.95 Ϯ 0.22 0.58 Ϯ 0.09 1.0 Ϯ 0.17 Shortening 2.3 Ϯ 0.58 2.0 Ϯ 0.53 1.9 Ϯ 0.43 1.4 Ϯ 0.44 1.6 Ϯ 0.37 1.4 Ϯ 0.21 Mean duration (min) Growing 1.7 Ϯ 0.52 1.3 Ϯ 0.18 1.5 Ϯ 0.29 1.7 Ϯ 0.30 1.7 Ϯ 0.38 2.36 Ϯ 0.34 Shortening 0.53 Ϯ 0.18 0.37 Ϯ 0.13 1.1 Ϯ 0.43 0.86 Ϯ 0.34 0.51 Ϯ 0.15 1.4 Ϯ 0.07 Total time (%) Growing 59.4 60.3 34.7 32.3 20.7 41.9 Shortening 27 22 27.8 17.3 8.1 6.8 Attenuated 13.7 17.7 37.6 50.5 71.2 51.3 Transition frequencies (events/min) Catastrophe 0.62 0.7 0.22 0.24 0.11 0.07 Rescue 0.79 2 0.38 1.03 1.2 0.8 Dynamicity 1.5 1.2 0.65 0.43 0.24 0.31 a Ϯ, SE. 1217

potent action of Taxol on dynamic instability is the suppression of the rate and extent of shortening in the absence of an appreciable effect on the rate or extent of growth (39), an effect we find is shared by dicoumarol. In addition, both compounds greatly increase the percentage of time the microtubules spend in the attenuated state. This is in contrast to the mechanism of other common microtubule-stabilizing agents, such as vinblastine and colchicine, which stabilize plus ends of the microtubules by suppressing both the rate and extent of growth and shortening (32, 34, 40). We believe that in this respect, the mechanism by which dicoumarol stabilizes microtubule dynamics is Taxol-like. There are, however, several notable differences in the action of dicoumarol and Taxol. Taxol exhibits differential effects on microtubule properties at different binding stoichiometries. Substoichiometric binding of Taxol to tubulin in microtubules suppresses the microtubule shortening rate but does not significantly affect the growing rate or polymer mass (39). However, at Taxol concentrations that are stoichiometric with respect to tubulin concentrations in solution, there is extensive stimulation of the rate and extent of microtubule polym- Fig. 5. Effects of dicoumarol on the shortening (A) and growing (B) rates of individual erization in vitro (41). It enhances nucleation and elongation phases of microtubules at their plus ends. microtubule polymerization reaction in vitro and reduces the critical subunit concentration (39). Unlike the actions of Taxol, high concentrations of dicoumarol do not produce an appreciable change in the microtubule dynamic instability in vitro. We found that at concentra- rate or extent of microtubule polymerization. An additional difference tions that did not affect the polymer mass, dicoumarol strongly sup- between dicoumarol and Taxol is their difference in tubulin-binding pressed microtubule dynamic instability. It suppressed the shortening properties. Taxol binds preferentially to the polymeric microtubule rate more significantly than the growing rate and decreased the form of tubulin and binds extremely poorly to soluble tubulin (42– catastrophe frequency, resulting in stabilized microtubules. 44). Dicoumarol, however, binds to tubulin dimers in 1:1 stoichiom- Direct evidence that dicoumarol binds to tubulin was obtained by etry with moderate affinity. This difference suggests that dicoumarol fluorescence spectroscopy. A dissociation constant of 22 ␮M and a and Taxol may have different binding sites, which may prove useful binding stoichiometry of 1 mol of dicoumarol per mol of tubulin were in formulating combination therapy approaches. determined by analyzing the ability of dicoumarol to quench the intrinsic fluorescence of tubulin. Dicoumarol decreased the intrinsic fluorescence of tubulin, most likely by causing a conformational change in tubulin in the region of a fluorophore, such as a tryptophan residue on binding. The conformational change in tubulin induced by dicoumarol binding may be an important factor in its ability to stabilize microtubule dynamics. It is also possible that in the presence of dicoumarol, the affinity between adjacent dimers is increased, making tubulin dissociation less favorable. Microtubules are composed of an unstable tubulin-GDP core and a stable tubulin-GTP or tubulin-GDP-Pi “cap” at the microtubule ends. It is believed that microtubule dynamic instability is caused by the gain and loss of this stabilizing cap at microtubule ends (38). Loss of the cap is thought to be required for initiation of a shortening phase. Dicoumarol may stabilize microtubules by binding to unstable GDP- liganded tubulin with a higher affinity than growing GTP-liganded tubulin or that the conformational change induced in tubulin by dicoumarol resembles that of the stable tubulin-GTP cap. It is also possible that dicoumarol reduces the rate of GTP hydrolysis or Pi release, preventing cap loss and a decrease in the frequency of catastrophe. Comparison of the Mechanism of Action of Dicoumarol and Taxol. Antimitotic agents that interact with microtubules are of interest because of their potential uses in the treatment of human neoplastic and inflammatory diseases. Although substoichiometric concentrations of diverse antimitotic drugs, such as Taxol, vinblastine, and colchicine, suppress microtubule dynamics in cells and inhibit cell proliferation, the mechanism by which these drugs affect dynamic instability parameters, such as growing and shortening rates, duration Fig. 6. Synergy between Taxol and dicoumarol. The combination of Taxol with of attenuated states, or transition frequencies (rescues and catastro- dicoumarol decreases the concentration of Taxol required to induce 50% inhibition of cell proliferation in the sea urchin cell division assay. The results are expressed as the phes), differs. Dicoumarol appears to be most similar mechanistically percentage of inhibition compared with vehicle and the means of five separate experi- to Taxol in its ability to stabilize microtubule dynamics. The most ments; bars, SE. 1218

Implications for Combination Therapy. Evidence suggests that structural class for synthetic elaboration that could lead to improved combination therapy with drugs that stabilize microtubule dynamics antineoplastic drugs. The simple chemical structure of the coumarins by different mechanisms may improve responses and minimize side allows great potential to clinically explore combinations of coumarin effects of the individual drugs. Although Taxol has shown good analogs with other microtubule-stabilizing agents in an attempt to efficacy against refractive ovarian cancer, metastatic breast cancer, improve efficacy. In addition, dicoumarol and possibly other couma- head and neck cancer, melanoma, and lung cancer, its use in the clinic rin compounds that act by stabilizing microtubule dynamics might is hampered by its side effects, low solubility, and narrow therapeutic have entirely new tumor specificity than the currently used antimitotic index (45). Recent clinical trials have indicated that combinations of agents. Their synergistic activity in combination therapy could be the antimitotic drugs result in increased antitumor activity and decreased basis for development of rational approaches to new forms of cancer toxicity (46–50), e.g., combinations of Taxol and estramustine, and chemotherapy. enhance cytotoxicity against human prostate carcinoma (47, 50). Estramustine, an antitumor drug used in the treatment of hormone- ACKNOWLEDGMENTS refractory prostate cancer, is similar to dicoumarol in that it weakly Ϸ ␮ binds to tubulin (Kd 30 M), does not appreciably affect the We thank Seyedeh Nasrin Sohrab for access to important Persian material microtubule polymer mass, but strongly reduces the overall dynam- medica used in developing conceptual aspects underlying this research. We icity of the microtubules (51). Thus, compounds with the same intra- also thank Herb Miller for excellent technical assistance and Mary Ann Jordan cellular target, but different binding sites and modes of action, may for advice and comments through the course of this work. have the potential to be used in combination for chemotherapeutic purposes. REFERENCES We have demonstrated that the combination of both drugs acts 1. Keating, G. J., and O’Kennedy, R. The chemistry and occurrence of coumarins. In: synergistically to the inhibit cell division in sea urchin embryos. This R. O’Kennedy and R. D. Thornes (eds), Coumarins: Biology, Applications and Mode of Action, pp. 23–64. West Sussex, England: John Wiley & Sons, 1997. observation is similar to that obtained in a study in which the com- 2. Marshall, M. E., Mohler, J. L., Edmonds, K., Williams, B., Butler, K., Ryles, M., bination of Taxol and coumarin on meristematic cells of Allium Weiss, L., Urban, D., Beuschen, A., and Markiewicz, M. An updated review of the sativum root tips increased antimitotic activity and was associated clinical development of coumarin (1, 2-benxopyrone) and 7-hydroxycoumarin. J. Cancer Res. Clin. Oncol., 120: 39–42, 1994. with fewer cytotoxic and clastogenic effects than when Taxol was 3. Moran, E., Prosser, E., O’Kennedy. R., and Thornes, R. D. The effect of coumarin and used alone (52). Together, these results suggest that dicoumarol and 7-hydroxycoumarin on the growth of human tumor cell lines. J. Irish Coll. Phys. Taxol have the potential to be used together to improve efficacy and Surg., 22: 41–48, 1993. 4. Siegers, C. P., and Bostelmann, H. C. Effect of coumarin on cell cycle proliferation lower toxicity. in human tumor cell lines. J. Irish Coll. Phys. Surg., 22: 47–50, 1993. It is also possible that the coumarin anticoagulants in combination 5. Myers, R. B., Parker, M., and Grizzle, W. E. The effects of coumarin and suramin on growth of malignant renal and prostatic cell lines. J. Cancer Res. Clin. Oncol., 120: with Taxol may improve cancer survival rates by other mechanisms. 11–13, 1994. Tumor cell invasion is dependent on angiogenesis and requires both 6. Finn, G. J., Creaven, B., and Egan, D. A. Study of the in vitro cytotoxic potential of cell migration and digestion of basement membrane by proteases (53). natural and synthetic coumarin derivatives by using human normal and neoplastic skin cell lines. Melanoma Res., 11: 461–467, 2001. Anticoagulants were thought to reduce the incidence of cancer me- 7. Feuer, G., Kellen, J. A., and Kovacs, K. Suppression of 7, 12-dimethylbenz(a)an- tastases by inhibiting formation of a fibrin matrix required for the thracene-induced breast carcinoma by coumarin in the rat. Oncology, 33: 35–39, fixation of circulating cancer cells. Because Taxol has been reported 1976. 8. Thornes, D., Daly, L., Lynch, G., Browne, H., Tanner, A., Keene, F., O’Loughlin, S., to inhibit migration of human ovarian and prostate carcinoma cells Corrigan, T., Daly, P., Edwards, G., Breslin, B., Browne, H. Y., Shine, M., Lennon, (54, 55) and the coumarin anticoagulants inhibit the thrombin-induced F., Hanley, J., McMurray, N., and Gaffney, E. Prevention of early recurrence of high risk malignant melanoma by coumarin. Eur. J. Surg. Oncol., 15: 431–435, 1989. release of matrix metalloproteinases that cause the breakdown of 9. Omarbasha, B., Fair, W. R., and Heston, W. D. Effect of coumarin on the normal rat extracellular matrix proteins (56), it is logical to assume that dicou- prostate and on the R-3327H prostatic adenocarcinoma. Cancer Res., 49: 3045–3049, marol might enhance the antiadhesive activity of taxanes. Coumarin 1989. 10. Raev, L. D., Voinova, E., Ivanov, I. C., and Popov, D. Antitumor activity of some anticoagulants inhibit the pathway involving tissue factor and factor coumarin derivatives. Pharmazie, 45: 696, 1990. VIIa. Tissue factor VIIa appears to be a major factor in the regulation 11. Maucher, A., and Von Angerer, E. Antitumor activity of coumarin and 7-hydroxy- of angiogenic growth properties of tumor cells, and in vivo studies coumarin against 7, 12-dimethylbenz[a]anthracene-induced rat mammary carcino- mas. J. Cancer Res. Clin. Oncol., 120: 502–504, 1994. have demonstrated a significant role for warfarin in the regulation of 12. Thornes, R. D., Daly, L., Lynch, G., Breslin, B., Browne, H., Browne, H. Y., the inhibition of angiogenesis (57). Taxol also has been shown to Corrigan, T., Daly, P., Edwards, G., and Gaffney, E. Treatment with coumarin to prevent or delay recurrence of malignant melanoma. J. Cancer Res. Clin. Oncol., 120: display antiangiogenic properties in association with the down-regu- S32–S34, 1994. lation of vascular endothelial growth factor, a pro-angiogenic factor 13. Marshall, M. E., Butler, K., and Fried, A. Phase I evaluation of coumarin (1, that can act directly on endothelial cells to promote vessel formation 2-benzopyrone) and cimetidine in patients with advanced malignancies. Mol. Biother., 3: 170–178, 1991. (58). Because these two compounds possess the ability to inhibit 14. Mohler, J. L., Gomella, L. G., Crawford, E. D., Glode, L. M., Zippe, C. D., Fair, angiogenesis by different mechanisms, it is possible that when com- W. R., and Marshall, M. E. Phase II evaluation of coumarin (1, 2-benzopyrone) in bined, they can produce a synergistic effect on the inhibition of tumor metastatic prostatic carcinoma. Prostate, 20: 123–131, 1992. 15. Smith, G. F., Neubauer, B. L., Sundboom, J. L., Best, K. L., Goode, Tanzer, L. R., growth by a mechanism other than stabilization of microtubule dy- Merriman, R. L., Frank, J. D., and Herrmann, R. G. Correlation of the in vivo namic instability. It is however important to stress that although anticoagulant, antithrombotic and antimetastatic efficacy of warfarin in the rat. Thromb. Res., 50: 163–174, 1988. coumarin comprises the substructure of dicoumarol and warfarin, it 16. Thornes, R. D. Anticoagulant therapy in patients with cancer. J. Irish Med. Assoc., does not contain any anticoagulant properties. It would be of great 62: 426–429, 1969. interest to further investigate the potential antiangiogenic effect of 17. Thornes, R. D. Warfarin as maintenance therapy for cancer. J. Irish Coll. Phys. Surg., 2: 41–42, 1972. coumarin anticoagulants and Taxol combinations in vitro and in vivo. 18. D’Souza, D., Daly, L., and Thornes, R. D. Levamisole. BCG and warfarin as In summary, our results indicate that dicoumarol may exert its adjuvants to chemotherapy for increased survival in advanced breast cancer. Ir. Med. antiproliferative effects by stabilizing spindle microtubule dynamics J., 11: 605–608, 1978. 19. Chahinian, A. P., Propert, K. J., Ware, J. H., Zimmer, B., Perry, M., Hirsh, V., Skarin, through a unique interaction with tubulin and microtubules. The A., and Kopel, S. A randomized trial of anticoagulation with warfarin and of structures of the known antimitotic chemotherapeutic drugs, such as alternating chemotherapy in extensive small-cell lung cancer by the Cancer and Leukemia Group B. J. Clin. Oncol., 7: 993–1002, 1989. the taxanes and Vinca alkaloids, are complex and not easily manipu- 20. Zacharski, L. R., Henderson, W. G., Rickles, F. R., Forman, W. B., Cornell, C. J. Jr., lated chemically. Dicoumarol and coumarins could provide a new Forcier, R. J., Edwards, R., Headley, E., Kim, S. H., O’Donnell, J. R., O’Dell, R., 1219

Tornyos, K., and Kwaan, H. C. Effect of warfarin on survival in small cell carcinoma 40. Panda, D., Jordan, M. A., Chu, K. C., and Wilson, L. Differential effects of vinblas- of the lung. Veterans Administration Study No. 75. JAMA, 245: 831–835, 1981. tine on polymerization and dynamics at opposite microtubule ends. J. Biol. Chem., 21. Zacharski, L. R., Henderson, W. G., Rickles, F. R., Forman, W. B., Cornell, C. J. Jr. 271: 29807–29812, 1996. Forcier, R. J., Edwards, R. L., Headley, E., Kim, S. H., and O’Donnell, J. F. Effect 41. Schiff, P. B., Fant, J., and Horwitz. S. B. Promotion of microtubule assembly in vitro of warfarin anticoagulation on survival in carcinoma of the lung, colon, head and by taxol. Nature, 277: 665–667, 1979. neck, and prostate. Final report of VA Cooperative Study #75. Cancer, 53: 2046– 42. Parness, J., and Horwitz, S. B. Taxol binds to polymerized tubulin in vitro. J. Cell 2052, 1984. Biol., 91: 479–487, 1981. 22. Schulman, S., and Lindmarker, P. Incidence of cancer after prophylaxis with warfarin 43. Horwitz, S. B. Mechanism of action of taxol. Trends Pharmacol. Sci., 13: 134–136, against recurrent venous thromboembolism. N. Engl. J. Med., 342: 1953–1958, 2000. 1992. 23. Podbielkowska, M., Kupidlowska, E., Waleza, M., Dobrzynska, K., Louis, S. A., 44. Caplow, M., Shanks, J., and Ruhlen, R. How taxol modulates microtubule disassem- Keightley, A., and Zobel, A. M. Coumarins as antimitotics. Int. J. Pharmacog., 32: bly. J. Biol. Chem., 269: 23399–23402, 1994. 262–273, 1994. 45. Rowinsky, E. K., Eisenhauer, E. A., Chaudhry, V., Arbuck, S. G., and Donehower, 24. Podbielkowska, M., Piwocka, M., Waszkowska, E., Waleza, M., and Zobel, A. M. R. C. Clinical toxicities encountered with paclitaxel (Taxol). Semin. Oncol., 20: 1–15, Effect of coumarin and its derivatives on mitosis and ultrasturcutre of meristematic 1993. cells. Int. J. Pharmacog., 33: 7–15, 1995. 46. Hudes, G. R., Greenberg, R., Krigel, R. L., Fox, S., Scher, R., Litwin, S., Watts, P., 25. Wilson, L., and Jordan, M. A. Microtubule dynamics: taking aim at a moving target. Speicher, L., Tew, K., and Comis, R. Phase II study of estramustine and vinblastine Chem. Biol., 2: 569–573, 1995. 26. Saxton, W. M., Stemple, D. L., Leslie, R. J., Salmon, E. D., Zavortink, M., and two microtubule inhibitors in hormone-refractory prostate cancer, 10: 1754–1761, McIntosh, J. R. Tubulin dynamics in cultured mammalian cells. J. Cell Biol., 99: 1992. 2175–2186, 1984. 47. Hudes, G., Nathan, F., Khater, C., Haas, N., Cornfield, M., Giantonio, B., Greenberg, 27. Mitchison, T. J. Poleward microtubule flux in the mitotic spindle; evidence from R., Gomella, L., Litwin, S., Ross, E., Roethke, S., and McAleer, C. Phase II trial of photoactivation of fluorescence. J. Cell Biol., 109: 637–652, 1989. 96-hour paclitaxel plus oral estramustine phosphate in metastatic hormone-refractory 28. Gliksman, N. R., Skibbens, R. V., and Salmon, E. D. How the transition frequencies prostate cancer. J. Clin. Oncol., 15: 3156–3163, 1997. of microtubule dynamic instability (nucleation, catastrophe, and rescue) regulate 48. Seidman, A. D., Scher, H. I., Petrylak, D., Dershaw, D. D., and Curley, T. Estramus- microtubule dynamics in interphase and mitosis: analysis using a Monte Carlo tine and vinblastine: use of prostate specific antigen as a clinical trial end point for computer simulation. Mol. Biol. Cell, 4: 1035–1050, 1993. hormone refractory prostatic cancer. J. Urol., 147: 931–934, 1992. 29. Wordeman, L., and Mitchison, T. J. Dynamics of microtubule assembly in vivo. In: 49. Cocconi, G., Mambrini, A., Quarta, M., Vasini, G., Bella, M. A., Ferrozzi, F., and J. S. Hyams and C. W. Z. Lloyd (eds.), Modern Cell Biology; Microtubules, pp. Beretta, M. D. Vinorelbine combined with paclitaxel infused over 96 hours (VI-TA- 287–301. New York: Wiley-Liss, Inc., 1994. 96) for patients with metastatic breast carcinoma. Cancer (Phila.), 88: 2731–2738, 30. Zhai, Y., Kronebusch, P. J., Simon, P. M., and Borisy, G. G. Microtubule dynamics 2000. at the G-2/M transition: Abrupt breakdown of cytoplasmic microtubules at nuclear 50. Kelly, W. K., Curley, T., Slovin, S., Heller, G., McCaffrey, J., Bajorin, D., Ciolino, envelope breakdown and implications for spindle morphogenesis. J. Cell Biol., 135: A., Regan, K., Schwartz, M., Kantoff, P., George, D., Oh, W., Smith, M., Kaufman, 201–214, 1996. D., Small, E. J., Schwartz, L., Larson, S., Tong, W., and Scher, H. Paclitaxel, 31. Jacobs, R., and Wilson, L. Fertilized sea urchin eggs as a model for detecting cell estramustine phosphate, and carboplatin in patients with advanced prostate cancer. division inhibitors. Drugs Pharm. Sci., 27: 481–493, 1986. J. Clin. Oncol., 19: 44–53, 2001. 32. Toso, R. J., Jordan, M. A., Farrell, K. W., Matsumoto, B., and Wilson, L. Kinetic 51. Panda, D., Miller, H. P., Islam, K., and Wilson, L. Stabilization of microtubule stabilization of microtubule dynamic instability in vitro by vinblastine. Biochemistry, dynamics by estramustine by binding to a novel site in tubulin: a possible mechanistic 32: 1285–1293, 1993. basis for its antitumor action. Proc. Natl. Acad. Sci. USA, 94: 10560–10564, 1997. 33. Bradford, M. M. A rapid and sensitive method for the quantitation of microgram 52. Zobel, A. M., and Schellenberger, S. E. Paclitaxel in combination with coumarin as quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem., a potentially effective anticancer agent. Pharm. Biol., 38: 192–196, 2000. 248–254, 1976. 53. Bae, S., Arand, G., Azzam, H., Pavasant, P., Torri, J., Frandsen, T. L., and Thompson, 34. Panda, D., Daijo, J. E., Jordan, M. A., and Wilson, L. Kinetic stabilization of E. W. Molecular and cellular analysis of basement membrane invasion by human microtubule dynamics at steady state in vitro by substoichiometric concentrations of breast cancer cells in Matrigel-based in vitro assays. Breast Cancer Res. Treat., 24: tubulin-colchicine complex. Biochemistry, 34: 9921–9929, 1995. 241–255, 1993. 35. Walker, R. A., O’Brien, E. T., Pryer. N. K., Soboeiro, M. F., and Voter, W. A., 54. Belotti, D., Nicoletti, I., Vergani, V., Rieppi, M., Drudis, T., Viale, G., Giavazzi, R., Erickson, H. P., and Salmon, E. D. Dynamic instability of individual microtubules analyzed by video light microscopy: rate constants and transition frequencies. J. Cell and Taraboletti, G. Paclitaxel (Taxol), a microtubule affecting drug, inhibits tumor Biol., 107: 1437–1448, 1988. induced angiogenesis. Proc. Am. Assoc. Cancer Res. Annu. Meet., 37: 57, 1996. 36. Sackett, D. Vinca site agents induce structural changes in tubulin different from and 55. Stearns, M. E., and Wang, M. Taxol blocks processes essential for prostate tumor cell antagonistic to changes induced by colchicine site agents. Biochemistry, 34: 7010– (PC-3 ML) invasion and metastases. Cancer Res., 52: 3776–3781, 1992. 7019, 1995. 56. Fernandez-Patrons, C., Zhang, Y., Radomski, M. W., Hollenberg, M. D., and Dav- 37. Chou, T. C., and Talalay, P. Quantitative analysis of dose-effect relationships: the idge, S. T. Rapid release of matrix metalloproteinase (MMP)-2 by thrombin in the rat combined effects of multiple drugs or enzyme inhibitors. Adv. Enzyme Regul., 22: aorta: modulation by protein tyronise kinase/phosphatase. Thromb. Haemost., 82: 27–55, 1984. 1353–1357, 1999. 38. Panda, D., Miller, H. P., and Wilson, L. Determination of the size and chemical nature 57. Mousa, S. A., and Mohamed, S. Anti-angiogenesis and anti-tumor efficacy of war- of the stabilizing “cap” at microtubule ends using modulators of polymerization farin in the chick chorioallantoic membrane (CAM) model. Blood, 96: 182b, 2000. dynamics. Biochemistry, 41: 1609–1617, 2002. 58. Lau, D. H., Xue, L., Young, L. J., Burke, P. A., and Cheung, A. T. Paclitaxel (Taxol): 39. Derry, W. B., Wilson, L., and Jordan, M. A. Substoichiometric binding of taxol an inhibitor of angiogenesis in a highly vascularized transgenic breast cancer. Cancer suppresses microtubule dynamics. Biochemistry, 34: 2203–2211, 1995. Biother. Radiopharm., 14: 31–36, 1999.

1220

Downloaded from cancerres.aacrjournals.org on September 28, 2021. © 2003 American Association for Cancer Research. Dicoumarol: A Unique Microtubule Stabilizing Natural Product that Is Synergistic with Taxol

Hamta Madari, Dulal Panda, Leslie Wilson, et al.

Cancer Res 2003;63:1214-1220.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/63/6/1214

Cited articles This article cites 51 articles, 14 of which you can access for free at: http://cancerres.aacrjournals.org/content/63/6/1214.full#ref-list-1

Citing articles This article has been cited by 10 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/63/6/1214.full#related-urls

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

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

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/63/6/1214. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.