[CANCER RESEARCH 50, 1231-1236. February 15. 1990]

Induction of Erythroid Differentiation in Human Leukemic K-562 Cells byMembrane-directed Action of Adriamycin Covalently Bound to Microspheres1

Pierre Jeannesson,2 Chantai Trentesaux, Brigitte Gerard, Jean-Claude Jardillier, Kevin L. Ross, and /.olíanA. Tokés

GIBSA, Laboratoire de Biochimie, Facultéde Pharmacie, 5l rue Cognacq-Jay, 51096 Reims Cedex, and Institut Jean Godinot, I rue du GénéralKoenig,51100 Reims, France ¡P.J., C. T., B. G., J-C. J.j, and Department of Biochemistry, Kenneth Morris, Jr., Comprehensive Cancer Center, University of Southern California,School of Medicine, Los Angeles, California 90033 fK. L. R., Z. A. T.J

ABSTRACT

Covalent coupling of Adriamycin (ADR) to polyglutaraldehyde micro-spheres focuses the drug action to the plasma membrane which leads togrowth inhibition and also to induction of erythroid differentiation inhuman leukemic K-S62 cells without any evidence of cellular internali-zation of the drug-microsphere complexes. As observed with the freedrug, a reduction in cell growth by the coupled drug correlated with arecruitment of differentiating cells. Treatment of sensitive cells withADR-microspheres results in 40% of cells containing hemoglobins asdetermined by benzidine staining at 87% growth inhibition. Similartreatment of ADR-resistant cells produces 24% of benzidine-positivecells at 72% growth inhibition. Furthermore, free and coupled forms ofADR stimulate heme synthesis 12- and 20-fold. Hemoglobin analysis ofADR-polymer induced cells demonstrates additional embryonic (Gower-2, X, Portland) and fetal (F) types of hemoglobin in comparison touninduced cells which synthesize only small amounts of Gower-1 insensitive cells and Gower-1 plus hemoglobin X in resistant cells. Inaddition, free and bound forms of Adriamycin differ markedly in therelative proportion of hemoglobin types that they induce. P'ree and

coupled forms of ADR produce an increase in the -y-globin mRNAsynthesis in sensitive K-562 cells. These results demonstrate that bothADR-sensitive and -resistant K-562 cells can be induced to differentiateat the cell surface by ADR-microspheres and that this induction differsqualitatively from that of free ADR.

INTRODUCTION

Anthracycline antitumor antibiotics have been shown to bepotent inducers of differentiation in human leukemic cells,particularly in vitro (1). We have recently reported that theanthracycline ADR' can induce the human multipotent K-562

cells to differentiate towards the erythroid lineage (2). Themechanism by which this occurs is largely unknown, althoughmost other differentiating agents act by intercalation into DNA,suggesting DNA to be the primary target (3). However, theplasma membrane has been demonstrated to be another targetfor inducers of erythroleukemic differentiation (4); thus, it isdifficult to ascertain where the initiation event for differentiation occurs. Using retinoic acid immobilized to a retainingsubstrate. Yen et al. (5) have shown that drug interaction thatis limited to the cell surface was sufficient for the induction ofdifferentiation.

Adriamycin may act as a differentiation-inducing agent whichcan also be covalently coupled to a solid support like micro-spheres (6, 7). The resulting drug complexes show a stronginteraction with the plasma membrane. Therefore it was intriguing to test whether or not ADR could act on erythroid differ-

Received 3/31/89; revised 10/30/89; accepted 11/13/89.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.

' This work was supported by A. R. C. France (Association pour la Recherche

sur le Cancer) and NATO Grant 86/314 for collaborative research.2To whom requests for reprints should be addressed.1The abbreviations used are: ADR. Adriamycin; ADR-PGL. Adriamycin

coupled to polyglutaraldehyde microspheres; PGL, polyglutaraldehyde micro-spheres; DMEM, Dulbecco's modified Eagle's medium; IC50. 50% inhibitoryconcentration; PBS. phosphate-buffered saline.

4 Unpublished results.

entiation through a membrane-directed interaction. In thisstudy, we demonstrate that ADR-PGL induce a progressiveincrease in the number of hemoglobin-producing cells in bothADR-sensitive and -resistant K-562 cells. The two modes oftreatment resulted in a concurrent increase in heme synthesisand in 7-globulin mRNA accumulation. The appearance inboth cell lines of new hemoglobin types that also appear duringnormal differentiation argues for a differentiating signal originating at the cell membrane.

MATERIALS AND METHODS

Cell Culture. K-562 cells were maintained in DMEM medium(GIBCO Laboratories, Grand Island, NY) supplemented with 10% fetalbovine serum in 5% CO2/95% air. K-562 cells resistant to ADR (K-562/ADR) were established in our laboratory by continuous exposureto increasing drug concentrations over a period of 8 mo and weremaintained at 50 n\i ADR. Both cell lines were kept in exponentialgrowth phase between 5 and 60 x IO4cells/ml by diluting the cultures

with fresh medium 3 times weekly.Preparation of Microsphere-Adriamycin Complexes. PGL micro-

spheres were synthesized by vigorously shaking for 12 h at roomtemperature 10% glutaraldehyde and 0.1% nonionic detergent of Aerosol 604 (American Cyanamide, Wayne, NJ) at pH 11.0 (6). Followingsynthesis, the microspheres were washed exhaustively with water toremove unreacted glutaraldehyde. Covalent coupling of ADR was accomplished by reacting the drug with PGL in a 1:20 ratio (w/w) at pH6.0 for 24 h. Subsequent to coupling, non-covalently bound ADR wasremoved by treatment with 0.05% Nonidet P-40 (Particle Data Laboratories, Elmhurst, IL) and by repeated washings with 30% ethanol inwater and with 0.9% saline solution. The average diameter of micro-spheres under these conditions was 4500 x A (6). The Schiff basecondensation between amino groups (daunosamine of ADR) and thealdehyde groups is, in general, reversible. However, an extra doublebond on the polymer carbon position provides resonance stabilizingenergy, making the coupled complex stable (8, 9). Microspectrofluo-rometry was performed by Dr. M. Manfait (Facultéde Pharmacie,Reims, France) on single ADR-PGL microspheres and on ADR solutions. Identical fluorescent spectra were obtained from the free andcoupled ADR preparations indicating that no chemical change occurredin the anthracycline ring structure due to the coupling reaction.

The amount of ADR covalently coupled to the microspheres wasdetermined by solubilizing ADR-PGL in one volume of ethanolaminefollowed by neutralization with approximately 5 volumes of 0.1 N HC1.ADR content was determined by measuring the absorbancc at 495 nmor by measuring the spectrofluorescence on an Aminoco-Bowmanspectrophotofluorometer (American Instrument Co., Silver Spring,MD) and comparing the obtained values with a standard curve established with similarly treated, equivalent amounts of plain microspheresand an appropriate range of free ADR. The molarity of ADR-microspheres is expressed as the total ADR concentration in a suspension ofADR-PGL. Since not all of the coupled drug molecules are equallyaccessible for cellular interaction, the "available" drug concentration is

slightly lower than the molarity of ADR-PGL.The stability of ADR-PGL during incubation wilh cells was assessed

with five different cell lines: L1210; S-180; CCRF-CEM; CCRF-CEM/Vbl 500; and CCRF-CEM/Vbl 100 (6). Cells were incubated for 72 hwith ADR-PGL. The culture supernatants were extracted with ethanol-HC1, and the concentrations of released ADR were determined spectro-

1231

on May 27, 2021. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

INDUCTION OF ERYTHROID DIFFERENTIATION IN HUMAN LEUKEMIC K-562 CELLS

fluorometrically and expressed as the percentage of total bound drug.In all of these incubations less than 0.09% of the drug was releasedafter 72-h incubations. No released fluorescent drug was detected in K-562 cells by fluorescent microscopy. After incubation of ADR-PGLwith cells for 72 h, the drug complexes retained their complete cytotoxicactivity when they were retested in a new cell culture system.

Cytostatic Assays. Cells in exponential growth were suspended infresh growth medium and diluted to 10 x IO4cells/ml per assay, andthe proper amount of free ADR, ADR-PGL, or plain microspheres

(PGL) was added in 1 ml of medium. The number of cells was determined at Day 3 in triplicate samples by using phase-contrast microscopy. Cell viability was assessed by the trypan blue exclusion test. Drugconcentrations for 50% growth inhibition, IC50,were determined froma plot of the logarithm of drug concentrations versus growth rateexpressed as the percentage of treated cells compared with control cells.

Assay for Hemoglobin-producing Cells. Erythroid differentiation wasscored by benzidine staining using a procedure which was reportedpreviously (2).

Assessment of Phagocytic Activity. Two ml of cells in exponentialgrowth phase were washed twice in PBS and mixed at a concentrationof 25 x IO4cells/ml in hemolysis tubes with ADR-PGL or with latex

particles (Bactolatex 0.81; DIFCO, Detroit, MI). Mouse peritonealcells were used as positive controls. They were obtained by washing theperitoneal cavity of BALB/c mice as described by Pages el al. (10).Cells and particles were incubated for 2 h at 37°Cwith gentle rotation.

The cells were then washed 3 times with PBS by centrifuging at 2100x g for 5 min and rigorously resuspended using a Pasteur pipet.Phagocytic activity was quantitated by determining the percentage ofphagocytic cells among 300 viable cells using blue fluorescence undera x40 water immersion objective for ADR-PGL or by staining withméthylèneblue for latex particles.

Incorporation of 55Fe into Heme. K-562 cells were labeled with[55Fe]transferrin as follows: 100 n\ of "Fe chloride (Amersham Corp.;

specific activity, 1 mCi/0.42 mg of iron) were mixed for 30 min atroom temperature with 400 ^1 of water and 500 n\ of 52 pM trisodiumcitrate. In order to obtain an exchange of Fe3+between the diciate andtransferrin, 1 ml of iron-citrate was incubated for 3 h at 37°Cwith 4

ml of DMEM containing 4.7 mg of NaHCO3 and 40 mg of humantransferrin obtained from Sigma Chemical Co. (11). Subsequently, cellswere incubated with [55Fe]transferrin at the following ratio: 250 M' of[55Fe)transferrin complex per 10 ml of cell suspension. The cells were

washed after 3 days of incubation and used either to measure hemesynthesis or to analyze the types of hemoglobins produced.

Determination of Heme Synthesis. Cells were washed 3 times in ice-cold PBS, and heme was extracted with acid butanone as describedearlier (12). Radioactivity was measured by /^-scintillation counting.Each experimental value is the mean of assays performed in triplicate.

Analysis of Hemoglobins. Five million cells were washed 3 times withice-cold PBS and lysed in 50 p\ of 10 mivi KCN using 3 cycles offreezing and thawing. Postmitochondrial supernatants, obtained aftercentrifugation for l h at 100,000 x g, were analyzed by electrophoresison 6% polyacrylamide gels in Tris-glycine buffer (25 mM Tris/200 mMglycine at pH 8.6). Gels were subjected to autoradiography on KodakX OMAT AR film (Eastman Kodak, Rochester. NY) at -80°C and

scanned at 550 nm on a Beckman DU-8 spectrophotometer.Cytoplasmic mRNA Determinations. Accumulation of -y-globin

mRNA was quantitated by dot blot hybridization of cellular cytoplasmicRNA preparations as described earlier (13). The DNA probe specificfor 7-globin mRNA was a generous gift of Dr. P. Charnay, Departmentof Biochemistry and Molecular Biology, Harvard University, Cambridge, MA.

The nick-translated DNA probe (14) used for hybridization wasplasmid pRA y carrying the 3.3-kilobase £coRIrestriction fragmentfrom the human fetal A-y-globingene (IS). The question of potentialcross-hybridization with the 7-globin DNA probe has been examinedwith {- and «-globinDNA probes. No cross-hybridization was observedbetween the 7-globin DNA probe and the a-globin DNA probe. However, a slight cross-hybridization, which was too low to be quantified,was noticed with the {-globin DNA probe.

RESULTS

The cytostatic activity of ADR-PGL and PGL on humanleukemic K-562 cells after 3 days of culture is shown in Fig. 1.Since PGL had displayed no detectable cytostatic activity atADR-PGL concentrations between 1.5 x 10~6 and 10~8 M,

growth inhibition can be said to be specific for the drug andnot to be due to the effect of the drug-supporting microspheres.The inhibition of cell growth produced by ADR-PGL is paralleled by an increase of differentiated (benzidine-positive) cellsin a dose-dependent manner. At the most inhibiting ADR-PGLconcentration, the percentage of hemoglobin-containing cells isthe same as that obtained after treatment with ADR, which isapproximately 40% after 3 days of incubation. By contrast, noinduction of differentiation is found with very high concentrations of PGL even at similar growth inhibitions (70 to 90%) ofPGL that would correspond to microsphere concentrationsfound in 10~5ADR-PGL-containing samples.

Table 1 summarizes the experiments of incubating free ADRwith K-562 cells in the presence of plain PGL. The resultsestablish that coupling of drug to PGL is essential for theinduction of benzidine-positive cells. It is of interest that plainPGL microspheres inhibit both the cytotoxic and the differentiating activity of free ADR, presumably due to binding competition for the free drug.

Fig. 2 shows the growth-inhibitory effect of ADR-PGL andfree ADR on K-562 cells resistant to ADR. The growth inhibition of resistant cells produced by either the coupled or thefree drug is accompanied by an increase in benzidine-positivecells as observed in sensitive cells.

The IC50values for free ADR with sensitive or resistant cellsare lower than those for the bound drug, but comparisons ofthe resistance indices for both drugs (Table 2) indicate that themicrosphere-coupled drug is about 8-fold more effective thanthe free drug on resistant cells. This confirms previous results

£50

-9

Fig. 1. Growth inhibition curves of K-562 cells after treatment with ADR,PGL, and ADR-PGL are shown ( ). Differentiation and hemoglobin synthesisas determined by benzidine staining are also shown ( ). Treatments areillustrated as follows: ADR (A); ADR-PGL (•);and PCL (O). ADR-PGLconcentrations are given as MADR and depict the total amount of ADR associatedwith microspheres. However, not all of the coupled drug molecules are availablefor interaction with cells. PGL concentrations are equated to the amount ofmicrospheres in ADR-PGL. Each point represents the mean of three experiments.ADM, Adriamycin.

Table 1 Incubation of K-562 cells with uncoupled ADR (40 nM)for 3 days in thepresence of plain PGL microspheres

Amount of PGLaddedExperiment

I: ADR (40nxi)5x10-'PGL1

x 10~!PGLExperiment

II: ADR (40n\i)5x 10-'PGL1x IO"' PGL»ofgrowth

inhibition907635764726%of benzidine-

positivecells25<3<!'33<3<1°

' Untreated K-562 cells had a simitar frequency of benzidine-positive cells.

1232

on May 27, 2021. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

INDUCTION OF ERYTHROID DIFFERENTIATION IN HUMAN LEUKEM1C K-562 CELLS

¡•«I

OKO

1CT5 IO""

ADM (M)

1«T7

Fig. 2. Growth inhibition ( ) and induction of differentiation ( ) ofADR-resistant K-562/ADR cells after treatment with ADR (A) and ADR-PGL(•).ADM, Adriamycin.

Table 2 , values for free and coupled ADR in sensitive and resistant K-562cells

IC,„(nM)ofADRFree

ADRADR PCLK-5624.5 130K-562/ADR7002500Index

ofresistance155

19

Table 3 Valuesfor phagocytic uptake of ADR-PGL microspheres in K-562 andperitoneal cells

Phagocytic activity was determined as described in "Materials and Methods."

Phagocytic cells(%)CellsK-562PeritonealCompoundLatex

particlesADR-PGL (9.2 X IO'6 M)ADR-PGL (9.2 X IQ-7M)ADR-PGL

(9.2 x IO"7M)Experi

ment10

053Experi

ment20ND°Experiment30

0046

" ND, not determined.

obtained on other resistant cell lines (16).Although K-562 cells have been reported to be nonphagocytic

(17), previous results have shown that a small percentage ofthese cells has the capacity to phagocytize inactivated cocci(18). Because PGL microspheres have a mean diameter of 0.45firn and are visible under a fluorescence microscope, we wereable to examine if K-562 cells could phagocytize ADR-PGL.As shown in Table 3, ADR-PGL were not taken up by K-562cells during an incubation of 2 h at 37°C,whereas in similar

experimental conditions about 50% of peritoneal cells haveingested the drug-microsphere complexes. This corresponds tothe usual percentage of phagocytic macrophages observed inthe peritoneal cavity of mice (10). In addition, when K-562 cellswere treated with ADR-PGL and examined for phagocytosis,no evidence of phagocytosis was found after 1, 2, or 3 days ofincubation. On the other hand, by contrast to previous resultsobtained with LI210 cells which bind ADR-PGL strongly totheir surfaces, K-562 cells did not firmly retain the micro-spheres at their surfaces, failed to develop bleb formations (6),and retained their characteristic cellular morphology (6). Atany one time during the incubation, an estimated 6 to 15microspheres were found in contact with the plasma membraneas opposed to LI210 cells which could carry several hundredmicrospheres at similar drug concentrations (not illustrated).

Because trace amounts (less than 0.09%) of free Adriamycin

might be released from the microspheres during incubationwith cells, it was important to test whether such a release couldaccount for the induction of erythroid differentiation. Comparing the concentration of free drug (40 nM) in sensitive cells withthat of coupled drug (1500 nM) necessary to result in 40% ofdifferentiated cells (Fig. 1), an estimated release of 2.7% oftotal ADR would be required to account for the induction. Acomparison of drug levels which induce 25% differentiated cells(l UM free drug and 4.6 /JM coupled drug) in resistant cellsindicates that a release of 22% of the bound drug would benecessary to produce this level of differentiation (Fig. 2). However, neither intracellular drug fluorescence which should beapparent at these levels of free ADR nor depletion of drugfluorescence from ADR-PGL occurred. Moreover, superna-tants obtained after mixing the highest differentiating ADR-PGL concentration (1.5 x 10~6M)with culture medium at 37°C

and then incubating it for 3 days with K-562 cells did notproduce any benzidine-positive cells. If 1% of the coupled drug(1.5 x 10~8M) was released from the microspheres, it would

have been detectable by fluorescent microscopy on incubatedcells. However, no fluorescence was observed after 2 to 3 daysof incubation. The recovered ADR-PGL preparations afterincubation with cells retained their full activity of growth inhibition, indicating that no drug release or chemical degradationoccurred during the incubation. Furthermore, free drug andplain microspheres fail to induce differentiation (Table 1),which indicates that intact ADR-PGL is responsible for theobserved biological effects. Previous experiments of incubationsof leukemic cells with ADR-PGL for similar lengths of timeresulted in a possible drug release of less than 0.09% (6). All ofthese observations argue against a possible role of drug releasein causing the induction of erythroid differentiation.

Labeling of K-562 cells with 55Fe permits the evaluation of

heme synthesis and the analysis of the types of hemoglobinsinduced after treatments with free or coupled drug. Hemesynthesis is strongly stimulated by both treatments; the incorporation of 55Fewas 1,693 ±172dpm/105 cells in the untreated

control samples, 20,888 ±774 dpm after treatment with ADR,and 34,858 ±1,929 dpm after ADR-PGL treatment. Furthermore, electrophoretic analysis of hemoglobin types reveals that,in uninduced sensitive cells, only the embryonic hemoglobinGower-1 was detected. When cells were induced with ADR-PGL or ADR, distinct differences were observed in the typesof hemoglobin synthesized (Fig. 3). Gower-1 was the predominant hemoglobin to be synthesized with ADR-PGL treatmentin addition to new embryonic hemoglobin types of Gower-2,Portland, and fetal (F). Similar types were produced with freeADR, but the major one produced was hemoglobin type X.Using resistant K-562/ADR cells, treatment with free or coupled drug leads to the synthesis of a new hemoglobin profile.In contrast to sensitive cells, uninduced resistant cells containedlow levels of Gower-1 and X. After treatment with free andcoupled drug, 3 additional bands could be identified as Gower-2, Portland, and hemoglobin F.

Analysis of 7-globin mRNA by cytoplasmic dot hybridizationin sensitive K-562 cells which were treated for 3 days with eitherfree or coupled drug is shown in Fig. 4. The results of thedensitometric scanning obtained by comparing untreated withtreated cells indicate a significant (approximately 2-fold) increase in 7-globin-specific mRNA after both types of drugtreatments.

DISCUSSION

It has been previously reported that ADR which has beencovalently bound to a solid-phase support can be used to restrict

1233

on May 27, 2021. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

INDUCTION OF ERYTHROID DIFFERENTIATION IN HUMAN LEUKEMIC K-562 CELLS

3A123456

I

G 2

G 1X

F

P

3B

II

MIGRATION

DIRECTION OF ELECTROPHORESIS : * —¿�-

Fig. 3. A, autoradiograph of "Fe-labeled hemoglobins of K-562 and K-562/ADR cells before and after induction to differentiate with free ADR and ADR-PGL. The positions of hemoglobin types Cower-1 (G I), Cower-2 (G 2), hemoglobin X (A'), fetal hemoglobin (F). and Portland (/") are indicated by arrows.Sample /, untreated K-562 cells; Sample 2, K-562 cells treated with 0.04 JIMADR; Sample 3, K-562 cells treated with 1.5 JIMADR-PGL; Sample 4, untreatedK-562/ADR cells; Sample 5. K-562/ADR cells treated with 1 JIM ADR; andSample 6, K-562/ADR cells treated with 4.6 MMADR-PGL. B, densitometricscanning of autoradiograph of "Fe-labeled hemoglobins. /, K-562 cells; //, K-562/ADR cells. •¿�ADR treatment; •¿�.ADR-PGL treatment; , untreatedcontrol cells.

drug action to the plasma membrane. This mode of drugpresentation results in a novel mechanism of cytotoxicity whichis not available to the free drug (7, 19, 20). This should bedistinguished from other applications of microspheres or polymeric supports for drug delivery where the fundamental mechanism of drug action is not altered and where the microspheresserve only as carriers for the slow release of therapeutic agents.In this study we demonstrate with sensitive and resistant humanleukemic K-562 cells that ADR-PGL complexes can producegrowth inhibition and an induction of cell differentiation.

As observed with the free drug, the kinetics of growth inhibition and the induction of hemoglobin-producing cells are doseand time dependent in both ADR-sensitive and -resistant cells(21, 22). However, obtaining a maximal differentiating effect

coi-

3

O

<O

2 0.5

<N

>X

0.5

RELATIVE CELL NUMBER

Fig. 4. Measurement of cyloplasmic mRNA for >-globin by dot hybridizationin K-562 cells induced to differentiate with either ADR (A) or ADR-PGL (•).Control cells were untreated (O). Inset, autoradiographic spots used for analysis:Lane 1. untreated; Lane 2, K-562 + 0.04 ?M ADR; Lane 3. K-562 + 1.5 >IMADR-PGL.

apparently requires higher concentrations of ADR-PGL. Thismay be explained by the fact that only a fraction of the micro-spheres at any one time is in contact with K-562 cells duringthe assay for erythroid differentiation. In addition, when micro-spheres are in contact with the cells, approximately one-twentieth of the microsphere surface is in direct contact with theplasma membrane (19). Such data suggest that immobilizedADR is a more effective entity than the free drug (7, 19).

It is unlikely that the differentiating effect could be mediatedby either internalization of microsphere or drug release fromthe microspheres (6). In studies made with fluorescence microscopy of K-562 cells treated with ADR-PGL during short-or long-term incubation, no examples for phagocytosis werefound (Table 3). We have obtained similar results by transmission electron microscopy.

A number of observations indicate that the induction ofdifferentiation is not due to free drug released from the micro-spheres. Comparisons of the inducing activity of free and boundADR show that a release of 2.7% in sensitive and 22% inresistant cells would be necessary to account for inducing thedegree of differentiation shown in Fig. 1. These levels of released drug would cause intracellular fluorescence and, furthermore, ADR-PGL would be gradually depleted of bound drugmolecules. However, none of these was observed. In addition,supernatants from cultures which were treated with the highestlevel of ADR-PGL did not induce differentiation. Earlier studies have demonstrated that less than 0.09% of the bound drugis released under similar conditions (6). Free ADR releasedfrom microspheres would also be expected to cause the induction of a similar hemoglobin profile as the one found in ADR-induced cells, which was not observed.

When plain PGL is tested in the same range of concentrationthat was used for ADR-PGL, neither differentiating nor

1234

on May 27, 2021. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

INDUCTION OF ERYTHROID DIFFERENTIATION IN HUMAN LEUKEMIC K-562 CELLS

growth-inhibitory effects are observed. Only at much higherlevels does PGL produce a steep growth inhibition curve; theIC50for PGL occurred at 60-fold higher levels than with ADR-PGL. At this concentration, the cells are surrounded by a largemass of microspheres, and toxic activity may be explained bynonspecific steric perturbations. In this kind of growth inhibition, no cell differentiation is observed; thus we conclude thatADR is essential for the induction of the differentiation process.

The induction of the erythroid differentiation pathway asjudged from the appearance of benzidine-positive cells is confirmed by the increase in heme biosynthesis. In separate experiments, we have shown that 10~3 M succinylacetone, a well-

known specific inhibitor of A-aminolevulinate synthetase, abolished both the incorporation of "Fe and the recruitment ofbenzidine-positive cells in the control and in drug-treated cells(data not shown). This experiment supports the view that 55Feincorporation into K-562 cells is significantly correlated withthe process of hemoglobin biosynthesis.

As demonstrated by electrophoretic analysis, free and coupledforms of ADR first induce a significant increase in the amountof hemoglobin found in either sensitive or resistant controlcells. Later they produce a modification in the hemoglobinpattern synthesized. Such an induction is characteristic of othertrue differentiating agents. New types of embryonic hemoglobins (Gower-2, X, Portland) and fetal hemoglobin F wereobserved in the patterns of electrophoretic migration (23). Freeand coupled forms of ADR also differ in their capacity toinduce different hemoglobin types. The microsphere-coupleddrug produces a specific and distinct induction of Gower-1,while ADR predominantly induces hemoglobin X. The reasonfor this differential expression of hemoglobins is unknown.

It has been reported earlier that growth inhibition alone isnot sufficient to induce differentiation (24-27). Interleukin 1induced a concentration-dependent inhibition of the proliferation in K-562 cells. It induced a decrease in transferrin receptordensity and an increase in Class I major histocompatibilitycomplex antigens (24). This cytostatic action of interleukin 1was not associated with direct cytotoxicity and was only partially reversible. Prostaglandin E or interferon did not appearto mediate these effects. Marcellomycin, an anthracycline an-tiobiotic with antineoplastic activity, was a potent inducer ofmaturation in the HL-60 promyelocytic leukemia cell line (25).However, ADR and carminomycin, which are potent cytostaticagents, were completely inactive as inducers of cell maturation.Actinomycin D was the most potent inducer of differentiationof HL-60 cells in another study (26). Vincristine, an inhibitorof mitosis, induced only a small increase in lysozyme which isa characteristic marker of differentiation (26). Arabinofurano-sylcytosine induced hemoglobin synthesis in K-562 cells maximally at concentrations of 5 x 10~7M (27). In contrast to the

reversible effects of hemin and hydroxyurea on globin synthesisin this cell line, the induction of hemoglobin synthesis byarabinofuranosylcytosine was irreversible. Aphidicolin, anotherinhibitor of S-phase DNA synthesis, was also a potent inducerof differentiation, but not vinblastine, an inhibitor of mitosis(27). Moreover, we have observed that the chemotherapeuticagents methylgag, cyclophosphazene, and gallium chloridewhich effect cell growth in a way similar to ADR and ADR-PGL did not induce K-562 cells to differentiate.4 The differen

tiating effect of free ADR and the bound drug was not reversible.If the free or bound drug was depleted from the medium,hemoglobin expression and growth inhibition persisted for 5days, after which the cell viability began to decrease. The lossof self-renewal capacity is consistent with terminal differentia

tion (27), but reservations apply to this system since the generation of reticulocytes is not observed in cell culture.

Dot blot analysis shows that both the ADR- and ADR-PGL-treated K-562 cells exhibited elevated levels of fetal 7-globin

mRNA, suggesting that both forms of drug are effective inchanging the erythroid phenotype, even though the coupledADR acts predominantly at the plasma membrane level. However, further experiments would be needed to ascertain if thismRNA accumulation is a consequence of stimulation in thetranscription process or is a result of an increase in the life spanof mRNA.

Restriction of ADR-PGL interaction to the cell surface dem

onstrates that the initiation of induction is occurring at thelevel of the cell membrane and suggests that intracellular secondmessengers are involved in the maturation of the signal. However, the location of the primary signal elicited by free ADRremains unclear. Although free ADR effects a great variety ofmembrane functions (21, 28), including the inhibition of ino-sitol lipid metabolism (29) which has been well documented inplaying a role in membrane signal transduction (30), the freedrug also intercalates into DNA and inhibits DNA function(31). The differences in induced hemoglobin types caused byfree and bound drug could reflect the free drug's broader distri

bution within the cell.The cell surface-mediated induction of differentiation by

ADR-PGL appears to be dependent upon the type of differentiation system being examined. We find that murine erythroleu-kemia Friend cells are not inducible by ADR-PGL. Others havereported that immobilized ADR is unable to induce differentiation in WEHI-3D D+ leukemia cells (32). However, both freeADR and the nonintercalative anthracycline AD-32 were activein the differentiation of these cells, which further suggests thatdrug-DNA intercalation is not a necessary step within theinduction mechanism. One important factor which remains tobe examined in the induction of differentiation by free andbound ADR is the involvement of signal-transducing events,such as the inositol lipid metabolism and the adenyl cyclase/cyclic AMP pathways (33). Relevant studies are in progress inour laboratories.

REFERENCES

1. Reiss, M., Gamba-Vitalo, C.. and Sartorelli, A. C. Induction of tumor celldifferentiation as a therapeutic approach: preclinical models for hemato-poietic and solid neoplasms. Cancer Treat. Rep., 70: 201-218, 1986.

2. Jeannesson, P., Ginot, L., Manfait, M., and Jardillier, J. C. Induction ofhemoglobin synthesis in human leukemic K 562 cells by Adriamycin. Anti-cancer Res., 4: 47-52, 1984.

3. Bloch, A. Induced cell differentiation in cancer therapy. Cancer Treat. Rep.,6«:199-205, 1984.

4. Reuben, R. C., Khanna, P. L., Gaziti, Y., Breslow, R., Rifkind, R. A., andMarks, P. A. Inducers of erythroleukemic differentiation. Relationship ofstructure to activity among planar-polar compounds. J. Biol. Chem., 253:4214-4218, 1978.

5. Yen, A., Reece, S. L., and Albright, K. L. Membrane origin for a signaleliciting a program of cell differentiation. Exp. Cell Res., 152: 493-499,1984.

6. Tokés,Z. A., Rogers, K. E., and Rembaum, A. Synthesis of Adriamycin-coupled polyglutaraldehyde microspheres and evaluation of their cytostaticactivity. Proc. Nati. Acad. Sci. USA, 79: 2026-2030. 1982.

7. Tritton, T. R., and Yee, G. The anticancer agent Adriamycin can be activelycytotoxic without entering cells. Science (Wash. DC), 217: 248-250, 1982.

8. Monsan, P., Puzo, G., and Mazarguil. H. Etude du mécanismed'établissement des liaisons glutaraldehyde-proteines. Biochimie (Paris), 57: 1281-1292. 1975.

9. Margel, S., Zisblatt, S., and Rembaum, A. Polyglutaraldehyde: a new reagentfor coupling proteins to microspheres and for labeling cell-surface receptors.J. Immunol. Methods. 26: 341-353. 1979.

10. Pages. J., Thiernesse, N., Bernard, J., Jeannesson, P.. Arnaud, D.. andBussard, A. E. Ultrastructure of mouse peritoneal cells engaged in spontaneous secretion of antibody against mouse and sheep erythrocytes andevolution of cell types during culture. Cell. Immunol., 35: 289-302, 1978.

1235

on May 27, 2021. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

INDUCTION OF ERYTHROID DIFFERENTIATION IN HUMAN LEUKEMIC K-562 CELLS

11. Bates, G. W., Billups, ('.. and Salt man, P. The kinetics and mechanism of

iron (III) exchange between chelates and transferrin. I. The complexes ofcitrate and nitrilotriacetic acid. J. Biol. Chem., 242: 2810-2815, 1967.

12. Krystal, G., Eaves, A. C., and Eaves, C. J. A quantitative bioassay forerythropoietin, using mouse bone marrow. J. Lab. Clin. Med., 97: 144-157,1981.

13. Mayeux, P., Felix, J. M., Billat, C., and Jacquot, R. Induction by hemin ofproliferation and of differentiation of progenitor erythroid cells responsiblefor erythropoietin. Exp. Hematol., 14: 801-808, 1986.

14. Rigby, P. W. J., Dieckmann, M., Rhodes, C., and Berg, P. Labeling deoxy-ribonucleic acid to high specific activity in vitro by nick translation with DNApolymerase I. J. Mol. Biol., 113: 237-251, 1977.

15. Slightom, J. L., Biechi, A. E., and Smithies, O. Human fetal G7- and *•>•-

globin genes: complete nucleotide sequences suggest that DNA can be exchanged between these duplicated genes. Cell, 21: 627-638, 1980.

16. Rogers, K. E., Carr, B. I., and Tokés,Z. A. Cell surface-mediated cytotoxicityof polymer-bound Adriamycin against drug-resistant hepatocytes. CancerRes., «.-2741-2748. 1983.

17. Lozzio, B. B., and Lozzio, C. B. Properties and usefulness of the original K562 human myelogenous leukemia cell line. Leuk. Res., 3: 363-370, 1979.

18. Rallet, A., Trentesaux, C., Jeannesson, P., Dufer, J., Bouchet, P., andJardillier, J. C. Phagocytosis of the streptococcal preparation OK-432 (Pi-cibanil) by nonphagocytic human leukemia K 562 cells and its modulationby tumor cell differentiation. Anticancer Res., 7: 535-540, 1987.

19. Ross, K. L., and Tökes,Z. A. Targeting of drug action to the cell surface bymicrospheres. In: A. Rembaum and Z. A. Tökes(eds.), Microspheres. Medical and Biological Applications, pp. 123-138. Boca Raton, FL: CRC Press,Inc., 1988.

20. Rogers, K. E., and Tökes,Z. A. Novel mode of cytotoxicity obtained bycoupling inactive anthracycline to a polymer. Biochem. Pharmacol., 33: 605-609, 1984.

21. Jeannesson, P., Ginot, L., Manfait, M., Jardillier, J. C., and Tökes,Z. A.Cell surface-mediated differentiation in human K 562 cells with polymer-bound Adriamycin. Anticancer Res., 5: 625, 1985.

22. Trentesaux, C., Gerard, B., Mayeux, P., Jeannesson, P., Jacquot, R., and

Jardillier, J-C. Differentiating activity of Adriamycin in human erythroleu-kemic cells: effect on globin and heme synthesis. Biochim. Biophys. Acta,1012: 161-165, 1989.

23. Vainchenker, W., Guerrasio, A., Testa, U., Titeux, M., Guichard, J., Beuzard,Y., Rosa, J., and Breton-Gorius, J. Multipotential capacities of differentiation and hemoglobin pattern of K 562 cell line. In: G. Stamatoyannopoulos(ed.). Hemoglobins in Development and Differentiation, pp. 507-518. NewYork: Alan R. Liss, 1981.

24. Lovett, D., Kozan, B., I ladani. M., Resch, K., and Gemsa, D. Macrophagecytotoxicity: interleukin 1 as a mediator of tumor cytostasis. J. Immunol.,136: 340-347, 1986.

25. Schwartz, E. L., and Sartorelli, A. C. Structure-activity relationship for theinduction of differentiation of HL-60 human promyelocytic leukemia cellsby anthracyclines. Cancer Res., 42: 2651-2655, 1982.

26. Lotem, J., and Sachs, L. Potential pre-screening for therapeutic agents thatinduce differentiation in human myeloid leukemia cells. Int. J. Cancer, 25:561-564, 1980.

27. Luisi-DeLuca, C., Mitchell, T., Spriggs, D., and Kufe, D. W. Induction ofterminal differentiation in human K 562 erythroleukemia cells by arabino-furanosylcytosine. J. Clin. Invest., 74: 821-827, 1984.

28. Tritton, T. R., and Hickman, J. A. Cell surface membranes as a chemother-apeutic target. In: F. M. Muggia (ed.). Experimental and Clinical Progressin Cancer Chemotherapy, pp. 81-131. Boston: Martin Nijhoff Publishers,1985.

29. Thompson, M. G., Chahwala, S. B., and Hickman, J. A. Inhibition of humanerythrocyte inositol lipid metabolism by Adriamycin. Cancer Res., 47: 2799-2803, 1987.

30. Berridge, M. J. Inositol triphosphate and diacylglycerol as second messengers. Biochem. J., 220: 345-360, 1984.

31. Arcamene, F. Antitumor anthracycline: recent developments. Med. Res.Rev., 4: 153:188, 1984.

32. Gamba-Vitalo, C, Blair, O. C., Tritton, T. R., Lane, P. A., Carbone, R., andSartorelli, A. C. Cytotoxicity and differentiating actions of Adriamycin inWEHI-3B D+ leukemia cells. Leukemia, /: 188-197, 1987.

33. Barnes, D. M. How cells respond to signals. Science (Wash. DC), 234: 286-288, 1987.

1236

on May 27, 2021. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1990;50:1231-1236. Cancer Res   Pierre Jeannesson, Chantal Trentesaux, Brigitte Gèrard, et al.   Bound to MicrospheresCells by Membrane-directed Action of Adriamycin Covalently Induction of Erythroid Differentiation in Human Leukemic K-562

  Updated version

  http://cancerres.aacrjournals.org/content/50/4/1231

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/50/4/1231To request permission to re-use all or part of this article, use this link

on May 27, 2021. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from