[CANCER RESEARCH 49. 5555-5560. October 15. 1989]

Induction of Erythroid Differentiation in K562 Cells by Inhibitors of InosineMonophosphate Dehydrogenase1

John Yu,2 Victor Lemas, Theodore Page, James D. Connor, and Alice L. Yu

Department of Molecular and Experimental Medicine, Research Institute of Scripps Clinic, La Jolla, California 92037 fj. Y., V. L.], and Department of Pediatrics,University of California at San Diego, La Jolla, California 92093 [T. P., J. D. C., A. L. Y. ]

ABSTRACT

The effects of three inhibitors of inosine monophosphate (IMP) de-hydrogenase on a human erythroleukemic cell line, K562, were studied.Following incubation with these inhibitors, K562 cells underwent differentiation and accumulated hemoglobins. The induction of hemoglobinaccumulation was dose dependent; maximum induction was observed at100, 25, and 3 MM,respectively, for ribavirin, tiazofurin, and mycophen-olic acid. The induction was associated with reduction of intracellularGTP content and was blocked by adding guanosine within 24 h afteradding inducer. The effective dose for half-maximum induction by ribavirin was 3 times less than that for 50% inhibition of K562 proliferation;however, for tiazofurin and mycophenolic acid, it closely approximatedthe concentrations which suppressed cellular proliferation. Ribavirin wassequestered preferentially inside the K562 cells, and the induction byribavirin had a greater than 30-fold increase in hemoglobin. Studies withisoelectric focusing, globin chain analyses, and immunochemical assaysindicated that both Ay and Gy were detected and that the hemoglobinproduced in the ribavirin-treated cells consisted of approximately 60%fetal hemoglobin and its acetylated equivalents. The adult-type a globinwas found, while no ,t globin chains were demonstrated. Thus, accumulation of fetal hemoglobin and production of a globin chain in ribavirin-treated cells are different from the pattern of hemoglobins induced byhernia.

INTRODUCTION

Ribavirin ( 1- ß-D- ribofuranosyl -1,2,4 - triazole - 3 -carboxam-ide), tiazofurin (2-/3-D-ribofuranosylthiazole-4-carboxamide),and mycophenolic acid (1-3) were shown to be inhibitors forIMP dehydrogenase and growth of tumor cells. Their primaryeffect on purine metabolism appeared to be a marked depletionof guanine nucleotide pools (3-6). In vivo studies have indicatedthat ribavirin and other inhibitors of IMP dehydrogenase mightinduce hematological changes. Administration of ribavirin inhigh doses frequently results in a reversible anemia, and theeffect is dose and time dependent (7). Tiazofurin was also foundto cause myelosuppression in Phase I clinical trials (8). Inaddition, it was found that several inhibitors of IMP dehydrogenase were potent inducers of myeloid maturation in the HL-60 cell line (9-11). These inhibitors were also reported topromote the terminal maturation and inhibit the self-renewalof normal myeloid progenitors (12, 13). Several previous studiesexplored the effects of ribavirin on mature erythrocytes (7); sofar, there have been few data on its effect on immature erythroidprecursor cells in vitro or in vivo. In this study we report theeffects of ribavirin, tiazofurin, and mycophenolic acid on theinduction of erythroid differentiation of a human erythroleukemic cell line, K562 (14). The types of hemoglobin produced

Received 6/13/88; revised 11/4/88, 4/17/89, 7/7/89; accepted 7/20/89.The costs of publication of this article were defrayed in part by the payment

of page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1This work was supported in part by grants from the NIH (DK 37039 and CA40186). This is Publication 5408-BCR from the Research Institute of ScrippsClinic, La Jolla, CA.

2To whom requests for reprints should be addressed at Department of Basic

and Clinical Research. BCR3. Research Institute of Scripps Clinic, 10666 NorthTorrey Pines Road. La Jolla. CA 92037.

by incubation with ribavirin were also analyzed. Preliminaryreports of this investigation have been presented (15, 16).

MATERIALS AND METHODS

Cell Culture. Stock cultures of K562 (American Type Culture Collection, Rockville, MD) were grown in RPMI 1640 medium supplemented with 50 lU/ml of penicillin, 50 Mg/ml of streptomycin, and15% fetal calf serum (Hyclone Laboratories, Inc., Logan, UT), aspreviously described (17). Cells were caused to differentiate by additionof inducers to final concentrations as specified in the text. Ribavirin(Viratek, Inc., Covina, CA), tiazofurin (gift from Dr. Roland Robins),and mycophenolic acid (Sigma, St. Louis, MO) were diluted to 10 mg/ml in RPMI 1640 medium and stored at 4°C.Experimental cultureswere grown for 3 to 4 days to densities less than 2x10'' cells/ml. Cell

numbers were counted with a hemocytometer. Cell viability was determined by trypan blue dye exclusion. The number of cells containinghemoglobin was assayed by benzidine staining (18). The clonogenicassay was performed using 0.3% agar in RPMI 1640 medium containing 15% fetal calf serum (19, 20). The amount of hemoglobin accumulation was assayed using a benzidine colorimetrie method (21).

GTP and ATP Assay. For the determination of intracellular GTPand ATP in K562 cells, samples were prepared as we described previously (22, 23). Briefly, K562 cells were incubated with ribavirin ortiazofurin. After a specified time of incubation, approximately 2 to 5 x10" cells were removed, counted with a hemocytometer, and centrifuged

for 5 min at 1000 rpm. The resulting cell pellets were immediatelyextracted with 100 ^1 of l M ice-cold HC1O4 at 4°C.After vortexing,samples were spun at 4°Cat 10,000 rpm, and the supernatant was

collected. The supernatant was immediately neutralized with 4 M KOH,and the pH was carefully titrated to 7.4 with l M KOH. The finalvolume was recorded and samples were quickly frozen and stored at-20°C until final assay for ATP and GTP concentrations by high-

pressure liquid chromatography as previously described (23).Isoelectric Focusing. Approximately 1 x IO8cells of ribavirin-induced

K562 were washed twice in phosphate-buffered saline (pH 7.4). Thenone volume of packed cells was mixed with one volume of deionizedwater and one-half volume of CC14. After mixing for 10 min andcentrifugation for another 10 min at 4°C,KCN was added to the

supernatant collected to a final concentration of 100 ^g/ml, and thesample was stored at —¿�20°C.Isoelectric focusing of the samples was

performed using 1% agarose containing ampholites at pH 6 to 8 inLKB Multiphor II apparatus at a constant 15 watts and 4°Cfor 20 to

30 min. The agarose gel was then fixed in 10% trichloroacetic acid andstained with benzidine for the presence of hemoglobins.

Globin Chain Analysis. After isoelectric focusing of cell lysates,individual gel slices were excised from gel without prior fixation andthen stored in 20 u\ of ß-mercaptoethanolat 4°C.These samples were

denatured with 100^1 of buffer containing 6.7 Murea, 8.3% acetic acid,8.3% /3-mercaptoethanol, and 0.3 mg/ml of pyronin Y and boiled for 2min prior to subsequent electrophoresis in 12% polyacrylamide gelcontaining 6 M urea, 2% Triton, and 5% acetic acid (24). Electrophoresis was carried out in 5% acetic acid for 17 h at a constant current of10 mA. Gels were stained using Coomassie blue. In some cases, theglobin chain composition of the induced K562 cells was further analyzed by a combination of Western-blotting and immunochemical assayas follows. The urea-Triton-acid polyacrylamide gel prepared as described was electroblotted to nitrocellulose paper (25) and probedsequentially with a 1:2000 dilution of rabbit anti-human hemoglobin For hemoglobin A serum and biotin-conjugated anti-rabbit antibodies

5555

Research. on August 17, 2021. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from

EFFECTS OF IMP DEHYDROGENASE INHIBITORS ON K562 CELLS

(Vector Laboratories, Burlingame, CA). The respective globin chainson the nitrocellulose paper were then analyzed with an avidin-biotin-

glucose oxidase assay (26).Quantitation of Ribavirin. Concentrations of ribavirin in cell extracts

and culture media were quantitated with a competitive binding radio-immunoassay as described previously (27). The antibody is specific forboth ribavirin and its phosphorylated nucleotides and can detect concentrations as low as 0.01 UM. The cell number was counted with ahemocytomer. The intracellular concentrations were calculated on thebasis that 1 ml of packed K562 cells contains 1.85 x 10s cells, whichtranslates to a volume of 5.4 x 10~9ml/cell.

RESULTS

Effects of Ribavirin and Other IMP Dehydrogenase Inhibitorson K562 Cells. Exposure of K562 cells to a wide range ofconcentrations of ribavirin resulted in a concentration-relateddecrease in cellular proliferation (Fig. I A). The concentrationof ribavirin which caused 50% inhibition of cell proliferationafter 3-day incubation was approximately 10 Mg/ml or 40 MM(Fig. 1B). The viability of these ribavirin-treated cells remainedgreater than 95% after 3 days of incubation (Fig. IB). In orderto ascertain whether growth restriction by high concentrationsof ribavirin reflects a cytotoxic effect on proliferation of K562

Ribavirincone.

24 48

Incubation(lus)

i£a

3

0.1 5000

10000.5 1.0 5.0 10.0

Ribavirin Concentration ,<n ml

Fig. 1. Effect of ribavirin on proliferation, viability, and differentiation ofK562 cells. In A, approximately 1.5 x IO5cells/ml of K562 were incubated withvarious concentrations of ribavirin, in triplicates. At the specified time intervals,aliquots were removed for determination of cell density (•)using a hemocytom-eter. In B, data on cell density (• •¿�)and the percentage of cells staining forhemoglobin with benzidine (D; Ref. 18) after 3 days of incubation were presented(points, mean; bars, SD) and plotted against concentrations of ribavirin used. Cellviability (• •¿�)was determined by trypan blue dye exclusion.

cells and/or a commitment to differentiation, clonogenic assay(19, 20) was performed after various periods of preincubationwith ribavirin (Table 1). The colonies derived from K.562 cellsreached an average size of 9 to 16 cells per colony, after 3 daysof cultures. With prior exposure to ribavirin for 1 day, the sizeof colonies was significantly smaller (Table 1), although thetotal number of colonies per plate remained similar to controlcultures. However, when K562 cells were preincubated withribavirin for more than 2 days before plating, the total numberof colonies decreased significantly, and most of the coloniesconsisted of less than 4 cells/colony. These results indicatedthat exposure to ribavirin led to growth restriction of K562cells.

As also shown in Fig. 1B was the response of K.562 cells afterincubation with ribavirin in liquid culture for 3 days, when theybecame benzidine positive in a dose-dependent manner. Maximum induction was observed with ribavirin at 25 Mg/ml (=100/UM).The effective dose for half-maximum induction by ribavirinwas approximately 4 Mg/ml (—16UM).

The effects of other inhibitors of IMP dehydrogenases onK562 cells were also investigated. The addition of tiazofurinand mycophenolic acid for 3 days induced K.562 cells to becomebenzidine positive in a dose-dependent manner (Fig. 2). Maximum induction was observed at 25 ^M for tiazofurin and 3 pMfor mycophenolic acid. The effective doses for half-maximuminduction by tiazofurin and mycophenolic were approximately2 and 0.3 pM, respectively. Both tiazofurin and mycophenolicacid restricted the growth of K562 cells in a dose-dependentmanner; the concentrations for 50% inhibition of K562 proliferation were 2 /¿Mfor tiazofurin and 0.3 UMfor mycophenolicacid (Fig. 2).

Sequestration of Ribavirin Inside K562 Cells. To determinethe pharmacokinetics of ribavirin during the incubation of K562cells, the intracellular concentrations of ribavirin were determined. After incubation of cells with 20 MM ribavirin, theintracellular concentration of ribavirin was approximately 1.5/¿M;this amounts to a 75-fold increase in concentration insidethe cells over the ribavirin concentration in the medium (Fig.3). The elevated intracellular concentration of ribavirin remained relatively constant for up to 96-h incubation. WhenK562 cells were incubated with 100 pM ribavirin for 24 h, theintracellular concentration increased to 11.5 MM,which reflectsa 115-fold increase in intracellular concentration. After longerincubation with ribavirin, the intracellular concentration gradually declined to approximately 50-fold of the extracellularconcentration by 96 h. These experiments indicate that ribavirinwas sequestered preferentially inside the K562 cells to at least50-fold.

Effect on Intracellular Concentrations of Purine Nucleotides.When K562 cells were incubated with ribavirin or tiazofurin,consistent alterations in purine metabolism were observed; theGTP level decreased within 24 h of incubation, and this decreasepersisted throughout the 4-day culture periods (Table 2). In thetreated cells, the decrease of intracellular GTP content becamemore apparent when the intracellular GTP concentrations werenormalized against the measured concentrations of intracellularATP (Table 2).

To further evaluate the role of GTP depletion in the inductionof hemoglobin accumulation in K562 cells, the effect of gua-nosine, which should prevent the depletion of GTP, was studied.In the presence of 25 MMguanosine, the number of benzidine-positive cells in the culture after incubation with ribavirin ortiazofurin decreased remarkably as compared with the culturesof each inducer and in the absence of guanosine (Table 3). A

5556

Research. on August 17, 2021. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from

EFFECTS OF IMP DEHYDROGENASE INHIBITORS ON K562 CELLS

Table 1 Clonogenic analysis ofK562 cells after preincubation with ribavirinK562 cells were preincubated in the RPMI 1640 medium in the presence or absence of 100 UMribavirin for a period of 5 h, l day, or 2 days. After preincubation,

cells were washed extensively and plated, in triplicates, at 500 cells per ml of 0.3% agar in the culture medium without the addition of ribavirin. After 3 days ofincubation, the number of colonies per plate and the cell number per colony were counted.

TreatmentsFive

h of preincubationControl medium

RibavirinOne

day of preincubationControl medium

RibavirinTwo

days of preincubationControl mediumRibavirinColony

no3-427.0

±3.0"(11.4)*

31.7 ±4.7(14.6)28.0

±13.3(11.2)88.3 ±5.8(36.3)26.7

±2.5(11.5)42.0 ±3.6 (61. 2).

with the following size distribution of colonies(cells/colony)5-877.0

±5.0 (32.6)103.6 ±7.5(47.7)58.3

±14.2 (23.4)130.3 +34.9(53.6)64.0

±4.6 (27.5)26.0 ±4.0 (37.8)9-1613

1.0 ±3.0 (55.4)82.0 ±12.0(37.7)154.5

±19.7(61.9)24.7 ±3.0(10.1)127.0

+ 5.7(54.7)0.7 ±1.2(1.0)>171.3

±2.3 (0.6)08.8

+ 6.1 (3.5)014.6

±8.3 (6.3)0Total

no. ofcolonies236.3(100)

217.3(100)249.6(100)

243.3(100)232.3

(100)68.7(100)

' Mean ±SD.*Numbers in parentheses, percentage of distribution of various sizes of colonies within each sample.

00.1 1.0 10 100

Tiazofurin Concentration I

1000

1000-

00.1 0.5 10

Mycophenolic Acid (¡M\

Fig. 2. Effect of tiazofurin and mycophenolic acid on proliferation and induction of K562 cells. K562 cells were incubated, in triplicates, with the specifiedconcentrations of tiazofurin (A) and mycophenolic acid (B) for 3 days. Aliquotswere then analyzed for percentage of cells staining for hemoglobin with benzidine(•)and for cell density (O). The variation of the data was within 10% amongtriplicate samples. Cell viability was more than 95% of the control.

decrease in the number of benzidine-positive cells was alsoaccompanied by a decline in the benzidine staining intensity ofthese cells. At higher concentrations (50 and 100 AIM),therewas no significant difference between control and ribavirin- ortiazofurin-treated samples. These results suggest that the inhibitory effect of guanosine on K562 induction by ribavirin ortiazofurin became less pronounced, owing to a moderate increase in hemoglobin-containing cells as induced by high con-

12.0

9.0

•¿�5-55 6.0to u

!= «,

B3.0

24 48 72

Hours of Incubation

96

Fig. 3. Accumulation of ribavirin in K562 cells upon incubation with ribavirin.K562 cells were incubated at I to 2 X 10s cells/ml with ribavirin at concentrationsof 20 UM(A) and 100 JIM(•).At the specific time interval of incubation, aliquotsof approximately 2 to 5 x IO6cells were taken, and the intracellular concentrationof ribavirin was determined as described in "Materials and Methods." Theintracellular "ribavirin" includes phosphorylated ribavirins, because the antibody

used detects both ribavirin and its phosphorylated nucleotides (27). The resultswere expressed as nmol/g of cell extract (assuming that 1 ml of packed cells yieldsl g of cell extract). The ribavirin concentrations in the culture medium to which100 >i\i (•)were initially added are also indicated in units of /-\i.

centrations of guanosine alone (Table 3). Furthermore, guano-sine had no influence on erythroid differentiation in K562 cellsif it was added 24 h postinduction or later. Neither does it affectthe induction of differentiation in K562 cells by hemin.

Characterization of Hemoglobin Accumulation. In ribavirin-or hemin-induced K562 cells, the amount of hemoglobin accumulation and the extent of induction were found to be comparable with a greater than 30-fold increase in hemoglobincontents (Table 4). The lysates of these cells were analyzed byisoelectric focusing in order to identify the types of hemoglobinsproduced. The individual hemoglobins produced in the ribavi-rin-induced K562 cells were designated as hemoglobins 1 to 5in Fig. 4A, Approximately 60% of hemoglobins, which accumulated in ribavirin-treated cells, was found to be hemoglobinF (Hbf) (Band 3 in Fig. 4A) and its acetylated form (Band 2)by comparison with hemoglobin standards (Fig. 4B). The hemoglobin Band 2 in Fig. 4A was more acidic than hemoglobinA (as demonstrated in the experiment of mixing samples withstandards) and had the same pi as the acetylated hemoglobin F

5557

Research. on August 17, 2021. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from

EFFECTS OF IMP DEHYDROGENASE INHIBITORS ON K562 CELLS

Table 2 Perturbation of GTP and ATP contents ¡nK562 cells after exposure toribavirin or tiazofurin

The log-phase K562 cells were seeded at 1.8 x 10' cells/ml in the presence or

absence of 100 /JM ribavirin or 25 >iM tiazofurin. After the specific time ofincubation, aliquots of cells were removed and analyzed for GTP and ATPcontents as described in "Materials and Methods." The assay method (22, 23)

used for quanlitation of 1 nmol of various purine compounds gave an average ofstandard deviation of ±0.04 nmol with an average coefficient of variation of3.84%, as previously reported from the same laboratory (43). The average valuesof GTP/ATP ratios (±SD)for 4 days of incubation were 0.18 (±0.05).0.07 (±0.02), and 0.04 (±0.01).respectively, for incubation with control medium, ribavirin, and tiazofurin.

A B

IncubationGTP

content, nmol/107cellsControl

mediumRibavirinTiazofurinATP

content, nmol/107cellsControl

mediumRibavirinTiazofurinGTP/ATP

ratioControlmediumRibavirinTriazofurinDay

0 Day18.07

4.641.971.5368.0

31.241.445.40.12

0.150.050.03Day

24.052.521.9228.945.652.20.140.060.04Day37.022.411.9028.427.146.30.250.090.04Day47.692.032.1041.525.744.10.190.080.05

Table 3 Effect ofguanosine concentrations and time of addition on theaccumulation of benzidine-staining cells (percentage} in K562 caused

by various inducenK562 cells were incubated, in triplicates, with 100 nM ribavirin, 25 nM

tiazofurin. 25 nM hemin, or without any inducer. To these samples, guanosinewas added to a concentration specified in the table simultaneously with induceror at a time of either 24 or 48 h postinduction. After 4-day incubation, aliquotsof samples were taken and analyzed for the percentage of cells stained withbenzidine reaction.

Guanosine concentration (JIM)

0 25 50 100

Guanosine added simultaneouslywith inducer

No inducer 6 ±1"

Ribavirin. 100/jM 85 ±1Tiazofurin, 25 MM 88 ±1Hemin, 25 »JM 94 ±I

Guanosine added 24 h postinduction

Ribavirin. lOO^M 77 ±4Tiazofurin. 25 »IM 78 ±1

Guanosine added 48 h postinduction

Ribavirin. 100 JIM 79 ±2Tiazofurin. 25 JIM 86 ±1

8 ±1 18 ±1 30 ±213 ±3 18 ±1 26 ±28±2 18 ±1 29 ±3

95 ±1 97 ±1 94 ±1

78 ±183 ±2

85 ±1

82 ±180 ±2

80 ±289 ±1

89 ±181 ±2

83 ±287 ±2

" Mean ±SD of triplicate samples.

Table 4 Hemoglobin contents in uninduced and induced A'562 cells

K562 cells were incubated with hemin, ribavirin, or medium alone for 3 daysand then cell lysates were prepared from these cells, as described below. Approximately 1 x IO7 cells were collected and washed twice in phosphate-buffered

saline. pH 7.4. and aliquots were taken for cell counts and benzidine staining(18). One volume of packed cells was then mixed with one volume of deionizedwater and 0.5 volume of CCI4 and centrifuged at 4'C. and the amount of

hemoglobin in the supernatants was determined as described (21).

InducersNone

Hemin. 25 fiMRibavirin. 100 , MHemoglobin

contentper 10* cells(fig)0.133.75

3.27Benzidine-staining

cells (%)7

9492

(HbF,) in the cord blood (picture not shown). As also indicatedin Fig. 4/1, the remaining hemoglobins consisted of 21% hemoglobin Portland (Band I), 8% Gower-1, and 11% Gower-2(Bands 4 and 5). On the other hand, similar experiments usinghemin-induced K562 cells (Fig. 4C) showed that only 14% of

0_ ^^ 254"

Fig. 4. Identification of hemoglobin phenotypes in the cell extracts of ribavirin-and Ili-min iniliiivil K562 cells. K562 cells were induced with 100 f<\i ribavirinand 25 ^M of hemin. respectively. After 3 days incubation, cell lysates wereprepared, and isoelectric focusing was performed as described in "Materials andMethods." A. ribavirin-trealed cell extract; B, hemoglobin standards (Isolab, Inc..Akron, OH); and C, hemin-treated cell extract. The individual hemoglobin bandsin the ribavirin-treated cell extract were designated as hemoglobins / to 5. Thepositions for hemoglobin standards. HbA. HhF, HhS. and HbC. arc indicated. Inhemin-treated cell extract (('), hemoglobins Portland, Bart's, and HbFarc labeled

as a. b. and c, respectively.

accumulated hemoglobins was the hemoglobin F, and the remaining 86% was hemoglobin Portland, Barts, and other unidentified hemoglobins (24, 28-30).

Analysis of Globin Chain Composition. The phenotypes of thehemoglobins that accumulated in the ribavirin-induced K562cells were confirmed by globin chain analysis. The individualhemoglobin bands in the isoelectric focusing gel (Fig. 5A) wereexcised. After elution and denaturation, the globin chains wereanalyzed in urea-Triton-acid polyacrylamide gel (24). As shownin Fig. 5Ä,the composition of globin chains for hemoglobinBand 1 in Fig. 5A is y globin chains (i.e., Gy and A7) and fchain, confirming the presence of hemoglobin Portland (fc y2).Hemoglobin Band 3, which is the major hemoglobin speciesaccumulated in the lysates, is composed of the y globin chainsalong with a globin, confirming the accumulation of hemoglobin F (ah 7:). Band 2 showed characteristic patterns of ace-tylated 7 chains (Fig. 5B, sample 6), confirming that thishemoglobin is hemoglobin F,. Hemoglobin Band 4 is mainlycomposed of two embryonic globin chains, fand £,while Band5 consists of «and £chains. They are thus shown to be thehemoglobins Gower-1 and -2, respectively. Therefore, the majorspecies of globin chains which accumulated in cells after ribavirin induction were the y globin chains and the «chains. Inno instances was hemoglobin A (a? ft) detected in globin chainanalysis. These results were also confirmed by immunoblottingof the gels in Fig. 5B with appropriate antibodies. Antibodiesdirected against hemoglobin F («272) detected the presence of

5558

Research. on August 17, 2021. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from

EFFECTS OF IMP DEHYDROGENASE INHIBITORS ON K562 CELLS

B1 2

-2-

HbC-

£-Ay-

Gr-

L5_ ~

Fig. 5. Globin chain analysis for the hemoglobins produced after ribavirininduction. A, isoelectric focusing. Cell lysates were prepared from ribavirin-induced K562, and isoelectric focusing was performed as described in Fig. 4. Theindividual hemoglobin bands were designated as hemoglobins / to 5. The isoelectric focusing gels, a and b, which had the same cell lysates were shown to indicatethat hemoglobins .< and •¿�/could be separated during a shorter duration ofisoelectric focusing (Gel a) and that both have a similar pi after completion ofisoelectric focusing (Gel b). The positions for hemoglobin standards are alsoindicated for comparisons. B, gel slices containing the individual hemoglobinbands (/ to 5 in A) were excised from the isoelectric focusing gel and labeled asSamples / to 5. Globin chain analysis was performed in urea-Triton-acid poly-acrylamide gel (24), as described in "Materials and Methods." As standards for

various globin chains. HbF, and HbA derived from cord blood preparation wereused (Sample 6).

7 chains (in Samples 1, 2, and 3 of Fig. 5B), and both antibodiesspecific to hemoglobin A (a: ß2)and hemoglobin F, respectively,demonstrated the production of «chains (in Samples 2, 3, and5) in the ribavirin-treated K562. In addition, the presence ofhemoglobin F in the ribavirin- or tiazofurin-treated K.562 cellswas independently confirmed by immunofluorescence studiesof the cytospin preparations of the induced cells using monoclonal antibody specific for 7 chains (31) (picture not shown).

DISCUSSION

Many studies have suggested that alterations in the regulationof IMP dehydrogenase may modulate the capacity of cells toundergo proliferation (32, 33). Malignant transformation wasfound to associate with an increase in the activity of IMPdehydrogenase (34, 35) and various drugs which inhibit thisenzyme displayed significant antitumor activities (32,33). Studies of fetal and regenerating hepatic tissues had demonstrateda direct relation between levels of this enzyme and proliferationof these cells (36). In line with these findings, our currentstudies indicated that the use of inhibitors of IMP dehydrogenase led to the suppression of K562 proliferation. On the otherhand, it also resulted in the induction of hemoglobin production. The optimal concentrations for the induction of differentiation in K562 cells by tiazofurin and mycophenolic acidclosely approximated the concentrations that suppressed cellular proliferation. The clonogenic assays also showed thatthese differentiating cells exhibited restricted potential for proliferation (i.e., terminal differentiation). The longer the exposure to the inducer prior to plating, the smaller the colony size,because more cells would have already undergone terminaldifferentiation prior to plating. While most of the colonies inthe drug-pretreated K562 are benzidine positive in clonogenicassay, not all the cells within the colonies are clearly stainedwith benzidine. These results indicate that K.562 cells may notfully be committed and may require the continuous presence of

an inducer. The issue of "commitment" of K562 cells to "terminal differentiation" remains to be further investigated (e.g.,

Ref. 20). However, these results also showed that ribavirincauses cytotoxic effects on K562 cells. The higher the concentration and the longer the exposure to the inducer prior toplating, the lower was the cloning efficiency in the clonogenicassay. In the studies of HL-60 cells (37), it was proposed thatthe observed differentiation in HL-60 might result from an

N-acetyl adaptive cellular response to a toxic stress imposed by inducers.

However, our data on ribavirin showed that the concentrationfor half-maximum induction by ribavirin was about 3 times lessthan that for 50% growth inhibition. This would suggest thatinhibition of cellular proliferation may not be a necessaryprerequisite of cellular differentiation.

It has been shown that ribavirin 5'-monophosphate is a

potent inhibitor of IMP dehydrogenase (4), which is involvedin the de novo synthesis of GMP. In agreement with previousreports (5), present studies showed that K562 cells exposed toribavirin displayed significant decline in the level of GTP.Mycophenolic acid and tiazofurin also promote similar alterations in inlrace II11lar levels of GTP (3, 6). The fact that all theseinhibitors of IMP dehydrogenase induced cellular differentiation suggests a possible role of GTP depletion in K562 differentiation. It is of interest to note that the induced maturationof myeloid leukemia cell line HL-60 is also accompanied by anintracellular depletion of guanosine ribonucleotides (10, 11),which reflects a down-regulation of guanylate synthesis fromIMP at the rate-limiting step mediated by IMP dehydrogenase(9, 38, 39). The role of GTP depletion in ribavirin-induceddifferentiation in K562 is further strengthened by the findingsthat guanosine, which could replenish the effect of GTP depletion, was able to block the differentiation of K562 induced byribavirin or tiazofurin but not by an unrelated inducer, hemin.Similar results were also reported in the blockage of inducedmaturation of HL-60 cells by guanosine (39). On the otherhand, the accumulation of hemoglobins in K562 cells waspreceded by a decrease in GTP concentration which occurredwithin 24 h of ribavirin treatment. Whether or not these findings suggest any association between depletion of GTP and"commitment" of K562 cells into differentiation (20) remains

to be investigated.Present studies indicate that ribavirin was sequestered pref

erentially inside the K.562 cells. The results confirm previousreports that there is a selective accumulation of ribavirin inmature human erythrocytes in vitro (40). They also agree withthe in vivo observations that approximately 10 to 15% of thetotal administered ribavirin was incorporated into mature redcells, while red cells accounted for only 3% of the total bodymass and that the ratio between plasma trough levels of ribavirinand red cell ribavirin was approximately 1 to 90 (40). Conceivably, ribavirin is entrapped within the mature erythrocytesbecause it is phosphorylated upon entry into these cells. Ribavirin was shown to be rapidly phosphorylated in mature redcells and mainly exists as the 5'-triphosphate (41).

Present studies using the same subclone of K562 cells showedthat hemoglobin F constitutes most of the accumulated hemoglobin in ribavirin-treated cells, in contrast to the productionof mainly embryonic hemoglobins in hemin-treated cells (28-31). In ribavirin-treated cells, Gy represents about 75% of the7 globin chains, which is similar to the value found in normalfetuses and newborns (42). The induction of hemoglobin accumulation with ribavirin was also marked by the accumulationof a globin chains and the lack of ßglobin production. The «/7 ratio was estimated to be 0.77, and no hemoglobin Bart's was

5559

Research. on August 17, 2021. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from

EFFECTS OF IMP DEHYDROGENASE INHIBITORS ON K562 CELLS

identified in these cells. In hemin-treated K562 cells, the a/yratio was 0.11 in our studies and was reported to be in therange of 0.08 and 0.30 by others (24, 28). Therefore, the patternof hemoglobin accumulation in ribavirin-treated cells was different from the a thalassemic imbalance observed in hemin-treated K562. These findings raise the possibility of usingribavirin to increase fetal hemoglobin production in patientswith sickle cell disease and 0-thalassemia syndromes, if similarresults could be demonstrated in vivo.

ACKNOWLEDGMENTS

We would like to thank Dr. T. Papayannopoulou. University ofWashington at Seattle, for performing immunofiuorescence staining ofribavirin-treated samples with monoclonal antibodies specific for fi and7 chains.

REFERENCES

1. Witkowski. J. T., Robins, R. K., Sidwell, R. W., and Simon, L. N. Design,synthesis, and broad spectrum antiviral activity of 1-beta-D-ribofuranosyl-l,2,4-triazole-3-carboxamide and related nucleosides. J. Med. Chem.. 15:1150-1153. 1972.

2. Srivastava. P. C, Pickering, M. V., Allen. L. B., Streeter, D. G., Campbell,M. T., Witkowski, J. T., Sidwell, R. W., and Robins, R. K. Synthesis andantiviral activity of certain thiazole C-nucleosides. J. Med. Chem., 20: 256-262, 1977.

3. Franklin, T. J.. and Cook. J. M. The inhibition of nucleic acid synthesis bymycophenolic acid. Biochem. J.. 113: 515-520, 1969.

4. Streeter, D. G., Witkowski, J. T., Khare, G. P., Sidwell. R. W., Bauer. R. J..Robins, R. K., and Simon, L. N. Mechanism of action of 1-beta-n-ribofura-nosyl-l,2,4-triazole-3-carboxamide (Virazole), a new broad-spectrum antiviral agent. Proc. Nati. Acad. Sci. USA, 70: 1174-1178, 1973.

5. Wray, S. K., Gilbert, B. E.. Noall, M. W., and Knight. V. Mode of action ofribavirin: effect of nucleotide pool alterations on influenza virus ribonucleo-protein synthesis. Antiviral Res., 5: 29-37, 1985.

6. Sokoloski, J. A., and Sartorelli, A. C. Effects of the inhibitors of IMPdehydrogenase, tiazofurin and mycophenolic acid, on glycoprotein metabolism. Mol. Pharmacol., 28: 567-573, 1985.

7. Shulman, N. R. Assessment of hématologieeffects of ribavirin in humans.In: R. A. Smith, V. Knight, and J. A. D. Smith (eds.). Clinical Applicationsof Ribavirin, pp. 79-92. New York: Academic Press, 1984.

8. Melink. T. J.. Von Hoff, D. D., Kühn,J. G.. Hersh. M. R.. Sternson, L. A..Patton, T. F., Siegler, R.. Boldt, D. H., and Clark, G. M. Phase I evaluationand pharmacokinetics of Tiazofurin (2-fM>-ribofuranosylthiazole-4-carbox-amide, NSC 286193). Cancer Res.. 45: 2859-2865. 1985.

9. Lucas, D. L.. Webster, H. K., and Wright, D. G. Purine metabolism inmyeloid precursor cells during maturation: studies with the HL-60 cell line.J. Clin. Invest., 72: 1889-1900, 1983.

10. Sokoloski, J. A., Blair, O. C., and Sartorelli, A. C. Alterations in glycoproteinsynthesis and guanosine triphosphate levels associated with the differentiation of HL-60 leukemia cells produced by inhibitors of inosine 5'-phosphatedehydrogenase. Cancer Res., 46: 2314-2319. 1986.

11. Lucas, D. L.. Robins. R. K., Knight. R. D., and Wright, D. G. Inducedmaturation of the human promyelocytic leukemia cell line, HL-60, by 2-beta-D-ribofuranosylselenazole-4-carboxamide. Biochem. Biophys. Res. Commun., 115:971-980, 1983.

12. Wright, D. G., LaRussa, V. F.. Salvado, A. J., and Meagher. R. C. Intracel-lular supplies of guanosine ribonuclcotides regulate the terminal maturationof normal human myeloid progenitors. Clin. Res., 34: 474A, 1986.

13. Tricot. G. J., Jayaram, H. N.. Nichols, C. R.. Pennington. K., Lapis, E.,Weber, G.. and Hoffman, R. Hematological and biochemical action oftiazofurin (NSC 286193) in a case of refractory acute myeloid leukemia.Cancer Res.. 47: 4988-4991, 1987.

14. Lozzio. C. B., and Lozzio. B. B. Human chronic myelogenous leukemia cell-line with positive Philadelphia chromosome. Blood, 45: 321-334, 1975.

15. Yu, A. L., Connor, J. D.. Page, T.. Lemas, V.. and Yu, J. Induction ofhemoglobin accumulation in human K562 cells by inhibitors of inosinemonophosphate dehydrogenase. Proc. Am. Assoc. Cancer Res., 29:32, 1988.

16. Yu, A. L., Connor, J. D., Page, T., Lemas, V., and Yu, J. Ribavirin causesinduction of erythroid differentiation in the human cell line, K562 (Abstract).Leukemia, 2: 198. 1988.

17. Singh, M. K., and Yu, J. Accumulation of a heat shock-like protein during

differentiation of the human erythroid cell line K562. Nature (Lond.), 309:631-633, 1984.

18. Orkin, S. H., Haros, P. 1., and Leder, P. Differentiation in erythroleukemiacells and their somatic hybrids. Proc. Nati. Acad. Sci. USA, 72: 98-102,1975.

19. Yu, J., and Smith, R. D. Sequential alterations in globin gene chromatinstructure during erythroleukemia cell differentiation. J. Biol. Chem., 260:3035-3040, 1986.

20. Rowley, P. T., Ohlsson-Wilhelm, B. M., and Farley. B. A. K562 humanerythroleukemia cells demonstrate commitment. Blood, 65: 862-868, 1985.

21. Tsiftsoglou, A. S.. Gusella, J. F., Vollock, V., and Houseman. D. E. Inhibitionby dexamethasone of commitment to erythroid differentiation in murineerythroleukemia cells. Cancer Res.. 39: 3849-3855. 1979.

22. Page, T., Bakay, B., Nissinen, E., and Nyhan, W. L. Hypoxanthine-guaninephosphoribosyltransferase variants: correlation of clinical phenotype withenzyme activity. J. Inherited Metab. Dis., 4: 203-206, 1981.

23. Bakay, B.. Nissinen. E.. and Sweetman, L. Analysis of radioactive andnonradioactive purine bases, nucleosides, and nucleotides by high-speedchromatography on a single column. Anal. Biochem.. 86: 65-77, 1978.

24. Alter, B. P., and Goff, S. C. Electrophoretic separation of human embryonicglobin demonstrates "alpha-thalassemia" in the human leukemia cell lineK562. Biochem. Biophys. Res. Commun.. 94: 843-848, 1980.

25. Towbin, H., Staehelin, T., and Gordon, J. Electrophoretic transfer of proteinsfrom polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Nati. Acad. Sci. USA, 76:4350-4354, 1979.

26. Robb, J. A. Enzyme-immunolectin histological techniques for surgical pathology. Introducing "GAB": a new glucose oxidase-avidin-biotin assay. In:

R. Nakamura, and D. Tucker (eds.). Clinical Laboratory Assays: New Technology and Future Direction, pp. 323-332. New York: Masson Publishing,Inc., 1983.

27. Austin, R. K.. Trefts. P. E., Hintz, M., Connor. J. D., and Kagnoff, M. F.Sensitive radioimmunoassay for the broad-spectrum antiviral agent ribavirin.Antimicrob. Agents Chemother.. 24: 696-701. 1983.

28. Rutherford, T. R., Clegg, J. B.. and Weatherall, D. J. K562 human leukaemiccells synthesise embryonic haemoglobin in response to haemin. Nature(Lond.). 280: 164-165, 1979.

29. Martin, P., and Papayannopoulou, T. HEL cells: a new human erythroleukemia cell line with spontaneous and induced globin expression. Science(Wash. DC), 216: 1233-1235, 1982.

30. Dean, A., Erard, F., Schneider, A. B., and Schechter, A. N. Induction ofhemoglobin accumulation in human K562 cells by hemin is reversible.Science (Wash. DC), 212: 459-461. 1981.

31. Benz, E. J., Murname, M. J., Tonkonow, B. L., Berman, B. W., Mazur, E.M.. Cavallesco, C., Jenko, T.. Snyder. E. L., Forget, B. G., and Hoffman, R.Embryonic-fetal erythroid characteristics of a human leukemic cell line. Proc.Nati. Acad. Sci. USA, 77: 3509-3513. 1980.

32. Weber, G. Biochemical strategy of cancer cells and the design of chemotherapy. G. H. A. Clowes Memorial Lecture. Cancer Res., 43: 3466-3492. 1983.

33. Jackson. R. C. Computer simulation of the effects of antimetabolites onmetabolic pathways. In: K. R. Harrap and T. A. Connors (eds.). New Avenuesin Developmental Cancer Chemotherapy, pp. 3-35. Orlando. FL: AcademicPress, Inc.. 1987.

34. Becher. H. J.. and Lohr, G. W. Inosine 5'-phosphate dehydrogenase activityin normal and leukemic blood cells. Klin. Wochenschr., 57: 1109-1115,1979.

35. Jackson, R. C.. and Weber. G. IMP dehydrogenase. an enzyme linked withproliferation and malignancy. Nature (Lond.), 256: 331-333. 1975.

36. Jackson, R. C., Morris, H. P., and Weber, G. Partial purification, properties,and regulation in inosine 5'-phosphate dehydrogenase in normal and malignant rat tissues. Biochem. J.. /66: 1-10, 1977.

37. Langdon, S. P., and Hickman, J. A. Correlation between the molecularweight and potency of polar compounds which induce the differentiation ofHL-60 human promyelocytic leukemia cells. Cancer Res.. 47:140-144, 1987.

38. Knight. R. D.. Mangum. J., Lucas. D. L., Cooney, D. A., Khan, E. C.. andWright, D. G. Inosine monophosphate dehydrogenase and myeloid cellmaturation. Blood. 69:634-639, 1987.

39. Wright, D. G. A role for guaninc ribonucleotides in the regulation of myeloidcell maturation. Blood, 69: 334-337, 1987.

40. Laskin, O. L., Longstreth, J. A.. Hart. C. C., Scavuzzo. D., Kaiman, C. M.,Connor, J. D., and Roberts. R. B. Ribavirin disposition in high-risk patientsfor acquired immunodeficiency syndrome. Clin. Pharmacol. Ther.. 41: 546-555, 1987.

41. Zimmerman, T. P., and Deeprose. R. D. Metabolism of 5-amino-l-beta-n-ribofuranosyl-imidazole-4-carboxamide and related five-membered hetero-cycles to 5'-triphosphates in human blood and L517-delta-Y cells. Biochem.Pharmacol., 27: 709-716. 1978.

42. Alter, B. P. The G-gamma:A-gamma composition of fetal hemoglobin infetuses and newborns. Blood, 54: 1158-1163. 1979.

43. Nissinen, E. Analysis of purine and pyrimidine bases, ribonucleosides, andribonucleotides by high-pressure liquid chromatography. Anal. Biochem.,706:497-505. 1980.

5560

Research. on August 17, 2021. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from

1989;49:5555-5560. Cancer Res   John Yu, Victor Lemas, Theodore Page, et al.   of Inosine Monophosphate DehydrogenaseInduction of Erythroid Differentiation in K562 Cells by Inhibitors

  Updated version

  http://cancerres.aacrjournals.org/content/49/20/5555

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/49/20/5555To request permission to re-use all or part of this article, use this link

Research. on August 17, 2021. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from