ARTICLE IN PRESS

123456789

10111213141516171819202122232425262728293031323334353637383940414243444546474849

50

www.elsevier.com/locate/cellbi

Cell Biology International -- (2005) ---e---

ECTEDPROOF

Metabolic alterations in K562 cells exposed to taxol andtyrphostin AG957: 1H NMR and biochemical studies

Arno Knijn a, Fabrizia Brisdelli b, Amalia Ferretti c, Egidio Iorio c,Donatella Marcheggiani b, Argante Bozzi b,*

a Data Management Service, Istituto Superiore di Sanita, 00161 Rome, Italyb Department of Biomedical Sciences and Technologies, University of L’Aquila, 67100 L’Aquila, Italy

c Laboratory of Cell Biology, Istituto Superiore di Sanita, 00161 Rome, Italy

Received 20 January 2005; revised 13 June 2005; accepted 13 July 2005

Abstract

K562 cells exposed for 3 h to taxol or taxol plus tyrphostin AG957 exhibited a significant variation in the concentration of the

water-soluble metabolites glutathione, myo-inositol and phosphorylcholine, as evaluated by 1H NMR up to 72 h incubation in drug-free medium. Cells treated with both drugs showed an increase of glutathione and glutathione reductase at 24 h and a sharp decreaseof myo-inositol between 8 and 24 h. Phosphorylcholine increased at 8 h both in taxol and taxol plus AG957-treated cells, which was

then abruptly inverted to a significantly lower concentration at 24 h, subsequently increasing again to values higher than those foundin taxol-treated and control cells. All the above reported effects were lacking in cells exposed to AG957 alone. These modifications,despite the enhancement of the overall apoptotic cascade in taxol plus AG957-treated cells, can be related to the activation of

cellular detoxification mechanisms, to the correct osmolarity maintenance, and to alterations of phospholipid metabolism.� 2005 International Federation for Cell Biology. Published by Elsevier Ltd. All rights reserved.

Keywords: Glutathione; Phosphorylcholine; Myo-inositol; Taxol; Tyrphostin AG957

5152535455565758596061626364656667

UNCORR

1. Introduction

Paclitaxel (taxol�) is a powerful antineoplastic agentespecially against advanced ovarian cancer, breastcancer and malignant melanoma (Holmes et al., 1991).Its cytotoxicity is due to stabilization of microtubuleseven under conditions that normally promote theirdisassembly (Nicolaou et al., 1993). Furthermore, taxolhas been reported to block progression in the G2/Mphase of the cell cycle (Hei and Hall, 1993). Neverthe-less, the multiple mechanisms of its cytotoxicity have not

* Corresponding author. Tel.: C39 0862 433472; fax: C39 0862

433433.

E-mail address: bozzi@cc.univaq.it (A. Bozzi).

YCBIR1387_proof � 1

1065-6995/$ - see front matter � 2005 International Federation for Cell B

doi:10.1016/j.cellbi.2005.07.004

68

yet been fully clarified (Wang et al., 2000). Taxol-resistance, on the other hand, has been attributed toa number of mechanisms including induction of themultidrug-resistant phenotype and over-expression of P-glycoprotein (Horwitz et al., 1993). In addition, changesin the composition of and mutations in b-tubulinisotypes have been identified in cells resistant to taxol(Ranganathan et al., 1998). As far as the effects of taxolon K562 cells concerns, the literature contains contrast-ing data: K562 cells have been described as resistant totreatment over 300 nM taxol (Jamieson et al., 1999), butare also sensitive to treatment with 20e100 nM taxol(Olah et al., 1996).

It has been reported that human melanoma cellsexposed to taxol exhibit decreased levels of glucose-1,6-bisphosphate and fructose-1,6-bisphosphate and

3 August 2005 � 1/8

iology. Published by Elsevier Ltd. All rights reserved. 6970

T

ARTICLE IN PRESS

7172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126

127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182

2 A. Knijn et al. / Cell Biology International -- (2005) ---e---

UNCORREC

increased detachment of phosphofructokinase from thecytoskeleton, with a consequent overall ATP decrease(Glass-Marmor and Beitner, 1999).

Tyrphostin AG957 (AG957) is a selective inhibitor ofthe p210BCR-ABL tyrosine kinase activity and its additionto taxol-treated cells enhances the sensitivity of thesecells to apoptosis induced by taxol (Beran et al., 1998).The molecular mechanisms underlying the effect ofAG957 on the apoptotic cascade induced by taxol inK562 cells require elucidation. Bcr/abl expression inleukemic cells is known to exert a potent effect againstapoptosis induced by chemotherapeutic drugs througha mechanism that is still under investigation. Theexpression of bcr/abl kinases exclusively in neoplasticcells makes these enzymes attractive targets for diseasespecific therapy (Daley et al., 1990). AG957 has showna high affinity for the selective inhibition of thetransforming p210bcr-abl activity with respect top140c-abl, although epidermal growth factor receptoractivity was also blocked. The drug is small enough topenetrate cells and reach the p210bcr-abl proteins in thecytoplasm. AG957 has been reported to be non-competitive by Svingen et al. (2000), but competitive(Sun et al., 2001) with regard to ATP. Tyrphostins seemto act synergistically with cytotoxic agents by decreasingresistance of cells towards drug-induced apoptosis. Inparticular, AG957 sensitised K562 cells committed toapoptosis by taxol (Jamieson et al., 1999). Recently, ithas been demonstrated that AG957 (at longer exposuretimes than in our experimental protocols) alters phos-phorylation of the Akt kinase and BAD, withoutaffecting the phosphorylation of Bcl-2 or BAX (Urbanoet al., 2002).

1H NMR spectroscopy studies usually report theappearance of intense signals due to isotropically mobilelipid chains in the spectra of apoptotic cell samples(Hakumaki and Kauppinen, 2000). We previouslyfocused on the formation of NMR-visible mobile lipids(ML) during apoptosis triggered by cells treated with taxol

Nomenclature

AG957 tyrphostin AG957ATP adenosinetriphosphateBSO L-buthionine-(S,R)-sulfoximineDAG diacylglycerolDTPA diethylenetriaminepentaacetic acidGSH reduced glutathioneGSSG oxidized glutathionePBS phosphate buffered salinePC phosphatidylcholinePCho phosphorylcholineSMIT NaC/myo-inositol co-transporterTMA trimethylammonium

YCBIR1387_proof � 13

EDPROOF

or taxol plus AG957. Our results showed a two-stepprocess characterised by an early stage increase inML (3e8 h), separated from another retarded increase(48e72 h) by an intermediate plateau at 24e48 h(Brisdelli et al., 2003).

We now report data on K562 cells which, after a 3 hexposure to taxol or taxol plus AG957 were analyzed upto 72 h mainly with regard to 1H NMR signals and somebiochemical determinations. The results displayed, aspreviously shown for ML (Brisdelli et al., 2003),a peculiar behaviour for spectral signals due to water-soluble metabolites. In particular, three signals arisingfrom myo-inositol, glutathione and the trimethylammo-nium group, mainly due to phosphorylcholine (PCho),were well enough resolved to allow the determination thetime-courses of these metabolites’. Other NMR signals inthe spectra were either due to inert metabolites (likelysine), already studied (ML), or inadequately resolvedto allow for their accurate quantification (the glutamineand glutamate signals). Furthermore, enzymatic activi-ties related to glutathione content were determined.

The kinetics of the metabolites under investigationexhibited peculiar patterns either individually andespecially in relation with the protocol of cell treatment.Interestingly, after 24 h of incubation, upon the initialpulse with taxol plus AG957, a significant increase ofglutathione level was detected. These modificationscould be related to the activation of cellular detoxifica-tion mechanisms, to alterations of lipid metabolism andpossibly to the correct osmolarity maintenance.

2. Materials and methods

2.1. Cell culture and treatment

K562 human erythroleukemic cells were obtainedfrom the American Type Culture Collection (ATCC)and were free of microorganisms and mycoplasma. Cellswere maintained in exponential growth at 37 �C ina humidified atmosphere containing 5% CO2, using anRPMI 1640 medium supplemented with 10% heat-inactivated foetal calf serum, 2 mM L-glutamine, 50 U/ml penicillin and 50 mg/ml streptomycin. The cells(4! 105 cells/ml) were exposed for 3 h to 200 nM taxol,either alone or in combination with 50 mM tyrphostinAG957. After two washes, the cells were resuspendedand incubated in drug-free medium for up to 72 h. Tostudy the involvement of glutathione synthesis on theobserved increase of GSH concentration, K562 cellswere also incubated with 20 mM L-buthionine sulphox-imine (BSO), a g-glutamylcysteine synthetase inhibitor.Cell counting and viability were determined at varioustimes by trypan blue exclusion method. Apoptotic cellswere identified by fluorescence microscopy upon doublestaining with ethidium bromide and acridine orange.

August 2005 � 2/8

T

ARTICLE IN PRESS

183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238

239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294

3A. Knijn et al. / Cell Biology International -- (2005) ---e---

UNCORREC

2.2. Glutathione and related enzymatic activitiesdetermination

Total and oxidized glutathione were measured as de-scribed by Anderson (1985). The glutathione reductase,peroxidase and transferase activities of cell extracts wereassayed spectrophotometrically as reported (CarlbergandMannervik, 1975; Paglia and Valentine, 1967; Habiget al., 1974). Protein concentration was determined aspreviously reported (Bradford, 1976).

2.3. NMR spectra and data analysis

One-dimensional (1D) 1H NMR experiments onintact K562 cell preparations (30e40! 106 in 0.5 mlPBS/D2O) were carried out at 400 MHz on a BrukerAVANCE 400 WB spectrometer (9.4 T), at 25 �C, using60 � pulses preceded by 1 s presaturation for water signalsuppression (spectral width 10 ppm), the total measure-ment time being 17 min for 320 scans. Experiments wereperformed spinning the 5 mm NMR tubes and cells didnot settle out during the measurements, since the samplewas still homogeneous afterwards. In effect, three spectracould be consecutively measured for 5 min each, withoutchanging pattern between one and another. Assignmentof peaks was performed as described elsewhere (Ferrettiet al., 2002). Spectral data analyses were performed asdescribed before (Brisdelli et al., 2003). Briefly, theBruker WIN-NMR software package was used to modelthe peaks of interest by deconvolution, assumingLorentzian line shapes. When spectra of treated andcontrol cell samples were matched, the lysine signal at1.7 ppm, the creatine signal at 3.04 ppm and somesmaller signals around 3.3e3.4 ppm and 4.0e4.2 ppmdue to (ipo-)taurine, creatine and unknown metabolitesappeared to be invariant. Amongst these signals, theresonance at 1.7 ppm due to the b and d methyleneprotons of lysine was chosen as the most suitable to beused as an internal reference. The creatine signal sufferedfrom overlap of the nearby varying unassigned signal at2.94 ppm, the resonances around 3.3e3.4 ppm were ofa too low intensity, as well as those around 4.0e4.2 ppmwhich also underwent the influence of water-suppression.On the other hand, the lysine peak at 1.7 ppm could beestimated in these spectra with a standard deviation ofabout 14% of the value estimated for the peak area itself(data not shown), as followed from quantifications withthe AMARES Maximum Likelihood non-linear least-squares program with the application of proper priorknowledge (Vanhamme et al., 1997).

2.4. Statistical analysis

Differences were considered statistically significant ifP-values determined by Student’s t-test were !0.05.

YCBIR1387_proof � 1

EDPROOF

3. Results

K562 cell samples were measured by 1H NMR at 3, 8,24, 48 and 72 h after or without pulsed treatment for 3 hwith 200 nM taxol, and alone or in combination with50 mM AG957. We focused our attention on signalsarising from myo-inositol (at 3.52 ppm due to its H1 andH3), glutathione (at 3.79 ppm due to its glycine residue),and the so-called ‘‘choline peak’’ (a complex signal duemainly to PCho at 3.22 ppm and, to much lower extentsglycerophosphorylcholine at 3.23 ppm and free cholineat 3.20 ppm). Examples of 1H NMR spectra at 24 h aregiven in Fig. 1. From these spectra, the area of the threeaforementioned signals was quantified in order to studythe time-courses of these metabolites.

3.1. Glutathione

From the glutathione concentration in Fig. 2, it canbe appreciated that in control cells its time-course is

Fig. 1. 1H NMR spectra of intact K562 cell preparations. Cells were

exposed for 3 h to treatments. After two washes, the cells were resus-

pended in drug-free medium and then observed after 24 h. (A) Control

cells; (B) cells treated with 200 nM taxol; (C) cells treated with

a combination of 200 nM taxol and 50 mM tyrphostin AG957. Some

peak assignments are given: CH2 (b, d) from lysine, NC(CH3)3 from

choline-containing compounds (‘Cho’-peak), H1 and H3 from myo-

inositol and the glycine residue from glutathione.

3 August 2005 � 3/8

ARTICLE IN PRESS

295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350

351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406

4 A. Knijn et al. / Cell Biology International -- (2005) ---e---

UNCORREC

largely stable after a mild, non-significant (PO 0.13),elevation at 8 h. Treatment with taxol did not induceany remarkable deviation from the glutathione level incontrol cells, and only after 48 h was the concentrationof glutathione slightly lower but within one standarddeviation limit (PO 0.14). Addition of AG957 to taxol-treated K562 cells induced a significant accumulation ofglutathione from 24 to 72 h with respect to treatmentwith taxol alone (P! 0.0024, !0.021, and !0.024).Compared to control cells, the glutathione level wasonly significantly elevated at 24 h (P! 0.0091, O0.47,O0.066 at 24, 48 and 72 h, respectively).

This increase, in parallel determined by the enzymaticmethod (Anderson, 1985), was observed both in reduced(GSH) and oxidized (GSSG) glutathione (w3-fold withrespect to untreated cells; Fig. 3). The GSH/GSSG ratiothat reflects the redox state of the cell remainedpractically unchanged.

3.2. Effect of L-buthionine-(SR)-sulfoximine

K562 cells treated with taxol plus AG957 wereincubated with L-buthionine-(SR)-sulfoximine (BSO),a selective inhibitor of g-glutamylcysteine synthetase,either for 3 h concomitantly with the other two drugsand for 24 h. The cells were analyzed with 1H NMR.Fig. 4 shows that within 3 h from the addition of BSO,a large fall (�79%) in the increment of the glutathionesignal, that reached C122% (with respect to control) incells exposed to taxol plus AG957. One day afteraddition of BSO together with taxol plus AG957, theglutathione NMR resonances were completely abolished(data not shown). These results suggest that AG957 may

0,30 24 48 72

0,8

1,3

1,8G

luta

thio

ne [

A. U

.]

Hours

Fig. 2. Time-course of the relative peak areaof the signal due to the glycine

residue from glutathione in K562 cells treated for 3 h with 200 nM

taxol (d-d), taxol plus 50 mM tyrphostin AG957 (e e : e e)

or untreated control ($ $ $ $ B $ $ $ $). The spectra were

recorded after 3, 8, 24, 48 and 72 h of incubation in drug-free

medium. Values are in arbitrary units with respect to the peak area

due to CH2 (b, d) from lysine and the results represent the mean

(GSD) values obtained from three independent experiments.

YCBIR1387_proof � 13

TEDPROOF

initially enhance the activity of g-glutamylcysteinesynthetase in K562 cells.

3.3. Myo-inositol

Fig. 5 shows the intracellular concentration of myo-inositol at different time-points for the various treat-ments. Control cells exhibited some accumulation ofmyo-inositol at 24 h, followed by a progressive lineardecrease to the initial level at 72 h. Treatment with taxoldid not affect the cellular content of myo-inositol at all,while the addition of AG957 together with taxol,induced an abrupt fall of its concentration between 8and 24 h (P! 0.0029). Consequently, myo-inositolsteadily accumulated to reach the level of control cellsbetween 48 and 72 h and further increased to a concen-tration significantly higher than that of control cells(P! 0.0079). At 72 h, in cells treated with AG957alone, the level of myo-inositol was lower than in cellstreated with taxol plus AG957, but higher than in cellstreated with taxol alone (data not shown, result ofa single experiment). This behaviour shows the impor-tance of monitoring the metabolic effects during a longtime-interval after cell exposure to the drug(s).

A

0

100

200

300

400

500

B

0

150

300

450

600

750

0 24 48 72

Hours

0 24 48 72Hours

[GSH

] (

o

f co

ntro

l)[G

SSG

] (

o

f co

ntro

l)

Fig. 3. Time-course of reduced (A) and oxidized (B) glutathione levels

variations in taxol (d-d) and taxol plus AG957-treated (e e: e e)K562 cells determined by enzymatic recycling assay. The values of

GSH and GSSG are the mean (GSD) percent of control (100%). The

means of three independent experiments are shown.

August 2005 � 4/8

T

ARTICLE IN PRESS

407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462

463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518

5A. Knijn et al. / Cell Biology International -- (2005) ---e---

UNCORREC3.4. Phosphorylcholine (PCho)

The levels of intracellular PCho (Fig. 6) increased incontrol cells at 24 h similarly to myo-inositol butsubsequently decreased sharply to about its initial level.Taxol-treated cells displayed an initial PCho concentra-tion similar to that of untreated cells, and then showedthe same increase at 24 h but in a linear mode witha non-significant lower intermediate concentration at8 h with respect to control (PO 0.14). Again, PChodecreased to its initial level with a significantly highervalue at 48 h with respect to control (P! 0.011). Thesimultaneous addition of AG957 and taxol to cellsincreased PCho at 8 h similar to cells treated with taxolalone, which was then abruptly inverted inducinga significant lower concentration at 24 h with respectto 8 h (P! 0.015) and taxol-alone treated cells at 24 h(P! 0.001). K562 cells maintained this level of PChofor the following 24 h, comparable with control cells.However, PCho then accumulated for the next 24 h andreached a level, somewhat but not significantly higherthan taxol-treated and control cells. One experiment oncells treated with AG957 alone showed at 72 h anintermediate level of PCho between cells treated withtaxol alone and cells treated with taxol plus AG957(data not shown).

Fig. 4. 1H NMR spectra of intact K562 cell preparations. Cells were

exposed for 3 h to treatment and subsequently observed after 3 h. (A)

Control cells; (B) cells treated with a combination of 200 nM taxol and

50 mM tyrphostin AG957; and (C) cells treated with a combination of

200 nM taxol, 50 mM tyrphostin AG957 and 20 mM L-buthionine-

(SR)-sulfoximine. Some peak assignments are given: CH2 (b, d) from

lysine, NC(CH3)3 from choline-containing compounds (‘Cho’-peak),

H1 and H3 from myo-inositol and the glycine residue from

glutathione.

YCBIR1387_proof � 1

EDPROOF

3.5. Glutathione-dependent enzymatic activities

No significant alterations in the activity of glutathi-one peroxidase and transferase were observed in taxoland taxol plus AG957-treated K562 cells in the time-interval of 3e72 h (data not shown). In contrast, theglutathione reductase activity appeared affected by bothtreatments, taxol alone and taxol plus AG957 (Fig. 7).Taxol-treated cells exhibited a significant increase tow60% higher than control cells at 3 h (P! 0.05) and8 h (P! 0.005), followed by a sharp drop to basal levelsbetween 8 and 24 h. The addition of tyrphostin AG957,

0,30 24 48 72

0,8

1,3

1,8

myo

- in

osit

ol [

A. U

.]

Hours

Fig. 5. Time-course of the relative peak area of the signal due to the H1

and H3 from myo-inositol in K562 cells exposed for 3 h to 200 nM

taxol (d-d), taxol plus 50 mM tyrphostin AG957 (e e : e e) oruntreated control ($ $ $ $ B $ $ $ $). The spectra were recorded after 3,

8, 24, 48 and 72 h of incubation in drug-free medium. Values are in

arbitrary units with respect to the peak area due to CH2 (b, d) from

lysine and the results represent the mean (GSD) values obtained from

three independent experiments.

0,70 24 48 72

1,2

1,7

2,2

2,7

‘Cho

’-pe

ak [

A.U

.]

Hours

Fig. 6. Time-course of the relative peak area of the signal due to

NC(CH3)3 from choline-containing compounds in K562 cells treated

for 3 h with 200 nM taxol (d-d), taxol plus 50 mM tyrphostin

AG957 (e e : e e) or untreated control ($ $ $ $ B $ $ $ $). The

spectra were recorded after 3, 8, 24, 48 and 72 h of incubation in drug-

free medium. Values are in arbitrary units with respect to the peak area

due to CH2 (b, d) from lysine and the results represent the mean

(GSD) values obtained from three independent experiments.

3 August 2005 � 5/8

T

ARTICLE IN PRESS

519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574

575576577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630

6 A. Knijn et al. / Cell Biology International -- (2005) ---e---

UNCORREC

in combination with taxol, induced a progressive in-crease, even if milder, of glutathione reductase activityfrom 8 h, reaching maximal values (56% greater thancontrol cell activity, P! 0.05) at 48 h after exposure tothe drugs.

4. Discussion

The results show that a major variation in solublemetabolites (glutathione, myo-inositol and TMA peaks)occurs when K562 cells are treated with taxol plusAG957, in particular 24 h after treatment.

The time-course of apoptosis reported in the previouspaper (Brisdelli et al., 2003) showed an earlier formationof some apoptotic markers (chromatin condensationand phosphatidylserine externalization), while othermarkers, such as the loss of mitochondrial membranepotential, became evident after 24 h. The G2M blockwas retarded and cell death, estimated on the basis ofhypodiploidy, was low in K562 cells treated with bothagents, while DNA laddering was detectable only after72 h of treatment.

An interesting feature observed under these condi-tions (cells treated 3 h with taxol plus AG957),concomitantly with the above reported apoptoticpattern, was the relevant increase of reduced glutathioneformation exhibited at 24 h, followed by a slow declineat 72 h. This result could imply that the cells are tryingto react to the AG957-mediated induction of apoptosiswith a ‘‘paradoxical’’ activation of detoxification mech-anisms, at a rather intermediate stage (24e48 h), whichslowly attenuates at longer incubation times (72 h).Experiments performed with BSO, to block the synthesisof glutathione, indicated that at 24 h the cells treatedwith taxol plus AG957 exhibited an early apoptotic level

00 24 48

0,03

72

0,06

0,09

0,12

glut

athi

one

redu

ctas

e ac

tivi

ty(m

mol

/min

/mg

prot

ein)

Hours

Fig. 7. Time-course of glutathione reductase activity in K562 cells

exposed for 3 h to 200 nM taxol (d-d), taxol plus 50 mM tyrphostin

AG957 (e e : e e) or untreated control ($ $ $ $ B $ $ $ $). Theactivities were evaluated after 3, 8, 24, 48 and 72 h incubation in drug-

free medium. The results represent the mean (GSD) values obtained

from three independent experiments.

YCBIR1387_proof � 13 A

EDPROOF

of 40% (against a 15% level of necrotic cells). Additionof BSO to drugs exposed cells decreased early apoptosis(20%), but an increment in late apoptosis and necrosis(47%). However, monitoring by 1H NMR the presenceof glutathione signals in K562 cells exposed to taxol plusAG957 and then incubated without or with BSO allowsone to observe a sharp reduction at 3 h, and a totaldisappearance at 24 h, of cellular glutathione.

Activities of glutathione peroxidase and glutathionetransferase, measured by biochemical methods, indrugs-treated cells were ineffective between 3 and 72 h.In contrast, glutathione reductase activity showed anenhanced level from 24 to 48 h. These results suggestthat, under our experimental conditions, an additionalmechanism of action of AG957 based on the increase ofg-glutamylcysteine synthetase and reductase activitiesand displaying its maximal effect 24 h after drugexposure.

Haematopoietic tissues express low levels of activecystathionase and g-glutamyltransferase (g-GT); thusinhibition of some steps of the glutathione cycledecreases intracellular glutathione (Bratton et al.,2000). Depletion of intracellular glutathione results inhigher levels of apoptosis (Jungas et al., 2002), becauseits physiological concentration inhibits sphingomyeli-nases, ceramide formation and subsequent activation ofeffector caspase-3 and/or 7 (Liu et al., 1998). Asa consequence, the increase (at 24 h) of glutathionelevels in K562 cells retarded the final apoptotic stepsinduced by taxol plus AG957. Moreover, we recentlyshowed that K562 cells exhibited high levels ofglutathione even in basal conditions, i.e. without drugtreatment (Luzi et al., 2004).

It has been reported, however, that exposure to BSOand inhibition of g-glutamylcysteine synthetase depletesglutathione concentration and concomitantly increasessphingomyelinase activity in Molt-4 cells (Liu et al.,1998). It is likely that a similar process also operated inK562 cells exposed to BSO, thus contrasting with theincreasing glutathione concentration induced by AG957in combination with taxol.

Decline in the effect of glutathione increase bysimultaneous treatment with both drugs, after 48e72 hcould be explained by considering that the intracellularconcentration of AG957 is decreased at these longincubation times, either by extrusion and/or by meta-bolic conversion, while taxol remains inside the cellsassociated with microtubules. As far as myo-inositol isconcerned, its pattern of modulation in K562 cellstreated with taxol plus AG957 follows that of the earlymarkers of apoptosis (Brisdelli et al., 2003) and could berelated with the necessity to maintain the normal cellvolume. The majority of cells counteract tonicityalterations by activating specific metabolic and trans-port mechanisms that restore normal cell volume.Among these mechanisms, a key role is exerted by

ugust 2005 � 6/8

T

ARTICLE IN PRESS

631632633634635636637638639640641642643644645646647648649650651652653654655656657658659660661662663664665666667668669670671672673674675676677678679680681682683684685686

687688689690691692693694695696697698699700701702703704705706707708709710711712713714715716717718719720721722723724725726727728729730731732733734735736737738739740741742

7A. Knijn et al. / Cell Biology International -- (2005) ---e---

UNCORREC

transport of inorganic ions and of compatible osmolytessuch as myo-inositol. The accumulation of cellular myo-inositol in hypertonic state(s) is mediated by the NaC/myo-inositol co-transporter (SMIT). Because proteinkinase C activation inhibits SMIT (Preston et al., 1995)and in hypertonic cells tyrosine kinase inhibitors, liketyrphostins, are reported to reduce the activity of SMIT(Atta et al., 1999), the myo-inositol decrease observed at24 h in K562 cells exposed to taxol plus AG957 could bea consequence of a partial inhibition of SMIT caused byAG957. Besides, myo-inositol is known to be aninhibitor of the MAPK signal transduction pathwayinvolved in induced apoptotic signals. The decrease inmyo-inositol could thus stimulate the cells to progress inthe apoptotic cycle. On the whole, the synthesis andturnover pathways of myo-inositol (also in relation toapoptosis) are complex due to its involvement insignaling and cellular protection, as myo-inositol ismore than a simple osmolyte (Ross, 2000). In fact,alterations in myo-inositol concentration may also affectthe biosynthesis and turnover of some phospholipids(Ferretti et al., 2002).

It is well known that a reduction in phosphatidyl-choline (PC) synthesis can trigger apoptosis, as demon-strated by experiments with several inhibitors of keyenzymes of the PC synthetic pathways, and by cholinedeprivation (Cui and Houweling, 2002). The high levelsof PCho found in some tumour cells indicate a markedPC synthesis, while arrest of growth and apoptosis canbe followed by a reduction of PCho (Bogin et al., 1998).In our experiments, the PCho-levels detected by NMRin apoptosis could be a balance between the decrease ofPC synthesis and the production of PCho together withDAG due to the activity of phospholipases activated bythe apoptotic steps, in addition to the modulation ofPCho during the phases of cell growth. In K562 cellstreated with taxol plus AG957, PCho signals begin todecrease at 8 h, and by 24 h there is a significantreduction of PCho probably due to the arrest of PCsynthesis. In the same time-interval, an increase ofphosphatidylserine externalization and a peak of apo-ptotic nuclei were observed (Brisdelli et al., 2003). Thesubsequent PCho modulations can be regarded asa consequence of the complex mechanisms in PChomeostasis following the induction of apoptosis.

Jurkat cells treated with taxol show two differentapoptotic pathways (Moos and Fitzpatrick, 1998).These authors hypothesized that the normal phosphor-ylation of Bcl-2 triggers one mechanism, but hyper-phosphorylation and inactivation of c-Raf-1, p44Erkand of other tyrosine kinases, promote the secondapoptotic pathway. These processes cause the appear-ance of two populations of cells with distinctivecytological and biochemical traits. The first one iscomposed of viable cells converted into ‘‘metastable’’cells; the formation of this metastable population is an

YCBIR1387_proof � 1

EDPROOF

intermediate step along an ordered apoptotic cascadeand is caspase-3 independent. The second populationdisplays the formation of apoptotic cells, which followsthe loss of the combined protective effects of Bcl-2, Raf-1, c-Lck, is caspase-3 dependent and results in DNAfragmentation (Moos and Fitzpatrick, 1998).

These patterns are probably common to the mecha-nism of action of taxol and also occur in K562 cells. Thesimultaneous addition of taxol plus AG957 (inhibitor oftyrosine kinase activity) rapidly diverts the taxol-treatedcells in the first parallel step of conversion of viable cellsin the metastable population, without Bcl-2 phosphor-ylation. Subsequently, the dilution and elimination ofAG957 from cells decrease the dual effect. In K562 cellstreated with taxol alone, both the distinct apoptoticfluxes are probably present in parallel. The apoptoticmarkers previously studied (Brisdelli et al., 2003),together with the metabolites and the enzymaticactivities investigated in the present paper, seem toindicate this dual pathway, and follow the fluxesthrough the above reported mechanisms.

In conclusion, we report the ability to detect andquantify by 1H NMR in intact cells the time-course ofalteration in the concentrations of the water-solublemetabolites, glutathione, myo-inositol and phosphoryl-choline. In agreement to previously published results(Au et al., 1998), the data show that the toxicity of taxol,alone or in combination with AG957, can be envisagedas a complex multifactorial phenomenon which needsfurther studies to be fully elucidated.

Acknowledgments

This work was supported by grants from PRIN 2004and from C.N.R. Target Project on Biotechnology(contract no. 01045.49). The authors are indebted withDr. Franca Podo and Prof. Roberto Strom for thefruitful suggestions and stimulating discussions through-out the manuscript preparation. We thank Mr. MassimoGiannini for excellent technical assistance and mainte-nance of NMR equipment.

References

Anderson ME. Determination of glutathione and glutathione disulfide

in biological samples. Methods Enzymol 1985;113:548e55.

Atta MG, Dahl SC, Kwon HM, Handler JS. Tyrosine kinase

inhibitors and immunosuppressant perturb the myo-inositol but

not the betaine cotransporter in isotonic and hypertonic MDCK

cells. Kidney Int 1999;55:956e62.

Au JL-S, Li D, Gan Y, Gao X, Johnson AL, Johnson J, et al.

Pharmacodynamics of immediate and delayed effects of paclitaxel:

role of slow apoptosis and intracellular drug retention. Cancer Res

1998;58:2141e8.

Beran M, Cao X, Estrov Z, Jeha S, Jin G, O’Brien S, et al. Selective

inhibition of cell proliferation and BCR-ABL phosphorylation

in acute lymphoblastic leukemia cells expressing Mr 190,000

3 August 2005 � 7/8

ARTICLE IN PRESS

743744745746747748749750751752753754755756757758759760761762763764765766767768769770771772773774775776777778779

780781782783784785786787788789790791792793794795796797798799800801802803804805806807808809810811812813814815816817

8 A. Knijn et al. / Cell Biology International -- (2005) ---e---

ORREC

BCR-ABL protein by a tyrosine kinase inhibitor (CGP-57148).

Clin Cancer Res 1998;4:1661e72.

Bogin L, Papa MZ, Polak-Charcon S, Degani H. TNF-induced

modulations of phospholipids metabolism in human breast cancer

cells. Biochim Biophys Acta 1998;1392:217e32.

Bradford MM. A rapid and sensitive method for the quantitation of

microgram quantities of protein utilizing the principle of protein-

dye binding. Anal Biochem 1976;72:248e54.Bratton SB, Lau SS, Monks TJ. The putative benzene metabolite 2,3,5-

tris(glutathion-S-yl) hydroquinone depletes glutathione, stimulates

sphingomyelin turnover, and induces apoptosis in HL-60 cells.

Chem Res Toxicol 2000;13:550e6.Brisdelli F, Iorio E, Knijn A, Ferretti A, Marcheggiani D, Lenti L,

et al. Two-step formation of 1H NMR visible mobile lipids during

apoptosis of paclitaxel-treated K562 cells. Biochem Pharmacol

2003;65:1271e80.

Carlberg I, Mannervik B. Purification and characterization of

flavoenzyme glutathione reductase from rat liver. J Biol Chem

1975;250:5475e80.Cui Z, Houweling M. Phosphatidylcholine and cell death. Biochim

Biophys Acta 2002;1585:87e96.

Daley GQ, Van Etten RA, Baltimore D. Induction of chronic

myelogenous leukemia in mice by the P210bcr/abl gene of the

Philadelphia chromosome. Science 1990;247:824e30.

Ferretti A, D’Ascenzo S, Knijn A, Iorio E, Dolo V, Pavan A, et al.

Detection of polyol accumulation in a new ovarian carcinoma cell

line, CABA 1: a 1H NMR study. Br J Cancer 2002;86:1180e7.Glass-Marmor L, Beitner R. Taxol (paclitaxel) induces a detachment

of phosphofructokinase from cytoskeleton of melanoma cells and

decreases the levels of glucose 1,6-bisphosphate, fructose 1,6-

bisphosphate and ATP. Eur J Pharmacol 1999;370:195e9.

Habig WH, Pabot MJ, Jarkoby WB. Glutathione S-transferase. The

first enzymatic step in mercapturic acid formation. J Biol Chem

1974;249:7130e9.Hakumaki JM, Kauppinen RA. 1H NMR visible lipids in the life and

death of cells. Trends Biochem Sci 2000;25:9e14.

Hei TK, Hall EJ. Taxol, radiation and oncogenic transformation.

Cancer Res 1993;53:1368e72.Holmes FA, Walters RS, Theriault RL, Forman AD, Newton LK,

Raber MN, et al. Phase II trial of taxol, an active drug in the

treatment of metastatic breast cancer. J Natl Cancer Inst 1991;83:

1797e805.

Horwitz SB, Cohen D, Rao S, Ringel I, Shen HJ, Yang CP. Taxol:

mechanism of action and resistance. [Review]. J Natl Cancer Inst

Monogr 1993;15:55e61.Jamieson L, Carpenter L, Biden TJ, Fields AP. Protein kinase Ci

activity is necessary for Bcr-Abl-mediated resistance to drug-

induced apoptosis. J Biol Chem 1999;274:3927e30.

Jungas T, Motta I, Duffieux F, Fanen P, Stoven V, Ojcius DM.

Glutathione levels and BAX activation during apoptosis due to

UNC

YCBIR1387_proof � 13

TEDPROOF

oxidative stress in cells expressing wild-type and mutant cystic

fibrosis transmembrane conductance regulator. J Biol Chem 2002;

277:27912e8.

Liu B, Andrieu-Abadie N, Levade T, Zhang P, Obeid LM,

Hannun YA. Glutathione regulation of neutral sphingomyelinase

in tumor necrosis factor-alpha-induced cell death. J Biol Chem

1998;273:11313e20.

Luzi C, Brisdelli F, Cinque B, Cifone G, Bozzi A. Differential

sensitivity to resveratrol-induced apoptosis of human chronic

myeloid (K562) and acute lymphoblastic (HSB-2) leukemia cells.

Biochem Pharmacol 2004;68:2019e30.

Moos PJ, Fitzpatrick FA. Taxanes propagate apoptosis via two cell

populations with distinctive cytological and molecular traits. Cell

Growth Differ 1998;9:687e97.

Nicolaou KC, Riemer C, Kerr MA, Rideout D, Wrasidlo W. Design,

synthesis and biological activity of protaxols. Nature 1993;364:

464e6.

Olah E, Csokay B, Prajda N, Kote-Jarai Z, Yeh YA, Weber G.

Molecular mechanisms in the antiproliferative action of taxol and

tiazofurin. Anticancer Res 1996;16:2469e78.

Paglia BE, Valentine WN. Studies on the quantitative and qualitative

characterization of erythrocyte glutathione peroxidase. J Lab Clin

Med 1967;1:158e69.Preston AS, Yamauchi A, Known HM, Handler JS. Activators of

protein kinase A and of protein kinase C inhibit MDCK cell

myo-inositol and betaine uptake. J Am Soc Nephrol 1995;6:

1559e64.Ranganathan S, Benetatos CA, Colarusso PJ, Dexter DW,

Hudes GR. Altered beta-tubulin isotype expression in paclitaxel-

resistant human prostate carcinoma cells. Br J Cancer 1998;77:

562e6.

Ross BD. Real or imaginary? Human metabolism through nuclear

magnetism. IUBMB Life 2000;50:177e87.

Sun X, Layton JE, Elefanty A, Lieschke GJ. Comparison of effects of

the tyrosine kinase inhibitors AG957, AG490, and STI571 on

BCR-ABL-expressing cells, demonstrating synergy between AG490

and STI571. Blood 2001;97:2008e15.

Svingen PA, Tefferi A, Kottke TJ, Kaur G, Narayanan VL,

Sausville EA, et al. Effects of the bcr/abl kinase inhibitors

AG957 and NSC 680410 on chronic myelogenous leukemia cells

in vitro. Clin Cancer Res 2000;6:237e49.Urbano A, Gorgun G, Foss F. Mechanisms of apoptosis by the

tyrphostin AG957 in hematopoietic cells. Biochem Pharmacol

2002;63:689e92.

Vanhamme L, van den Boogaart A, van Huffel S. Improved method

for accurate and efficient quantification of MRS data with use of

prior knowledge. J Magn Reson 1997;129:35e43.

Wang T-H, Wang H-S, Soong Y-K. Paclitaxel-induced cell death.

Where the cell cycle and apoptosis come together. Cancer 2000;88:

2619e28.

August 2005 � 8/8