Leukemia Research Vol. 16, No. 12, pp. 1165-1173. 1992. 0145-2126/92 $5.00 + .

0(I
Printed in Great Britain. 1992 Pergamon Press Ltd

T H E MTT C E L L V I A B I L I T Y A S S A Y F O R C Y T O T O X I C I T Y T E S T I N G
IN M U L T I D R U G - R E S I S T A N T H U M A N L E U K E M I C CELLS

DENESE C. MARKS, LARISSA BELOV,* MARY W. DAVEY, ROSS A. DAVEY* and

ANTONY D. KIDMAN
Neurobiology Unit, School of Biological and Biomedical Sciences, University of Technology, Sydney,
Westbourne Street, Gore Hill, N.S.W., 2065, Australia, and *Department of Clinical Oncology,
Royal North Shore Hospital, St Leonards, N.S.W., 2065, Australia

(Received 20 February 1992. Revision accepted 25 July 1992)

Abstract--The MTT cell viability assay is widely used in determining drug sensitivity profiles for
patients with hematological malignancies and in primary screening of potential chemotherapeutic
drugs. Because the multidrug resistance (MDR) phenotype is associated with these malignancies, and
since many vital dyes are effluxed from MDR expressing cells, we have investigated whether the
MDR phenotype interferes with the MTT assay. In CCRF-CEM and K562 human leukemic cell lines
and drug-resistant sub-lines developed from them, comparison of the MTT assay with other cell
viability assays showed significant variation in IC50concentrations, although the resistance relative to
the sensitive parent cell was correlated. Inclusion of verapamil, an inhibitor of drug efflux activity,
had no effect on the MTT assay.

Key words: MTI', multidrug resistance, P-glycoprotein, cell viability assay, cytotoxicity assay,
CCRF-CEM, K562 leukemic cells.

INTRODUCTION screening of cells and drugs. The formazan crystals

are dissolved and the optical density measured using
CYTOTOXIC drug therapy remains the main method
a multiwell plate reader. The use of M T T has thus
of treatment for hematological malignancies. While
become the method of choice because of its simplicity
malignancies such as acute leukemia have initial
and adaptability to automation.
response rates exceeding 50%, many patients relapse
Recently the retention of vital dyes (Rhodamine-
and the long-term response rate is significantly
123, a mitochondrial specific dye, hydroethidine,
reduced [1]. Blast cells from patients with hema-
Hoechst 33342) was shown to correlate inversely with
tological malignancies have been tested for drug sen-
the expression of the multidrug resistance ( M D R )
sitivity using a tetrazolium based assay to determine
phenotype [5, 6]. M D R is characterized by cross-
cell viability [2]. This assay is also being used for
resistance to a wide variety of natural product drugs
first stage drug screening in cell lines [3]. Previous
and is mediated by a membrane protein called P-
methodologies which involve colony formation or
glycoprotein [7]. Studies using fluorescently labeled
dye exclusion assays for determining cell viability are
drugs support the hypothesis that the drugs bind to
limited by their labor intensive nature. However, the
P-glycoprotein which then acts as an A T P - d e p e n d e n t
development of the rapid colorimetric assay [4] which
drug efflux pump, so reducing the intracellular drug
relies on the ability of mitochondrial dehydrogenase
concentration and confering resistance [8]. A number
enzymes to convert 3,-4,5 dimethyithiazol-2,5
of compounds, including verapamil, a calcium chan-
diphenyl tetrazolium bromide (MTT) to a purple
nel blocker, can reverse M D R and inhibit P-gly-
formazan precipitate, has simplified large scale
coprotein mediated drug efflux [7]. More recently,
Abbreviations: MTT, 3,-4,5 dimethylthiazo!-2,5 di- multidrug-resistant cells have been described which
phenyl tetrazolium bromide; MDR, multidrug resistance; exhibit drug efflux, but which do not express P-
IC5o, 50% inhibitory concentration of drug; DMSO, glycoprotein. These have been termed atypical or
dimethylsulfoxide. non-P-glycoprotein multidrug resistant [9].
Correspondence to: Mary Davey, Ph.D., Senior Since P-glycoprotein expression has been reported
Research Officer, Neurobiology Unit, School of Biological
and Biomedical Sciences, University of Technology, in the peripheral blood leukemic cells of patients
Sydney, Westbourne Street, Gore Hill, N.S.W., 2605, being treated for acute non-lymphoblastic leukemia
Australia. [10] and P-glycoprotein exhibits broad specificity, it
1165
1166 D.C. MARKSet al.

is important to establish that MTT accumulation is Drugs

not affected by the presence of P-glycoprotein or
similar drug efftux mechanisms. Optimal drug incu-
bation times for chemosensitivity assays were estab-
lished and the MTT assay was compared with the
more traditional methods of [3H]leucine incor-
poration and trypan blue exclusion in a series of
human leukemic cell lines with a range of levels of
resistance and P-glycoprotein expression. Some of
the resistant cell lines did not express detectable
levels of P-glycoprotein, although they demonstrated
drug efflux.
M A T E R I A L S AND M E T H O D S
Cell lines
The human leukemic cell line CCRF-CEM (CEM;
[11 ]), its multidrug-resistant vinblastine-selected sub-
line (VLB~00; [12]) and CEM sublines selected with
epirubicin (16, 25, 50, 100 and 200 ng/ml, named
the E series) or vinblastine at 6 ng/ml followed by
epirubicin (16, 25, 50,100 and 200 ng/ml, named the
VE series) were maintained as suspension cultures
in alpha-MEM (ICN Flow, Australia) supplemented
with sodium hydrogen carbonate (26 mM), penicillin
(100 ~tg/ml), streptomycin sulphate (60 ~tg/ml) and
10% fetal calf serum (Cytosystems, Sydney,
Australia). The human myeloleukemic cell line K562
was obtained from the American Tissue Collection
and a K562 subline which expressed P-glycoprotein
was selected with 8 ng/ml vinblastine (V8). These
were maintained in RPMI 1640 (ICN Flow) con-
taining Hepes (20 mM) supplemented with 10% fetal
calf serum, penicillin (100 IU/ml), streptomycin
sulphate (100~tg/ml) and sodium hydrogen car-
Vinblastine, vincristine and methotrexate were
purchased from David Bull (Melbourne, Australia),
epirubicin and doxorubicin from Farmitalia (Mel-
bourne, Australia), daunorubicin from May and
Baker (Melbourne, Australia), VP-16 from Bristol
(Sydney, Australia), colchicine and actinomycin D
from Sigma (St. Louis, MO) and verapamil from
Knoll (Ludwigshafen on Rhine, Germany).
Cytotoxicity assay
Exponentially growing cells were plated in trip-
licate in flat-bottomed 96-well plates (Nunc,
Roskilde, Denmark) at 5 x 10 4 cells/well for CEM
cells and 3 x 104 cells/well for K562 cells. Drug was
added in serial dilution to give a final volume of
200 ~tl/well. Control wells contained medium without
drug. Plates were incubated for 3 (CEM) or 4 (K562)
days in a humidified 5% CO2 incubator and assayed
for cell viability. The 50% inhibitory concentration
0('5o) for a particular drug was defined as the con-
centration producing 50% decrease in cell growth.
Relative resistance was defined as los0 cell line/lOs0
sensitive parent cell line.
Cell viability
(1) MTT assay: MTI" dissolved in phosphate-buf-
1.4
a
1.2
1.0
0.8
J~
e~
0.6

bonate (10 raM). All cultures were mycoplasma free. ~ - 0.4

it-)
0.2
Western-blot analysis o
3
Membrane fractions were prepared, solubilized,
subjected to electrophoresis, transferred to a nitro-
cellulose membrane and probed with C219 mono- 0.6
clonal antibody (Centocore, Malvern, PA) according ,.Q
<
to the method previously described [13].
0.4

Drug uptake and effect of verapamil

Exponentially growing cells were adjusted to a 0.2
density of 2.5 105 cells/ml in complete growth
medium buffered with 18 mM HEPES and containing
0
the fluorescent drug daunorubicin (200 ng/ml). Cells 100 200

were incubated at 37C for 150 min or for 60 min Time (min)
followed by 90 min in the presence of 9 mM ver-
FIG. 1. Effect of verapamil on MTT assay. Cells (105 well)
apamii and the intracellular drug concentration were incubated in the presence ( ) or absence (. . . . )
measured in an Epics C flow cytometer (Coulter, of 10juM verapamil. (a) CEM U, [~; VLBL00O, (2); (b)
Hialeah, FL) with an argon laser operated at 488 nm. K562 II, D; V8 O, O.
 MTT assay in multidrug-resistant cells 1167

fered saline, pH 7.2, at 2.5 mg/ml was added to the blue (0.5%) and the cells excluding the dye were
cells (20 ~tl/well). After incubation at 37C for 2 h, counted in a haemocytometer. Cell viability (%) =
cells were centrifuged at 800g for 5 rain and the average number of cells/average number of cells in
medium aspirated. The formazan crystals were dis- control wells.
solved in 100 ~1 dimethyl sulfoxide (DMSO) and the
A540nm read in a microtitre plate reader. Results were RESULTS
calculated as: cell viability ( % ) = average O.D. of
wells/average O.D. of control wells. Optimisation of M T T assay
(2) Leucine incorporation: after 48 h incubation of Absorbance increased linearly with cell number
cells with drug, the medium was replaced with drug from 104 cells per well to 5 x 105 cells per well
free medium and the plates incubated for 5 h to (r > 0.97). This is within the range of cell growth
allow recovery of surviving cells. Plates were then during a cytotoxicity test. Inclusion of verapamil
centrifuged for 5 min at 700 g, the medium carefully (10 ~tM) in the assay had no effect on the production
removed and replaced with 200 ~tl medium containing of formazan (Fig. 1). However the formazan pro-
0.6 ~Ci L-[4,5-3H]leucine (Amersham, Sydney, Aus- duction was slightly lower in the drug-resistant sub-
tralia). After overnight incubation, cells were washed lines (VLB100 and V8) than in the parental drug-
once with Hank's balanced salt solution (HBSS; sensitive cell lines (CEM, K562). The absorbance
Commonwealth Serum Laboratories, Melbourne, was proportional to MTT concentration up to 5 mg/
Australia), resuspended in 100 ~tl HBSS and 80 ~tl ml for CEM and VLB100 and 2.5 mg/ml for the K562
spotted onto G F / A glass filters (Whatman, Maid- and V8 cells. The optimal incubation time in MTT
stone, England) and the incorporated radioactivity was chosen to give a linear, increase and an
measured as described previously [14]. Cell viability absorbance value of 0.7-1.3 units in control wells.
(%) = average cpm of wells/average cpm of control
wells. Cytotoxicity assays
(3) Dye exclusion assay: cells from triplicate wells Cell viability was determined by the MTT assay
were pooled, mixed with an equal volume of trypan for two unrelated drugs vinblastine (a Vinca alkaloid,

A .A
100 100

50 50
~ A ^ , / x A
>.
[..
10 101 102 10 ~ 10 ~ 102 10 J 10 4

<
7, 100 c 1C d
,-2
,-2
L~
"~Q. ZX
50

,, "O-
........ , ........ i ........ , ........ i

10 10 ~ 10 2 10 3 102 103 104 105

VlNBLASTINE (nM) EPIRUBICIN (nM)
FIG. 2. Effect of incubation time on cytotoxicity assay.
Cells were incubated in serial dilution of drug and the cell
viability determined by the MTT assay. (a,b) CEM cells,
day 1 (A), day 2 (~2), day 3 (O), day 4 (O); (c,d) K562
cells, day 1 (A), day 2 (~2), day 3 (), day 4 (O), day 5
(x). (The time chosen for the MTY is shown as solid
circles.)
1168 D . C . MARKS et al.

VERAPAMU.

~ 60 ~ CONTROL

FIG. 4. Effect of verapamil on the accumulation of dauno-

rubicin in drug-resistant cells. The cell associated dauno-
rubicin in control and treated cells was determined by flow
cytometry after 150 rain incubation at 37C in daunorubicin
containing medium. The final 90 min of the treated cell
incubation was in the presence of 9 btM verapamil.

Fig. 2a, c) and epirubicin (an anthracycline, Fig.
2b, d). The los0 initially decreased with increasing
incubation time, however, after 2 days for CEM and
3 days for K562 cells, the ICs0 remains constant. Thus
times chosen for cytotoxicity testing were 3 days for
CEM and 4 days for K562 cells (indicated by the
solid lines in Fig 2).
The Ics0 of any drug was not affected by performing
the MTT incubation in fresh media. The addition of
0.1 M glycine in 0.1 M NaCI, pH 10.5 to the DMSO
solubilisation of formazan did not have any effect on
the Ics0 determination.
Characterization of drug-resistant cells
Figure 3 shows the Western-blot analysis of a series
of multidrug-resistant cells of increasing drug resist-
ance. The VE series expressed increased levels of
P-glycoprotein which correlated with the level of
resistance to epirubicin and to the level of cross-
resistance to other drugs (not presented). The level
of P-glycoprotein in the VE series also correlated
with the decrease in daunorubicin accumulation by
these cells (Fig. 4). The E series did not express
detectable levels of P-glycoprotein (Fig. 3), but they
had similar levels of anthracycline resistance as their
VE series counterparts. An example of this is given
for the VE200 and the E200 subline in Table 1.
The El00 and E200 sublines also showed a similar
decrease in daunorubicin accumulation over 150 min
incubation (Fig. 4). However, this decreased
accumulation was only partly reversed by verapamil
in the El00 and E200 sublines while reversal was
more complete for the VE100 and VE200 sublines.
As expected, the highly drug-resistant VLBm0 subline
accumulated much less daunorubicin (approximately
one fifth) than the CEM cell line and verapamil
caused a 4-fold increase in drug accumulation (Fig.
4).
Comparison of MTT and [3H]leucine incorporation
assays
The ICs0 for 9 different cytotoxic drugs including
anthracyclines, Vinca alkaloids, antibiotics, epi-
podophyllotoxins and methotrexate was determined
for the CEM cells and the multidrug-resistant sub-
lines. The Ics0s for the sensitive CEM cells, the
vinblastine resistant VLB100 and two of the multi-
drug-resistant sublines E200 and VE200 which are
representative of the E and VE series are shown
in Table 1. The MTT assay generally gave an ICs0
approximately 3 times higher than the [3H]leucine
incorporation assay with some notable exceptions.
For example the lc50 concentrations obtained for
daunorubicin for all cell lines were 10 to 20 times
higher by the MTT assay. Although the IC50 values
were dependent on the assay used, the relative resist-
ances (compared to the parent cell line) were highly
correlated by the two assays, as demonstrated in Fig.
5 and P-glycoprotein expression (closed circles) did
not significantly alter this relationship.
Comparison of MTT and trypan blue exclusion assay
The Ics0 for epirubicin and vinblastine was deter-
mined in VLBI00 and K562 ceils by M T T and trypan
blue exclusion assays. The results (Table 2) showed
that for both cell lines the M T T assay gave an Its0
 0
0
0 0
m o 0 0 m
IJ.I LI,I ILl UJ LU

-200kD

- 9 7kD
.... : ~ ,qi!iiiiji~
iiii~ iiii
!i~<i

FIG. 3. Western-blot analysis for P-glycoprotein

expression. Membrane fractions were solubilized, electro-
phoresed, blotted onto nitrocellulose and probed for P-
glycoprotein using C219 monoclonal antibody. Positions
of molecular weight markers are indicated.

1169
 MTT assay in multidrug-resistant cells 1171

TABLE 1. COMPARISON OF IC50 BY M T T ASSAY AND [3H]- DISCUSSION

LEUCINE INCORPORATION ASSAY
Different cell lines have different rates of formazan
IC50 (nM) production. This is demonstrated by the C E M cells
which convert M T T to formazan at approximately
Ratio
Drug Cell line MTT Leu (MTT/Leu) twice the rate of K562 cells (Fig. 1). H o w e v e r , this
does not affect the versatility of the M T T cell viability
Epirubicin CEM 140 38 3.8 assay since most determinations of drug toxicity are
VLBt00 11,900 4500 2.7 made within a cell line. They are internally consistent
E200 3600 1600 2.3 since they use the same cells incubated in the absence
VE200 5100 1600 3.3
Doxorubicin CEM 97 19 5.1 of drug as a control. In both cases the M D R - e x p r e s s -
VLB10o 3800 1050 3.6 ing sublines (VLB100 and V8) had a slightly lower
E200 2800 530 5.2 rate of formazan production than the parental cell
VE200 3100 670 4.6 line. As with the difference between the C E M cells
Daunorubicin CEM 170 13 13.1 and the K562 cells, this probably reflects differences
VLBl00 12,400 620 20.0
E200 2660 140 18.5 in cell physiology or metabolic rate (the drug resistant
VE200 3550 320 11.1 VLB100 cells grow more slowly than the parental
Vinblastine CEM 0.80 0.41 2.0 C E M cell, [12] or changes in cell m e m b r a n e per-
VLB100 480 150 3.2 meability for the MTT. It is unlikely that P-gly-
E200 0.31 0.41 0.8 coprotein actively removes MT-I" from resistant cells
VE200 165 54 3.1
Vincristine CEM 0.12 0.24 0.5 and that this accounts for the lower rate of formazan
VLB10~I 3035 1030 3.0 production since verapamil, an inhibitor of efflux
E200 5.1 3.8 1.3 activity [7], had no affect on formazan production
VE200 705 240 3.0 but increased daunorubicin accumulation 4-fold in
VP-16 CEM 1865 255 7.3 the VLB100 subline (Fig. 4). The results presented in
VLB10 0 23,090 8285 2.8
E200 40,745 19,830 2.1 Fig. 5 substantiate this by showing that the presence
VE200 1 0 , 1 8 5 5775 1.8 of P-glycoprotein had no effect on the relationship
Actinomycin D CEM 17.5 0.62 28.0 between the MTI" assay and the [3H]leucine incor-
VLB~0o 2390 400 6.0 poration assay. This suggests that M T T does not
E200 27 3.0 9.0 interact with P-glycoprotein unlike some other vital
VE200 110 23 4.8
Colchicine CEM 7.5 13.3 0.6 dyes such as Rhodamine-123, hydroethidine [5] and
VLB100 1705 875 1.9 Hoechst 33342 [6].
E200 25 45 0.6 Many modifications have been made to the M T T
VE200 325 95 3.4 assay [3, 15, 16]. We found that D M S O was the best
Methotrexate CEM 42 14.5 2.9 solvent for the formazan crystals provided care was
VLBio0 265 33 8.0
taken in the removal of the culture supernatant,
E200 26 14.5 1.8
VE200 29 12 2.4 as more than 40 ~tl remaining in the well prevents
solubilization of the formazan crystals [16].
Although the M T T m e t h o d is a reliable way of
determining cell viability, care is needed when using
TABLE 2. COMPARISON OF IC50 BY M T T ASSAY AND DYE it for drug cytotoxicity assays. It is necessary to
EXCLUSION ASSAY standardize the cell density and incubation time in
drug as these affect the Ics0 (Fig. 2). The total incu-
1Cs0 (nM) bation time must also be chosen to maintain log phase
Ratio growth of the untreated control cells. Vestica et al.
Drug Cell line MTT Dye (MTT/Dye)
[17] reported that glucose concentration in culture
EPR VLB100 11,900 1900 6.3 media affects the IC50determination. H o w e v e r in our
K562 520 51 10.2 culture conditions, replacement of drug-containing
VLB VLBlo0 460 360 1.3 media with fresh media at the time of cell viability
K562 5.2 5.5 0.95 assay had no effect on the IC50. Therefore glucose
depletion is not important in these cell lines. O u r
assay conditions were similar to those described by
Plumb and workers [18], but unlike these workers
for epirubicin of 6-10-fold higher than by trypan blue we did not find any change in ICs0 by performing the
exclusion assay while vinblastine gave similar ICs0 formazan solubilization at p H 10.5.
concentrations. The IC50 for a given cell line and drug was depen-
1172 D.C. MARKSet al.

EPIRUBICIN ////OO ""'c'N

/
r=0.97 0
40
20
30

20

10

0 i i
10 20 30 40 50 10 20
40 80

DOXORUBICIN //q VP-16

10 20

o o
0 10 20 30 40 0 20 40 60 80
Z 300 ~,o /
,a~ VINBLASTINE / ACTINOMYCIN
[" r=0.98 /r=0 96

,~LIA 200

;>
~ 100
<

100 200 300 0 20 40 60

8000 30
VINCRISTINE COLCHICINE
r=0.98 r=0.98
6000
2o

4000

2000

0 ~ i 0 A
2000 4000 6000 8000 Jo 20
RELATIVE RESISTANCE (MTT ASSAY)
FIG. 5. Comparison of ]c5(~of CEM cells and drug-resistant
sublines determined by MTT assay and [3H]leucine incor-
poration assay. Cells were incubated for 3 days in serial
dilution of drug and assayed for cell viability by the two
methods. Correlation coefficients (r) are given; E cell series
( O ) , P-glycoprotein expressing VE cell series ( O ) .

dent on the type of cell viability assay used (Tables
1 and 2). This may be explained by the drug directly
affecting the cell function involved in the viability
assay. For instance, the ]c50 for daunorubicin deter-
mined by the MTT assay was 10-20-fold higher
than that determined by the [3H]leucine incor-
poration assay (Table 1). It is k n o w n that anthra-
cyclines interact with the mitochondrial membrane
[19] and also cause free radical formation. H o w e v e r
it has been demonstrated that daunorubicin is
unable to bind ferric iron and cannot produce free
radicals by this pathway due to the absence of a
ketol group present on doxorubicin and epirubicin
[20]. Thus different anthracyclines while exhibiting
structural similarities, do exert different effects on
cell metabolism and this could result in changes in
mitochondrial activity involved in the MTT assay.
Similarly, the lcs0 for actinomycin D for C E M cells
 MTT assay in multidrug-resistant cells
is 28 times higher by the M T T assay c o m p a r e d to
the [3H]leucine incorporation assay. This could be
due to the fact that actinomycin D inhibits protein
synthesis and hence leucine incorporation. It must
be borne in mind that all cytotoxicity assays rely
on a particular metabolic activity and some of these
activities may be preferentially inhibited by the
drug under test.
The availability of a series of cell lines exhibiting
different levels of resistance and containing different
amounts of P-glycoprotein (Fig. 3) has allowed us to
compare methods for determining relative resistance.
Although the M T T assay and the [3H]leucine incor-
poration assay gave different I5o values, relative
resistance was correlated for both assays (Fig. 5). It
is important to note that both methods were able
to detect the low levels of increased resistance or
collateral drug sensitivity in these cell lines. Thus the
MT-F assay can detect the d e v e l o p m e n t of increased
resistance at clinically relevant drug concentrations
(e.g. epirubicin, 16 ng/ml; vinblastine, 6 ng/ml).
Therefore use of the M T T assay, which is much
simpler than other assays for cytotoxicity testing, is
appropriate for determining the relative resistance of
cells expressing P-glycoprotein (VE series) or other
drug efflux mechanisms (E series). H o w e v e r , it is
important in the initial screening of drugs to deter-
mine any direct effect the drug may have on formazan
production as this could interfere with the cyto-
toxicity assay and may lead to potentially useful drugs
being rejected. It is also important to be aware of
effects of the drug on Its0 determinations by the
MT-I" assay when screening clinical samples for drug
sensitivity.
REFERENCES
1. Ma D. D. F. & Bell D. R. (1989) Multidrug resistance
and P-glycoprotein expression in human cancer. Aust.
N.Z. J. Med. 19, 736.
2. Sargent J. M. & Taylor C. G. (1989) Appraisal of the
MTI" assay as a rapid test of chemosensitivity in acute
myeloid leukaemia. Br. J. Cancer 60, 206.
3. Alley M. C., Scudiero D. A., Monks A., Hursey M.
L., Czerwinski M. J., Fine D. L., Abbott B. J., Mayo
J. G., Shoemaker R. H. & Boyd M. R. (1988) Feasi-
bility of drug screening with panels of human tumor
cell lines using a microculture tetrazolium assay. Cancer
Res. 48, 589.
4. Mosmann T. (1983) Rapid colorimetric assay for cellu-
lar growth and survival: application to proliferation
and cytotoxicity assays. J. Immun. Meth. 65, 55.
5. Bucana C. D., Giavazzi R., Nayar R., O'Brien C. A.,
Seid C., Earnest L. E. & Fan D. (1990) Retention
of vital dyes correlates inversely with the multidrug-
resistant phenotype of adriamycin-selected murine
fibrosarcoma variants. Expl Cell Res. 190, 69.
1173
6. Morgan S. A., Watson J. V., Twentyman P. R. &
Smith P. J. (1989) Flow cytometric analysis of Hoechst
33342 uptake as an indicator of multi-drug resistance
in human lung cancer. Br. J. Cancer 60, 282.
7. Endicott J. A. & Ling V. (1989) The biochemistry
of P-glycoprotein-mediated multidrug resistance. Ann.
Rev. Biochem. 58, 137.
8. Yusa K. & Tsuruo T. (1989) Reversal mechanism of
multidrug resistance by verapamil: direct binding of
verapamil to P-glycoprotein on specific sites and trans-
port of verapamil outward across the plasma membrane
of K562 cells. Cancer Res. 49, 5002.
9. Mirski S. E. L., Gerlach J. H. & Cole S. P. C. (1987)
Multidrug resistance in a human small cell lung cancer
cell line selected in adriamycin. Cancer Res. 47, 2594.
10. Ma D. D. F., Davey R. A., Harman D. H., Isbister J.
P., Scurf R. D., Mackertich S. M., Dowden G. &
Bell D. R. (1987) Detection of a multidrug resistant
phenotype in acute non-lymphoblastic leukemia. Lan-
cet i, 135-137.
11. Foley G. E., Lazarus H., Farber S., Uzman B. G.,
Boone B. A. & McCarthy R. E. (1965) Continuous
culture of human lymphoblasts from peripheral blood
of a child with acute leukemia. Cancer 18, 522.
12. Beck W. T., Mueller T. J. & Tanzer L. R. (1979)
Altered surface membrane glycoproteins in Vinca alka-
loid-resistant human leukemic lymphoblasts. Cancer
Res. 39, 2070.
13. Haber M., Norris M. D., Kavallaris M., Bell D. R.,
Davey R. A., White L. & Stewart B. W. (1989) Atypi-
cal multidrug resistance in a therapy-induced drug-
resistant human leukemia cell line (LALW-2): resist-
ance to Vince alkaloids independent of P-glycoprotein.
Cancer Res. 49, 5281.
14. Davey R., Bowen R. & Cahill J. (1983) The analysis of
soluble galactosyltransferase isoenzyme patterns using
high resolution agarose isoelectricfocusing. Biochem.
Int. 6, 643.
15. Carmichael J., DeGraff W. G., Gazdar A. F., Minna J.
D. & Mitchell J. B. (1987) Evaluation of a tetrazolium-
based semiautomated colorimetric assay: assessment
of chemosensitivity testing. Cancer Res. 47, 936.
16. Twentyman P. R. & Luscombe M. (1987) A study of
some variables in a tetrazolium dye (MTT) based assay
for cell growth and chemosensitivity. Br. J. Cancer 56,
279.
17. Vistica D. T., Skehan P., Scudiero D., Monks A.,
Pittman A. & Boyd M. R. (1991) Tetrazolium-based
assays for cellular viability: a critical examination of
selected parameters affecting formazan production.
Cancer Res. 51, 2515.
18. Plumb J. A., Milroy R. & Kaye S. B. (1989) Effects
of the pH dependence of 3-(4,5-dimethylthiazol-2-yl)-
2,5-diphenyl-tetrazolium bromide-formazan absorp-
tion on chemosensitivity determined by a novel tetra-
zolium-based assay. Cancer Res. 49, 4435.
19. Goormanhtigh E., Chatelain P. Caspers J. & Ruys-
schaert J. M. (1980) Evidence of a specific complex
between adriamycin and negatively-charged phospho-
lipids. Biochim. Biophys. Acta 597, 1.
20. Zweier J. L., Gianni L., Muindi J. & Myers C. E.
(1986) Differences in oxygen reduction by the iron
complexes of adriamycin and daunomycin: the import-
ance of the side chain hydroxyl group. Biochim.
Biophys. Acta 884, 326.