Targeting of eEF1A with Amaryllidaceae isocarbostyrils as a strategy to combat melanomas

The FASEB Journal • Research Communication

Targeting of eEF1A with Amaryllidaceae isocarbostyrilsas a strategy to combat melanomas

Gwendoline Van Goietsenoven,* Jenna Hutton,� Jean-Paul Becker,§

Benjamin Lallemand,† Francis Robert,¶ Florence Lefranc,*,# Christine Pirker,**Guy Vandenbussche,§ Pierre Van Antwerpen,‡ Antonio Evidente,†† Walter Berger,**Martine Prevost,§ Jerry Pelletier,¶ Robert Kiss,*,1 Terri Goss Kinzy,�

Alexander Kornienko,‡‡ and Veronique Mathieu**Laborarory of Toxicology, †Laboratory of Bioanalytical Chemistry, Toxicology, and AppliedChemistry, and ‡Laboratory of Pharmaceutical Chemistry, Institute of Pharmacy, and §Laboratory forthe Structure and Function of Biological Membranes, Center for Structural Biology andBioinformatics, Universite Libre de Bruxelles, Brussels, Belgium; �Department of Molecular Genetics,Microbiology, and Immunology, University of Medicine and Dentistry of New Jersey Robert WoodJohnson Medical School, Piscataway, New Jersey, USA; ¶Biochemistry Department, McGill University,Montreal, Quebec, Canada; #Department of Neurosurgery, Erasme University Hospital, Brussels,Belgium; **Department of Medicine I, Institute of Cancer Research, Medical University Vienna,Vienna, Austria; ††Dipartimento di Scienze del Suolo, della Pianta, dell’Ambiente e delle ProduzioniAnimali, Universita di Napoli Federico II, Portici, Italy; and ‡‡Department of Chemistry, New MexicoInstitute of Mining and Technology, Socorro, New Mexico, USA

ABSTRACT Melanomas display poor response ratesto adjuvant therapies because of their intrinsic resis-tance to proapoptotic stimuli. This study indicates thatsuch resistance can be overcome, at least partly,through the targeting of eEF1A elongation factor withnarciclasine, an Amaryllidaceae isocarbostyril control-ling plant growth. Narciclasine displays IC50 growthinhibitory values between 30–100 nM in melanoma celllines, irrespective of their levels of resistance to pro-apoptotic stimuli. Normal noncancerous cell lines aremuch less affected. At nontoxic doses, narciclasine alsosignificantly improves (P�0.004) the survival of micebearing metastatic apoptosis-resistant melanoma xeno-grafts in their brain. The eEF1A targeting with narcic-lasine (50 nM) leads to 1) marked actin cytoskeletondisorganization, resulting in cytokinesis impairment,and 2) protein synthesis impairment (elongation andinitiation steps), whereas apoptosis is induced at higherdoses only (>200 nM). In addition to molecular dock-ing validation and identification of potential bindingsites, we biochemically confirmed that narciclasine di-rectly binds to human recombinant and yeast-purifiedeEF1A in a nanomolar range, but not to actin orelongation factor 2, and that 5 nM narciclasine issufficient to impair eEF1A-related actin bundling activ-ity. eEF1A is thus a potential target to combat melano-mas regardless of their apoptosis-sensitivity, and thisfinding reconciles the pleiotropic cytostatic of narcic-lasine.—Van Goietsenoven, G., Hutton, J., Becker,J.-P., Lallemand, B., Robert, F., Lefranc, F., Pirker, C.,Vandenbussche, G., Van Antwerpen, P., Evidente, A.,Berger, W., Prevost, M., Pelletier, J., Kiss, R., GossKinzy, T., Kornienko, A., Mathieu, V. Targeting ofeEF1A with Amaryllidaceae isocarbostyrils as a strategy

to combat melanomas. FASEB J. 24, 4575–4584 (2010).www.fasebj.org

Key Words: narciclasine � growth inhibition � apoptosis resis-tance � actin

The incidence of melanoma is steadily increasingamong the Caucasian population and affects 1 in 55cancer patients in the United States (1). One-third ofearly stage melanoma patients will develop metastases,and metastatic melanomas are associated with dismalprognosis. Such patients have a median overall survivalof 6 to 8 months despite significant efforts to developadjuvant therapies (1). The response rate to the stan-dard FDA-approved treatment dacarbazine, adminis-tered as a single agent therapy, is as low as 16%. Theresponse rates to other cytotoxic drugs, such as cispla-tin, nitrosurea, vinca alkaloids, or temozolomide, givenalone or in combination, are even lower (1, 2). Signif-icantly, combination treatments with IL-2 or IFN�2bimprove the response rate without any impact on theoverall survival of metastatic melanoma patients (1).

These highly disappointing statistics are likely relatedto the resistance of melanoma cells to apoptosis. In-deed, while in many other cancer types chemoresis-tance is acquired from radiation and chemotherapy,melanomas belong to cancer types that are intrinsicallyresistant to apoptosis (3). This feature has been attrib-

1 Correspondence: Laboratory of Toxicology, Institute ofPharmacy, Universite Libre de Bruxelles, Campus de laPlaine, Blvd. du Triomphe, 1050 Brussels, Belgium. E-mail:[email protected]

doi: 10.1096/fj.10-162263

45750892-6638/10/0024-4575 © FASEB

uted to numerous molecular changes in proliferation,survival, and death signaling pathways (3–6), includingdeath receptor signaling impairment (5), activation ofproliferative and survival pathways Raf/MAPK, PI3K/Akt and NF-�B pathways (3, 4, 6), and modifications tothe mitochondrial apoptosis regulating proteins includ-ing IAPs and Apaf-1 (3, 4, 6). Actin cytoskeleton andassociated proteins have been shown more recently toinduce and participate in mitochondrial apoptosis viapathways that are yet to be fully deciphered (7–9). Thetargeting of actin cytoskeleton is an approach that hasnot received its due attention (10), despite its signifi-cant potential to affect both cell proliferation (cytoki-nesis) and migration. eEF1A is an abundant evolution-arily conserved protein that binds to and deliversaa-tRNA to the empty A site of elongating ribosomes.However, it has been shown more recently that �60%of eEF1A can be bound to fibrillar actin, promotingbundling and inhibiting both polymerization and de-polymerization. Therefore, eEF1A plays critical roles inactin cytoskeleton organization and functions involvedin cell migration, cell morphology, protein synthesis,and cell death (11–13). Targeting eEF1A could thus bea promising strategy to combat apoptosis-resistant can-cers and melanomas in particular.

For �2000 yr, plants belonging to the Amaryllidaceaefamily have been used in traditional medicine forvarious anticancer applications. Over 100 alkaloids andisocarbostyrils, exhibiting diverse biological activities,have been recently isolated from various Amaryllidaceaespecies (14). Lycorine was the first alkaloid isolatedfrom these plants (15) and found to possess anticanceractivities both in vitro and in vivo (16–18). More re-cently, several isocarbostyril constituents of the Amaryl-lidaceae, such as narciclasine, pancratistatin, trans-dihy-dronarciclasine, and their congeners (for structureexamples, see Fig. 1), have been isolated or semisyn-thesized (19, 20). Narciclasine, known for �40 yr, isisolated from narcissus bulbs and serves as an inhibitorof plant growth (21), which was first characterized forits antimitotic effects (22). Pancratistatin, whose chem-ical structure is similar to that of narciclasine, inducesrapid apoptosis in neuroblastoma cells but not innormal cells. The apoptosis induction has been shownto occur via mitochondrial targeting (23), which couldfurther synergize with other anticancer agents, such astamoxifen (24). Comparable apoptosis induction hasbeen also described for narciclasine in breast cancercells (25). Although it has been shown that the Amaryl-lidaceae isocarbostyrils (AIs) inhibit peptide bond for-mation step and thus protein synthesis (26, 27), recentdata obtained in glioblastoma and prostate cancer cellsindicate that narciclasine could also target actin cy-toskeleton and related proteins (28, 29). The presentstudy demonstrates that eEF1A is a target for the AIs attheir growth inhibitory IC50 values, reconciling theireffects on protein synthesis and actin cytoskeleton.Thus, using AIs as a chemical tool, we demonstratedthat eEF1A targeting represents a useful strategy tobypass apoptosis resistance of melanoma cells. Conceiv-

ably, eEF1A targeting also occurs in plants, since theobserved effects in plants are similar, and eEF1A is ahighly conserved protein in the biological kingdom.

MATERIALS AND METHODS

Molecular docking

Docking was performed using the Glide 4.5 software (Schro-dinger LLC, Mannheim, Germany) (30), which positions the3-D structure of flexible ligands in the 3-D structure of rigidor partially flexible receptors. First, docking runs were per-formed to explore the whole surface of the rigid proteinexhaustively. From this exploration step, the locations of theAIs with the best affinity scores were examined to select theirpotential binding regions on the protein surface. In thesecond step, induced fit docking (30) experiments wereperformed in these putative binding regions to account forprotein conformational changes induced by the presence ofthe ligands. For each docking run, the binding zones weredefined as cubes with a length of 20 Å. The initial 3-Dstructures of the ligands were generated with the CORINAprogram (31).

Narciclasine binding to eEF1A

Human eEF1A isoform 2 recombinant protein was purchasedfrom Novus Biologicals (Littleton, CO, USA), while endoge-nous yeast eEF1A was prepared as described previously (32).Various concentrations of narciclasine were incubated inphosphate buffer with 100 nM of eEF1A protein overnight at37°C. The samples were centrifuged through 5-kDa filters(Sartorius, Vilvoorde, Belgium) at 14,000 rpm for 15 min toretain the proteins with bound narciclasine. The concentra-tions of unbound narciclasine in the ultrafiltrates were mea-sured by fluorescence (excitation: 360 nm; emission: 480 nm)at pH 12 and compared to a standard curve of narciclasine-filtered solutions.

The cell assay was performed as follows: 2 melanoma celllines, VM-21 and VM-48 (see below), were treated for 16 hwith narciclasine at 50 and 100 nM. Cells were collected inlysis buffer (20 mM Tris-HCl, pH 7.5; 150 mM NaCl; 1%Triton X-100; and 1 mM EDTA). The totality of each samplewas loaded in a 96-well plate coated with 3 �g/ml of ananti-eEF1A antibody (Millipore, Brussels, Belgium) and incu-bated for 2 h with agitation at room temperature. After 4washes with lysis buffer, eEF1A-bound narciclasine was re-leased by methanol and measured using the fluorescenttechnique described above. Each assay was performed intriplicate.

Cell cultures and compounds

The human SKMEL-28 (ATCC HTB-72) and the mouseB16F10 (ATCC CRL-6475) melanoma cell lines were ob-tained from the American Type Culture Collection (ATCC;Manassas, VA, USA) and maintained in our laboratory asdetailed previously (33). Primary melanoma cell cultures(VM-21, VM-47, and VM-48) were established at the Instituteof Cancer Research, Medical University Vienna, as describedpreviously (33, 34). Their melanocyte origin was confirmedby the presence of melanosomes in electron microscopy, byimmunohistochemical S100 and HMB-45 stainings, and by

4576 Vol. 24 November 2010 VAN GOIETSENOVEN ET AL.The FASEB Journal � www.fasebj.org

melanin and tyrosinase determination. Temozolomide (TMZ)was purchased from Schering Plough (Brussels, Belgium).Narciclasine was extracted from Narcissus bulbs as de-scribed previously (35).

In vitro overall growth determination

Overall cell growth was assessed using the 3-[4,5-dimethylthia-zol-2yl]-diphenyltetrazolium bromide (MTT) colorimetric as-say (Sigma, Bornem, Belgium), as detailed elsewhere (29,33). All determinations were performed in sextuplicate.

Determination of apoptosis by flow cytometry

Detection of apoptosis was performed by flow cytometricanalysis of double-staining with propidium iodide and an-nexin-V FITC as detailed elsewhere (25, 33) using the APOAF apoptosis detection kit (Sigma).

Alteration of the outside mitochondrial membrane poten-tial was monitored with the fluorescent dye JC-1 (Calbiochem,Nottingham, UK) as described by Dumont et al. (25). The dyeforms red fluorescent aggregates inside the mitochondria,while its monomeric cytoplasmic form fluoresces green (25).Mitochondrial outside membrane permeabilization (MOMP)is therefore characterized by a shift from red to greenfluorescence (25). Fluorescence was analyzed immediatelyafter the staining procedures on an Epics XL MCL flowcytometer (Beckman Coulter, Fullerton, CA, USA). All exper-iments were conducted in triplicate.

Computer-assisted phase-contrast microscopy

Quantitative videomicroscopy was performed as detailed pre-viously (28). Each experiment was conducted in triplicate.

Actin fluorescent staining

Actin cytoskeleton organization was visualized in formol-fixedcells using fluorescent probes with high affinities for eitherfibrillar actin (Alexa Fluor 488-conjugated phallacidin;Molecular Probes; Invitrogen, Merelbeke, Belgium) orglobular actin (Alexa Fluor 594-conjugated DNase I; Mo-lecular Probes), as detailed elsewhere (28).

Actin bundling

Evaluation of eEF1A bundling of F-actin was assayed bylight scattering. The effect of narciclasine on the ability ofpurified rabbit eEF1A to bundle F-actin was analyzed usingright angle light scattering as described by Liu et al. (36).Briefly, G-actin purified from S. cerevisiae was allowed topolymerize for 2 h at 25°C. A FluoroMax-3 fluorescencespectrophotometer (Horiba Jobin Yvon, Longjumeau,France) was used with 600-nm emission/excitation wave-lengths combined with a slit width of 5 nm. Prior to eEF1Aaddition baseline, readings of buffer and F-actin alonewere obtained. Data were collected and analyzed usingFluorEssence (Horiba Jobin Yvon).

Protein synthesis assays

Polysome organization analysis was conducted as described byPelletier et al. (37). Briefly, cells left untreated (negativecontrol), treated with puromycin (100 �g/ml; positive con-trol) or narciclasine (50 or 100 nM) were collected in PBSand centrifuged. The cell pellets were resuspended in hypo-tonic lysis buffer. After centrifugation, the supernatants wereloaded onto 10–50% sucrose gradients and centrifuged in aSW 60 Ti rotor (Beckman) at 39,000 rpm for 2 h. Gradientswere divided into 33 fractions for absorbance measurement at254 nm.

Protein elongation assay was conducted using the Click-iTAHA kit following the manufacturer’s instructions (Invitro-gen). Cells were treated with puromycin (100 �g/ml; positivecontrol) or narciclasine (50 nM or 100 nM) for 5 and 24 h.After 1 h incubation in a methionine-free medium (Invitro-gen), the cells were incubated 5 h with methionine-freemedium supplemented with 100 �M l-azidohomoalanine.Cells were harvested after 30 min incubation in a lysis bufferand centrifuged 5 min at 14,000 g at 4°C. Azide/alkynereaction was performed on 200 �g of azide-labeled protein byaddition of CuSO4 in the reaction buffer to obtain biotinyl-ated triazole conjugates. Proteins were precipitated and res-olubilized in the loading buffer for further electrophoreticanalysis.

In vivo orthotopic xenografts

Human melanoma brain metastatic cells (106; VM-48) werestereotactically implanted into the brains of nude mice (6-wk-old female nu/nu mice; 21–23 g; Iffa Credo; Charles River,Arbresle, France) as described previously (33). Each experi-mental group contained 11 mice. This experiment was per-formed on the basis of Authorization LA1230509 of theAnimal Ethics Committee of the Federal Department ofHealth, Nutritional Safety and the Environment (Belgium).

RESULTS

Docking of the AIs to the yeast eEF1A

The known effects of altered eEF1A activities closelyresemble those induced by AIs. Therefore, we hypoth-esized that the latter could target this elongation factor.Docking experiments revealed that AIs are indeedpotential eEF1A ligands. Three binding pockets werefound in independent docking experiments conductedwith the crystallographic structure of the yeast eEF1A(PDB code 1G7C) (Fig. 1A). One of the pocketscorresponds to the GTP binding site (Fig. 1A, pocket a)and another to the binding region of the nucleotideexchange factor eEF1B� (Fig. 1A, pocket c). The mo-lecular structures of the docked AIs and alkaloid lyco-rine are presented in Fig. 1B, together with theirbinding free energy score in each pocket. Narciclasinefeatures affinity scores ranging from �12.3 to �10.8kcal/mol. For comparison, the GDPNP, an analog ofthe natural ligand GTP, has a free energy score of �9.8kcal/mol (Fig. 1). Lycorine, characterized by higherIC50 growth inhibitory values than narciclasine andpancratistatin, (i.e., �5000 vs. �50–100 nM; refs. 11,

4577NARCICLASINE TARGETS eEF1A

21; see Fig. 4A), also displays weaker affinity to eEF1A inthis theoretical model (Fig. 1).

eEF1A is a target for AIs

To validate these modeling results, we incubated 100nM human eEF1A recombinant protein (commercialisoform 2 coupled with GST) and authentic yeast

eEF1A with narciclasine at 37°C at various concentra-tions in vitro. The percentage of bound narciclasine wascalculated based on the amount of unbound narcic-lasine measured by fluorescence. Figure 2A indicatesthat human eEF1A protein binds nearly all narciclasinemolecules until equal molarity between the ligand andits target is reached. At higher narciclasine concentra-tions, the binding appears to be saturated. The assaywas also conducted with other potential binding part-

Figure 1. Potential binding sites for the Amaryllidaceae isocarbostyrils on eEF1A protein. A) Three potential binding pockets ofnarciclasine were found in the 3-D structure of yeast eEF1A in independent docking experiments. They are shown here on thesame structure. Pocket a corresponds to the GTP binding site. Pocket c is located in the binding region of the nucleotideexchange factor. The 3 eEF1A domains are colored in gray, red, and green. Positions of residues lining each bindingpocket are indicated as red spheres, and their numbers are indicated. Inset: same view of eEF1A structure in associationwith the nucleotide exchange factor (blue). B) Molecular structures of various AIs and alkaloid lycorine, together withtheir theoretical affinity scores, calculated for all compounds in each binding site.

Figure 2. eEF1As is a target for the AIs. A) Narciclasine binding to eEF1A, actin andEF-2 in vitro: various concentrations of narciclasine were incubated overnight withrecombinant yeast eEF1A or human eEF1A isoform 2, human actin, and yeasteEF-2. Bound fraction of narciclasine (expressed as percentage) is calculated on thebasis of free narciclasine measured by fluorescence. B) Narciclasine binding toeEF1A in mammalian melanoma cells. C) Summary of the various roles of eEF1A inprotein synthesis and actin cytoskeleton organization.

4578 Vol. 24 November 2010 VAN GOIETSENOVEN ET AL.The FASEB Journal � www.fasebj.org

ners; e.g., human actin (37°C) and yeast elongationfactor 2 (EF2; 30°C), but no significant binding couldbe detected (Fig. 2A). Similarly, narciclasine did notbind to GST alone (internal control, data not shown).The binding affinity for the eEF1A yeast isoform islower than that for its human counterpart; however,this protein was not coupled with GST, and thus itsconformation could be unstable. Moreover, these pro-teins are only 80% identical, which may account fordifferences in binding as well. Two melanoma cell lineswere treated with narciclasine at 50 and 100 nM for16 h. Measurement of eEF1A-bound narciclasine afterimmunoimmobilization of the protein confirmed thatnarciclasine penetrates melanoma cells and binds toeEF1A intracellularly (Fig. 2B).

No evidence indicated that narciclasine competeswith the GDP binding in pocket a, at least up toconcentrations of 250 nM of narciclasine (GDP displaysa calculated Kd value of 300 nM; data not shown). This

finding suggests that the GDP binding site is not wherenarciclasine binds eEF1A. Further ongoing tests involv-ing mutational studies and/or crystallography are re-quired to determine more precisely the binding site ofnarciclasine to eEF1A.

AIs affect actin cytoskeleton organization and impairprotein synthesis at both initiation and elongationsteps

The main eEF1A activities are summarized in Fig. 2C.Ample data have shown that eEF1A modulates actincytoskeleton by stimulating actin bundling as well asaltering polymerization (12, 13). Narciclasine markedlyimpairs eEF1A-mediated actin bundling at doses as lowas 5 nM (Fig. 3A). Actin bundling impairment has beenconfirmed further via actin cosedimentation assays(data not shown). Narciclasine has been shown to

E

Figure 3. Narciclasine-induced effects on actin cytoskeleton and protein synthesis.A) Right-angle light scattering was measured for 1.5 �M of preassembled F-actin.Then 0.5 �M rabbit eEF1A was added in the presence of 0–50 nM narciclasine, andscattering was measured for another 400 s. a, b) Representative graphs at 0 nM (a)and 20 nM narciclasine (b). Asterisk indicates emission shutter closed, eEF1A �narciclasine added. c) Increased fluorescence intensity on addition of eEF1A toF-actin is reduced in the presence of narciclasine. B) Narciclasine-induced effectson the actin cytoskeleton of VM-21 and VM-48 human melanoma cells treated with50 nM (IC50 value) narciclasine for 15 min and 3 h, respectively, highlighted byfluorescence. Fibrillar actin is represented in green and globular actin in red.

C, E) Spectrometry analysis of polyribosome status in VM-21 (C) and VM-48 (E) human melanoma cells treated with 0 nM (solidcircles), 50 nM (open circles), and 100 nM narciclasine (open squares) for 1 h (C) and 5 h (E), of cell extracts centrifuged insucrose gradients and separated into 33 fractions. D, F) Protein synthesis in VM-21 (D) and VM-48 (F) human melanoma cellsleft untreated or treated with 50 or 100 nM of narciclasine for 5 and 24 h. After treatment, cells were incubated with amethionine analog that was incorporated in nascent proteins and further biotinylated. Lane 1: cells incubated with themethionine analog alone (negative control). Lanes 2–7: cells incubated for 5 h with the methionine analog. Lane 2: cells treatedwith 100 �g/ml puromycin for 2 h (positive control). Lane 3: untreated cells. Lanes 4 and 5: cells treated for 5 h with 50 nMand 100 nM narciclasine, respectively. Lanes 6 and 7: cells treated for 24 h with 50 nM and 100 nM narciclasine, respectively.Bottom panels correspond to the Coomassie blue staining of the membranes assessing equal total (labeled and nonlabeled)protein loading and integrity.

4579NARCICLASINE TARGETS eEF1A

interfere with actin cytoskeleton in glioblastoma andpancreatic cancer cells (28, 29). We therefore analyzedthe effects of narciclasine on actin cytoskeleton inapoptosis-sensitive VM-21 (Fig. 4B) and apoptosis-resis-tant VM-48 (Fig. 4B) human melanoma cells by fluo-rescent staining. After incubation with 50 nM narcic-lasine for 15 min, VM-21 cells displayed increasedfibrillar actin, whereas these effects occurred only after3 h incubation in VM-48 cells (Fig. 3B).

eEF1A is essential for maintaining the translationalstatus of the cell, while specific mutants of eEF1A separatethis canonical activity from the role of the protein in actinorganization (38). Furthermore, the intact cytoskeletonhas been shown to be essential for appropriate polysomeorganization in an eEF1A-dependent manner (38). Fig-ure 3C, E clearly shows that narciclasine induces dose-dependent modifications of polysome organization:When cells were treated with narciclasine at 50 or 100 nM,polysomes were reduced in favor of 80S ribosome com-plexes (VM-48 cells, 5h of treatment; Fig. 3E) and 40Sribosomes (VM-21 cells, 1 h of treatment; Fig. 3C). Narci-clasine-induced actin cytoskeleton disorganization, dem-onstrated in the present study, occurred earlier and couldtherefore lead to impairment of cytokinesis as well aspolysome disorganization, which could in turn impairprotein synthesis at the initiation step, as previously re-ported (38).

Effects of narciclasine on the elongation step of

protein synthesis were evaluated by the incorporationof l-azidohomoalanine, a methionine analog, in nas-cent proteins of VM-21 and VM-48 melanoma models.Figure 3D, F shows that narciclasine at 50 and 100 nMinduces a dose-dependent decrease in newly synthe-sized protein in VM-21 cells and VM-48 cells, occurringseveral hours after the disorganization of polysomalstructures. However, the effects were more marked (50vs. 100 nM) and occurred earlier in VM-21 cells than inVM-48 cells (5 vs. 24 h, respectively). The mechanismsunderlying these effects can be associated with eEF1A-dependent actin organization and altered initiation(38), as well as a direct inhibition of peptide bondformation, as described by Carrasco et al. (26). Theseresults are consistent with our hypothesis of eEF1A as atarget for AIs because eEF1A is an elongation factorthat delivers aa-tRNA to the ribosomes.

Antitumor activities of AIs are mainly related to theircytostatic rather than proapoptotic effects

Three AIs, narciclasine and cis- and trans-dihydronarci-clasine, as well as alkaloid lycorine, were evaluated fortheir in vitro growth inhibitory activities by means of thecolorimetric MTT test. Figure 4A illustrates the IC50values obtained on 5 melanoma cell lines, whose sensi-tivities to the compounds were found to be similar.

Figure 4. AI-induced effects on melanoma cell death and proliferation. A) Mean IC50 values obtained with 5 cell lines for 4 AIsby means of the colorimetric MTT test. B) Percentage of cells in a late apoptotic status (double-positive cells for propidium andannexin V stainings analyzed by flow cytometry) when treated with various concentrations of narciclasine. C, E) Flow cytometricanalysis of JC-1 staining. Open columns represent green fluorescence of the dye; solid columns represent red fluorescence.D) Western blot analysis of PARP cleavage: PARP full protein (116 kDa) is cleaved during apoptosis in fragments of 85 kDa, whilenecrotic fragments are of 55 kDa. Tubulin blots assess equal loading and protein integrity. F) Mitosis number per cell over a 72-hperiod was counted in 5 melanoma cell lines by the videomicroscopic device. Solid bars, control; open bars, 50 nM narciclasine.Results are expressed as means � se.

4580 Vol. 24 November 2010 VAN GOIETSENOVEN ET AL.The FASEB Journal � www.fasebj.org

Narciclasine exerted the most potent anticancer activi-ties, with IC50 � 40 nM, which is comparable to the NCIdata (NCI database).

Induction of apoptosis was studied by means ofdouble-annexin V and propidium iodide staining. Fig-ure 4B shows that narciclasine induces typical lateapoptotic changes after 72 h of treatment in a dose-dependent manner in 4 of the 5 melanoma modelsinvestigated. These results were confirmed further by aWestern blot analysis of PARP (poly-ADP-ribose poly-merase) cleavage in VM-21 cells (displaying apoptoticfeatures at 200 and 500 nM) and VM-48 cells (resistantto apoptotic induction in Fig. 4B at doses �500 nM; Fig.4D). Intact PARP expression (116 kDa) decreased whilethe apoptotic fragments (83–89 kDa) increased in adose-dependent manner from 200 to 500 nM in VM-21cells; nevertheless this effect was observed only athigher doses (from 500 to 1000 nM) in VM-48 cells(Fig. 4D). These events have been related further to thechanges in mitochondrial outside membrane perme-abilization, analyzed by flow cytometry with JC-1 dye:Fig. 4C, E shows a dose-dependent increase in greenfluorescence paralleled by a dose-dependent decreasein red fluorescence in VM-21; these features were muchless pronounced for VM-48 treated cells. However,VM-21 and VM-48 display similar IC50 values (�40 nM),not high enough to induce apoptosis even in theapoptosis-sensitive VM-21 model (Fig. 4B). We theninvestigated by means of quantitative videomicroscopywhether narciclasine induces cytotoxic vs. cytostaticeffects when exerting growth inhibition of humanmelanoma cells. This experimental approach enabledus to quantify the number of mitoses during the 72 h ofthe experiment: It was found that narciclasine clearlyimpairs proliferation at 50 nM (Fig. 4F), without induc-ing death of melanoma cells during this 72 h-period ofobservation (data not shown).

In vivo therapeutic benefits of eEF1A targeting bynarciclasine in an apoptosis-resistant brain metastaticmelanoma orthotopic model in immunocompromizedmice

Pharmacokinetic studies performed previously in micerevealed that high doses of narciclasine that can inducecancer cell death (200–500 nM) could not be reachedin plasma due to toxicity effects (29). However, narci-clasine is bioavailable either via intravenous or oralroutes, and safe repeated administrations of 1 mg/kg ofnarciclasine lead to plasmatic concentrations close tothe in vitro IC50 value (�40 nM; ref. 29). These con-centrations were sufficient to provide significant thera-peutic benefits in the case of primary and secondarybrain tumors (28, 29).

We grafted human brain metastatic and apoptosis-resistant melanoma cells (VM-48) directly into thebrains of immunocompromized mice to mimic mela-noma brain metastasis (Fig. 5A). The mice were treatedwith narciclasine or the drug temozolomide, chosen as

a reference. Kaplan-Meier survival analyses revealedthat narciclasine provided significant therapeutic ben-efits in this aggressive melanoma model and that itseffects were similar to temozolomide (Fig. 5B).

DISCUSSION

The connection between actin cytoskeleton and pro-tein synthesis has become more and more evident inrecent years: While microtubules link to mRNA in largecell types, such as oocytes, neurons, or oligodendro-cytes, for transport to specific locations, actin cytoskel-eton plays this role in smaller, and most somatic, celltypes (39, 40). Association of mRNA and/or polysomeswith actin cytoskeleton has been evidenced by fraction-ation experiments; elongation factors, such as EF-2,

Figure 5. Narciclasine improves the survival of mice bearingbrain melanoma metastatic xenografts. A) Typical hematox-ylin eosin staining of a tumor that developed in the brain ofa nude mouse 1 mo after the stereotactical graft of humanmelanoma brain metastatic cells (50). T, tumor; NB,normal brain tissue; black arrows indicate invasive islets).B) Kaplan-Meier survival analysis of the tumor-bearingmice left untreated, treated with narciclasine (1 mg/kgp.o.; 2 administrations/wk for 3 wk) or with temozolomide(TMZ, 40 mg/kg p.o.; 3 administrations/wk for 3 wk).

4581NARCICLASINE TARGETS eEF1A

colocalize with polysomes and actin fibers (39). Muta-genic studies of elongation factors, in particular eEF1A,which binds actin or aa-tRNA, have been recentlyconducted in yeast (38). The results clearly indicatethat improper actin cytoskeleton organization leads totranslation impairment, while elongation inhibition isnot sufficient to affect cytoskeleton organization (38).Actin disruption, through modifications of actin-bind-ing proteins, actin itself, or elongation factor interac-tions, does not affect protein translation at the elonga-tion step, but rather does so at the initiation step, as itinduces changes in the ratio of 80S free ribosomes andpolysome translation complexes (38). The presentstudy validates this hypothesis in mammalian cells asevidenced by the effects of narciclasine on actin cy-toskeleton (Fig. 3), polysome organization, and proteinsynthesis (Fig. 3). These effects could occur, at leastpartly, due to the interaction of narciclasine (and likelyother AIs) with eEF1A (Fig. 2C). Targeting this proteinseems, therefore, to be a promising strategy to bypassthe intrinsic resistance of melanoma cells to apoptosis(3). Indeed, the AIs used in the current study displayedpotent cytostatic effects toward melanoma cells whenused at their respective IC50 growth inhibitory values,with similar efficacy between apoptosis-resistant andsensitive melanoma cell lines. Narciclasine providedsignificant in vivo therapeutic benefits in an orthotopicbrain model related to apoptotis-resistant melanomabrain metastasis (Fig. 5). Narciclasine induced apopto-tic features in melanoma cells, but at nonpharmaco-logical doses; i.e., doses that ranged between 5 and10 the IC50 growth inhibitory values, leading tomitochondria-dependent apoptosis (Fig. 4). At theintracellular level, these events could be related partlyto their prominent effects on actin cytoskeleton. In-deed, the actin-severing protein cofilin, shown previ-ously to be inactivated by narciclasine resulting inincreased fibrillar actin (27), has been demonstrated totranslocate to mitochondria under apoptotic stimuliand to play crucial roles in the induction of apoptosis(8). CAP-1, a cofilin binding protein, also translocatesto mitochondria and appears to assist cofilin in itsproapoptotic functions (41). For both proteins, theiractin-binding domains are required to exert their pro-apoptotic functions: actin could be the effector, sinceactin fragments are involved in spontaneous mitochon-drial apoptosis induction (9). While eEF1A is involvedprofoundly in actin cytoskeleton organization (13) andeEF1A expression level has been linked to modulationof apoptosis-sensitivity (42), the proapoptotic effects ofnarciclasine cannot be directly and exclusively relatedto eEF1A targeting because the saturation of this pro-tein with this compound occurs at doses of 100 nM(Fig. 2A), which are in turn not high enough to induceapoptosis (Fig. 4). Therefore, when used at cytotoxicdoses, narciclasine (and perhaps AIs in general), prob-ably interacts with other lower affinity targets, which arenot necessarily the same for all these natural products.Indeed, correlations of the differential cellular sensitiv-ities to these compounds in the NCI 60-cell line screen

indicate that the AIs exert a uniform mode of action atcytostatic concentrations; i.e., at their mean GI50 valuescharacteristic of growth inhibition (e.g., the correlationcoefficient of �0.7 for 15 nM trans-dihydronarciclasineand 41 nM pancratistatin, Fig. 6). However, the unifor-mity of the mechanism among these compounds disap-pears if the correlation is performed at their mean LC50values, characteristic of cell death (e.g., the correlationcoefficient of 0.12 for 5.9 �M trans-dihydronarciclasineand 23.3�M pancratistatin). In addition, differentialcellular sensitivity profiles associated with these com-pounds at their GI50 do not match with any othercompounds in the NCI database, emphasizing theunique and new mode of action for AIs in general, andnarciclasine in particular. The present study demon-strates that the new mode of action is related to thetargeting of eEF1A, and this finding might constitutean important new strategy to combat apoptosis resistantmelanomas. In addition, the selective growth inhibitoryactivities of AIs toward cancer cells as compared to theirnormal counterparts (23, 25, 29) can be explained bythe overexpression of eEF1A in cancer cells, which hasbeen demonstrated in the case of ovarian (43), breast(44), lung (45), and liver cancers (46) with respect toits isoform eEF1A-2. This marks eEF1A as a putativeoncogene. In addition, selectivity can stem from the pHdifference between cancer and normal cells (47). In-

Figure 6. Correlations of the differential cellular sensitivitiesin the NCI 60 cell line screen using the COMPARE algorithm:compare correlation coefficients (CCCs) were generated by acomputerized pattern-recognition algorithm and serve as anindication of similarities in differential cellular sensitivities orcharacteristic fingerprints for each compound. Pancratistatin(circles), trans-dihydronarciclasine (triangles), 7-deoxynarcic-lasine (squares), cis-dihydronarciclasine (large diamonds),and paclitaxel (small diamonds) were each used as a seed tofind significant correlations with the anticancer agents in theNCI Standard Compound Database, containing pancratista-tin as a representative of the AIs, at the GI50, TGI, and LC50levels (for definitions of these parameters, see DTP humantumor cell line screen; http://dtp.nci.nih.gov/branches/btb/ivclsp.html). At the GI50 level, all correlations, with theexception of paclitaxel identified pancratistatin, ranked firstamong all the compounds in the database with CCC � 7(open markers). Correlations with pancratistatin at the TGI(shaded markers) and LC50 levels (solid markers) weresignificantly inferior and virtually nonexistent, respectively(CCC 0.4–0.5 and 0.1–0.2). Correlations of paclitaxel withpancratistatin were poor and similar to the rest of the seedcompounds at the LC50 level.

4582 Vol. 24 November 2010 VAN GOIETSENOVEN ET AL.The FASEB Journal � www.fasebj.org

deed, binding of eEF1A to aa-tRNA or to actin could betuned by the local pH level (36). The affinity of the AIsfor eEF1A could thus also be dependent on the pH.

eEF1A is a large protein composed of 3 domains, and itdisplays considerable flexibility (48). While the nucleotide(either GTP or GDP) binding pocket is well characterizedin both bacterial EF-Tu and eukaryotic eEF1A, the struc-ture of the aa-tRNA-bound form is known only in pro-karyotes. The binding sites for fibrillar actin seem to beeven more complex (32, 36). These two binding sites, aswell as for eEF1B�, overlap so that the binding of theseligands could be mutually exclusive (32, 36). The compe-tition assay with GDP, one of its natural ligands, leads us topostulate that the nucleotide binding site could not be theone for the AIs. Mutational studies of eEF1A and investi-gation of effects of pH are currently ongoing to deter-mine and characterize the binding site of the AIs oneEF1A, which we plan to validate by crystallography.

The authors thank Jean-Francois Gaussin and SebastienSauvage for the in vivo experiment. The present work hasbeen supported partly by grants awarded by the Fonds YvonneBoel (Brussels, Belgium) and by the Fonds National de laRecherche Scientifique (FNRS; Brussels, Belgium). F.L. is aclinical research fellow, V.M. is a senior research assistant,and R.K. is a director of research with the FNRS.

REFERENCES

1. Hamm, C., Verma, S., Petrella, T., Bak, K., and Charette, M.(2008) Melanoma Disease Site Group of Cancer Care Ontario’sProgram in evidence-based care. Biochemotherapy for thetreatment of metastatic malignant melanoma: a systematic re-view. Cancer Treat. Rev. 34, 145–146

2. Atallah, E., and Flaherty, L. (2005) Treatment of metastaticmalignant melanoma. Curr. Treat. Options Oncol. 6, 185–193

3. Soengas, M. S., and Lowe, S. W. (2003) Apoptosis and mela-noma chemoresistance. Oncogene 22, 3138–3151

4. Eberle, J., Kurbanov, B. M., Hossini, A. M., Trefzer, U., andFecker, L. F. (2007) Overcoming apoptosis deficiency of mela-noma-hope for new therapeutic approaches. Drug Resist. Updat.10, 218–234

5. Ivanov, V. N., Bhoumik, A., and Ronai, Z. (2003) Death recep-tors and melanoma resistance to apoptosis. Oncogene 22, 3152–3161

6. La Porta, C. A. (2007) Drug resistance in melanoma: newperspectives. Curr. Med. Chem. 14, 387–391

7. Boldogh, I. R., and Pon, L. A. (2006) Interactions of mitochon-dria with the actin cytoskeleton. Biochim. Biophys. Acta 1763,450–462

8. Chua, B. T., Volbracht, C., Tan, K. O., Li, R., Yu, V. C., and Li,P. (2003) Mitochondrial translocation of cofilin is an early stepin apoptosis induction. Nat. Cell Biol. 5, 1083–1089

9. Franklin-Tong, V. E., and Gourlay, C. W. (2008) A role for actinin regulating apoptosis/programmed cell death: evidence span-ning yeast, plants and animals. Biochem. J. 413, 389–404

10. Hayot, C., Debeir, O., Van Ham, P., Van Damme, M., Kiss, R.,and Decaestecker, C. (2006) Characterization of the activities ofactin-affecting drugs on tumor cell migration. Toxicol. Appl.Pharmacol. 211, 30–40

11. Yang, F., Demma, M., Warren, V., Dharmawardhane, S., andCondeelis, J. (1990) Identification of an actin-binding proteinfrom Dictyostelium as elongation factor 1a. Nature 347, 494–496

12. Murray, J. W., Edmonds, B. T., Liu, G., and Condeelis, J. (1996)Bundling of actin filaments by elongation factor 1 alpha inhibitspolymerization at filament ends. J. Cell Biol. 135, 1309–1321

13. Gross, S. R., and Kinzy, T. G. (2005) Translation elongationfactor 1A is essential for regulation of the actin cytoskeleton andcell morphology. Nat. Struct. Mol. Biol. 12, 772–778

14. Kornienko, A., and Evidente, A. (2008) Chemistry, biology, andmedicinal potential of narciclasine and its congeners. Chem. Rev.108, 1982–2014

15. Gheorghiu, A., and Ionescu-Matiue. (1962) Presence of lycorineand galanthamine in Leucojum aestivum L. Anatomical study ofthe aerial parts of the plant and the corresponding powders.Ann. Pharm. Fr. 20, 531–538

16. Liu, J., Li, Y., Tang, L. J., Zhang, G. P., and Hu, W. X. (2007)Treatment of lycorine on SCID mice model with human APLcells. Biomed. Pharmacother. 61, 229–234

17. Liu, X. S., Jiang, J., Jiao, X. Y., Wu, Y. E., Lin, J. H., and Cai, Y. M.(2009) Lycorine induces apoptosis and down-regulation ofMcl-1 in human leukemia cells. Cancer Lett. 274, 16–24

18. Lamoral-Theys, D., Andolfi, A., Van Goietsenoven, G., Cimmino,A., Le Calve, B., Wauthoz, N., Megalizzi, V., Gras, T., Bruyere, C.,Dubois, J., Mathieu, V., Kornienko, A., Kiss, R., and Evidente, A.(2009) Lycorine, the main phenanthridine amaryllidaceae alka-loid, exhibits significant anti-tumor activity in cancer cells that areresistant to proapoptotic stimuli: an investigation of SAR andmechanistic insight. J. Med. Chem. 52, 6244–6256

19. Pettit, G. R., Gaddamidi, V., Herald, D. L., Singh, S. B., Cragg,G. M., Schmidt, J. M., Boettner, F. E., Williams, M., and Sagawa,Y. (1986) Antineoplastic agents, 120. Pancratium littorale. J. Nat.Prod. 49, 995–1002

20. Pettit, G. R., Ducki, S., Eastham, S. A., and Melody, N. (2009)Antineoplastic agents. 454. Synthesis of the strong cancer cellgrowth inhibitors trans-dihydronarciclasine and 7-deoxy-trans-dihydronarciclasine (1a). J. Nat. Prod. 72, 1279–1282

21. Bi, Y., Guo, J., Zhang, L., and Wong, Y. (2003) Changes in someenzymes of microbodies and plastid development in excisedradish cotyledons: effect of narciclasine. J. Plant. Physiol. 160,1041–1049

22. Ceriotti, G. (1967) Narciclasine: an antimitotic substance fromNarcissus bulbs. Nature 213, 595–596

23. McLachlan, A., Kekre, N., McNulty, J., and Pandey, S. (2005)Pancratistatin: a natural anti-cancer compound that targetsmitochondria specifically in cancer cells to induce apoptosis.Apoptosis 10, 619–630

24. Siedlakowski, P., McLachlan-Burgess, A., Griffin, C., Tirumalai,S. S., McNulty, J., and Pandey, S. (2008) Synergy of Pancratista-tin and Tamoxifen on breast cancer cells in inducing apoptosisby targeting mitochondria. Cancer Biol. Ther. 7, 376–384

25. Dumont, P., Ingrassia, L., Rouzeau, S., Ribaucour, F., Thomas, S.,Roland, I., Darro, F., Lefranc, F., and Kiss, R. (2007) The Amaryllidaceaeisocarbostyril narciclasine induces apoptosis by activation of the deathreceptor and/or mitochondrial pathways in cancer cells but not innormal fibroblasts. Neoplasia 9, 766–776

26. Carrasco, L., Fresno, M., and Vazquez, D. (1975) Narciclasine:an antitumour alkaloid which blocks peptide bond formation byeukaryotic ribosomes. FEBS Lett. 52, 236–239

27. Jimenez, A., Santos, A., Alonso, G., and Vazquez, D. (1976) Inhibitorsof protein synthesis in eukarytic cells. Comparative effects of someamaryllidaceae alkaloids. Biochim. Biophys. Acta 425, 342–348

28. Lefranc, F., Sauvage, S., Van Goietsenoven, G., Megalizzi, V.,Lamoral-Theys, D., Debeir, O., Spiegl-Kreinecker, S., Berger,W., Mathieu, V., Decaestecker, C., and Kiss, R. (2009) Narcic-lasine, a plant growth modulator, activates Rho and stress fibersin glioblastoma cells. Mol. Cancer Ther. 8, 1739–1750

29. Ingrassia, L., Lefranc, F., Dewelle, J., Pottier, L., Mathieu, V.,Spiegl-Kreinecker, S., Sauvage, S., El Yazidi, M., Dehoux, M.,Berger, W., Van Quaquebeke, E., and Kiss, R. (2009) Structure-activity relationship analysis of novel derivatives of narciclasine(an Amaryllidaceae isocarbostyril derivative) as potential antican-cer agents. J. Med. Chem. 52, 1100–1114

30. Friesner, R. A., Murphy, R. B., Repasky, M. P., Frye, L. L.,Greenwood, J. R., Halgren, T. A., Sanschagrin, P. C., and Mainz,D. T. (2006) Extra precision glide: docking and scoring incor-porating a model of hydrophobic enclosure for protein-ligandcomplexes. J. Med. Chem. 49, 6177–6196

31. Sadowski, J., and Gasteiger, J. (1993) From atoms and bonds to3-dimensional atomic coordinates – automatic model builders.Chemical Rev. 93, 2567–2581

32. Pittman, Y. R., Kandl, K., Lewis, M., Valente, L., and Kinzy, T. G.(2009) Coordination of eukaryotic translation elongation factor 1A(eEF1A) function in actin organization and translation elongationby the guanine nucleotide exchange factor eEF1Balpha. J. Biol.Chem. 284, 4739–4747

4583NARCICLASINE TARGETS eEF1A

33. Mathieu, V., Pirker, C., de Lassalle, E. M., Vernier, M., Mijatovic,T., DeNeve, N., Gaussin, J-F., Dehoux, M., Lefranc, F., Berger,W., and Kiss, R. (2009) The sodium pump alpha1 subunit: adisease progression-related target for metastatic melanomatreatment. J. Cell. Mol. Med. 13, 3960–3972

34. Berger, W., Hauptmann, E., Elbling, L., Vetterlein, M., Kokos-chka, E. M., and Micksche, M. (1997) Possible role of themultidrug resistance-associated protein (MRP) in chemoresis-tance of human melanoma cells. Int. J. Cancer 71, 108–115

35. Evidente, A. (1991) Narciclasine: 1H- and 13C-NMR Data and anew improved method of preparation. Planta. Med. 57, 293–295

36. Liu, G., Tang, J., Edmonds, B. T., Murray, J., Levin, S., andCondeelis, J. (1996) F-actin sequesters elongation factor 1alphafrom interaction with aminoacyl-tRNA in a pH-dependent reac-tion. J. Cell Biol. 135, 953–963

37. Pelletier, C. L., Maggi, L.B. Jr., Brady, S. N., Scheidenhelm, K.,Gutmann, D. H., and Weber, J. D. (2007) TSC1 sets the rate ofribosome export and protein synthesis through nucleophosmintranslation. Cancer Res. 67, 1609–1617

38. Gross, S. R., and Kinzy, T. G. (2007) Improper organization ofthe actin cytoskeleton affects protein synthesis at initiation. Mol.Cell. Biol. 27, 1974–1989

39. Jansen, R. P. (1999) RNA-cytoskeletal associations. FASEB J. 13,455–466

40. Lopez de Heredia, M., and Jansen, R. P. (2004) mRNA localiza-tion and the cytoskeleton. Curr. Opin. Cell Biol. 16, 80–85

41. Wang, C., Zhou, G. L., Vedantam, S., Li, P., and Field, J. (2008)Mitochondrial shuttling of CAP1 promotes actin- and cofilin-dependent apoptosis. J. Cell Sci. 121, 2913–2920

42. Duttaroy, A., Bourbeau, D., Wang, X. L., and Wang, E. (1998)Apoptosis rate can be accelerated or decelerated by overexpres-sion or reduction of the level of elongation factor-1 alpha. Exp.Cell. Res. 238, 168–176

43. Pinke, D. E., Kalloger, S. E., Francetic, T., Huntsman, D. G., andLee, J. M. (2008) The prognostic significance of elongationfactor eEF1A2 in ovarian cancer. Gynecol. Oncol. 108, 561–568

44. Edmonds, B. T., Wyckoff, J., Yeung, Y. G., Wang, Y., Stanley,E. R., Jones, J., Segall, J., and Condeelis, J. (1996) Elongationfactor-1 alpha is an overexpressed actin binding protein inmetastatic rat mammary adenocarcinoma. J. Cell Sci. 109, 2705–2714

45. Li, R., Wang, H., Bekele, B. N., Yin, Z., Caraway, N. P., Katz,R. L., Stass, S. A., and Jiang, F. (2006) Identification of putativeoncogenes in lung adenocarcinoma by a comprehensive func-tional genomic approach. Oncogene 25, 2628–2635

46. Grassi, G., Scaggiante, B., Farra, R., Dapas, B., Agostini, F., Baiz,D., Rosso, N., and Tiribelli, C. (2007) The expression levels ofthe translational factors eEF1A 1/2 correlate with cell growthbut not apoptosis in hepatocellular carcinoma cell lines withdifferent differentiation grade. Biochimie (Paris) 89, 1544–1552

47. Gerweck, L. E., and Seetharaman, K. (1996) Cellular pHgradient in tumor versus normal tissue: potential exploitationfor the treatment of cancer. Cancer Res. 56, 1194–1198

48. Noble, C. G., and Song, H. (2008) Structural studies of elonga-tion and release factors. Cell. Mol. Life. Sci. 65, 1335–1346

Received for publication June 2, 2010.Accepted for publication July 8, 2010.

4584 Vol. 24 November 2010 VAN GOIETSENOVEN ET AL.The FASEB Journal � www.fasebj.org