Characterization of differentiation-inducer-resistant HL-60 cells

Leukemia Research Vol. 9, No. 8, lap. 96"/-986. 1985

0145-2126/85S3.tX) ÷ OAK) :c, 1985 Pergamon P r e ~ Ltd.

Printed in Great Britain.

CHARACTERIZATION OF DIFFERENTIATIONINDUCER-RESISTANT HL-60 CELLS* ROBERT E. GALLAGHER, PATRICIA A. BILELLO, ANNA C. FERRARI, CHIN-SEN CHANG, RAY-WHAY CHIU YEN, WAYNE A. NICKOLS a n d EMIL C. MULY, III University of Maryland Cancer Center and Baltimore Cancer Research Program, NCI, NIH, Baltimore, MD 21201, U.S.A.

(Received 30 November 1984. Accepted 29 December 1984) Abstract--Sub-lines of the cultured human promyelocytic leukemia cell line HL-60 were individually selected for their ability to sustain exponential growth in the presence of 3 structurallyunrelated inducers of granulocytic differentiation - - retinoic acid (RA), dimethylsulfoxide (DMSO), and 6-thioguanine (6TG). Selections were made by step-wise augmentation to final drug concentrations of 10-3mM RA, 169raM (1.2070) DMSO and 0.12raM (20 I.tg ml-') 6TG. In addition to growth resistance, cells in each sub-line displayed variable cytodifferentiation resistance to each of the 3 selective agents, which was quantitated as the ratio of the concentration of drug required to induce differentiation in 50070 of the cells in each resistant sub-line versus comparably-passaged wild-type HL-60 cells. The levels of resistance/cross-resistance were as follows: RA-resistant (res) sub-line > 2700-fold to RA, 1.3-fold to DMSO and > 1.5-fold to hypoxanthine (HXN; the noncytotoxic purine base inducer analogue of 6TG); DMSO-res sub-line 2.5-fold to DMSO, 137-fold to RA and > 1.5-fold to HXN; and 6TG-res sub-line > 1.5-fold to HXN, 9-fold to RA and 1.6-fold to DMSO. These sub-lines were not cross-resistant to sodium butyrate (NaBut), a monocyte inducer, or to 12-0-tetradecanoylphorbol 13-acetate (TPA), a macrophage inducer. HL-60 sub-lines selected by exposure to a single high concentration of 5-bromo-2 '-deoxyuridine (BUdR; 3.3 x 10-~mM) or oubain (Ou; 5 x 10-~mM) were not or were slightly cross-resistant to either granulocyte or monocyte inducers. Although some variations in line/sub-line phenotype were observed, this was minor compared to the quantitative variations in response to individual inducing agents. The RA-res and 6TG-res sub-lines contained numerous double minute chromosomes (indicators of amplified genes) which were either absent or present in much smaller numbers in the parental wild-type cells or in the other drug-resistant sub-lines. There was little change or a decrease in the amplification level of the known amplified oncogene c-myc in the various drug-resistant sub-lines compared to wild-type HL-60 cells. These results (a) confirm that the neutrophilic granulocytic and monocytic/macrophagic differentiation programs in HL-60 cells are mechanistically different and separable; (b) suggest that both agent-specific and common quantitative alterations contribute to the mechanism(s) for resistance to granulocyte differentiation; and (c) suggest that the latter quantitative defects could be related to amplification of genes other than c-myc.

Key words: Myeloid leukemic cell differentiation, drug/differentiation resistance, double minute chromosomes, gene amplification, oncogenes. INTRODUCTION ACUTE m y e l o g e n o u s leukemia ( A M L ) is characterized b v the a c c u m u l a t i o n of persistently replicatively-active leukemic p r o g e n i t o r cells which fail to c o m p l e t e the g r a n u l o c y t i c - m o n o c y t i c m a t u r a t i o n sequence [27, 35, 37, 51, 60]. In m a n y cases it has been possible to at least

partially correct this i m b a l a n c e o f replicative vis-/~-vis differentiative f u n c t i o n s by t r e a t m e n t o f A M L cells in vitro with exogenous natural or chemical d i f f e r e n t i a t i o n inducers [6, 28, 40, 51]. In some myeloid l e u k e m i a culture systems evidence indicates t h a t the i n d u c t i o n of leukemic cell d i f f e r e n t i a t i o n involves alterations in levels

*This work was supported in part by PHS Grant R01-CA34793 awarded by the National Cancer Institute, DHHS. Abbreviations: RA, retinoic acid; DMSO, dimethylsulfoxide; 6TG, 6-thioguanine; BUdR, 5-bromo-2'-deoxyuridine; Ou. oubain; HXN, hypoxanthine; NaBut, sodium butyrate; TPA, 12-0-tetradecanoylphor~ol 13-acetate; AML, acute myelogenous leukemia; DM, double minute (chromosomes); HSR, homogeneously staining region; NBT, nitroblue

tetrazolium; NSE, non-specific esterase; ECsoW, effective conoentration to induce 5007o differentiation by Wright stain evaluation; ECsoN, effective concentration to induce 50°70 NBT-positive cells; EC~oE, effective concentration to induce 50°7o NSE-positive cells; ICso, concentration to inhibit 50070 cell growth. Correspondence to: Dr R. E. Gallagher, University of Maryland Cancer Center, Room 9-043 Bressler Research Bldg., 655 W. Baltimore St., Baltimore, MD 21201, U.S.A.

967

968

ROBERTE. GALLAGHERel al.

of intracellular physiologic effectors (stimulators or inhibitors) of growth and differentiation a n d / o r of responsiveness to extracellular mediators of growth and differentiation [24, 26, 37, 40, 44, 51, .58]. Thus, the development of A M L likely involves the development of resistance to physiologic effectors which function in normal myelopoiesis to direct orderly terminal maturation. A l t h o u g h little is known about the fundamental genetic mechanisms involved in such resistance, they might be analogous to those utilized by cells for the development of resistance to pharmacological cytotoxic agents. This seems plausible, since many pharmacological agents interact with specific physiological cellular targets and since the net effect of cytotoxic agents and inducers of terminal myeloid differentiation is the same - - the production of cell death. One well-documented mechanism by which tumor cells develop resistance to cytotoxic agents is by gene amplification [.52]. In several model systems, the amplified genes have been determined to specify an augmentation of a critical metabolic target for the cytotoxic agent [52]. However, in other systems involving multidrug-resistance the relationship between the amplified genes and the resistance mechanism appears to be complex and less direct, since it involves cross-resistance to chemically unrelated compounds [5, 3 2 1 . Amplified genes are frequently associated with pathognomonic cytogenetic indicators, viz., extrachromosomal double minutes (DM) or intrachromosomal homogeneously staining regions (HSR). Thus, it seems notable that DM or HSR have frequently been noted in primary cultures of various types of natural tumors, including untreated leukemias [3, 31]. Recently, such indicators have been associated with amplification of endogenous oncogenes, e.g. [1, 53], which, in turn, have been postulated to have a role in the regulation of normal cell growth and differentiation [251. The established human promyelocytic leukemia cell line HL-60 [l 1, 201 provides an in vitro model for investigating some of the above concepts. HL.-60 cells can be induced to terminally differentiate along 2 distinct differentiative pathways to either neutrophilic granulocytes or macrophages by a variety of alternative exogenous inducers [9, 13, 17, 19, 28, 49]. In addition, eosinophilic granulocytes have been described in some HL-60 cultures [18, 33, 391, and some inducers, e.g. vitamin D3 and sodium butyrate, induce differentiation to cells resembling monocytes [54, 8]. Factors regulating these various differentiative potentials are .poorly understood. However, the fact that terminal differentiation can be induced implies that the most critical genetic defects are quantitative and dysregulatory. Notably, the early culture generations of HL-60 cells were observed to contain a few DM [20], and, subsequently, the line was discovered to harbor the amplified oncogene c-myc [12, 14]. These findings could indicate a propensity of HL-60 ceils to use the genetic mechanism of gene amplification to selective advantage [12, 14, 21]. Thus, for the present experiments, sub-lines of HL-60 were selected for resistance to 3 chemically-disparate inducers of granulocytic differentiation - - RA, D M S O and 6TG - - using multi-step selection procedures as previously

utilized to select for resistance to cytotoxic agents b~ gene amplification [5, 32. 52]. The results suggest that these sub-lines will be useful for dissecting molecular genetic mechanisms involved in the dysregulation of differentiation in these cultured human leukemic cells. MATERIALS

AND METHODS

Cell culture conditions

Cells were routinely cultured in RPMI 1640 medium (GIBCO. Grand Island, NY) containing 10% heat-inactivated fetal calf serum (Sterile Systems, Inc., Logan UT, Lot 10027 or Flow Laboratories. Rockville. MD, Lot 29t01066t. in the absence of antibiotics [20]. For differentiation induction experiments, late log phase cells were washed x 2 with 5 mt of warm RPMI 1640 to remove spent medium and maintenance drugs, The washed cells were resuspended at 1.25 × 10" cells, 5 ml in complete medium containing appropriate concentrations of inducer and cultured in plastic T25 flasks (Falcon, Oxnard. CA) as indicated under Results. For r P A induction experiments, the washed cells were instead suspended at 2.5 > 10t cells/2 ml and cultured in plastic 6 well cluster dishes (Falcon). Master freeze stocks were prepared in RPMI 1640 containing 15°70 fetal calf serum and 7.5% DMSO (M A Bioproducts, Walkersville, MD) and stored at -170°C. All master freezes were demonstrated to be free of contaminating microbiological agents including culturable and non-culturable mycoplasma. Master stocks of wild-type (wt) HL-60 cells were utilized from freezes prepared at passage generations 12-14 and 48-50 from the inception of the cell line [I 1.20] and were. respectively, designated low passage (p 12 to 30) or high passage 1p50 to 70). All drug-resistant sub-lines were also utilized from within 20 passage generations of master freeze stocks. Selection o f drug-resistant sub-lines

The selection of the 6TG-res sub-line on 0.12raM (20v.g/ml) 6TG has been previously described [21]. The DMSO-res sub-line was selected by initiating wt HL-60 cells, passage 28, on 0.1°7.,0 DMSO and doubling the DMSO concentration every 2 weeks up to 1.2% DMSO. Cell growth was considerably retarded on concentrations of DMSO aho~e 0.8%, and a several week interval of irregular, slow cell growth occurred on 1.2% (I 69mM) DMSO before the relativelv stable sub-line used in the present studies emerged. During selection on 0.8 and 1.2% DMSO, numerous differentiated cells were noted, which gradually decreased in the population. Master stocks were prepared at 23 weeks from inception. As noted under Results, this sub-line has a marked tendency to form binucteate tetraploid cells but this is not necessary for resistance to DMSO. The RA-res sub-line was selected from passage 35 wt HL-60 cells by initial exposure to 10-+M RA. followed by sequential increases of one-half log concentration every 2 weeks. Some retardation of growth, attended by the presence of a variable fraction of differentiated cells, was noted from the inception. This was most marked at 10-~M RA at which concentration 4 weeks were required to secure a progressively expanding cell population. At higher RA concentrations up to 10-%1, progressive cell growth was noted, although a total of 16 weeks elapsed before ma~ter freeze stocks were prepared from cells t~ith the characteristics shown in Tables I and 2. For preparation of the BUdR-res and Ou-res sub-lines, wt

Differentiation-inducer-resistant

HL-60

cells

H L-60 cells 0745) were exposed in mass culture (50 ml of cells at

film

IO* cells/ml)

immersion

objective.

In TPA

induction

IO-‘mM

to either 3.3 x IO-‘mM

Ou. After

mined by trypan

(IO kg/ml)

BUdR

or 5 x

4 days most of the cells were dead as deterblue dye exclusion,

and the remaining

viable

cells were collected by centrifugation over a pad of FicollPaque (Pharmacia. Piscataway, NJ). The recovered viable cells horn each culture

(<

well in complete

medium

10’) were each placed in a microtiter

anded to mass culture.

plate

in the absence of drug and re-exp-

These cells were then plated at 10’ cells

per ml in I % methylcellulose.

as previously

\\a\ amicipated

clones of resistant cells could be

picbed:

that individual

however,

the plates were confluent

Nld it was possible to immediately on full dosages of either BUdR doubly-resistant

to BUdR

\omatic

cell hybrids;

\clecting

BUdR-res

generally

similar

\ub-lme

and Ou (universal

fails

MD).

(31. the Ott-res

HL-60

by

lacks

to the weak

Slides for cytologic

crntrifuge

evaluation,

which contains a-naphthyl counterstain

(Sigma

instructions.

I’crentiation

(Table

separate

in

r‘erentiation. e\aluatrd Photographs

appeared

at

Cytologic

1000x

magnification,

-i-\,I,

from

evaluation

using an Olympus were taken

kit

L

SI I LC‘TlOh

estimated

tests

difof 3

at

400x

model

black

<‘OY,~,,lONS

of adherent microscope

following

to a published

using the inverted

semi-quantitative

+ + ( 2s50%;

+ + +,

determined

by

cells after removal procedure

for TPA-

of clumped

cells was

microscope,

scale: 0.0;

so-75%;

&

10%;

+ ‘+ + +,

using the

+,

IO-25%;

>75%.

CvroReneric procedures Cytogenetic DM

content

determinations were

based

as previously

described

of

chromosome

on counts

metaphase cells/specimen,

of

number

a minimum

which were prepared

and of

50

and evaluated

1211.

Procedures for All

oncoyene

procedures

for

analysis

the preparation

indicated S.

BH2.

and white

AN,)

(‘UI

of cellular

DNA,

analysis

Astrin

have

DNA,

for

been

(Philadelphia,

PA):

previously

c-N-ras,

from avian M. Erisman

from

probe pHB-I

both gifts from D. Lowy,

pKX162

human

(1551; a gift from M. Wigler,

NY); v-Ha-ras, ([IS];

v-mnvb probe MD);

((301; a gift from

cells, probe p6al

ras, probe pKBE-2

for

molecular

plasmid probes with their sources

virus, MC29

Cold Spring Harbor,

((461;

a gift

from

I and v-KiBethesda. S.

Baylin,

and v-abl probe pAbl sub 9 (1571; a gift from

R. G. Smith and B. Ozanne,

Dallas,

TX).

RESULTS Sub-line

selection

and suspension

culture

characteristics

in Materials and Methods, HL-60 sublines were selected for resistance to RA, DMSO and 6TG by initially exposing wt HL-60 cells to a low drug concentration and then step-wise augmenting the drug concentration. On the other hand, sub-lines resistant to BUdR and Ou were selected from rare surviving ceils As

TUKE

detailed

CHAKACTEKISTIC‘S

DKLI(;-KLSISTANT

HL-60 CEI.I.S

Maintenance drug concentration

Vital culture appearance?

OF

Cell doubling time (11)

(mM)

WILD-TYPE

AND

Maximum cell density ( x lO”/ml)

\\7.

p<30

-

-

ss

33.9f2.7

2.5f0.3

\\I.

p50-70

-

-

ss

30.5k4.2

2.6kO.6

DMSO-res

M

1.7x

10:

SC

44.4k7.93

2.2kO.4

hTC-res

hl

1.2 x lo-’

ss

41.4+4.6$

2.0f0.3

K.&-l?\

XI

1.0x

ss

24.9zt3.0:

2.8ztO.6

HUdR-rcr

S

3.3 Y 10-z

ss

30.3zh3.4

2.7kO.5

Ott-res

S

5.0 x lo-’

SC

25.4k2.1:

3.ozbo.9

lo-‘

and

were used in the present studies: v-myc,

neuroblastoma

MD);

probe

and

[21]. The following

to

from at least 5

a minimum

Selection mode*

according

1000 oil-

the percentage

cells 1431. The percentage

dif-

and NSE tests were

NBT

Panatomic-X

Line/ cub-line

as minute

of induced

microscope,

with

1.

optically

Baltimore,

to the

of spontaneous

For evaluation

200 cells were evaluated

mclgnifisatton.

dye 1211.

on cyto-

according

2), 500 cells were evaluated

Lcpat-ate experiments.

treated HL-60

myelocytomatosis

described

positivity

experiments.

by trypsinization

Apo

using an inverted

cell counts of suspension cells and adherent

detailed

acetate esterase wjith a Hematoxylin

NSE

an SPlan

and was quantitatively

reactions

test was performed

dots. For evaluation

IM)

hybridization

stains, and NBT

MO),

model

data with D.

results).

St Louis,

procedures,

of oncogene

with a commercially-available

Corp.,

large brownish-black

Wright

with

estimated

radiolabelling

does not differentiate

as previously

esterase (NSE)

slide preparations

(Olympus,

and

tests were performed

The non-specific

cells was optically

NY)

differ-

Friend erythroleukemia

Ott (unpublished

Rochester,

they have

sub-line

(unpublished

sub-line

prepared

to the BUdR-res

BUdR-res

Ou-res

cells for

of cell differenriarion

reduction

entirely

donor

Ou and,

to respond

Unlike

response to up to IO-*mM

suppliers

The

effect of BUdR

Koss. Baltimore,

Erolrrar/ot~

IO-‘mM properties

results).

colonies)

I and 2). Cells

145)) have been similarly

and

entiation-inducing cell\

or Ou (Tables

differentiative

kinase

(>2500

1211. It

harvest and grow these cells

cells in 5 x

(unpublished

thymidine

described

(Kodak,

969

*“rl. multi-step \elestton: S. single step selection. +Fir
970

ROBERT l~. GALLAGHER et aL

following exposure to a single high drug concentration. After the preparation of master freeze stocks, the various drug-resistant sub-lines were maintained on the drug concentrations indicated in Table l and were used over a frame of 20 subsequent passage generations for comparative studies with low passage (p12-30) and high passage (p50-70) wt HL-60 cells. Cells from all sub-lines grew freely in suspension without adherence to the plastic culture vessel. A minor portion of cells in the DMSO-resistant and Ou-resistant sub-lines formed small clumps of 2-10 cells; cells from the other sub-lines grew as single cells. The sub-lines showed approximately the same degree of cell size heterogeneity as wt HL-60 cells [20], except the DMSOresistant sub-line, which contained more large cells. The RA-resistant cells tended to be somewhat more irregular in shape and to shed more fragments into the culture medium. The Ou-resistant sub-line consistently produced more acidification o f the culture medium, which could not be accounted for by its high growth rate (see below) or by mycoplasma contamination (all lines and sub-lines are mycoplasma-free): perhaps it is related to a Ou-induced change in ion transport. In agreement with previously published results [201, the low passage wt cells had an average cell doubling time of 33.9 h; this was somewhat less in the high passage wt cells (Table l). The RA-resistant and Ouresistant sub-lines had significantly reduced cell doubling times, while the DMSO-resistant and 6TG-resistant sub-lines had significantly increased cell doubling times compared to either low or high passage wt cells (Table l). The peak cell density achieved over a 7-day culture period in RPMI 1640 medium containing 10070 fetal calf serum varied in parallel with the cell growth rate (Table 1).

Cytologic characteristics under routine culture conditions The predominant immature myeloblasts-promyelocytes present in the drug-resistant sub-lines closely resembled those previously described in wt HL-60 cell cultures [20]. That is, as cytologically analysed after Wright staining, the cells were generally round with a centrally-located nucleus occupying greater than 75070 of the stained cell surface. The nucleus typically consisted of fine chromatin material and contained 1--4 nucleoli. The peripheral rim of cytoplasm was quite basophilic. Numerous coarse azurophilic granules were usually observed in the cytoplasm and over the nucleus. Irregular cytoplasmic membrane projections or pseudopodia were more frequently observed in the RA-, Ouand 6TG-resistant sub-lines. The DMSO-resistant subline was exceptional in that it was more heterogeneous in appearance, the majority of ceils being larger than wt cells. This included an increased number of binucleate cells (12.2+9.6°70 vs 2.6-1-1.6070), of multinucleate cells (2.9-+-2.507o vs 0.3-1-0. ! 07o)and of giant cells (about half of which contained > 1 nucleus) with a diameter greater than 40 ~M (3.8-1-2.4°7o vs 1.1"+'0.3O7o). Also, the DMSO-resistant cells contained particularly conspicuous azurophilic granules and intensly basophilic cytoplasm. These characteristics resemble those recently

reported by another laboratory for DMSO-resistant HL-60 cells selected by an alternative procedure [56]. As summarized in Table 2, all of the drug-resistant sub-lines showed some degree of spontaneous granulocytic differentiation which was the same or only slightly reduced compared to high passage wt HL-60 cells. Granulocytes containing large, bright-red eosinophilic granules at the myelocyte to the mature granulocyte stage were the predominant identifiable form in the wt and drug-resistant cultures. Eosinophilic granules were also frequently observed in HL-60 cells without other obvious signs of maturation. This observation is contrary to the original description of the cell line [20] but is consistent with some more recent reports [18, 39]. Mature segmented neutrophils were only infrequently observed. A minor sub-population of monocytoid forms was also observed as reported by others [8]. A marker stain for monocyte-macrophage differentiation, the unaphthyl acetate non-specific esterase (NSE) stain, was weakly positive in a low percentage of cells from all of the lines/sub-lines tested (Table 2), although this did not always coincide with cells showing monocytoid differentiation.

Studies of induced granulocyte and monocyte/macrophage maturation The observed differentiation response of the various drug-resistant sub-lines depended primarily on the differentiation inducer employed, and, therefore, the results are considered below under individual agents. The most notable differences between the variant and wt HL-60 cells were quantitative. Such measurements depended on our observations, in agreement with previous reports, e.g. [24], that the fraction of cultured myeloid cells which responded to an inducer was directly-related to the inducer concentration. Thus, our standard procedure was to perform an inducer doseresponse curve for each of the sub-lines and wt HL-60 cells near the time of maximal differentiation induction. From a minimum of 3 entirely separate experiments, the concentration of inducer producing differentiation in 50070 of cells as judged by Wright stain cytology (effective concentration-50W; EC~0W) was determined, and the EC~0W was used to compare the sensitivity to differentiation induction of the various resistant cells and both low and high passage wt HL-60 cells. Cytologic evaluation of Wright-stained cells was routinely used as the most sensitive, reliable and informative procedure. The nitroblue tetrazolium (NBT) dye reduction test and the NSE stain were used, respectively, as adjunctive tests for granulocytic and monocyte/macrophage functional differentiation. As described individually below, the results of these 3 tests did not always conform to anticipated results for the various inducers, suggesting some qualitative variations in the induced phenotypes of the lines/sub-lines. Dimethylsulfoxide (DMSO). Figure I shows the determination of the EC~oW values for the various lines/sublines after 7 days of treatment with various concentrations of DMSO. These results, with their standard deviations are summarized in Table 3 and indicated the following: (a) the EC,oW doses for the low and high

971

Differentiation-induCer-resistant HL-60 cells TABLE 2. SPONTANEOUSDIFFERENTIATIONCHARACTERISTICSOF WILD-TYPEANDDRUG-RESISTANTHL-60 CELLS Differentiated characteristic (% observed/positive) Wright-stain cytology

Line/ sub-line

Total differentiated

Neutrophilic cells

Eosinophilic cells

Wild-type, p12-30

15.3"+7.8

1.8"+'0.5

Wild-type, p50-70

5.1-1-3.2

0.3±0.4

1.0±0.7

RA-resistant

3.2±4.3

0.3±0.3

0.7±0.4

DMSO-resistant

4.8±3.8

6TG-resistant

4.7-+-4.7

0.4±0.5

0.7"t-0.4

BUdR-resistant

8.9±4.1

0.4±0.6

1.2-1-0.7

Ou-resistant

2.6±1.5

<0.1

2.8"+2.5

1.8±0.9

<0.1

0.2±0.2

Monocytic cells 0.8::1:0.9 0.6±0.9 <0.1 0.2±0.2 0.2±0.2 <0.1 0.8±1.0

NBT

NSE

9.2--I-5.4

3.1-1-2.7

5.3±2.9

2.2±2.4

4.0±0.9

3.4±2.5

2.6±1.8

2.8"t-3.8

1.2__.2.0

4.1 ±2.9

4.6±2.7

5.0±6.0

3.0±3.6

2.7±3.4

t~ -,-J i o c

d ac

0.4 1561

0.8 (12"Z)

1.2 116~

1,6 12251

2.0 12~)

PERCENT OMSO(mM) F](;. 1. Cytologic differentiative responses of HL-60 lines/sublines to various concentrations of DMSO after 7 days treatment. Differentiated cells were determined by cytologic evaluation of Wright-stained preparations, and they included all cells beyond the basal myeloblast-promyelocyte stage of differentiation. Data points represent the average of at least triplicate independent experiments. The ECs0W data reported in Table 3 provide an index of the level of experimental variation. ( 0 ) wt HL-60, low passage; (C:) wt HL-60, high passage; (®) DMSOres sub-line; (A) 6TG-res sub-line; (BI) RA-res sub-line; (V) BUdR-res sub-line; (x) Ou-res sub~-Iine. passage wt HL-60 cells and the BUdR-res and Ou-res sub-lines did not significantly differ; (b) the EC,oW of the DMSO-res sub-line was approximately 2.5-fold increased in comparison to wt cells; and (c) the EC,oW doses of the 6TG-res and RA-res sublines were also significantly increased, albeit to a lesser degree. The ratios of the EC,0W's between the sub-lines and wt cells was taken to represent the relative sensitivities/resistances to the inducer, and they are re-summarized in Table 4

(column 1) with statistical indications of the significance of the differences. These ratios are presumed to represent minimum sensitivity/resistance differences, since all cells judged to demonstrate intermediate levels of granuIocytic differentiation, i.e. myeiocytes or beyond, were included in these determinations. ECso determinations using the NBT test (EC~oN) as the differentiation marker paralleled the cytologic determinations, except in the BUdR-res sub-line in which the EC,oN was disproportionately elevated (Table 3). Finally, the sensitivity to differentiation induction did not appear to be closely coupled to the inhibition of cell growth compared to cells cultured in the absence of drug (inhibitory concentration-50; IC,0, Table 3). Significant variations in cytologic appearance were observed amongst the different lines/sub-lines after the induction of terminal differentiation with DMSO. In the l o w passage wt line the predominant mature differentiated cells were metamyelocytes and segmented neutrophilic granulocytes as previously described [13, 20]. In the higher passage wt cells and the variant sub-lines, segmented neutrophils were seen less frequently, the predominant mature cells being metamyelocytes or banded forms with variable amounts of incomplete nuclear lobulation (Fig. 2A). In some instances, most notably in the RA-res and Ou-res sub-lines, the differentiated nuclei of many cells failed to show evidence of nuclear segmentation or of the 'rope-like' chromatin characteristic of mature neutrophils. The latter were not infrequently present in an abundant pale staining cytoplasm with fine vacuoles and extended processes, giving these cells a monocytoid appearance (Fig. 2B). Both the RA-res and Ou-res cells were, however, observed to undergo more typical granulocyte differentiation either spontaneously or in response to alternative inducers (Fig. 2C).

972

ROBERT E . GALLAGHER etaL TABLE 3. DIMETHYLSULFOXIDE (DMSO): DIFFERENTIATION AND GROWTH RESPONSES OF WILD-TYPE AND DRUG-RESISTANT H L - 6 0 CELLS

Dose-response data (mM) Line/sub-line EC,0W

EC~oN

wt, p < 3 0

90±11

84

wt, p50-70

91+-t4

109 >280

DMSO-res

224± 14

6TG-res

140±7

RA-res

147

Differentiated/positive cells (%) in 169mM DMSO Total

IC~o

Mature

NBT

NSE

157-1--0 9 5 - 1 - 5

53__+26 76--t-12

2--I-2

104-t-3

14±10

1±1

91-1-6

>280

69±20

7±4

1-4-1

1± 1

4±4

197±20

8 0 ± 17

35 ± 11

6 0 ± 13

61 ±34

120-1-10

179

>210

93±5

2 6 ± 11

49±24

71 ___25

BUdR-res

97-t-13

189

179±15

94±5

33±16

36±19

6±4

Ou-res

94-1-6

90

>210

97±2

7±10

85+'16

4±3

TABLE 4. RELATIVE RESISTANCE TO INDUCTION OF GRANULOCYTE AND MONOCYTE/ MACROPHAGE DIFFERENTIATION OF DRUG-RESISTANT HE-60 SUB-LINES COMPARED TO HIGH AND LOW PASSAGE WILD-TYPE HL-60 CELLS

Line/sub-line

DMSO-res

6TG-res

RA-res

BUd R-res

OU-res

Passage of wt HL-60 control

EC~oW relative to control in response to: DMSO

HXN

High

2.5*

>1.5"

Low

2.5*

>1.9"

High

1.6*

> 1.5*

Low

1.6"

>1.9"

High

1.3"

>1.5"

Low

1.3"

>1.9*

High

1.1

1.0

Low

I. 1

1.3

High

1.1

0.8

Low

I. I

1.0

RA

NaBut

TPA

137.3"

0.2*

0.3*

824.0*

0.9

3.7*

8.6*

0.4

0.1"

64.6*

1.9

1.2

> 2 . 7 × l0 s*

0.9

0.1"

>2.0 x 10"*

3.6*

1.2

1.6

0.8

0.4*

11.9*

3.6*

4.4*

1.3

1.4*

0.3*

9.8*

5.9"

3.1 *

*p <0.05. As previously indicated, the N B T test was positive in responding cells, seemingly i n d e p e n d e n t l y o f the cited cytologic variations in terminal differentiation. Surprisingly, the NSE test was consistently positive in the RAres a n d 6TG-res sub-lines (Table 3). A l t h o u g h not always associated with any characteristic cell type, irregular m o n o c y t o i d cells with a n a p p a r e n t i n t e r m e d i a t e level of d i f f e r e n t i a t i o n seemed m o s t frequently and heavily NSE stained; the most d i f f e r e n t i a t e d cells were almost always negative. Tests at shorter time intervals a n d with lesser a n d greater a m o u n t s o f D M S O did not give higher NSE d e t e r m i n a t i o n s in any o f the sub-lines (not shown]. Cells c o n t a i n i n g eosinophilic granules were decreased in all lines/sub-lines on all doses o f D M S O tested c o m p a r e d to u n t r e a t e d cultures (Table 2).

Hypoxanthine (HXN). T h e non-cytotoxic purine inducer H X N was used in place o f the specific-selective purine base 6 - t h i o g u a n i n e (6TG), which, in o u r h a n d s , is a n effective granulocytic inducer o f HL-60 cells only after acquisition o f resistance to the cytotoxic effect o f 6TG (21). As s u m m a r i z e d in Tables 4 ( c o l u m n 2) a n d 5, the pattern o f sensitivity/cross-resistance to H X N was the same as noted for D M S O . T h a t is, the low a n d high passage wt HL-60 cells a n d the B U d R - r e s and Ou-res sub-lines s h o w e d little difference in sensitivity, while the 3 sub-lines selected for resistance to 6TG, D M S O a n d RA were cross-resistant. T h e ability to test for sensitivity at higher doses o f H X N was restricted by the solubility limit o f H X N (6mM). Hence, a t t e m p t s were made to c o m p a r e the inducing effects of 6 m M H X N a f t e r up to

B

A

W 4

C

FK:. 2. Photographic examples of differentiative responses of wt and variant HL-60 cells to DMSO and HXN. All photomicrographs at 1000 x magnification, as described in Materials and Methods. (A) high passage wt I--IL*60cells treated 7 days with 1.2aT0 DMSO; (B) RA-res cells treated 7 days with 1.2~0 DMSO; (C) Ou*res cells treated "7 days with 6 mM HXN.

973

V

B

A

i ¸¸ ~

. . . .

~ ~

s ¸

C

Fk;. 6. Photographic examples of wt and variant HL-60 cells after treatment with monocyte/macrophage inducers. (A) low passage wt cells after treatment with 0.5 mM NaBut for 7 days; (B) low passage HL-60 cells after treatment .with 0.25 mM NaBut for 7 days; (C) RA-res cells after treatment with 1.6 × 10-' m T P A for 2 days.

974

Kbp

Kbp

I

2

3

4

5

6

7

I

2

3

4

5

6

7

8

21.2 ~23.1 7.4B

5.8~ Ill 5.6~ 4.9--

/

~9.4 ~6.T ~4.3

3.5~

~2.3 "--?-.0

Fit;. 9. Variable c-myc gene copy numbers in wild-type and drug-resistant HL-60 cells. The radiolabelled DNA used was pBR322 plasmid DNA containing a 2.9 kb Barn HI restriction fragment of the avian myelocytomatosis virus, (MC29), the v-myc oncogene (see Materials and Methods). (A) Cellular DNAs digested with Eco RI; Lanes: (1) 10 copies of pBR325; (2) normal human liver (single copy/haploid genome); (3) wildtype HL-60 p59; (4) DMSO-res; (5) RA-res; (6) BUdR-res; (7) 6TG-res. (B) Cellular DNAs digested with Sst-l. Lanes: (1) 10 copies of pBR325 DNA; (2) normal liver; (3) wild-type HL-60, p31; (4) DMSO-res; (5) RA-res; (6) BUdR-res; (7)6TG-res; {8) Ou-res.

975

Differentiation-inducer-resistantHL-60 cells

977

"C

o

2O

I

4 o (/~

I i IOi

o

i Z

i

i

,

i

i

,

I100 0

a.

25

8

D

..~ 2o

2o ~

W

I,LI t5

*0

rO

60

11: W

80

~)

~,

6

8

IOI00

DAYS

Fl~,. 3. Effect of 6 mM HXN on the growth and differentiation of HL-60 lines/sub-lines as a function of time. Differentiation is reported as the percentage of positive cells by the NBT dye reduction test, although parallel results were observed by cytologic evaluation. The cells were diluted by an equal volume of fresh medium "+ HXN on culture day 5, dotted vertical line. (A) wt HL-60, high passage; (B) 6TG-res sub-line; (C) RA-res sub-line; (D) DMSO-res sub-line; (El BUdR-res sub-line; (F) Ou-res sub-line. (Q) cell count in absence of HXN; (C)) cell count in presence of HXN; (A) % NBT positive in absence of HNX; ( A ) % NBT positive in presence of HXN. T\III

F 5. H ' ~ P O X A N T H I N E

(HXN):

DIFI-ERENTIATION

DRU(.i-RESIS'fAN'I

Line/sub-line

Dose-response data (mM) EC,oW

EC~oN

wt, p<30

3.2"+1.3

wt. p50-70 DMSO-res 6TG-res RA-res

AND GROWTH

HE-60

RESPONSES OF W I L D - T Y P E

AND

CELLS

Differentiated/positive cells in 6mM HXN

IC,o

Total

Mature

NBT

NSE

4.0"+0.3

4.7"+1.4

79"+15

35"+6

88"-I-4

0

4.0"+0.7

4.0"+0.2

>6

78.+17

35.+13

78.+6

0

>6

>6

>6

14"+ 11

1± 1

4"+4

0

>6

>6

>6

24"+18

1"+1

22"+16

0

>6

>6

>6

21-1-12

1"+1

16:t:11

0

BUdR-res

4.2.+0.9 4.6.+0.3

5.4_+0.8

86"1-12

32"+15

82:t:10

0

Ou-res

3.1___0.2 4.0-1-0.4

>6

94"+4

32-t-12

86"+15

0

1{I days o f treatment. No indication o f significant diffete,aces in the sensitivity to differentiation induction ~erc observed a m o n g s t the 6TG-res, D M S O - r e s and RA-res sub-lines (Fig. 3). It is a p p a r e n t that the actual lcxcl o f resistance for these sub-lines is m u c h greater than 1.5-fold. These results, shown for the NBT test, ,x ere c o n f i r m e d by cytological examination in which the NBT tests correlated best with the total fraction o f dif-

ferentiated cells rather than metamyelocytes and segmented f o r m s (Table 5). Cells f r o m all o f the cell lines/ sub-lines, a p p e a r e d c o m m i t t e d to neutrophilic granulocyte m a t u r a t i o n , since no increase in m o n o c y t o i d or eosinophilic form~ xxas observed on any tested concentration o f H X N . Tile NSE test was uniformly negative. In experiments involving u n d i s t u r b e d culture, 50% or more inhibition o f cell growth by 6mM H X N was

978

ROBERT E. GALLAGHER

observed at 5 or 7 days only in cultures of wt HL-60 cells and BUdR-res cells (Table 5 or not shown). However, in experiments in which cells were diluted with HXNcontaining medium after 5 days culture, significant growth inhibition by HXN occurred irrespective of the level of terminal maturation (Fig. 3). Retinoic acid (RA). The overall pattern of sensitivity/cross-resistance to RA-induced granulocyte differentiation amongst the lines/sub-lines was similar to that noted for DMSO and HXN (Fig. 4 and Tables 4, column 3 and 6). Some notable differences were obtO ..I ..I

I

tO

0

IO"°

IO4

IO"r

IO4

IO"5

RETINOIC ACID [ M ] FIG. 4. Cytologic differentiative response of H L - ~ lines/sublines to various concentrations of RA after 5 days treatment.

Details and symbols as in Fig. 1. served, however. First, the low passage HL-60 cells were exquisitely sensitive to RA induction with an average ECsoW of 4.8 x l0-'° M, while the higher passaged cells had a 7-fold higher ECsoW of 3.6 x l0 -° M. Second, in contrast to the DMSO-res and 6TG-res cells (also see Ref. [21]), the homologous RA-res cells were absolutely resistant to the differentiation-inducing effect of RA, since little evidence of differentiation was observed up to acutely cytolytic RA concentrations,.5 x 10-' M (Fig. 4 and not shown). Third, the response to RA was noted over a 3-log concentration range in contrast to other granulocyte inducers which exerted their full effect over a single log range. This broad activity range probably accounts for the greater variation in experimental results in different experiments compared to those with other differentiation inducers. The terminally differentiated product was the neutrophilic granulocyte, although the nuclear structure of these cells was more atypical than following DMSO- or HXN-induced neutrophilic differentiation. No stimulation of eosinophilic or monocytic differentiation was observed. The NSE test was negative or very weakly positive in a few cells. Inhibition of cell growth was variable and did not seem tightly-coupled to the differentiative response (Table 6). Sodium butyrate (NaBut). As for RA-induced differentiation, the high passage wt HL-60 cells were several-fold (4.2-fold) less responsive to NaBut than low passage wt HL-60 cells, as determined by the ECsoW (Table 7). The drug-resistant sub-lines fell within this

et aL

response range except for the Ou-res cells which were slightly but significantly less responsive (1.4-fold) than comparably-passaged (high passage) wt cells (Table 4, column 4). These variations in cytologic differentiative response were generally paralleled by variations in the NBT test (EC,oN; Table 7). A notable exception was the DMSO-res sub-line in which the NBT test was negative or only weakly positive in a majority of cells, despite its relatively high responsivity by cytological criteria. Since the 6TG-res sub-line had an intermediate level of responsivity (Table 7), it was selected to demonstrate the type of data generated to substantiate the results presented above and below (Fig. 5). The terminal differentiative response of the lines/subtines to NaBut markedly differed from that in response to the granulocyte inducers described above. Segmented neutrophilic granulocytes were rarely observed. Instead, as previously reported by Boyd and Metcalf [8], the predominant terminally-differentiated cell after treatment with the most effective concentrations of NaBut (0.5--0.75mM) rese/nbled a monocyte (Fig. 6A). Typically, this cell was reduced in size compared to untreated cells and had a reduced nuclear:cytoplasmic ratio ('<0.75). The nuclear chromatin, while partiallycondensed, was less so than following neutrophilic granulocyte differentiation. The nucleus, which was usually eccentrically-located, varied in shape from being round or oblong, frequently with a nuclear fold, to having 2-4 bulbous lobes. The cytoplasm was frequently irregular with some extended pseudopodia. It stained pale blue-gray and sometimes contained fine vacuoles and a few granules. These differentiated cells did not clump or adhere to the culture vessel. Notably, on less than optimal inducing concentrations of NaBut (0.1-0.25mM), many cells contained prominent eosinophilic granules (Fig. 6B). Although these eosinophilic cells never reached the same high percentage of the total cells as the monocytic cells, since less inhibition of cell growth occurred at the lower NaBut concentrations, the total number of these cells accumulated was substantial, in some instances exceeding the absolute numbers of accumulated monocytoid cells (Fig. 7). Such an accumulation of eosinophilic-granute-containing cells was not observed with any of the other inducers tested. The percentage of cells showing NSE positivity increased as a function of monocytic differentiation (Fig. 5 and Table 7). Although this positivity in all likelihood reflected the monocytic differentiation lineage, it is notable that cells which cytologically appeared most mature frequently had a lesser degree of NSE positivity than less mature-appearing cells. Similarly, in kinetic studies the NSE positivity frequently reached higher levels before maximal monocytic differentiation (Fig. 5) and in some of the sub-lines, e.g. the Ou-res and RA-res sub-lines, the induction of NSE positivity was antecedent to and exceeded terminal monocytic differentiation (Table 6). These results imply that NSE expression is maximal in sub-mature cells and/or that NSE expression and monocytic differentiation are not tightly coupled. 12-O-tetradecanoylphorbol 13-acetate (TP,4). TPA was tested for its macrophagic inducing effect bet~een

Differentiation-inducer-resistant HL-60 cells TABI.E 6.

RETINOIC ACID

(RA):

979

DIFFERENTIATION AND GROWTH RESPONSES OF WILD-TYPE AN[) I)RUG.RESISTANT HL-60 CEI.LS

Dose-response data ( × 10-gM) Line/sub-line

EC~oW

EC~oN

wt, p < 3 0

0.5-1"-0.1

wt, p50-70

3.6+__2.0

DMSO-res

411.7-t-272.4

6TG-res

31.2"t-5.0

Differentiated/positive cells in 10-6M RA

IC~0

Total

Mature

48

210"4-180

97-t-4

62"t-16

695-3

13

30"1-32

95__.9

69"+21

76"t-19

~ 103

- 103

72"+-19

4"t-3

3--t-2

NSE 1.5 <0.5 1.5

480

2505-80

94__.8

375-28

67"1-13

>10'

>10'

>10"

9--t-6

2-4-1

4"+-4

3.5

BUdR-res

5.7"+-3.9

>103

98"+106

97__.2

665-17

285-10

<0.5

Ou-res

4.7"1-4.6

775-12

865-22

<0.5

Ra-res

TABLE 7. SODIUM BUTYRATE

(NaBut):

3

75-5

100_0

<0.5

DIFFERENTIATION AND GROWTH RESPONSES OF WILD-TYPE AND DRUG-RESISTAN'I

Dose-response data (raM) Line/sub-line

NBT

HL-60

CELLS

Differentiated/positive cells in 0.5mM NaBut

ECsoW

ECsoN

EC~oE

IC5o

Total

Mature

NBT

NSE

x~ap<30

0.14-t-.01

0.38-1-.09

0.36-t-.13

0.39-'1-.09

97"+2

52"+'11

76--1-22

75-t-21

wt. p50-70

0.59:::i:.13 0.69::t:.06

0.25-1-.09

0.67-t-.05

34-+-17

16-+-7

13-t-6

94::t::6 625-31

DMSO-res

0.13"+.04

>0.75

0.38-t-.23

0.42±.06

98±2

665-14

34"t-25

6TG-res

0.26-t-.08

0.585-.24

0.405-.19

0.605-.13

84"+13

48:1::21

43__.31

815-21

R A-res

0.51 -I-.20

0.67"1-.09

0.16-1-.04

0.68"+.22

51 -I-20

135-5

235-13

100_0

BUdR-res

0.505-.20

0.75__..11

0.44-1-.11

0.695-.08

505-27

155-6

11±8

Ou-res

0.825-.16

0.78

0.09"1-.03

0.785-.09

24"+16

I+1

185-17

1.6 × 10-" to 1.6 × 10-' M. The m a j o r differentiative response, as determined by cytologic evaluation occurred within a 1 log concentration range. However, the full response as determined by a variety of parameters, viz. cell growth inhibition, adherence, cell clumping, cytologic evaluation of Wright stained cells, and the NSE and NBT tests, occurred over a 2-to-3-fold dose range, which, as for RA, may account for some of the variability observed. The high passage wt HL-60 subline was significantly less responsive (12-fold) to T P A than the Io~ passage cells by cytologic evaluation, but not by the other parameters (Fig. 8). This defined the effective dose range for induction of cytologic differentiation by T P A . since all of the sub-lines were more sen~itive than high passage HL-60 cells and less sensitive than lox~ passage HL-60 cells (Table 4, column 5, and Table 8). Adherence to the plastic culture vessels tended to quantitatively parallel the cytologic response in the high passage wt cells and the drug-resistant sub-lines t-l-able 8). An exception was the RA-res sub-line in

665-25 1005-0

which the adherence response was weak despite manifestation of other elements of the differentiative response, e.g. see Fig. 6C. The formation of cell clumps, as estimated semi-quantitatively, very closed followed the adherence response (Fig. 8 and not shown). In contrast to results with the other inducers, the NBT response was weak, being positive in less than 25°70 of the cells even at 1.6 x 10-' M T P A after up to 72 h (Table 8 and not shown). NSE activity induction occurred at quite variable doses in the different lines/sublines, although a relatively uniform percentage of positive cells was noted at the highest tested T P A concentration (Table 8). As with NaBut, the Ou-res cells were particularly sensitive to NSE induction. The percentage of NSE positive cells tended to correlate with the portion of highly differentiated cells, but, as for NaBut induction, the presence of NSE positivity in individual cells frequently did not correlate with cytologic evidence of terminal maturation.

ROBERTE. GALLAGHERet al.

980 .J i~lOC ¢J

-

UJ

_~ sc

z.siB IE zo,-

~ ,'/

~,oo"C u so

x

I-'7

,.s

W 8¢ m W t,

J,. Z W

~O.g

z 3

u~ ...J

m S

x

?

IOC

so

tU

20

7

3

5

7

lO0/lO

>

I"Z

5

~

_

!i

4¢

~. 2o

,,% ~

N ,I

5

1

7

~'-

n

7

FIG. 5. Various response parameters of 6TG-res HL-60 cells to different concentrations of NaBut. (A) cell count; (B) cytologic differentiation by Wright-stain; (C) NBT dye reduction test; (D) NSE test; (E) percentage and accumulated number of monocytes; (F) percentage and accumulated number of eosinophils. NaBut concentrations: ( 1 ) 0; ( 0 ) 0.25 mM; (A) 0.5 mM; (X) 0.75 m M E and F; open symbols, percentage monocytes or eosinophils; closed symbols, accumulated number of monocytes or eosinophils. 157

~

IZ5

80/8

Cytogenetic studies

CELL LINE

/ SUB-LINE

FI(~, 7. Maximum percentage and accumulated total number of monocytes and eosinophils in HL-60 lines/sub-lines after treatment with NaBut. Conditions for maximum monocyte determinations were somewhat variable, occurring after treatment

with 0.5-0.75 mM NaBut for 5-7 days. Maximal eosinophilic determinations invariably occurred with 0.25 mM NaBut for 7 days. Open bars, maximum % monocytes; closed bars, maximum % eosinophils; left-~ided hatched bars, maximum number of monocytes; right-sided hatched bars, maximum number of eosinophils.

As summarized in Table 9, c h r o m o s o m e counts were performed on 50 or more metaphase cells from all of the lines/sub-lines. The modal c h r o m o s o m e member was 45 for the low and high passage wild-type cells and the 6TG-res, RA-res and DMSO-res cells. The modal c h r o m o s o m e number for the BUdR-res cells was 48, and for the Ou-res ceils it was 44. Hyperploid cells were infrequent except in the DMSO-res sub-line in which 10-20°70 of the metaphase cells were tetraploid. These results related to tetraploidy were independently confirmed on at least 2 occasions for all lines/sub-lines by flow cytometric analysis of propidium iodide-stained cells (not shown). The DMSO-res sub-line's propensity to form tetraploids increased as a function of passage generation and one such higher-passaged DMSO-res sub-line (not used in the present studies) had decreased responsivity to both granulocyte and macrophage inducers. Two of the three sub-lines selected by multi-step pro-

981

Differentiation-inducer-resistant HL-60 cells

'°°FA

re/",,,

~"

Io~'

l

io"

,6"*

o

Io-'

TPA [Mxl.6] FI(~. 8. Response parameters of low and high passage wt HL-60 cells to various concentrations of TPA after 2 days treatment. The data points represent the averages of triplicate independent determinations; see Table 8 for indices of experimental variation. (in) percent of control (untreated) cell growth; (Q) percenl cytologic differentiation by Wright stain; {X) estimated percentage of clumped cells; (:~.) counted percentage of adherent cells; (A) percent NSE-positive cells.

(TPA): DIFFERENTIATION RESISTANT HL-60 CELLS

TABLE 8. I2-O-TETRADECANOYLPHORBOL 13-ACETATE

Dose-response data ( × 10-gM) Line/sub-line

EC~0W

EC+oAd

AND GROWTH RESPONSES OF WILD-TYPE AND DRUG-

Differentiated/positive cells (%) in 1.6 x 10-'M TPA

EC~oE

IC~o

Total

Mature

Adherent

NBT

NSE

wt,p<30

0.5-'1-0.7

6.9--I-4.3

25

5.3+1.9

99__.I

60--t-30

57--I-11

15

63

wi, p50-70

6.4-+1.5

4.8"+2.8

140

6.5.+2.5

91.+6

13.+8

87.+7

16

52

DMSO-res

2.0-1-1.6

3.8.+2.8

130

9.4.+9.6

97::1:5

30.+15

80-+12

20

51

6TG-res

0.6"+.01

1.4

1.3

1.9-1-2.3

93.+12

48.+18

82-+27

12

72

RA-res

0.6-+.06

>102

10

99-+1

54-+12

33-+11

18

69

BUdR-res

2.4-+2.4

9.3

>160

9.5-+2.2

94-+6

21-+16

65-+20

17

42

Ou-res

1.7-+1.0

3.6

0.6

1.7-+1.0

100-+ 1

60-+22

64-+33

18

97

26.0-+12

¢cdures - - the 6TG-res and RA-res sub-lines - - contamed n u m e r o u s d o u b l e m i n u t e c h r o m o s o m e s {DM; Table 9; Ref. [21]). Neither the multi-step selected DMSO-res sub-line n o r the 2 single-step selected B U d R re~ and Ou-res sub-lines c o n t a i n e d DM. Similarly, n o 1)'Xl were observed in repeated cytogenetic e x a m i n a t i o n s of the low or high passage wt HL-60 cells m a i n t a i n e d in our l a b o r a t o r y , when e x a m i n e d at the time t h a t the \ a r i a n t sub-lines x~ere selected [21]. Subsequently, a report a p p e a r e d in which substantial n u m b e r s of DM x~erc observed in wt HL-60 celts [2]. This p r o m p t e d us to

re-examine the low and high passage wt HL-60 cells and, as s h o w n in T a b l e 9, low n u m b e r s o f DM were observed in each. These results are similar to those originallyreported in low passage HL-60 cells [20] a n d are consistent with findings in a n o t h e r recent report [18].

Oncogene studies As previously reported [12, 14, 21], the o n c o g e n e

c-myc was amplified 20-40-fold in our wt H L - 6 0 cells c o m p a r e d to n o r m a l h u m a n D N A . O n repeated testing using different extractions o f cellular D N A a n d 2 dif-

982

ROBERT E. GALLAGHER et al. TABt.E 9. CYTOGENETIC STUDIES OF WlLD-T',PE AND DRUG-RESISTANT HL-6,0 CELLS

Chromosome distribution* Line/subline

<44

44

45

46

47

48

Double minutes chromosomes >48

Percent pos. cells (A)

Average/ pos. cell (B)

DM index (A x B)

Wild-type, p27

0.12

0.17

0.50

0.21

0.0

0.0

0.0

5

7.0+5.0

35

Wild-type, p45

0.30

0.28

0.30

0.09

0.01

0.0

0.0

0

0

0

Wild-type, p57

0.08

0.20

0.62

0.08

0.0

0.0

0.02

20

3.7___ 1.8

74

DMSO-res

0.32

0.14

0.29

0.10

0.0

0.0

0.12

0

0

0

6TG-res

0.09

0.02

0.62

0.24

0.0

0.0

0.02

60

18.8-1-16.8

1128

RA-res

0.10

0.17

0.62

0.10

0.0

0.0

0.0

78

13.7-+-8.4

1068

BUdR-res

0.02

0.0

0.0

0.30

0.26

0.40

0.02

0

0

0

Ou-res

0.20

0.52

0.16

0.08

0.0

0.0

0.04

0

0

0

*Modal chromosomal number is underlined.

ferent restriction endonuclease enzymes, the c-myc gene copy number was either approximately the same as in wt HL-60 ceils, viz. in the 6TG-res, RA-res, BUdR-res and Ou-res cells, or consistently reduced, viz. in the DMSOres cells. An example of these results is shown in Fig. 9. Tests with several other oncogene probes, including N-ras, Ha-ras, Ki-ras, myb and abl, showed no evidence o f amplification or gross re-arrangement of these genes in D N A from any line/sub-line (not shown). DISCUSSION The central finding of this study is that HL-60 cells selected for resistance to one granulocyte inducer are partially cross-resistant to 2 chemically-unrelated, alternative granulocyte inducers, while maintaining sensitivity to inducers of m o n o c y t e / m a c r o p h a g e differentiation. There was no a priori reason for expecting this result. If the resistance mechanism primarily involved a change in the way in which the chemical or its biologically-active metabolite interacted with cellular target molecules to induce differentiation, it might have been anticipated that high specificity would be found for resistance to these chemically-disparate inducers. These results imply either that there are no highly specific cellular targets for the action of these differentiation inducers or that, although the 3 classes of granulocyte in-

ducers initially interact with cells in different ways, ultimately they all impact on some central regulatory element(s) of the neutrophilic granulocyte pathway that has been perturbed in the resistant cells. Although not a predictable result, there is considerable precedent for finding cross-resistance in cytodifferentiation-inducer-resistant sub-lines of cultured leukemic cells, including HL-60 cells. For example, 6TG-res HL-60 cells have previously been reported to be less sensitive to differentiation induction by D M S O [7], and, reciprocally, DMSO-res HL-60 cells have been reported to be resistant to hypoxanthine [56], although in neither of these systems was the cross-resistance level quantitated. While the majority of the reported D M S O res cells were sensitive to macrophage induction by T P A , some sub-clones developed resistance [10]. In the present studies, we also observed that a sub-population of DMSO-res cells which seemed to fail to respond to T P A induction, and, subsequently, we correlated this ~ith the emergence of tetraploid cells (our unpublished results). This would agree with another report on the association of increased ploidy of HL-60 cells with TPA-resistance [36], although hyperploidy is not essential for T P A resistance [2]. Notably, TPA-resistant HL-60 cells were reported not to be resistant to DMSO or RA [41]. In sum, these results suggest that there is no close connection between the loss of responsivity to

Differentiation-inducer-resistant HL-60 cells granulocyte inducers and monocyte/macrophage inducers. The most extensive previous study of crossresistance using Friend murine erythroleukemia cells tFELC) seems to add further support for this conclusion. In the FELC system it was reported that DMSO and HXN belong to 1 cross-resistance 'complementation group' while NaBut belongs to a separate complementation group 150]. Although these observations are, obviously, distinguished from the HL-60 system results by. the different nature of the terminally differentiated cells, erythrocyte vs granulocytes and monocytes, the fact that the resistance pattern is the same in these 2 systems suggests that a shared mode of action (and resistance mechanism) may exist for the same complementation groupings of chemical inducers in these 2 leukemic cell culture systems. Another possible explanation for the granulocyteinducer-resistant HL-60 sub-lines could be the selection in each case of a sub-population of cells arrested at an earlier stage of differentiation. This possibility is raised by the observation, using a variety of less mature human leukemic cells of which the established human" myeloid leukemia cell line KGI is the prototype, that less differentiated human myeloid leukemic cells can respond to macrophage inducers but not granulocyte inducers [28, 29]. However, this does not seem to be the case in the current HL-60 variants, since cells from all of the resistant sub-lines undergo a low level of spontaneous granulocyte differentiation which does not significantly differ from that of high passage wt HL-60 cells (Table 2), and since all of the granulocyte inducers enhance the lexel of granulocyte differentiation in these sub-lines to some extent (Tables 3-6). Clearly, these sub-lines differ from a reported blastic variant of HL-60 cells, which fail to respond to either monocytic or granulocytic inducers [341. When examined in detail, our results suggest that a variety of factors contribute to the loss of sensitivity to inducers of terminal differentiation. Most conspicuousl.v, the level of resistance shows considerable agent specificity in the sense that, while the resistance/crossresistance tends to be in the same range amongst the sublines for an individual agent, it also tends to be maximal in the homologous resistant sub-line. This is particularly evident for RA, since the RA-res sub-line was 1000-fold tnore resistant to RA than to alternative granulocyte inducers (Table 4, horizontal file) and since, compared to wt HE-60 control cells, the RA-res sub-line was 20-300fold more resistant to RA than were the 6TG-res and DMSO-res sub-lines (Table 4, column 3). For DMSO, specificity was also observed in the sense that the highest resistance to the agent DMSO was observed in the DMSO-res sub-line (Table 4, column 1). However, compared to wt cell controls, the DMSO-res sub-line was >50-fold more resistant to RA than to DMSO and was in all probability also comparatively more resistant to HXN (Table 4, horizontal file and see Results). The latter quantitative differences could have more to do with the sensitivity of common molecular factors affecting cross-resistance to critical concentrations of individual agents than with quantitative or qualitative differences specific for individual agents. For HXN, no evidence of

983

specificity of resistance could be defined for its effect amongst the 3 resistant sub-lines (Table 4, column 3, and Fig. 3), although this could be related to the inability to test at sufficiently high HXN concentrations to overcome the resistance. From previous estimates that the resistance level of the 6TG-res sub-line used in this study is equal to or greater than 5-10-fold for 6TG compared to wt HL-60 cells [21], it would seem that the 6TG-res sub-line was approximately as or more resistant to HXN than to RA (Table 4, horizontal file). In sum, these observations suggest that the cytodifferentiative resistance mechanism(s) must be complex involving both agent-specific and differentiation-activation-specific elements, which will require detailed molecular studies to dissect. While the above considerations of detail indicate that the development of cytodifferentiation resistance to the 3 classes of granulocyte differentiation inducers is to a degree complex and variable, they do not detract from the possibility that a common quantitative aberration affecting the regulation of the granulocyte differentiation pathway could be involved. Overall, the data are most consistent with the concept that the neutrophilic granulocytic and monocytic differentiative pathways are activated by different stimuli and exist as 2 closely parallel differentiation sequences with some points of interdigitation related to some common elements of these 2 differentiation programs. Such parallel programs related to cell growth and differentiation have been well documented in yeast through the use of conditional lethal mutants [23] and have been postulated to account for variant differentiative responses of murine myeloid leukemic cells [51]. The simplest explanation for the decreased granulocytic differentiative response of the variant HL-60 cells would be an extension of alterations, which, during the natural history of the disease, prevented these leukemic cells from responding to endogenous physiologic effectors of terminal differentiation. In this regard, we wondered whether further alterations in 2 previously-demonstrated activated oncogenes in wt HL-60 cells - - amplified c - m y c [12, 14] and mutant N-ras [22, 42] - - might be further altered in the variants. In particular, the finding of numerous acquired DM in the 6TG-res and RA-res sub-lines (Table 9) raised the possibility of further c - m y c gene amplification in the variants. However, our results using specific probes versus genomic DNA indicated that neither oncogene was grossly re-arranged nor was the amplification level of c - m y c increased (Fig. 9). Further tests are necessary to determine if alterations of the transcription of these oncogenes has occurred in these variants compared to wt HL-60 cells, which shut-off c - m y c expression following exposure to differentiation inducers [47, 59]. Detailed, serial cytogenetic studies of the 6TG-res sub-line indicated that cytogenetic indicators of amplified genes quantitatively varied as a function of both 6TG selection level and the level of cytodifferentiative resistance to 6TG [21]. Additionally, we have found that both the 6TG-res and RA-res HL-60 sublines have surface membrane glycoprotein changes resembling those associated with the multi-drug (pleiotropic) mode of cytotoxic drug resistance [16].

ROBERT E. GALLAGHERet al.

984

Since multi-drug-resistance has been associated both with gene amplifications [5, 32, 48] and alterations in cell growth and differentiative behavior [5], the possibility must be considered that a related or analogous resistance m e c h a n i s m could be operative in the differentiation-inducer-resistant HL-60 cell variants. Such a m e c h a n i s m might function either by p r o d u c i n g physico-chemical changes in the cell m e m b r a n e which affect the ability o f a variety o f chemical inducers to trigger terminal granulocytic d i f f e r e n t i a t i o n or might act m o r e centrally by disturbing the transmission o f regulatory signals to the genetic granulocytic d i f f e r e n t i a t i o n program.

A c k n o w l e d g e m e n t s ~ T h e authors thank L. Muelter and Y. Logan for preparation of the manuscript.

REFERENCES 1. Alitalo K., Schwab M., Lin C. C., Varmus H. E. & Bishop J. M. (1983) Homogeneously staining chromosomal regions contain amplified copies of an abundantly expressed cellular oncogene (c-myc) in malignant neuroendocrine cells from a human colon carcinoma. Proc. natn. Acad. Sci. U.S.A. 80, 1707. 2. Au W. W., Callahan M. F., Workman M. L. & Huberman E. (1983) Double minute chromatin bodies and other chromosome alterations in human myeloid HL-60 leukemia cells susceptible or resistant to induction of differentiation by phorbol-12-myristate-13-acetate. Cancer Res. 43, 5873. 3. Barker P. E. (1982) Double minutes in human tumor cells. Cancer genet. Cytogenet. 5, 81. 4. Bernstein A., Hunt D. M., Crichley V. & Mak T. (1976) Induction by oubain of hemoglobin synthesis in cultured Friend erythroleukemia cells. Cell 9, 375. 5. Biedler J., Chang T-D., Meyers M. B., Peterson R. H. F. & Spengler B. A. (1983) Drug resistance in Chinese hamster lung and mouse tumor cells. Cancer Treatment Rpts 67, 859. 6. Bloch A. (1984) Induced cell differentiation in cancer chemotherapy. Cancer Treatment Rpts. 68, 199. 7. Bodner A. J., Tsai S., Ting R. C., Collins S. J. & Gallo R. C. (1980) Isolation and characterization of thioguanineresistant human promyclocytic leukemia cells. Leukemia Res. 4, 151. 8. Boyd A. W. & Metcalf D. (1984) Induction of differentiation in HL-60 leukemic cells: A cell cycle dependent all-ornone event. Leukemia Res. 8, 27. 9. Breitman T. R., Selonick S. E. & Collins S. J. (1980) Induction of differentiation of human promyelocytic leukemia cell line (HL-60) by retinoic acid. Proc. hath. Acad. Sci. U.S.A. 77, 2936. 10. Bunce C. M., Fisher A. G., Tokosoz D. & Brown G. (1983) Isolation and characterization of dimethylsulfoxide

resistant variants from the human promyeloid cell line HL60. E.rpl. Hemat. I1,828. 11. Collins S. J., Gallo R. C. & Gallagher R. E. (1977) Continuous growth and differentiation of human myeloid leukemia cells in suspension culture. Nature, Lond. 270, 347. 12. Collins S. J. & Groudine M. (1982) Amplification of endogenous myc-related DNA sequences in human myeloid leukemia cell line. Nature, Lond. 298, 679. 13. Collins S. J., Ruscetti F. W., Gallagher R. E. & Gallo R. C. (1978) Terminal differentiation o.f human promyelocytic leukemia ceils (HL-60) induced by dimethylsulfoxide and other polar compounds. Proc. hath. Acad. Sci. U.S.A. 75, 2458. 14. Dalla Favera R., Wong-Staal F. & Gallo R. C. (1982) One gene amplification in promyelocytic leukemia cell line HL-60 and in primary leukemic cells of the same patient. Nature, Lond. 299, 61. 15. Ellis R. W., De.Feo D., Shilo T. X., Gonda M, A., Young H. A., Tsuchida N., Lowy D. R. & Scolnick E. M. (1981) The p21 src genes of Harvey and Kirsten sarcoma viruses originate from divergent members of a family of normal vertebrate genes. Nature, Lond. 292, 506. 16. Felsted R. L., Gupta S. K., Glover C. J. & Gallagher R. E. Surface membrane glycoproteins of wild-type and differentiation-inducer-resistant HL-60 cells. Blood, in press. 17. Fibach E., Peled T., Treves A., Kornerg A. & Rachmilewitz E. A. (1982) Modulation of the maturation of human leukemic promyelocytes (HL-60) to granutocytes or macrophages. Leukemia Res. 6, 781. 18. Fischkoff S. A., Pollak A., Gleich G. J., Testa J. R., Misawa S. & Reber T. J. (1984) Eosinophilic differentiation of the human promyelocytic leukemia cell line HL-60. J. exp. Med. 160, 179. 19. Fontana J. A., Colbert D. A. & Deisseroth A. B. (1981~ Identification of a population of bipotent stem cells in the HL-60 human promyelocytic leukemia cell line. Proc. hath. Acad. Sci. U.S.A. 78, 3864. 20. Gallagher R., Collins S., Trujillo J., McCredie K., Ahearn M., Tsai S., Metzgar R., Aulakh G., Ting R., Ruscetti F. & Gallo R. (1979) Characterization of the continuous, differentiating myeloid cell line (HL-60) from a patient with acute promyelocytic leukemia. Blood 54, 713. 21. Gallagher R. E., Ferrari A. C., Zulich A. W., Yen R-W. C. & Testa J. R. (1984) Cytotoxic and cytodifferentiative components of 6-thioguanine resistance in HL-60 cells containing acquired double minute chromosomes. Canc. Res. 44, 2642. 22. Hall A. H., Marshall C. J., Spurr N. K. & Weiss R. A. (1983) Identification of transforming gene in two human sarcoma cell lines as a new member of the ras gene family located on chromosome 1. Nature, Lond. 303,396. 23. Hartwell L. H. ( 19781 Cell division from a genetic perspective. J. Cell Biol. 77, 627. 24. Hayashi M., Okabe-Kado J. & Hozumi M. 11981~ Mechanisms controlling the kinetics in proliferation and

Differentiation-inducer-resistantHL-60 cells differentiation of populations of mouse myeloid leukemic cells in vitro. J. Cell Ph.vsioL 108, 123. 25. Heldin C-H. & Westermark B. (1984) Growth factors: Mechanism of action and relation to oncogenes. Cell 37, 9. 26. Honma Y., Kasukabe T. & Hozumi M. (1978) Production of differentiation-stimulating factor in cultured mouse myeloid leukemia cells treated by glucocorticoids. ExpL Cell Res. !11,261. 27. Killman S-A. (1970) A hypothesis concerning the relationship between normal and leukemic hemopoiesis in acute myeloid leukemia. In Hematologic Cell Proliferation (F. Stohlman. Ed.). pp. 267-277. Grune and Stratton, New York. 28. Koeffler H. P. (1983) Induction of differentiation of human acute myelogenous leukemia cells: Therapeutic implications. Blood 62, 709. 29. Koeffler H. P., Bar-el M. & Territo M. C. (1981) Phorbol ester effect on differentiation of human myeloid leukemic cell lines blocked at different stages of maturation. Cancer Res. 41, 919. 30. Lautenberger J. A., Schulz R. A., Garon C. F., Tsichilis P. N. & Papas T. S. (1981) Molecular cloning of avian myelocytomatosis virus (MC 29) transforming sequences. Proc. natn. Acad. Sci. U.S.A. 78, 1518. 31. Li Y-S. (1983) Double minutes in acute myeloid leukemia. Int. J. Canc. 32,455. 32. Lin~ V., Karmer N., Sudo T., Siminovitch L. & Riordan J. R. (1983) Multi-drug-resistance phenotype in Chinese hamster ovary cells. Canc. Treatment. Repts. 67, 869.

985

duction of myeloid leukemic cell differentiation by endogenous cytokines and vitamin analogues. In Leukemia Research: Advances in Cell Biology and Treatmem (Murphy S. B. & Gilbert J. R., Eds.), pp. 31--44. Elsevier, New York. 41. Murao S-l, Gemmell M. A., Callaham M. F., Anderson N. L. & Huberman E. (1983) Control of macrophage cell differentiation in human promyelocytic HL-60 leukemia cells by 1,25-dihydroxyvitamin D and phorbol-12myristate-I 3-acetate. Cancer Res. 43, 4989. 42. Murray M. J., Cunningham J. M., Panada L. F., Dautry F., Lebonitz P. & Weinberg R. A. (1983) The HL-60 transforming sequence: A ras oncogene co-existing with altered m y c genes in hematopoietic tumors. Cell 33, 749. 43. Nakaki T., Mita S., Yamamota S., Nakadate T. & Kato R. (1984) Inhibition by palmitoylcarnitine of adhesion and morphological changes in HL-60 cells induced by 12-0tetradecanoylphorbol-13-acetate. Cancer Res. 44, 1908. 44. Okabe J., Honna Y., Hayashi M. & Hozumi M. (1979l Actinomycin D restores in vivo sensitivity to differentiation induction of non-differenting mouse myeloid leukemia cells. Int. J. Cancer 24, 87. 45. Ozer H. L. & Jha K. K. (1977) Malignancy and Transformation: Expression in somatic cell hybrids and variants. Adv. Cancer Res. 25, 37. 46. Perbal B. & Baluda M. A. (1982) Avian myeloblastosis virus transforming gene is related to unique chicken DNA regions separated by at least one intervening sequence. J. Virol. 41,250.

33. Lu L., Broxmeyer H. E., Pelus L. M., Andreff M. & 47. Roninson 1. B., Abelson H. T., Housman D. E., Howell N. & Varshavsky A. (1984) Amplification of specific DNA Moore M. A. S. (1981) Detection of Luxol-fast-blue sequences correlates with multi-drug resistance in Chinese positive cells in human promyelocytic leukemia cell line hamster cells. Nature, Lond. 309, 626. HL-60. Expl Hemat. 9, 887. 49. Rovera G., Santoli D. & Damski C. (1979) Human pro34. Major P P.. Griffin J. D., Minden M. & Kufe D. W. myeloctic leukemia cells in culture differentiate into (1981) A blast subclone of the HL-60 human promacrophage-like cells when treated with phorbol esters. myelocytic cell line. Leukemia Res. 5,429. Proc. ham. Acad. Sci. U.S.A. "/6, 2779. 35. Maria J. P., Izaguirre C., Civin C., Mirro J. & McCulloch 50. Rovera G. & Surrey S. (1978) Use of resistant or hypersenE. (1981) Granulopoietic differentiation in AML blasts in sitive variant clones of Friend cells in analysis of mode of culture. Blood 58, 670. action of inducers. Cancer Res. 38, 3737. 36. Mascioli D. W. & Esterson R. D. (1984) Analysis of human promyelocytic leukemic cell (HL60) variants insen- 51. Sachs L. (1978) Control of normal cell differentiation and phenotypic reversion of malignancy in myeloid leukemia. sitive to phorbol ester tumor promotors. Canc. Res. 44, Nature, Lond. 274, 535. 328O. 37. Metcalf D. (1979) Clonal analysis of the action of GM- 52. Schimke R. T. (1984) Gene amplification, drug resistance and cancer. Cancer Res. 44, 1735. CSF on the proliferation and differentiation of myelomonocytic leukemic cells. Int. J. Cane. 24, 616. 38. Metcalf D. (1980) Clonal analysis of proliferation and differentiation of paired daughter cells: Action of granulocyte-macrophage colony-stimulating factor on granulocyte-macrophage precursors. Proc. natn. Acad. Sci. U.S.A. 77, 5327.

53. Schwab M., Alitalo K., Varmus H. E. & Bishop J. M. (1983) A cellular oncogene (c-Ki-ras) is amplified, overexpressed, and located with karyotypic abnormalities in mouse adrenocorical tumor cells. Nature, Lond. 303,497.

39. Metcalf D. (1983) Clonal analysis of the response of HL-60 human myeloid leukemia ceils to biological regulators. Leukemia Res. 2, 117.

54. Tanaka H., Abe E., Miyaura C., Shiina Y. & Suda T. (1983) 1, 25-dihydroxy vitamin D, induces differentiation of human promyelocytic cells (HL-60) into monocytemacrophages but not into granulocytes. Biochem. bioph.vs. Res. C o m m u n . 117, 86.

40. Moore M. A. S., Gabrilove J. & Sheridan A. P. (1983) In-

55. Taparowsky E., Shimizu K., Goldfarb M. & Wigler M.

986

ROBERT E. GALLAGHERet al.

(1983) Structure and activation of the human N-ras gene. Cell 34, 581. 56. Toksoz D., Bunce C. M., Stone P. C. W., Michell R. H. & Brown G. (1982) Variant cell lines from the human promyelocytic line HL-60. L e u k e m i a Res. 6, 491. 57. Wang J. Y. J., Ledley F., Goff S., Lee R., Groner Y. & Baltimore D. (1984) The mouse c-abl locus: Molecular cloning and characterization. Cell 36, 349. 58. Weiss B. & Sachs L. (1978) Indirect induction of differentiation in myeloid leukemic cells by lipid A. Proc. hath. Acad. U.S.A. 75, 1374.

59. Westin E. H., Wong-Staal F., Gelmann E. P., Dalla Favera R., Papas T. A., Lautenberger J. A., Eva A., Premkumar Reddy E., Tronick S. R., Aaronson S. A. & Oallo R. E. (1982) Expression of cellular homoiogues of retroviral onc genes in human hematopoietic cells. Proc. natn. Acad. Sci. U.S.A. 79, 2490. 60. Wouters R. & Lowenberg B. (1984) On the maturation order of AML cells: A distinction on the basis of selfrenewal properties and immunological phenotypes. Blood 63, 684.