Peloruside A enhances apoptosis in H- ras -transformed cells and is cytotoxic to proliferating T...

Apoptosis 2004; 9: 785–796C© 2004 Kluwer Academic Publishers

Peloruside A enhances apoptosis in H-ras-transformedcells and is cytotoxic to proliferating T cells

J. H. Miller, B. Rouwe∗ , T. N. Gaitanos, K. A. Hood†, K. P. Crume, B. T. Backstrom, A. C. La Flamme,M. V. Berridge and P. T. Northcote

School of Biological Sciences (J. H. Miller, B. Rouwe, T. N. Gaitanos, K. A. Hood, K. P. Crume, A. C. La Flamme);School of Chemical and Physical Sciences (P. T. Northcote); Victoria University of Wellington, P.O. Box 600, Wellington,New Zealand; Malaghan Institute of Medical Research, P.O. Box 7060, Wellington South, New Zealand(B. T. Backstrom, M. V. Berridge)

Peloruside A (peloruside), a compound isolated from themarine sponge Mycale hentscheli, inhibits growth of hu-man (HL-60) and mouse (32D-ras) myeloid leukemic cells,as well as non-transformed 32D cells. Using the MTT cellproliferation assay and trypan blue dye exclusion tests,little difference was seen in growth inhibition between32D and 32D-ras cells; however, peloruside was morecytotoxic to the oncogene-transformed cells. Pelorusidealso blocked 32D-ras cells more readily in G2/M of thecell cycle, leading to apoptosis. Annexin-V/propidium io-dide staining of 32D and 32D-ras cells showed that 1.6 µMpeloruside induced significant cell death by 36 hours in32D cells (16% survival), but to comparable levels as earlyas 14 hours in 32D-ras cells (11% survival). There was noevidence for activation of either of the initiator caspases-8or -9 by 0.1 µM peloruside following 12 hours of exposure.Peloruside inhibited T cell proliferation and IL-2 and IFNγproduction in both the mixed lymphocyte reaction andfollowing CD3 cross-linking, and this effect was shownto be a non-specific cytotoxic effect. It is concluded thatpeloruside preferentially targets oncogene-transformedcells over non-transformed cells by inducing transformedcells to undergo apoptosis.

Keywords: caspase; immunosuppression; mycalamide;pateamine; peloruside; Ras.

Introduction

Peloruside A (peloruside), a secondary metabolite iso-lated from the New Zealand marine sponge My-cale hentscheli,1 induces apoptosis in transformed and

Correspondence to: J. H. Miller, School of Biological Sciences,Victoria University of Wellington, P.O. Box 600, Wellington,New Zealand. Tel: 64-4-463-6082; Fax: 64-4-463-5331; e-mail:[email protected]∗Current address: Sarphatipark, 121-D, 1073 CX Amsterdam,The Netherlands.†Current address: Dairy Biotechnologies Group, AgResearchLtd., Ruakura Research Centre, Private Bag 312 3, Hamilton,New Zealand.

untransformed cells in culture.2 Structural similarities be-tween peloruside and another potent marine toxin, bryo-statin, suggested a possible common mode of action in-volving protein kinase C pathways, but further tests indi-cated that this was not the case.2 Peloruside has sincebeen shown to be a potent stabilizer of microtubules,acting in a similar manner to paclitaxel and other mi-crotubule stabilizing drugs.3 In addition to peloruside,two other structurally unrelated sponge metabolites iso-lated from the same species of sponge, mycalamide A(mycalamide) and pateamine, are also cytotoxic,4−7 andinduce apoptosis8; although peloruside is a less potenttoxin than the other two.2,8 The mechanism of actionof mycalamide and pateamine differ from that of peloru-side. Mycalamide is a protein synthesis inhibitor7 thatis reported to selectively inhibit Ras biosynthesis.9 Themechanism of action of pateamine is not known; however,both mycalamide10 and pateamine11 are reported to haveimmunosuppressive properties, acting to inhibit T cellreceptor signal transduction.

Apoptosis involves the activation of a cascade of pro-teases called caspases. Caspases can be separated into threemain classes: initiators (caspases 8, 9, and 10), effectors(caspases 3, 6, and 7) and non-apoptotic caspases (caspasesnot involved in programmed cell death).12 There are twomain apoptotic pathways that lead to caspase activation,an external death receptor pathway in which a Fas ligand-receptor complex activates caspase-8, triggering a caspasecascade, or alternatively, an intrinsic pathway involvingcytochrome-c release from mitochondria leading to theactivation of caspase-9, and a similar caspase cascade.

The selective enhancement of malignant cell apoptosisis a major goal of anti-cancer research. Conventional cancerchemotherapy blocks cell division but lacks selectivity foroncogenic cells which can result in serious cytotoxic sideeffects for the patient.13 The key to specificity in cancerchemotherapy may be found in pharmacological target-ing of specific molecules,14 including the ras oncogene.

Apoptosis · Vol 9 · No 6 · 2004 785

J. H. Miller et al.

The ras gene is mutated in about 30% of human malig-nancies, including lung, colon, pancreatic, and thyroidcarcinomas as well as leukemias.13,15 Ras plays an impor-tant integrative role in relaying extracellular signals togene transcriptional systems; in particular those systemsinvolved in cell cycle progression, differentiation, apop-tosis, and T cell activation. There are different isoforms ofhuman ras oncogenes, including H, K, and N-ras.16 Mu-tant Ras protein displays reduced hydrolytic activity, isinsensitive to GTPase activating proteins,16 and is consti-tutively activated, leading to loss of cell cycle control thatresults in enhanced cell proliferation.15 Ogawara et al.9 re-ported that mycalamide down-regulates ras expression ina myeloid cell line, suggesting that this sponge metabolitemay specifically target Ras and inhibit cell proliferation.In our laboratory, both mycalamide and pateamine havebeen shown to have enhanced cytotoxic activity in rasand bcr/abl transformed murine myeloid leukemic cells(32D cells).8 To test the possibility that peloruside isalso more effective in ras-transformed cells, the effectsof peloruside on 32D and 32D-ras cells were compared.The ability of the three sponge metabolites to activatecaspases in a human myeloid leukemic cell line, HL-60,and to block mouse splenocyte activation through MHCand T cell receptor (TCR) signaling pathways were alsoexamined.

Materials and methods

Cell culture and mouse strains

The murine myeloid precursor parental 32D clone-3 cellline (32D), its ras-transformed derivative 32D-Ts-H-ras(32D-ras), and the human myeloid leukemic cell lineHL-60 were cultured as previously described.2,8,17 The32D parental cell line but not its ras-transformed deriva-tive requires cytokine IL-3 for growth and survival, andthis cytokine was provided to the cultures by additionof 10% WEHI3-conditioned medium. The myeloid celllines were cultured at 37◦C in a 5% CO2 atmospherein RPMI-1640 medium (GibcoBRL, Life Technologies),supplemented with 10% heat-inactivated fetal calf serum(FCS) (GibcoBRL, Life Technologies), 25–50 U/mL peni-cillin, and 25–50 µg/mL streptomycin. Mouse spleno-cytes were cultured in a 96-well plate (106 cells/well)in DMEM medium supplemented with 10% FCS,100 U/mL penicillin, 100 µg/mL streptomycin, 1 mML-glutamine, 10 mM HEPES, and 5× 10−5 M2-mercaptoethanol. Adult C57BL/6J mice were obtainedfrom the Wellington School of Medicine Biomedical Re-search Unit. Mice from an outbred strain were obtainedfrom the Victoria University Animal Facility. All proce-dures involving animals were approved by the VictoriaUniversity Animal Ethics Committee.

Cell proliferation assays

For cell proliferation, either an MTT cell proliferationassay8 or 3H-thymidine incorporation assay2 was used.In both assays, cells were seeded at 5 × 103 cells perwell in a 96-well plate (Nunc, Roskilde, Denmark). Foranti-CD3, stimulation 106 cells were seeded per well. In-dividual cell viability was assessed by measuring eithertrypan blue dye exclusion or Annexin-V-FLUOS binding(Roche Diagnostics NZ Ltd. Auckland, New Zealand).8

Flow cytometry

Distributions of annexin-V-propidium iodide (PI) stainedcells were examined in a FACSort flow cytometer (BectonDickinson, Franklin Lakes, NJ).8 In brief, 10,000 cellevents were collected for each assay, and the data wereanalyzed by CELLquest software (Becton Dickinson). Eachtest sample was accompanied by a time-matched controlsample (cells not treated with sponge metabolite).

Cell cycle block

To assess the effect of peloruside on the cell cycle, cellswere fixed in 70% ethanol at 4◦C, then stained for2 h at room temperature in 45 µg/mL PI containing10 µg/ml RNaseA and 0.1% glucose, and analyzed byflow cytometry.3 The numbers of cells in each PI fluores-cence peak were calculated by aligning the cell cycle pro-files with ModFitL V1.00 (MacIntosh Computers). Sincethis program is designed for human cell lines, not all datawere able to be aligned to the pre-programmed profile. Inthese cases, the FL2-area was plotted against SSC-heightto estimate the number of cells in different phases of thecell cycle.

Measurement of caspase activities

A fluorescent substrate assay was used following the man-ufacturer’s recommendations (Calbiochem, San Diego, CAand Bio-Rad Laboratories, Auckland, New Zealand) tomeasure caspase-8 and -9 activities in cells treated withpeloruside, mycalamide, and pateamine. In brief, 2× 106

cells were added to a culture dish, treated with 100 nMtest compound, and cultured as described above for a settime. Cell lysates were then prepared, and caspase ac-tivity was determined in a black, 96-well fluoroNuncplate (Nunc, Roskilde, Denmark) as the change in fluo-rescence at 37◦C over a 1 h period. Fluorescence was mea-sured using a microplate spectrofluorometer (SpectraMaxGemini, Molecular Devices, Sunnyvale, CA). Protein con-tent was determined using a Bradford assay, modified for

786 Apoptosis · Vol 9 · No 6 · 2004

Peloruside A enhances apoptosis in H-ras-transformed cells

96-well plates,18 and absorbance was measured in a mi-croplate reader (VersaMax, Molecular Devices). Caspaseactivity is presented as the Vmax (increase in relative fluo-rescence units (RFU) per min) per µg protein, relative toa control sample incubated without sponge metabolite.

Mixed lymphocyte reaction

The spleens from outbred or C57BL/6J mice were pre-pared in IMDM medium (Gibco-BRL, Grand Island, NY)containing 5% FCS as previously described.2 Splenocytecell suspensions from outbred mice (“stimulators”) wereexposed to 27.5 Grays of radiation in a “gammacell” irra-diator to stop cell proliferation. To distinguish between re-sponder and stimulator populations of splenocytes, stim-ulators were stained with 5,6-carboxyfluorescein diacetatesuccinimidyl ester (CFSE) (Fluka, Buchs, Switzerland) be-fore plating. Briefly, irradiated stimulators were washed inPBS and resuspended at 2×107 cells/mL in PBS. An equalvolume of CFSE (2.5 µM in PBS) was added to each cellsuspension, and the cells were incubated at room tempera-ture for 8 min, mixing occasionally. An equivalent volumeof FCS was added to the tube, and the cells were washed4× with IMDM growth medium. Cells were then seededin 96-well plates at 4× 105 stimulators/well and 5× 105

responders/well in a final volume of 0.2 mL and exposed toa range of concentrations of sponge metabolites. At 54 h,50 µL of supernatant was removed for determination ofIL-2 production by the responder cells, and the removedmedium was replaced with growth medium containing3H-thymidine (0.5 µCi/well). Cells were harvested af-ter a total of 72 h of culture, as previously described,2

and their incorporated radioactivity determined in a liq-uid scintillation counter (Betaplate Scint, Wallac, PerkinElmer, Boston, MA). To assess cell viability, annexin-Vstaining of the cells was ascertained by flow cytometry.Non-dividing stimulator cells that had been labeled withCFSE were gated out using a sample of stimulators with-out responders to confirm the position of this populationof cells in the flow scan, and only cells outside of this gatedregion were included in the data analysis.

Anti-CD3 T cell receptor stimulation

Splenocytes from C57BL/6J mice were isolated as de-scribed above. T cells were activated by cross-linking ofTCRs to plate-bound anti-CD3e antibody (0.5 µg/well)(hamster mAb 145-2C11, BD Bioscience, Franklin Lakes,NJ).19 Cultured cells in triplicate wells were pooled forflow cytometry. CD4 helper T cells and CD8 cytotoxicT cells were identified by Cy-Chrome-labeled antibodies(BD Bioscience, Franklin Lakes, NJ).

IL-2 and IFNγ ELISA

Two cytokines important in T cell signaling, IL-22 andIFNγ 20 (Yang et al. 1993) were assayed by sandwichELISA following the manufacturer’s recommendations(BD Bioscience, Franklin Lakes, NJ).

Statistical analysis

Data were analyzed using GraphPad Prism software(GraphPad Software, Inc, San Diego, CA). All values fordrug effects are presented as a percentage of untreatedcontrol cultures. Data from dose-response curves were an-alyzed for significant differences using a two-way repeatmeasures ANOVA (ANOVA-RM) followed by Bonfer-roni’s multiple comparisons test. The Student’s t-test wasused to determine whether IC50 values, calculated fromthe logit-log equation, were significantly different fromeach other. A Dunnett test was used to analyze normalizedcaspase activity data for significant increases in activityfollowing drug exposure.

Results

Cytotoxic effects of peloruside

The effects of peloruside on growth of 32D and 32D-rascells were tested using the MTT assay (Figures 1 and 2) andtrypan blue dye exclusion (Figure 3). Based on the MTTcell proliferation assay, at concentrations below 0.4 µMpeloruside, 32D cells were more sensitive to growth in-hibition than 32D-ras cells; however, this effect was re-versed at the highest concentration tested (1.6 µM). TheIC50 value, calculated after 2 days of peloruside exposure,was 4-fold greater with 32D-ras cells than with 32D cells(Figure 2). Using trypan blue dye exclusion to measure theactual number of viable cells, there was no difference inthe growth of 32D-ras cells relative to non-transformedcells (data not shown). It seems likely that the differ-ence between the MTT results and the trypan blue resultsis an artifact of the differing capacities of the two celllines to reduce the MTT substrate. The percent of viablecells was calculated from the trypan blue dye exclusiontests (Figure 3), and the viability of ras-transformed 32Dcells was significantly lower than that of 32D cells atpeloruside concentrations of 0.4 µM or higher. No differ-ences in viability were seen at peloruside concentrationsof 0.1 µM or less. After a 36 h treatment with 1.6 µMpeloruside, almost all 32D-ras cells were dead; whereas,the viability of untransformed 32D cells was largely unaf-fected. This indicates that whereas the growth inhibitoryeffect of peloruside on 32D cells is largely cytostatic, 32D-ras cells undergo cell death.


J. H. Miller et al.

Figure 1. Peloruside effects on MTT response of 32D and 32D-ras cells. Cells were incubated in 96-well plates for 0 to 48 h withpeloruside concentrations ranging from 12.5 nM to 1.6 µM af-ter which MTT responses were determined. Wild-type 32D cells(open circles) were compared with transformed 32D-ras cells(closed circles). Data are presented as the mean ± SEM (n = 6wells from two experiments). Differences were significant for allconcentrations tested (p < 0.001; repeat measure ANOVA) ex-cept 400 nM peloruside. Bonferroni post hoc test p values at dif-ferent incubation times are ∗ p < 0.05; ∗∗ p < 0.01; ∗∗∗ p < 0.001.

Apoptotic and necrotic effects of peloruside

Annexin-V and PI staining of 32D cells treated with1.6 µM peloruside were examined by flow cytometry(Figure 4A). Most cells were viable after 22 h exposure,although small populations of apoptotic (2%) or late apop-totic/necrotic (14%) cells were present by 36 h exposure.These results are consistent with the data of Figure 3 usingtrypan blue to assess viability. Cultures of 32D-ras cells, onthe other hand, treated with 1.6 µM peloruside showed asmall population of late apoptotic/necrotic cells by as earlyas 3 h (data not shown) and by 14 h, a larger populationof apoptotic cells appeared (7%) (Figure 4B). The size ofthe pooled apoptotic/necrotic population increased withtime, and by 24 h most cells (85%) were either apoptoticor necrotic. Tails from the upper-right quadrant to both

Figure 2. Pooled 48 h IC50 data from Figure 1. Data are presentedas the mean± SEM (n = 9 wells from 3 preparations). IC50 valueswere calculated from the logit-log equation. The 4-fold higher IC50value for 32D-ras cells (filled squares) over 32D cells (open circles)was significant (p < 0.025, Student’s t-test).

the lower-right (late apoptotic) and lower-left quadrants(necrotic) suggest that 1.6 µM peloruside can directlyinduce both apoptosis and necrosis in 32D-ras cells.

Effects of peloruside on caspase activity

Peloruside, mycalamide, and pateamine have been previ-ously shown to induce apoptosis in a number of differ-ent cell lines as seen by DNA laddering and annexin-Vbinding.2,8 To determine the apoptotic pathways inducedby these sponge metabolites, the effects of all three com-pounds on caspase activities were tested using the humanmyeloid leukemic HL-60 cell line (Figure 5). Pelorusidefailed to activate either caspase-8 or caspase-9 to any sig-nificant extent over the 12-h exposure period (Figure 5A).Mycalamide and pateamine, however, caused significantincreases in both caspase-8 and -9 activities in the first12 h of exposure (p < 0.0001; one-way ANOVA-RM).Treatment of HL-60 cells with mycalamide induced max-imum caspase-8 activity (5-fold increase) within 1.5 h ofexposure, with activity dropping back to control levels by2 h (Figure 5B). Caspase-9 was not significantly activateduntil 6 h exposure to mycalamide (5-fold increase), andthe activity remained high for at least the next 6 h ofdrug exposure. Pateamine, on the other hand, initially ac-tivated caspase-9 five-fold after 1.5 h of exposure; whereas,caspase-8 activity increased 2.5-fold by 2 h (Figure 5C).



Figure 3. Peloruside effects on % viability of 32D and 32D-rascells. Cells from the experiments described in Figure 1 were as-sessed for cell viability using trypan blue dye exclusion. The resultsare presented as the mean ± SEM (n = 6 wells from two exper-iments). The differences between the two cell types were signifi-cant for 400 and 1600 nM (p < 0.001; repeat measure ANOVA).Bonferroni post hoc test p values at different incubation times are∗∗∗p < 0.001.

The activation of caspase 9 was biphasic, falling to a lowlevel by 4 h after its initial rise, before increasing again.

Cell cycle block by peloruside

To determine whether the cytostatic, cytotoxic, and apop-totic effects of peloruside were cell cycle stage specific,32D cells were treated with peloruside (1.6 µM) and thecell cycle status monitored by flow cytometry. Althoughpeloruside did not affect 32D cell viability significantly,it had significant effects on cell proliferation. After 36 hof exposure, peloruside blocked 32D cells in the G2/M

phase of the cell cycle (Figure 6A), causing a concurrentloss of cells in G1/G0and S phase. The sub-diploid A0 peakincreased after peloruside exposure, increasing to 40% ofthe total by 48 h. These results are quantitatively simi-lar to those obtained with annexin-V/PI flow cytometry(Figure 4).

With 32D-ras cells, the G2/M cell cycle block occurredearlier (Figure 6B). Apoptosis was first evident between12–18 h, with a large increase in the A0 peak, and by24 h, 72% of the cells were in the A0 peak. These resultsshow that 32D-ras cells exhibit an early G2/M arrest andenhanced presence of apoptotic bodies containing sub-diploid DNA.

Effect of peloruside on T cell activation

Because peloruside appears to preferentially target ras-transformed cells, it may also show immunosuppressiveactivity since Ras activation is required for the TCR signalcascade. Additionally, mycalamide and pateamine, whichalso target ras-transformed cells,8 have been reported to beselective immunosuppressants. Peloruside, mycalamide,and pateamine were therefore tested for their effects onthe proliferation of splenocytes (3H-thymidine incorpo-ration or MTT assay), cell viability (annexin-V staining),and IL-2 or IFNγ production, using the mixed lympho-cyte reaction (MLR) and anti-CD3-mediated TCR acti-vation. Cells treated with cyclosporin A, a known im-munosuppressant, were used as positive controls for spe-cific immunosuppressive activity. Figure 7A shows thatall three sponge metabolites inhibited IL-2 productionand 3H-thymidine incorporation in a similar manner tocyclosporin A in the MLR. Only cyclosporin A, however,was non-toxic to spleen cells at concentrations that in-hibited proliferation and IL-2 production. Peloruside wasslightly less potent against T cells (IC50 of 40–50 nM)(Figure 7) compared to 32D cells (10 nM) (Figure 2). TheIC50 values for mycalamide and pateamine for T cells (0.3–3 nM) (Figure 7) were similar to those reported previouslyin 32D myeloid leukemic cells for mycalamide (0.7 nM)and pateamine (0.8 nM).8 The IC50 value for lymphocyteviability (LCV), assessed as the proportion of cells notstaining with annexin-V, was divided by the MLR IC50

value (MLR), assessed by 3H-thymidine incorporation, todetermine an LCV/MLR ratio. LCV/MLR ratios were 1.7,0.9, and 1.3 for peloruside, mycalamide, and pateamine,respectively, and >32 for cyclosporin A. Thus, only cy-closporin A was able to inhibit MLR-stimulated lympho-cyte proliferation and IL-2 production without causingsignificant cell death.

To more selectively target the TCR signaling pathway,T cells were also stimulated by plate-bound anti-CD3 an-tibody. The MTT cell proliferation assay was used to as-sess effects of the compounds on T cell proliferation, andannexin-V binding was used to determine the proportion


J. H. Miller et al.

Figure 4. Flow cytometry of annexin-V and PI staining. Both 32D cells (A) and 32D-ras cells (B) were treated for different lengths oftime with 1.6 µM peloruside. Annexin-V fluorescence is plotted along the x-axis and PI fluorescence on the y-axis. Representativeexperiments are presented.

of viable cells. IFNγ production was used as a marker ofTCR pathway activation. Peloruside and pateamine inhib-ited T cell proliferation and IFNγ production (Figure 7B),but killed the cells in a similar manner as in the MLR. Inthe case of mycalamide and cyclosporin A, however, IFNγ

production was completely inhibited at the higher con-centrations tested with T cell viability never falling below30% of the untreated controls after a 24 h treatment. My-calamide inhibited cell proliferation to a greater extentthan cyclosporin A. Longer exposures of 72 h completelyinhibited cell proliferation (MTT assay); however, cell vi-ability remained high (36% for mycalamide at 10 nMand 28% for cyclosporin A at 1000 nM). CD4 helper andCD8 cytotoxic T cell viabilities were tested separately andshowed the same trends (data not shown).

Discussion

Cytostatic and cytotoxic effects of pelorusidein 32D and 32D-ras cells

Previous studies in our laboratory8 using trypan blue dyeexclusion showed that transformation of 32D cells with

ras, and to a lesser extent bcr/abl, oncogenes made thecells more sensitive to the cytotoxic effects of the spongemetabolites mycalamide and pateamine. In the presentstudy with peloruside, 32D-ras cells were also shown tobe more susceptible to the cytotoxic action of peloruside(Figure 3). All three sponge metabolites, however, failedto show increased potency in ras-transformed cells basedon MTT cell proliferation assays which tend to be cell line-specific. In fact, the inhibition of MTT dye reduction bypeloruside in 32D cells was actually less than that of 32D-ras cells (Figures 1 and 2). Since the numbers of viable cellsat specific peloruside concentrations were similar betweentransformed and non-transformed cells (data not shown)and given that the MTT assay measures the number ofviable cells based on their metabolic status,21 it may notbe appropriate to directly compare MTT dye reductionbetween the two cell lines since this comparison ignorestheir inherent differences in MTT metabolism. Peloru-side seems to inhibit 32D cell metabolism more than32D-ras cell metabolism. Thus, the ras gene may offersome protection against partial inhibition of metabolismthat leads to a reduced MTT response. Cancer cells areknown to proliferate better than non-transformed cells



Figure 5. Effect of sponge metabolites on caspase-8 and -9activities. HL-60 cells were cultured with 100 nM peloruside(top), 100 nM mycalamide (middle), or 100 nM pateamine (bot-tom) for 0–12 h, lysed, and their caspase-8 (filled circles) andcaspase-9 (open circles) activities measured as relative fluores-cence units per min per µg protein. Untreated controls were in-cluded, and activity is presented as the fraction of the controlvalue. Data are presented as the mean ± SEM. Sample sizesare 4–6 preparations with triplicate wells for peloruside (12–18wells total), 3–4 preparations in triplicate for mycalamide (9–12wells total), and 2 preparations in triplicate (6 wells total) forpateamine. ∗ p< 0.05; ∗∗ p< 0.001 (Dunnett multiple comparisonstest).

in hypoxic environments22 and may also have unique cellsurface energy-generation that supports oxidative systemsand bypasses mitochondrial aerobic glycolysis.23 Reactiveoxygen species (ROS) may also contribute to this anoma-lous MTT effect since ROS are involved in Ras down-stream signaling.24,25 Ras-transformed NIH/3T3 fibrob-lasts, for example, have enhanced basal levels of intra-cellular ROS24,26 relative to non-transformed cells. Al-though neither cell surface oxidative activity nor intra-cellular ROS levels were measured in the present study,increased peloruside sensitivity of 32D cells compared to32D-ras cells in the MTT assay may reflect intrinsic dif-ferences in cell metabolism between non-transformed andtransformed cells.

Agents that selectively induce cell death in oncogeniccells are a major focus of anti-cancer research. UsingDNA laddering, we have previously shown in HL-60cells that peloruside causes cell death in part by induc-ing apoptosis.2 In the present study, DNA laddering wasexamined in 32D and 32D-ras cells treated with peloru-side concentrations ranging from 25 nM to 1 µM (resultsnot shown). After 2 days, typical apoptotic DNA ladderswere seen at all concentrations in both cell lines; however,after shorter exposures of only 1 day, DNA laddering wasonly seen in the ras-transformed cell line at concentra-tions of 0.5 µM or above. These higher concentrationsdid not cause DNA laddering in 32D cells after 1-dayexposures. These results are consistent with the percentviability data of Figure 3 and indicate that the cytotoxiceffects of peloruside in ras-transformed 32D cells are inpart due to induction of apoptosis.

Flow cytometry results in the present study indicatethat peloruside induces cell death by both necrosis andapoptosis. The time at which cell death occurred differed,with 32D-ras cells dying much earlier than 32D cells. In-duction of apoptosis can occur within as short a time as1–3 hours, as reflected in the caspase activation results inthe present study and by others.27,28 The lack of a largeapoptotic population at any one time in peloruside-treatedcells may relate to the mechanism of mitotic inhibitionby which peloruside acts.3 Since cells may not becomeapoptotic until they arrest at G2/M,29–31 in a mixed, asyn-chronous cell population, individual cells would block atdifferent times after exposure to peloruside as they pro-ceed through the cell cycle. 32D-ras cells have a shortergeneration time than 32D cells (approximately 12 h and16 h, respectively), and this may contribute to their in-creased susceptibility. Cell cycle analysis clearly demon-strated that peloruside arrests both 32D and 32D-ras cellsin G2/M.

An interesting question is whether ras transforma-tion causes the cells to become more sensitive in gen-eral to stresses imposed on them, or whether pelorusidespecifically targets the Ras pathway to induce apoptosis.


J. H. Miller et al.

Figure 6. Cell cycle analysis of 32D and 32D-ras cells. 32D cells (A) and 32D-ras cells (B) were treated with 1.6 µM peloruside for 48 hand 24 h, respectively. After exposure, the cells were stained with PI and analysed by flow cytometry. DNA content (FL2-A) is plottedon the x-axis against cell counts on the y-axis. The table at the bottom summarises the % of cells in each phase of the cell cycle. Thesub-diploid A0 peak represents apoptotic cell bodies.

Although other drugs, including general protein synthe-sis inhibitors like cycloheximide and puromycin, as wellas cytosine-β-D-arabinofuranoside, a drug that inducesDNA strand breaks, are also more potent in cells trans-

formed by the ras or bcr/abl oncogenes,8 transformationwith these oncogenes actually protects 32D cells from glu-cose deprivation-induced apoptosis (unpublished results).In addition, although the cytokine GM-CSF, which shares



Figure 7. Effects of sponge metabolites on T cell activation. (A)Lymphocyte proliferation in the MLR (open circles) was measuredby 3H-thymidine incorporation over the last 16 h of a 72 h in-cubation with peloruside, pateamine, mycalamide, or cyclosporinA. Cell viability (open squares) was measured by flow cytome-try as the % of annexin-V-unstained cells. IL-2 production (filleddiamonds) was measured by IL-2 ELISA of MLR supernatants.Tests were performed in triplicate for each experiment. The datapoints of the graphs represent the combined mean ± SEM valuesfrom 3 separate experiments. Data are expressed as the % of un-treated control cultures. (B) Anti-CD3-stimulated T cell responses.Lymphocyte proliferation (open triangles) was measured by MTTassay over a 24 h or 72 h period. The T cell viability (open squares)was measured by flow cytometry as the proportion of annexin-V-unstained cells. IFNγ production (open diamonds ) was measuredby ELISA. Tests were performed in triplicate for each experimentand presented as the mean ± SEM from 3 separate experiments.Data are expressed as the % of untreated control cultures.

a common subunit with IL-3, a growth factor needed for32D cell proliferation, inhibits proliferative responses in32D-ras cells, another cytokine, G-CSF, has no effect.17

Role of Ras protein in cytotoxicity

Ras-targeting may be a useful approach in the de-velopment of anti-cancer drugs.14,32,33 The ras geneplays a major role in G1 checkpoint events34 via theRaf/MEK/MAPK pathways as well as by the PI3K path-way. Ras signaling leads to production of cyclin D1 aswell as CDK inhibitors p21 and p27, all of which areinvolved in control of G1 progression.16 Little is knownhowever about the effect of mutant Ras protein on theG2/M transition. Microtubule disrupting agents activatethe cJun N-terminal kinase/stress-activated protein ki-nase (JNK/SAPK) and the apoptosis signal-regulating ki-nase (ASK1).30 Thus, both Ras and the ASK1 pathwaysmay be involved in the apoptotic response to pelorusidethrough activation of the JNK/SAPK cascade and Bcl-2.Because of the high sequence identity and similar abilitiesof the different forms of ras (H, K and N) to cause celltransformation, it is generally assumed that all ras muta-tions are equivalent, and most studies have used the H-rasisoform.15 However, a number of differences in structureand function of the three isoforms have been reported,16

and most leukemic mutations involve N-ras.16 Inacti-vation of p53 has also been shown to be important inTaxol-induced apoptosis.29,31 Cells deficient in p53 tendto block in G2M; whereas, p53 intact cells are able to passthrough mitosis and block in G1. The G2M block itself,however, does not seem to correlate with the loss of p53function that causes increased sensitivity to microtubulestabilizing agents.31

Comparison of peloruside to other metabolitesfrom sponges of Mycale sp.

Many bioactive compounds isolated from marine organ-isms arrest cells in the G2/M phase of the cell cycle.35,36

These compounds can be roughly divided into DNAdamaging agents and agents that interfere with the mi-totic spindle.37 Peloruside, mycalamide, and pateamineare all produced by the same sponge species.1,4,5 My-calamide and pateamine are 3-5-fold more toxic thanpeloruside2,8and have different mechanisms of action,3

but all three compounds induce apoptosis.

Caspase activation

Peloruside failed to induce biologically significantcaspase-8 or -9 activity in the first 12 h of exposure. Longer


J. H. Miller et al.

exposures for up to 28 h also failed to induce significantinitiator caspase activities (data not shown). It seems likelythat the apoptotic effects of peloruside are secondary toG2/M block, although p53 dysfunction may also play arole independently of cell cycle arrest.29,31 Asynchronousentry of cells into apoptosis would prevent a large enzymeactivation response from being seen at any one time. SinceHL-60 cells have a doubling time of about 30 h, the ma-jority of cells would not enter mitosis during the 12-hexposure used in our study. This might explain the failureof peloruside to increase caspase activity. An alternativeexplanation is that the peloruside concentration used inthis study (100 nM) was too low, although the concen-tration was 14-fold higher than the 7 nM IC50 previouslyreported for HL-60 cells.2

Experiments with mycalamide and pateamine on ini-tiator caspase-8 and -9 activities confirmed that apoptosisoccurs within 12 h of exposure of HL-60 cells to theseagents; however, the pathways by which apoptosis was in-duced differed. Mycalamide activated the death receptorpathway (caspase 8) relatively quickly following exposure(within 1 h), with caspase-9 activity increasing later. Itis not known if mycalamide acts directly or indirectly onthe death receptor in the plasma membrane, or whetherit activates caspase 8 downstream of the death receptor.Pateamine, on the other hand, first stimulated caspase-9activity, the initiator caspase for the mitochondrial apop-totic pathway. Due to cross-talk between the death recep-tor and mitochondrial pathways,12 it is possible that thedelayed increase in activity of the second caspase in eachcase was a downstream effect from activation of the firstcaspase.

We conclude that peloruside is unable to activate eithercaspase to a significant extent in the first 12 h, and thatmycalamide and pateamine activate alternative caspasepathways within the first 2 h of exposure.

Lack of immunosuppressive activity of peloruside

The immunosuppressive activity of peloruside hasnot been tested previously; however, mycalamide andpateamine are both reported to inhibit lymphocyteactivation.10,11 The in vitro mechanism of action of my-calamide is linked to inhibition of protein translation,7

and protein synthesis inhibitors may be effective immuno-suppressants through their ability to inhibit the produc-tion of cytokines, antibodies, adhesion molecules, andother proteins that are required for activation and recruit-ment of cells to inflammatory sites.38 The question thathas been asked is whether these sponge metabolites di-rectly inhibit TCR signal transduction as is the case withthe immunophilins, cyclosporin A and FK506.10,11 Cy-closporin A prevents TCR signal transduction by bindingto cyclophilin. The cyclosporin A-cyclophilin complex

then binds with calcineurin and prevents the calcineurinphosphatase activity that is required for transcriptionfactor activation and subsequent gene transcription.39,40

Galvin et al.10 and Romo et al.11 reported that mycalamideand pateamine have little cytotoxicity towards lympho-cytes. In addition, mycalamide was reported to inhibit Tcell proliferation induced by both TCR and CD28 as wellas PMA and CD28.10 With pateamine, which showedan LCV/MLR ratio of 1000 in their study, some specificTCR targeting was suggested because of its inhibition ofTCR-mediated IL-2 transcription through inhibition ofIL-2 promoter activation.11 There is some evidence thatpateamine also acts as a protein synthesis inhibitor, andthus, it may suppress immune cell responses in a simi-lar manner to mycalamide. Cell viability was assessed bytrypan blue dye exclusion in their studies, but since thisassay measures a loss of membrane integrity, an event thatoccurs late in apoptosis, it is possible that a proportion ofcells were counted as viable despite already being commit-ted to cell death. In the present study, we used annexin-Vbinding to measure an early event in apoptosis, that ofphosphatidyl serine translocation.

To test the immunosuppressive activity of pelorusideand to compare published data on the inhibitory effects ofmycalamide and pateamine on immune cell activation, allthree sponge metabolites were examined for their abilityto inhibit the MLR and anti-CD3 mediated TCR acti-vation pathways. None of the three sponge metaboliteswere selectively immunosuppressive in the MLR. Lym-phocyte signaling in the MLR is dependent on the pres-ence of antigen-presenting cells bearing MHC molecules;however, the anti-CD3 cross-linking activation of T cellsspecifically measures TCR activation pathways.19 Peloru-side and pateamine were cytotoxic against CD4 and CD8cells, but mycalamide showed a similar profile to cy-closporin A, with cell viability (28–59% viable based onannexin-V binding) at concentrations that completely in-hibited IFNγ production by the activated T cells. This isinconsistent with the MLR results in which mycalamidecaused similar cell death to the other two sponge metabo-lites. It is not known why mycalamide gave different re-sponses in the two assays; however, the complex signalingpathways involved in immune cell signaling may pro-vide an explanation. We conclude from these results thatpeloruside and pateamine are not specific immunosup-pressive drugs, but that mycalamide may be an effectiveimmunosuppressant toward T cell activation through par-ticular signaling pathways. Not all of the complexities ofthe immune system can be simulated in vitro, however,and the best test of immunosuppressive activity is to per-form allografts in mice in vivo. These experiments havebeen carried out with mycalamide,7 but no specific de-tails of the allograft rejection experiments were given intheir study.



Conclusion

We conclude that peloruside, although lacking specificimmunosuppressive activity, is more potent at causingcell death in myeloid leukemic cells that have been trans-formed with the ras oncogene than in non-transformedcells. This selective targeting of a cancerous cell by peloru-side indicates that further development of peloruside as apotential anti-cancer agent is warranted.

Acknowledgments

This work was supported in part by the Cancer Societyof NZ, the NZ Foundation for Research, Science, andTechnology (NERF contract number C01X0001), theWellington Medical Research Foundation, and VictoriaUniversity of Wellington.

References

1. West LM, Northcote PT, Battershill CH. Peloruside A: A po-tent cytotoxic macrolide isolated from the New Zealand marinesponge Mycale sp. J Org Chem 2000; 65: 445–449.

2. Hood KA, Backstrom BT, West LM, Northcote PT, BerridgeMV, Miller JH. The novel cytotoxic sponge metabolite peloru-side A, structurally similar to bryostatin-1, has unique bioac-tivity independent of protein kinase C. Anti-Cancer Drug Design2001; 16: 155–166.

3. Hood KA, West LM, Rouwe B, et al. Peloruside A, a novelantimitotic agent with paclitaxel-like microtubule-stabilizingactivity. Canc Res 2002; 62: 3356–3360.

4. Perry NB, Blunt JW, Munro MH. Mycalamide A, an antiviralcompound from a New Zealand sponge of the genus Mycale.J Am Chem Soc 1988; 110: 4850–4851.

5. Northcote P, Blunt J, Munro M. Pateamine: A potent cytotoxinfrom the New Zealand marine sponge, Mycale sp. TetrahedronLett 1991; 32: 6411–6414.

6. West LM, Northcote PT, Hood KA, Miller JH, Page MJ.Mycalamide D, a new cytotoxic amide from the New Zealandmarine sponge Mycale species. J Nat Prod 2000; 63: 707–709.

7. Burres NS, Clement JJ. Antitumor activity and mechanism ofaction of the novel marine natural products Mycalamide-A and-B and Onnamide. Canc Res 1989; 49: 2935–2940.

8. Hood KA, West LM, Northcote PT, Berridge MV, Miller JH.Induction of apoptosis by the marine sponge (Mycale) metabo-lites, mycalamide A and pateamine. Apoptosis 2001; 6: 207–219.

9. Ogawara H, Higashi K, Uchino K, Perry NB. Change of ras-transformed NRK-cells back to normal morphology by my-calamides A and B, anti-tumor agents from a marine sponge.Chem Pharmacol Bull 1991; 39: 2152–2154.

10. Galvin F, Freeman GJ, Razi-Wolf, Benacerraf B, Nadler L,Reiser H. Effects of cyclosporin A, FK506, and mycalamide Aon the activation of murine CD4+ T Cells by the murine B7antigen. Eur J Immunol 1993; 23: 283–286.

11. Romo D, Rzasa RM, Shea HA, et al. Total synthesis and im-munosuppressive activity of (-)-pateamine A and related com-pounds: Implementation of a β-lactam-based macrocycliza-tion. J Am Chem Soc 1998; 120: 12237–12254.

12. Strasser A, O’Conner L, Dixit VM. Apoptosis signalling. AnnRev Biochem 2000; 69: 217–245.

13. Buolamwini JK. Novel anticancer drug discovery. Curr OpinChem Biol 1999; 3: 500–509.

14. Workman P, Kaye SB. Translating basic cancer research intonew cancer therapeutics. Trends Molec Med 2002; 8 (No. 4Suppl): S1–S9.

15. Shields JM, Pruitt K, McFall A, Shaub A, Der CJ. Under-standing Ras: ‘It ain’t over ’til it’s over’. Cell Biol 2000; 10:147–154.

16. Malumbres M, Pellicer A. Ras pathways to cell cycle controland cell transformation. Frontiers Biosci 1998; 3: 887–912.

17. Ahmed N, Anderson SM, Berridge MV. IL-3 induces apoptosisin a ras-transformed myeloid cell line. Apoptosis 1999; 4: 1–10.

18. Bradford, MM. Rapid and sensitive method for the quantitationof microgram quantities of protein utilizing the principle ofprotein-dye binding. Analy Biochem 1976; 72: 248–254.

19. Kruisbeek AM, Shevach EM. Proliferative assays for T cell func-tion. In: Coligan J, Kruisbeek AM, Margulies D, Shevach EM,Strober W, eds. Current Protocols in Immunology. New York: JohnWiley and Sons, 1991: 3.12.1–3.12.14.

20. Yang X, Hayglass KT. A simple, sensitive, dual mAb basedELISA for murine gamma interferon determination: Compari-son with two common bioassays. J Immunoassay 1993; 14: 129–148.

21. Berridge MV, Tan AS, McCoy KD, Wang R. The biochemicaland cellular basis of cell proliferation assays that use tetrazoliumsalts. Biochemica 1996; 4: 14–19.

22. Seddon B, Kelland LR, Workman P. Bioreductive prodrugs forcancer therapy. Methods Mol Med 2004; 90: 515–542.

23. Berridge MV, Tan AS. High-capacity redox control at theplasma membrane of mammalian cells: Trans-membrane, cellsurface, and serum NADH-oxidases. Antioxid Redox Signal2000; 2: 231–242.

24. Irani K, Xia Y, Zweier JL, et al. Mitogenic signalling mediatedby oxidants in ras-transformed fibroblasts. Science 1997; 275:1649–1652.

25. Davis W Jr, Ronai Z, Tew KD. Cellular thiols and reactive oxy-gen species in drug-induced apoptosis. Persp Pharmacol 2001;296: 1–6.

26. Liou JS, Chen C-Y, Chen JS, Faller DV. Oncogenic Ras mediatesapoptosis in response to protein kinase C inhibition throughthe generation of reactive oxygen species. J Biol Chem 2000;275: 39001–39011.

27. Scaffidi C, Fulda S, Srinvivasan A, et al. Two CD95 (apo-1/fas)signalling pathways. EMBO 1998; 17: 1675–1687.

28. Zhuang S, Simon G. Peroxynitrite-induced apoptosis involvesactivation of multiple caspases in HL-60 cells. Am J Physiol2000; 279: C341–C351.

29. Wahl AF, Donaldson KL, Fairchild C, et al. Loss of normalp53 function confers sensitization to Taxol by increasing G2/Marrest and apoptosis. Nature Med 1996; 2: 72–79.

30. Wang T-H, Wang H-S, Ichijo H, et al. Microtubule-interferingagents activate c-Jun N-terminal kinase/stress-activated pro-tein kinase through both Ras and apoptosis signal-regulatingkinase pathways. J Biol Chem 1998; 273: 4928–4936.

31. Rakovitch E, Mellado W, Hall EJ, Pandita TK, Sawant S, GeardCR. Paclitaxel sensitivity correlates with p53 status and DNAfragmentation, but not G2/M accumulation. Int J Rad OncolBiol Phys 1999; 44: 1119–1124.

32. Reuter CWM, Morgan MA, Bergmann L. Targetting the Rassignaling pathway: A rational, mechanism-based treatment forhematologic malignancies? Blood 1996: 96: 1655–1669.

33. Gallangher AP, Burnett AK, Bowen DT, Darley RL. MutantRAS selectively promotes sensitivity of myeloid leukemia cells


J. H. Miller et al.

to apoptosis by a protein kinase C-dependent process. Canc Res1998; 58: 2029–2035.

34. Ewen ME. Relationship between Ras pathways and cell cyclecontrol. Prog Cell Cycle Res 2000; 4: 1–17.

35. Bergamaschi D, Ronzoni S, Taverna S, et al. Cell cycle per-turbations and apoptosis induced by isohomohalichondrin B(IHB), a natural marine compound. Br J Cancer 1999; 79: 267–277.

36. Zhang X, Minale L, Zampella A, Smith CD. Microfilamentdepletion and circumvention of multiple drug resistance bysphinxolides. Cancer Res 1997; 57: 3751–3758.

37. Shapiro GI, Harper JW. Anticancer drug targets: Cell cycle and

checkpoint control. J Clin Invest 1999; 104: 1645–1653.38. Sharma VK, Li B, Khanna A, Sehajpal PK, Suthanthiran M.

Which way for drug-mediated immunosuppression? Curr OpinImmunol 1994; 6: 784–790.

39. Fruman DA, Klee CB, Bierer BE, Burakoff SJ. Calcineurinphosphatase activity in T lymphocytes is inhibited by FK506and cyclosporin A. Proc Natl Acad Sci USA 1992; 89: 3686–3690.

40. Emmel EA, Verweij CL, Durand DB, Higgins KM, Lacy E,Crabtree GR. Cyclosporin A specifically inhibits function ofnuclear proteins involved in T cell activation. Science 1989;246: 1617–1620.


Peloruside A enhances apoptosis in H- ras -transformed cells and is cytotoxic to proliferating T...

Documents

Transcript of Peloruside A enhances apoptosis in H- ras -transformed cells and is cytotoxic to proliferating T...