(CANCERRESEARCH54,5387-5393,October15,1994J

ABSTRACT

The thlopurines 6-thioguanine (6TG) and 6-mercaptopurine (6MP) arecytotoxic to proliferating cells by a mechanism involving incorporationinto DNA via the purine salvage pathway, and resistance to these agentscan be conferred by lack of the salvage pathway enzyme hypoxanthineguanine phosphoribosyltransferase. However, human and murine hypoxanthine-guanine phosphoribosyltransferase-deficient leukemia cell lineshave been shown to respond to 6TG by growth arrest and differentiationby a mechanism apparently not involving incorporation of6TG into DNA.If so, leukemia cells resistant to 6MP should still respond to 6TG bygrowth arrest via an undescribed epigenetic mechanism. To test this,polyclonal 6MP-resistant variants were produced from three human Ieukemia cell lines, HL-60, U937, and CCRF-CEM. Treatment of both san

sitive and resistant celia with 6TG induced growth arrest The effect of6TG In the 6MP-sensitive HL-60 and U937 cells was associated withsi@cant loss of viability and DNA fragmentation. In contrast, the6TG-treated 6MP-resistant cells exhibited a slower decline in viability andno DNA fragmentation. To identify the mechanism by which 6TG mayInduce growth arrest, tRNA was Isolated from 6MP-reslstant cells cmitured for 48 h with 6TG. 6TG was found to be Incorporated into tRNAsnormally containing queuine In the anticodon wobble position. Thesestudies may provide a basis for the development of new therapeuticregimens for the treatment of leukemia.

INTRODUCTION

Despite much activity in the development and application of chemotherapeutic agents, new chemotherapeutic compounds and therapeutic strategies are needed that will provide effective treatment ofboth rapidly and slowly proliferating cancers, while minimizing undesirable side effects. One approach to this problem is to use existingchemotherapeutic agents in novel ways. We have used this approachto identify an apparent epigenetic mechanism of growth control thatmay provide the basis for a new application of known agents as wellas development of new compounds for cancer chemotherapy.

The thiopunnes 6TG3 and 6MP have been part of chemotherapeuticregimens for the treatment of childhood and adult leukemias since the1950s (1). These agents are cytotoxic to proliferating cells by amechanism which requires conversion to 6TG nucleotides via thepurine salvage pathway, followed by incorporation into DNA in placeof dGTP during the course of replication. The incorporation of thiopurines into DNA is not toxic per se but appears to prevent subsequentreplication. 6TG-containing DNA is a poor template for replication bybacterial and human DNA polymerases and a poor substrate forligation by several DNA ligases; hence, incorporation of 6TG intoDNA may interfere with replication of both the leading and laggingstrands of DNA (2).

Received 4/6/94; accepted 8/17/94.Thecostsof publicationof thisarticleweredefrayedin partby the paymentof page

charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1 This study was supported by American Cancer Society Grants CH-396 and DHP-37

(to R. W. T. and C. J. M.).2 To whom requests for reprints should be addressed, at Department of Surgery, Tulane

University School of Medicine, 1430 Tulane Avenue, New Orleans, LA 70i12-2699.3 The abbreviations used are: 6TG, 6-thioguanine; 6MP, 6-mercaptopurine; HGPRT,

hypoxanthine-guanine phosphoribosyltransferase; FBS, fetal bovine serum; CMF-PBS,calcium, magnesium-free phosphate-buffered saline; TCA, trichloroacetic acid.

The cytotoxic effects of 6MP and 6TG depend on the conversion ofthe purine bases to nucleotides by the salvage pathway enzymeHGPRT; absence of this enzyme is the best characterized mechanismof thiopurine resistance (reviewed in Ref. 3). Nevertheless, Frienderythroleukemia cells lacking HGPRT respond to treatment with 6TGin vitro by growth arrest and cellular differentiation, in the absence ofincorporation of 6TG nucleotides into DNA (4). A similar effect of6TG in HGPRT-deflcient HL-60 human acute promyelocytic leukemia cells has been shown to be induced by the free base 6TG ratherthan other metabolites (5).

We have demonstrated previously that 6TG becomes incorporatedinto the anticodon of specific tRNA isoacceptors of HGPRT-deficientHL-60 cells in place of the unusual base queuine (6). This highlymodified purine is found only in the anticodon wobble position oftRNA isoacceptors for asparagine, aspartic acid, histidine, and tyrosine. The modification, which is highly conserved evolutionarily,occurs via a base exchange mechanism catalyzed by the enzymequeuine tRNA-phosphoribosyltransferase (tRNA-GRT; tRNA-guanine:queuine tRNA ribosyltransferase; EC 2.4.2.29). We have shownthat treatment of HGPRT-deflcient HL-60 cells with 6TG in thepresence of additional exogenous queuine (1 pM supplementing the10—30nM present in culture medium) prevented the 6TG-inducedgrowth inhibition (7). This evidence implicates an epigenetic mechanism for growth arrest involving the incorporation of 6TG in place ofqueuine into tRNA.

We have proposed that if 610 causes growth arrest and differentiation in leukemia cell lines by a mechanism not involving HGPRT,then cells that have acquired resistance to 6MP due to lack of salvagepathway activity should nevertheless be sensitive to 6TG-inducedgrowth arrest. To test this, we produced 6MP-resistant variants ofthree human leukemic cell lines and assessed the effect of 6TG on

tRNA modification and cell proliferation. The results could haveimportant implications regarding the development of new therapeuticregimens for leukemia.

MATERIALS AND METhODS

Cell Llnes Thepromyelocyticleukemiacellline HL-60was obtainedfromDr. RObert Gab (National Cancer Institute, Bethesda MD); the HGPRTdeficient HL-60 variant was provided by Dr. Linda F. Thompson (ScrippsClinic and Research Foundation, La Jolla CA); U937 cells were provided byDr. Kathleen Clouse (Georgetown University, Washington DC); and CCRFCEM cells were Obtainedfrom the American Type Culture Collection (Rockvile MD). All cells were cultured in RPMI 1640 (GIBCO/BRL, Grand Island,NY) supplemented with 10% FBS (Hyclone Laboratories, Logan, Ui') at 37°Cin 7% CO2. Cells were subcultured every 5 to 7 days to a density of 2 X 10@cells/mi.

Selection of HGPRT-deflcient Cells. Cells were cultured in 25-cm@ tissue

culture flasks (Corning) in RPMI 1640 medium supplemented with 10% PBSand increasing concentrations of 6MP (0.4, 0.8, 2, 4, 10, 40, 80, and 350 @M;1.4, 2, and 3 mM).The concentration of 6MP in the cultures was increasedwhen the growth rate of cells in the presence of 6MP was comparable to thatof the parent cells. After several passages at the highest 6MP concentration,salvage pathway activity was assessed as the ability of cells to incorporate[3H]hypoxanthine into cellular nucleic acids.

5387

6-Thioguamne-induced Growth Arrest in 6-Mercaptopurine-resistant HumanLeukemia Cells1

Claudia J. Morgan,2 Rehan N. Chawdry, Andrea R Smith, Gina Siravo-Sagraves, and Ronald W. TrewynTulane University School of Medicine, Department of Surgery, New Orleans, Louisiana 70112 (C. J. MI, and The Ohio State University College of Medicine, Department ofMedical Biochemist,y (R. N. C., A. R. S., G. S-S., R. W. TI, ComprehensiveCancer Center (R. W. TI, Columbus, Ohio 43210

on April 17, 2018. © 1994 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

6TG ARREST IN 6MP-RESISTANT LEUKEMIA CELLS

[3HjHypoxanthlne Incorporation. Log phase cells (2 X 10@/assay) were

pelleted and resuspended in 1 ml RPMI 1640—10%FBS containing 1 @Ci[3H]hypoxanthine (1.66 GBq/mmol) and incubated for 1 h at 37°Cin 7% CO2.Cells were washed with 14 ml CMF-PBS. To the cell pellets was added 0.5 ml

10% TCA. Samples were mixed by vortexing, and 5 ml of 5% TCA wereadded. After incubation on ice for 30 minutes to 1 h, TCA-insoluble materialwas collected by vacuum on Whatman GF/A glass fiber filters, which were

rinsed with 5% TCA and 95% ethanol, and analyzed by scintillation counting

in Universol (ICN Biomedicals, Costa Mesa, CA).Cell Growth and Viability. Cells in late log phase (1—2X 10@cells/ml)

were subcultured in 25-cm3 flasks to 2 X 10@cells/mi in RPM! 1640-10%PBS. Eachday, 0.2 ml was removed,andthe cell densitywas assessed usinga hemacytometer. Viability was recorded as the percentage of cells able toexclude 0.2% trypan blue.

Analysis of DNA Fragmentation. Cells in late log phase were subculturedat 2 X 10@cells/ml in the presence or absence of 0.4 mM 6TG in RPM!1640—10% FBS for 24 h. Cells were then washed with CMF-PBS andresuspended in 1 ml T0.4NE [10 mMIris-Ha (pH 7.6), 0.4 MNaCI, and S m@EDTAJ. Samples were incubated for 2 h at 37°Cin the presence of sodiumdodecyl sulfate (0.4%) and proteinase K (100 @.tg/ml),followed by a 1.5-hincubation with ribonuclease A (50 @Wml),then an extraction twice with anequal volume of phenol, twice with an equal volume of phenol:chloroform

(1:1), and twice with chloroform. Nucleic acids were precipitated overnight at—70°Cwith 0.3 M NaOAc (pH 5) and 2.5 volumesethanol.Precipitateswerecollected by centrifugation for 20 mm at 12,000 rpm in a Sorvall microfuge,dried under vacuum, and resuspended in sterile H2Ofor measurement of yield(absorbance, 260 nm). DNA (5 @g)was separated by electrophoresis on a 2%agarose gel containing ethidium bromide.

Isolation of tRNA. Cells were collected and washed twice with coldCMF-PBS(pH 7.4). RNA was isolatedby a modificationof the procedureofBuck et a!. (8). Briefly, cells were lysed in reticulocyte standard buffer [0.01M Tris (pH 7.4), 0.01 M NaCI, and 1.5 mM MgC12] containing 0.5% Nonidet

P-40. Lysates were centrifuged in a Beckman Model 12 microfuge for 5 minat 5000 rpm, and the supernatants were centrifuged again for 15 min at 12,000rpm. Supernatants were extracted by mixing with an equal volume of phenol:water (5:1) and incubating for 20 mm on ice, and the phases were separated bycentrifugation at 12,000 rpm for 5 mm. The aqueous phase was extracted againwith an equal volume of phenol:water (5:1), and the combined organic phaseswere back extracted with one-half volume of reticulocyte standard buffer for10 mm on ice. To the combinedaqueousphaseswas added0.1 volume 20%(w/v) KOAc (pH 4.5) and LiCl (12 M)to a final concentration of 2 M.Afterstorage at 4°Covernight, samples were cleared by centrifugation (12,000 rpmfor 15 min at 4°C),and tRNA was precipitated from the supernatants byaddition of glycogen to 25 p@g/ml and ethanol to 70%. Samples were stored at

—70°Covernight, and the precipitated tRNA was collected by centrifugation(12,000 rpm for 25 mm at 4°C)and washed with 70% ethanol.

RPC-5 Chromatography and Dot Blotting. tRNA (5 to 15 @g)wasdeacylated for 30 mm at 37°Cin 0.1 MTris-HC1(pH 8.0), 50 mt@iKC1,and 10mM MgC12. Deacylated tRNA was applied to a 1 x 7 cm column of n-octylquaternary ammonium polychlorotrifluoroethylene-coated resin (RPC-5), (9)at 37°C using a Rainin A-60-S pump (Woburn, MA). tRNA was eluted with a

linear gradient (25 ml) of NaCl (0.475 to 1.5 M) in 10 mM NaOAc (jH 4.5).Fractions (95; 0.25 ml) were collected using a Gilson FC 203 fraction collectorinto tubes containing 50 @.dNaH2PO4 (pH 6.5).

For dot blotting, fractions were denatured in 1 Mdeionized glyoxal, 20 mMNaPO4 (pH 7.0) at 50°Cfor 1 h (10), and applied to a nylon membrane(Magna; MSI, Westborough MA) using a 96-well slot blot manifold(Schleicher and Schuell, Keene, NH). The membrane was baked at 80°Cfor 1 h, and the glyoxal was removed by washing in 20 mMTris-HCI (pH 8.0)at 100°Cfor 10 mm.

Hybridization with Oligonucleotide Probes. tRNA isoacceptors for histidine and tyrosine were detected by hybridization with specific DNA oligonucleotide probes as described (11). Briefly, 50 picomoles of probe (synthesized by Oligos, Etc., Wilsonville, OR) was 5' end-labeled using 20U T4polynucleotide kinase in 50 mM Tris-HCI (pH 7.5), 10 mM MgCl2, S mMdithiothreitol, 50 p@g/mlbovine serum albumin, and 150 @Ci[y-32P]ATP(>4000 Ci/rnmol)for 90 min at 37°C.Membraneswere prehybridizedwith 200g.@g/mldenatured, sonicated herring testicle DNA in 5X SSPE buffer [0.9 MNaCl, 50 mM NaPO4 (pH 7.4), and 5 mM EDTA], 50% formamide, 5X

Denhart's reagent, and 0.1% sodium dodecyl sulfate for 2 h at room temperature and hybridized overnight at room temperature with labeled probes andcarrier DNA in fresh hybridization solution. Filters were washed three times in5x SSPEbufferat 37°Candexposedto XAR-5X-OMATfilm(Kodak,Rochester, NY) at —70°Cwith an intensifying screen for less than 5 days.Autoradiographies were scanned using a Pharmacia LKB Laser DensitometerUltroscan XL (Piscataway, NJ).

RESULTS

Derivation of Purine Salvage-deficient Cell Lines. HGPRT-deficient HL-60 cells are growth inhibited and induced to differentiateby 6TG (4, 6, 7). 6TG and 6MP are thought to use a single pathway,requiring this enzyme, to cause cytotoxicity. We wished to determinespecifically whether cells resistant to 6MP retained responsiveness to6T0. Three human leukemia cell lines were used in these studies:HL-60, derived from a patient with acute promyelocytic leukemia(12); CCRF-CEM, a T-lymphoblast line derived from a patient withacute lymphoblastic leukemia (13); and U937, a monocyte-like cellline derived from a patient with histiocytic lymphoma (14). 6MPresistant variants of these cell lines were produced by culture in RPMI1640 containing 10% PBS and increasing concentrations of 6MP (Fig.1). After several passages in 6MP, resistant cells were cultured in theabsence of 6MP. To approximate the in vivo acquisition of resistanceto therapeutic 6MP, which may be polyclonal in origin, we did notattempt to isolate clonal cell lines from these cultures.

Purine salvage capability was assessed by measuring the incorporation of [3H]hypoxanthine into TCA-precipitable material (Table 1).Incorporation of [3H]hypoxanthine in 6MP-deficient HL-60 (HL-60―)cells was less than 1% of that in the 6MP-sensitive HL-60 cells. Thisvalue was very similar to that for incorporation of [3H]hypoxanthineinto nucleic acid by an independently obtained HGPRT-deficientHL-60 cell line (HL-6O@). Similarly, HGPRT activity in U937' cellswas less than 1% of the parent U937 cell line. In contrast, 6MPresistant CCRF-CEM (CEMd) cells retained a significant fraction

10

0.1E0@

@ 0.01

0.001

0.0001_lo 20 30

WeeksIn Passage1)937 CEM HL-60—e-—.-..o

40 50

Fig. 1. Selection of 6MP-resistant leukemia cell lines. CUltUresof HL-60 (Lx), U-937(0), andCCRF-CEM(0) cells were treatedwith increasingconcentrationsof 6MP inRPMI 1640 supplemented with 10% PBS. When cultures proliferated similarly in 6MPand in control medium, the concentration of 6MP was increased, as indicated.

5388

on April 17, 2018. © 1994 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

eo

fOO

J40

20

J40

20

0 1 2 3 4 5 e 7

6TG ARRESTIN 6MP-RESISTANTLEUKEMIACELLS

E

@0

0 1 2 3 4 5 6 7DayDay

I

Day Day

I--@

‘a-.

@EMd@, —.---

@9M+113---0---

0 1 2 3 4 5 6 7

DayDay

Fig. 2. Growth (left panels) and viability (right panels) of 6MP-resistant cell lines in the absence or presence of 6T0. Cells were subcultured from late log phase to a density of2 X 10' cells/mi into RPM! 1640—10%FBS in the absence or presence of 0.4 mt.i 6TG. Cell number was assessed by counting aliquots of each culture daily using a hemacytometer.Viability was evaluated using 0.2% trypan blue. These curves are representative of many repeated experiments. A and E, HL-6O@;B and F, HL@60d;C and G. U-937―;D and H, CEMd.

5389

Pt.eO(-) +e'rG---0---

I8

I3

on April 17, 2018. © 1994 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Table 1Incorporation of(3llJhypoxanthine―Cell

linerpmHL-6015085HL-6tFi24HL@60d116CEM46846CEMd3632U-93749466U@937d417

\40 ,, _____

“ Iaa7wt +81320 ‘@ ---0---

@@@‘0.@. ç@. e pr@t I

0 1 2 3 4 5 8 7

610 ARREST IN 6MP-RESISTANTLEUKEMIA CELLS

incorporation of [3H]hypoxanthine into nucleic acids by CEMd cellshas not declined to the level of HL-6(Y' or U937@'.

Effect of 6TG on Proliferation and Viabifity of 6MP-sensitiveand Resistant Cells To determine if 6TG could inhibit the growth ofleukemia cells resistant to 6MP, cells were subcultured from latelogarithmic growth into medium containing or lacking 0.4 m@ 6T0(Fig. 2). In the presence of 6TG, the growth of HL-6O@(Fig. 2B),U937d (Fig. 2C), and CEM@'(Fig. 2D) cells was arrested within 72 to96 h of treatment. The proliferation of HL-60―cells was affectedsimilarly to that of the HGPRT-deficient cell line HL-60 (Fig. 2A).In contrast, 6MP-resistant cells cultured in medium alone proliferatednormally. The viability of 6MP-resistant cells cultured in 6T0 remained high (>60%) for at least 3 days, declining gradually thereafterin the j@j,@d (Fig. 2F) and CEM―(Fig. 2H) cells but remaining at 60to 70% through day 7 in the U937@'cells (Fig. 2G). In numerousexperiments, day 3 viabilities of these cells in 6TG remained high, and

a Cells (2 X i0@) were incubated for 1 hour with 1 pCi [3Hjhypoxanthine in RPM

1640-10%FBS. Cells were disrupted, and nucleic acids were precipitated with 10% TCAas described in “Materialsand Methods.―Shown are rpm per 2 x i0@ cells from onerepresentative experiment.

(8%) of the HGPRT activity of the parent cell line. The salvagedeficient phenotypes are stable over many passages in 6MP-freemedium (data not shown); currently, cells are passaged in 1 mM 6MPat 4-week intervals. However, despite continued culture in 6MP,

D

‘I, @&40Moof*@1

I 0

I. I@8.40M +113@ ---0---

@‘

20

20

0 1 2 3 ‘@ I ‘6DayDay

II

DayDay

F

80 ‘@ [àB,I*toorSml@ .‘ 0‘@40 \

0@@

20 ____

0 1 2 3 4 5 8 7

DayDay

Fig. 3. Growth (left panels) and viability (right panels) of parent cell lines in the absence or presence of 6TG. Cells were cultured and evaluated as described in Fig. 2. A and D,HL-60; B and E, U-937; C and F, CEM.

5390

on April 17, 2018. © 1994 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Table 2 Effect of queuineon cell growth andviability―Control6TG6TG/OHL-6O'@

U-937―CEMd9.0―(91)

10.9(83)17.2(96)4.4(39)

3.6(58)8.5(88)7.2(81)

6.4(80)13.3(96)

6T0 ARR@T IN 6MP-RESISTANT LEUKEMIA CELLS

day 5 viabiities varied from 10 to 80%. In the absence of 6TG,6MP-resistant cells remained greater than 90% viable until saturationdensities were reached (days 5—6).

To compare the effect of 6TG on 6MP-resistant cells with that oncells with an intact purine salvage pathway, the parent cell lines werecultured in parallel in the presence and absence of 0.4 mt@i6TG (Fig.3). The proliferation of the parent cell lines in the absence of 6T0 wasvery similar to that of the untreated 6MP-resistant variants. However,in the presence of 6TG, little accumulation and extensive cell deathwas observed. The decline of cell viability of HL-60 and U937 cells(Fig. 3, D and E) in 6TG was precipitous, failing below 10% by 48 h.

The response of the CCRF-CEM cells to 6T0 differed qualitativelyfrom that of the HL-60 and U937 cells. Although little accumulationof CCRF-CEM cells was observed in 6TG (Fig. 3C), viability dedined more slowly (Fig. 3F). Nevertheless, these cells were moresensitive to 6TG cytotoxicity than their 6MP-resistant counterparts.

Induction of DNA Fragmentation by 6TG. The loss of cellviability in response to 6TG may result from cell death due toapoptotic or necrotic mechanisms. To characterize further the differences in the action of 6TG on the parent and 6MP-resistant cells, cellswere cultured for 24 h in the presence of 6TG and evaluated for theinduction of apoptotic cell death as revealed by DNA fragmentation(Fig. 4). DNA prepared from 6TG-treated salvage-competent HL-60and U937 cells was fragmented into nucleosomal ladders character

istic of apoptosis within 24 h of treatment. Neither the CCRF-CEMcells nor any of the 6MP-resistant cells analyzed after 24 h of culturein 6TG exhibited DNA fragmentation. Because significant cell deathwas not observed in the salvage-competent CCRF-CEM cells untilday 3, we analyzed DNA prepared from 3-day cultures of these cells,as well as HL-6(V cells and the 6MP-resistant HL-60, U937, andCEM cells. No DNA fragmentation was observed, despite viabiitiesas low as 50% (data not shown).

Incorporation of 6TG into tRNA. The site of action of 6TG incells which cannot incorporate this purine analogue into DNA isunclear. We have shown that in HL-60 cells, 6TG incorporates intothe anticodon of tRNA isoacceptors in place of the highly modifiedpurine queuine and that 6TG-induced growth inhibition in HL-60cells was prevented by addition of exogenous queuine (6, 7). In

a Cells (2 x iO:I/ml) were incubated for 4 days in control medium or in medium

containing 6TG (400@ in the absence or presence of exogenous queuine (1 @M,added daily). Shown are the average of duplicate counts from representative experiments.

Call number (X 10@); numbers in parentheses are viability as determined by trypanblue exclusion.

6MP-resistant cells, exogenous queuine was also able to reverse theeffect of 6TG (Table 2). To determine whether 6TG was incorporatedinto these tRNA isoacceptors in 6MP-resistant cells, we prepareddeacylated tRNA from cells cultured for 48 h in the absence orpresence of 0.4 mM 6TG. tRNA was fractionated by reverse-phasechromatography on an RPC-5 column, and column fractions were dotblotted onto nylon membranes. tRNA isoacceptors for histidine, tyrosine, asparagine, and aspartic acid were identified by consecutivehybridization of each blot with antisense oligonucleotides derivedfrom the 3' ends of known human and murine tRNA sequences (11).Elution profiles of tRNA isoacceptors for histidine and tyrosine generated by laser densitometry of the resulting autoradiographies areshown in Fig. 5. A shift of the tRNA profiles of treated cells to highersalt concentrations for all four tRNA isoaccepting species indicatesthe incorporation of 6T0 into tRNA (11). The extent of 6TG incorporation into tRNAs depended upon the cell line and tRNA analyzed.Incorporation of 6TG into tRNA isoacceptors for histidine, tyrosine,and asparagine (data not shown) was observed in all three 6MPresistant cell lines and was most evident in CEMd cells. In contrast,less incorporation of 6T0 into aspartate isoacceptors was observed inthe HL-60@',CEM―,or U937―cells (data not shown).

DISCUSSION

The chemotherapeutic agents 6MP and 6T0 have been presumed tocause cytotoxicity by a common mechanism involving incorporationof 6TG nucleotides into DNA via the purine salvage pathway. However, the ability of 6TG to inhibit the growth of HGPRT-deficientFriend erythroleukemia cells without incorporation into DNA suggests that an alternative, epigenetic mechanism of 6TG-induced arrestexists that may be operational in cells lacking purine salvage capability. If so, then cells that had acquired resistance to 6MP, often vialoss of purine salvage activity, should retain sensitivity to 6TG. Totest this, we determined the effect of 6TG on proliferation in 6MPresistant leukemia cells of the myeloid, monocytic, and lymphoidlineages. In each lineage, resistance to 6MP was associated with lossor diminution of purine salvage capability, indicated by inability toincorporate [3H]hypoxanthine into DNA. Treatment of 6MP-resistantcells with 6T0 resulted in growth inhibition, which differed from thetoxic effect of 6TG in the parental, purine salvage competent cell linesin several ways: (a) while 6TG treatment induced growth arrest within

24 h in the parent cells, the kinetics of arrest was delayed in theresistant cells, becoming apparent only after 2 to 4 days of culture; and(b) growth inhibition in the HL-60 and U937 cell lines was associatedwith a precipitous loss of cell viability, whereas viability remainedhigh for as much as a week in the 6MP-resistant cells treated with6TG. 6TG arrest in the purine salvage-competent HL-60 cell line wasassociated with DNA fragmentation. Treatment of salvage-competentU937 cells for 24 h also resulted in DNA fragmentation. The natureof the gradual loss of viability of the CCRF-CEM and 6MP-resistantcells is less clear. DNA fragmentation was not readily identifiable inthe salvage-competent CCRF-CEM cells or in the 6MP-resistantHL-60, U937, and CEM cells treated for 3 days with 6TG, despite

1 2 3 4 5 6 7 8 9 10 11 12

V

- +++ - + + - ++

Fig. 4. DNA fragmentation in 6TG-treated leukemia cell lines. Cells were subculturedfrom late log phase to a density of 2 X i0@cells/mI in RPMI 1640—10%PBS in theabsence or presence of 0.4 nmi 6TG. After 24 h, DNA was isolated, separated byelectrophoresis on a 2% agarose gel, and stained with ethidium bromide. Lanes I and 9,i-kilobase size markers (GIBCOIBRL); Lanes 2 and 3, HL-60; Lane 4, HL-60―;Lane 5,HL-60@; Lanes 6 and 7, CEM; Lane 8, CEMd; Lanes 10 and 11, U-937; Lane 12, U-937';÷,culturestreatedwith6TG.

5391

i@-II...

on April 17, 2018. © 1994 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

A f@

C@ —

6T0 ARREST IN 6MP-RESISTANT LEUKEMIA CELLS

D , •1I

1@ @1it'i

Ej:1@!\@@@

F

IL;@

Fig. 5. Incorporation of 6TG into tRNA of Ieukemia cells. Cells were subcultured as described inFig. 2. After 48 h, tRNA was isolated, descylated,and analyzedby RPC-5chromatographyand dotblotting as described in “Materialsand Methods.―Dot blots were probed by successive hybridizationwith oligonucleotide probes directed against histidine (leftpanels) and tyrosine (rightpanels) tRNAisoacceptors. Profiles were generated by laser densitometty of the resulting autoradiographies.A andD, HL-60'@;B and E, U-937―;Cand F, CEM―.Solidlines, control cultures; dashed lines, culture in6TG-containing medium.

viabilities as low as 66%. We also assessed the extent of apoptotic celldeath in treated and untreated cultures stained with acridine orange;the results were consistent with those obtained by analysis of DNAfragmentation (data not shown). Studies are in progress to clarify therole of programmed cell death in the loss of viability observed in6MP-resistant cell lines treated with 6TG.

The concentration of 6TG used in these studies, 0.4 m@s,is higherthan required for killing of purine salvage-competent cells in vitro butconsiderably less than the highest concentration of 6MP to whichthese cells are known to be resistant (2 mM). It is difficult to relatethese figures to therapeutic 6TG levels. Oral or i.v. 6TG is commonlyadministered at a dose of 55 to 100 mg/m2 (0.5 to 1 mmol in anaverage size patient) every 12 h (15, 16). Peak plasma thioguaninelevels are highly variable, in the range of 0.03 to 0.94 ,.LMin one study(16). In children receiving comparable levels of 6MP, peak erythrocyte accumulation of 6TG nucleotides is highly variable, ranging from100 to 600 pmol per 8 X 108 cells (17). It is unclear how these valuesreflect concentrations of 6TG at cellular targets. Thus, it is difficult tocompare the concentration of 6TG used here with actual therapeuticlevels. We note, however, that the concentration of 6TG used in theseexperiments is similar to the concentration of 6MP required to inhibitby 60% mixed lymphocyte responses in cells obtained from HGPRTdeficient Lesch-Nyhan patients (18).

Both CEM and CEMd cells responded to treatment with 6MP and6T0 differently from the myeloid and monocytic leukemia cells.CEM@'cells resistant to high concentrations of 6MP retained significant purine salvage capability. This may reflect the greater use of the

de novo pathway of purine synthesis,rather than purine salvage,thought to be characteristic of T lymphocytes.

We demonstrated previously that 6TG became incorporated intotRNA isoacceptors of the queuine-containing family in HGPRTdeficient HL-60 cells and that 6TG-induced growth inhibition in thesecells could be prevented by the addition of exogenous quewne (6, 7).Since queuine has been found only in the wobble position of specifictRNAs (19), 6TG-induced growth inhibition in these cells may occurvia a tRNA-mediated mechanism. Analysis of tRNA isoacceptors forhistidine and aspartic acid indicates that 6TG is variably incorporatedinto these tRNAs in the 6MP-resistant cells within 48 h. The extent ofincorporation of 6TG in these cells should depend on several factors.tRNA isoacceptors for aspartate are the preferred substrate for tRNAORT (20) and thus are expected to be relatively more queuinemodified than the other queuine-containing tRNAs. Since queuineincorporation into tRNA is irreversible, only unmodified and newlysynthesized tRNAs would be expected to be substrates for 6TGincorporation. Therefore, preexisting tRNA isoacceptors for asparticacid would be least sensitive to 6TG modification, and histidine tRNAisoacceptors, which are poorer substrates for tRNA-GRT, would beexpected to incorporate 6TG relatively well. Among newly synthesized tRNAs, aspartic acid isoacceptors would be expected to bepreferentially 6TG modified. However, since the incorporation of6TG into tRNA is a reversible process, incorporated 6TG shouldultimately be replaced by queuine from the serum. Such transient 6TGmodification has been observed in HL-60 cells (6). The reason forthe greater 6TG modification in CEM―cells, compared to HL@60dand

5392

on April 17, 2018. © 1994 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

6T0 ARREST IN 6MP-RESISTANT LEUKEMIA CELLS

U937d cells, is under investigation. CEM cells may be queuine hypomodified to a greater degree (see below) or may be more activelysynthesizing new tRNAs.

Treatment of the 6MP-resistant cells with 6TG in the presence ofexogenous queuine prevented the 6TG-induced growth arrest (Table2), suggesting that in 6MP-resistant cells, 6TG-induced growth arrestmay occur by a tRNA-mediated mechanism. In view of the relativelyhigh viability of the CCRF-CEM cells early after arrest in 6TG, wespeculate that early arrest in these cells may also involve incorpora

tion of 6TG into tRNA.The role of queuine modification in tRNA is still unclear. Abundant

evidence suggests that queuine modification is important for normalcell function (reviewed in Ref. 20). tRNAs in normal tissues tend tobe completely queuine-modified, despite the fact that eukaryotes mustobtain queuine from the diet and gut flora. tRNAs of numerous humanand experimental tumors have been shown to be hypomodified forqueuine, and introduction of excess exogenous queuine can causetumor regression in animal models. In human leukemias and lymphomas, there is a strong correlation between the extent of hypomodification and prognosticstage(21). tRNA isolatedfrom a T-lymphoblastic lymphoma was found to be highly undermodified, consistentwith our observations of the relatively efficient incorporation of 6TGinto the CEM―cells. Interestingly, in patients with such B-cell malignancies as CLL stage A and hairy cell leukemias, queuine hypomodification was not restricted to rapidly proliferating leukemias,suggesting that it is not generally a reflection of cell proliferation. Invitro, numerous transformed cell lines are queuine hypomodified (20,22), and treatment of methylcholanthrene-initiated primary hamsterfibroblasts with 7-methylguanine, which inhibits tRNA-GRT, generated cells exhibiting anchorage-independent growth and queuinehypomodification (23).

The molecular consequences of 6TG incorporation into tRNA areunknown. tRNA'@'@@rmodified with 6-thioqueuine can act as an ambersuppressor, suggesting that queuine modification may be importantfor translation fidelity (24). Queuine-containing tRNAs exhibit lesspreferential codon recognition than G-containing tRNAs (25), andspecific proteins are not expressed in queuine-deficient D. discoideumand Escherichia coli, suggesting that queuine modification may playa role in the posttranscriptional regulation of protein expression (26,27). Such a regulatory mechanism would be most important for theregulation of synthesis of proteins that are rapidly synthesized anddegraded, such as those involved in cellular signaling. Interferencewith the appropriate expression of these proteins could cause thegrowth arrest and perhaps the gradual cell death observed here. Inaddition, we and others have observed induction of differentiation inHGPRT-deflcient leukemia cell lines treated with 6TG, an effectwhich could also be triggered by alterations in the appropriatesynthesis of regulatory proteins.

These data suggest that growth arrest can be induced in leukemiacells by an epigenetic mechanism, possibly involving alterations intRNA queuine modification. Although the use of 6TG for this purposein purine salvage-competent individuals would be complicated by theincorporation of 6TG into DNA, 6TG may be an effective therapeuticstrategy in patients that have acquired resistance to 6MP. Moreover,the identification of purine analogues capable of acting as substratesfor tRNA-GRT but not HGPRT may provide a new class of chemotherapeutic agents which are not DNA synthesis-directed. The effectiveness of such agents may not be restricted to rapidly proliferatingtumors but may also be effective in the treatment of slowly proliferating, queuine hypomodified leukemias.

5393

ACKNOWLEDGMENTS

We thank Marsha Stalker and Carol LeMay for their secretarial assistancewith this manuscript.

REFERENCES

1. Burchenal, J. H., Murphy, M. L, Ellison, R. R., ci al. Clinical evaluation of a newantimetabolite, 6-mercaptopurine, in the treatment of leukemia and allied diseases.Blood, 8: 965—999,1953.

2. Ling, Y-H, Chan, J. Y., Beattie, K. L, and Nelson, J. A. Consequences of 6-thioguanine incorporation into DNA on polymerase ligase, and endonuclease reactions.Mol. Pharmacol., 42: 802—807,1992.

3. Lennard, L The clinical pharmacology of6-mercaptopurine. Eur. J. Oin. Pharmacol.,43: 329—339,1992.

4. Gusella, J. F., and Houseman, D. Induction of erythroid differentiation in vitro by

purines and purine analogs. Cell, 8: 263—269,1976.5. Ishiguro, K., Schwartz, E. L, and Sartorelli, A. C. Characterization of the metabolic

forms of6-thioguanine responsible for cytotoxicity and induction ofdifferentiation ofHL-60 acute promyelocytic leukemia cells. J. Cell Physiol., 121: 383—390,1984.

6. Kretz, K. A., Katze, J. R., and Trewyn, R. W. Guanine analog-induced differentiationof human promyelocytic leukemia cells and queuine hypomodification of transferRNA. Mol. Cell. Biol., 7: 3613—3619,1987.

7. French, B. T., Patrick, D. W., Graver, M. R., and Trewyn, R. W. Queuinc, a tRNAanticodon wobble base, maintains the proliferative and pluripotent potential of HL-60cells in the presence ofthe differentiating agent 6-thioguanine. Proc. NatI. Acad. Sd.USA, 88: 370—374,1991.

8. Buck, M., Connick, M., and Ames, B. N. Complete analysis of tRNA-modifiednucleosides by high-performance liquid chromatography: the 29 modified nucleosidesof Salmonella typhimurium and Escherichia coli tRNA. Anal. Biochem., 129: 1—13,1983.

9. Kelmers, A. D., and Heatherly, D. E. Columns for rapid chromatographic separationof small amounts of tracer-labeled transfer ribonucleic acids. Anal. Biochem., 44:486—495,1971.

10. Thomas, P. S. Hybridization of denatured DNA fragments transferred to nitrocellulose. Proc. Natl. Acad. Sd. USA, 77: 5201—5205,1980.

11. Utz, E. D., Siravo-Sagraves, 0. M., Trewyn, R. W., and Morgan, C. J. Detection ofhuman tRNAs with antisense oligonucleotides. Anal. Biochem., 216: 110—117,1994.

12. Collins, S. J., Gab, R. C., and Gallagher, R. E. Continuous growth and differentiation of human myeloid leukaemic cells in suspension culture. Nature (Lond.), 270:347—349,1977.

13. Foley, G. E., Lazarus, H., Farber, S., Uzman, B. G., Boone, B. A., and McCarthy,

R.E.Continuouscultureofhumanlymphoblastsfromperipheralbloodofa childwithacute leukemia. Cancer (Phila.), 18: 522—529,1965.

14. Sunstrom, C., and Nilsson, K. Establishment and characterization of a human histiocytic lymphoma cell line (U-937). mt. J. Cancer, 17: 565—577,1976.

15. Avvisati, G., Petsi, M. C., Petrucci, T., Falconi, E., Tirindelli, M. C., and Mandelli, F.Treatment of recurrent promyelocytic leukemia with a combination regimen utilizingamsacrine, cytosine arabinoside and 6-thioguanine (AAT). Haematologica, 74:279—282,1989.

16. Brox, L W., Birkett, L, and Belch, A. Clinical pharmacology of oral thioguanine inacute myelogenous leukemia. Cancer Chemother. Pharmacol., 6: 35—38,1981.

17. Leonard, L, and Lilleyman, J. S. Variable mercaptopunne metabolism and treatment

outcome in childhood Iymphoblastic leukemia. J. Clin. Oncol., 7: 1816—1823,1989.18. Szawlowski, P. W., Al-Safi, S. A., Dooley, T., and Maddocks, J. L. Azathioprine

suppresses the mixed lymphocyte reaction of patients with Lesch-Nyhan syndrome.Br.J. Clin.Pharmacol.,20: 489—491,1985.

19. Kasai, H., Ohashi, Z., Harada, F., ci aL Structure of the modified nucleoside 0isolated from E. coli transfer ribonucleic acid, 7-(4,5-cis-dihydmxy-1-cyclopenten3-ylaminomethyl)-7-deazaguanosine. Biochemistry, 14: 4198—4208,1975.

20. Katze, J. R., Beck, W. T., Cheng, C. S., and McCloskey, J. A. Why is tumor tRNAhypomodifiedwithrespectto 0 nucleoside?CancerRes.,84: 146—159,1983.

21. Emmerich, B., Meinhardt, G., Maubach, P. A., Zubrod, H., Rastetter, J., and Kersten,w. ModificationoftRNAanditsapplicationfortheassessmentofprognosis,stateofdifferentiation, and clonality in human leukemias and lymphomas. Haematol. BloodTransfus., 30: 241-250, 1987.

22. Nishimura, S. Structure, biosynthesis, and function of queuosine in transfer RNA.Prog.NucleicAcidRes.MoLBiol.,28: 49—80,1983.

23. Muralidhar, G., Ochieng, J., and Trewyn, R. W. Altered queuine modification oftransfer RNA involved in the in vitro transformation of Chinese hamster embryo cells.CancerRes.,49:7110—7114,1989.

24. Shindo-Okada, N., Akimoto, H., Nomura, H., and Nishimura, S. Recognition of UAGtermination codon by mammalian tyrosine tRNA containing 6-thioqueuine in the firstposition of the anticodon. Proc. Jpn. Acad., 61: 94—98,1985.

25. Meier, F., Suter, B., Grosjean, H., Keith, 0., and Kubli, E. Queuosine modification ofthe wobble base in tRNA-His influences “invivo―decoding properties. EMBO J., 4:823—827,1985.

26. Schachner, E., Achhoff, H-J., and Kersten, H. Specific changes in lactate levels,lactate dehydrogenase patterns and cytochrome b559 in Dictyostelium discoideumcaused by queuine. Eur. J. Biochem., 139: 481—487, 1984.

27. Janel, G., Michelsen, U., Nishimura, S., and Kersten, H. Queuosine modification intRNA and expression of the nitrate reductase in Escherichia coli. EMBO J., 3:

1603—1608,1984.

on April 17, 2018. © 1994 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1994;54:5387-5393. Cancer Res   Claudia J. Morgan, Rehan N. Chawdry, Andrea R. Smith, et al.   6-Mercaptopurine-resistant Human Leukemia Cells6-Thioguanine-induced Growth Arrest in

  Updated version

  http://cancerres.aacrjournals.org/content/54/20/5387

Access the most recent version of this article at:

   

   

   

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

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/54/20/5387To request permission to re-use all or part of this article, use this link

on April 17, 2018. © 1994 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from