expression of the mrp and mdr1 multidrug resistance genes...
TRANSCRIPT
Vol. 3, 115-122, January 1997 Clinical Cancer Research 115
Expression of the MRP and MDR1 Multidrug Resistance Genes in
Small Cell Lung Cancer’
Barbara G. Campling,2 Leah C. Young,
Kathy A. Baer, Yuk-Miu Lam, Roger G. Deeley,
Susan P. C. Cole, and James H. Gerlach
Cancer Research Laboratories [B. G. C., L. C. Y., K. A. B., R. G. D.,S. P. C. C., J. H. G.] and Departments of Oncology [B. G. C.,R. G. D., S. P. C. C., J. H. G.l, Pathology [B. G. C., L. C. Y., R. G. D.,S. P. C. C.], Community Health and Epidemiology [Y-M. L.J, andBiochemistry [R. G. D., J. H. G.], Queen’s University, Kingston,
Ontario, K7L 3N6 Canada
ABSTRACT
Acquired multidrug resistance is a major obstacle to acure for small cell lung cancer (SCLC). Overexpression ofthe MDR1 gene occurs infrequently in multidrug-resistant
SCLC cell lines. The multidrug resistance protein (MRP)can confer multidrug resistance, but its role in clinicallyacquired drug resistance is unknown. The purpose of thisstudy was to measure expression ofMRP and MDR1 mRNA
in cell lines and clinical samples from SCLC patients and tocorrelate the results with drug sensitivity profiles. Twenty-
three SCLC cell lines and iO tumor samples from SCLC
patients were examined. Samples expressing MRP andMDR1 were identified by reverse transcription-PCR, andlevels of MRP mRNA in the cell lines were measured byquantitative reverse transcription-PCR. One of 23 cell lines(4%) expressed MDR1 mRNA, whereas MRP expression was
detected in i9 of 23 cell lines (83%). There was a significant
correlation between doxorubicin resistance and MRP ex-pression levels (r = 0.422; P = 0.045). Of the 10 clinicalsamples, 3 expressed only MRP, 2 expressed only MDR1, and4 expressed both drug resistance genes. In summary, MRP is
frequently expressed in clinical samples and cell lines fromSCLC patients, and the levels correlate with doxorubicinresistance in unselected SCLC cell lines. Expression ofMDR1 can be detected in clinical samples of SCLC but israrely found in cell lines from drug-resistant patients. Thesemultidrug resistance proteins may contribute to the multi-
Received 5/24/96; revised 10/9/96; accepted 10/23/96.The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to
indicate this fact.
I Supported by the Medical Research Council of Canada. B. G. C. is aClinician Scientist, and S. P. C. C. and J. H. G. are Career Scientists of
the Ontario Cancer Treatment and Research Foundation. L. C. Y. issupported by the Ontario Cancer Treatment and Research Foundation.R. G. D. is the Stauffer Professor of Cancer Research at Queen’s Uni-
versity.
2 To whom requests for reprints should be addressed, at The Ontario CancerTreatment and Research Foundation, Kingston Regional Cancer Centre, 25
King St. West, Kingston, Ontario, K7L 5P9 Canada. Phone: (613) 545-6357; Fax: (613) 544-9708; E-mail: [email protected].
factorial problem of clinically acquired drug resistance inSCLC.
INTRODUCTION
Chemotherapy is the primary treatment modality for
SCLC,3 but despite high initial response rates, most patients
eventually die with drug-resistant tumors (1). In many in vitro
systems and in some clinical settings, resistance to multiple
chemotherapeutic agents is caused by overexpression of P-gp, a
membrane transport protein encoded by the MDR1 gene, which
actively effluxes many natural product-type drugs from cells (2,
3). However, data from large panels of cell lines indicate that
P-gp overexpression occurs infrequently in multidrug-resistant
SCLC cell lines (4-7). Despite these findings, P-gp overexpres-
sion cannot been ruled out as a factor contributing to multidrug
resistance in SCLC. Because most studies have used established
cell lines rather than tumor samples, it is possible that P-gp
expression may have been present in the original tumor but was
lost during culture in vitro. Alternatively, other mechanisms of
resistance may be more important in SCLC tumors. A novel
transport protein, termed MRP, that is overexpressed in a doxo-
rubicin-selected SCLC cell line has now been cloned (8). Like
P-gp, MRP is a member of the ATP-binding cassette superfam-
ily of transport proteins, and cDNA transfection studies dem-
onstrate that MRP can confer multidrug resistance (9, 10). It is
not known whether this protein is involved in clinically acquired
drug resistance in SCLC.
To investigate the role of MRP and MDRI in SCLC, we
examined 23 SCLC cell lines and 10 cryopreserved tumor
samples obtained from SCLC patients at various stages of
treatment. Six of these cryopreserved samples were the original
specimens from which cell lines in this collection were derived.
Expression of MRP and MDRI mRNA was examined using
RT-PCR, and levels of MRP mRNA were quantitated by Q-PCR. Results were correlated with the drug sensitivity profiles
and the treatment histories of the patients from whom the cell
lines were derived.
MATERIALS AND METHODS
Cell Lines. A collection of 23 unselected SCLC cell lines
was established from patients at various stages of treatment.
Table I summarizes the features of these cell lines, including the
source of the tumor tissue from which the lines were derived, the
treatment received by the patients at the time the line was
established, as well as subsequent treatment and response. The
conditions for establishing and culturing the cell lines have been
3 The abbreviations used are: SCLC. small cell lung cancer; P-gp.P-glycoprotein; MRP, multidrug resistance protein; RT-PCR, reverse
transcription-PCR; Q-PCR, quantitative RT-PCR; VP-l6, etoposide;
TFRR, transferrin receptor; AUC, area under the curve.
on May 9, 2018. © 1997 American Association for Cancer Research.clincancerres.aacrjournals.org Downloaded from
Cell line Source” Prior treatment” Subsequent treatment’ Response” MDR1 MRP
NCI-H69 Pleural effusion CMC-VAP Unknown - - -
NCI-H209 Bone marrow None Unknown - - +
SHP-77 Primary tumor None None - + +
AD-A Subcutaneous lesion� CA, VP/CP, RT Mitox PD - +
BK-T Primary tumor None CAV NA - -
LG-T Lymph node None CAV, VP/CP CR - -
HG-E Pleural effusion None None - - +
JO-E Pleural effusion CAV, VP/CP, RT None - - +
WL-E Pleural effusion VP/CP, CAV None - - +
JN-M’ Bone marrow None CAV, VP/Carb PD - +
SH-A Lymph node� CAV. VP/CP CAy, VP/CP PR - -
MM-I Pleural effusion CAV, VP/CP Unknown - - +
LD-T Primary tumor None CAV, VP/CP NA - +
MO-A Lymph node� CAy, VP/CP CAV, VP/CP CR - +
OS-A Subcutaneous lesion” CAV, VP/CP None - - -
SV-E Pleural effusion CAV VP/CP PD - +
LV-E Pleural effusion Oral VP None - - +
JS-E Pericardial effusion CAV, VP/CP None - - +
GL-E Pericardial effusion CAy, VP/CP, RT Oral VP PD - +
SM-E Pleural effusion CAV VP/CP, CAV PD - +
TY-E Pleural effusion CAV, VP/CP None - - +
YR-A Subcutaneous lesion’ None Oral VP PD - +
HA-E Pleural effusion CAV, VP/CP. RT, Oral VP None - - +
“ Tumor cells from which the cell line was derived.
“ Therapy that the patient had received at the time that the cell line was established. CMC-VAP, cyclophosphamide, methotrexate, CCNU,vincristine, Adriamycin, procarbazine: CAV, cyclophosphamide, Adriamycin. and vincristine: VP/CP. VP-16 and cisplatin; RT, radiotherapy; Carb,
carboplatin; Mitox, mitoxantrone.
‘ Therapy that the patient received after the cell line was established.
“ The response of the patient to further chemotherapy (if given) after the cell line was established. CR, complete response: PR, partial response:
PD, progressive disease; NA, not assessable; -, patient was not treated or treatment status is unknown.‘. Needle aspirate.
116 Multidrug Resistance in Small Cell Lung Cancer
Table I SCLC cell lines
described previously (1 1 ). Nineteen of the cell lines were de-
rived from patients treated at this center, and the origin of those
lines obtained from other investigators has been noted previ-
ously ( 12). The doxorubicin-selected multidrug-resistant my-
eloma cell line, 82261Dox40, obtained from Dr. W. S. Dalton,
was used as a positive control for MDRI expression (13), and
the doxorubicin-selected multidrug-resistant SCLC cell line,
H69AR, was used as a positive control for MRP expression (14).
The cell lines were harvested in exponential growth phase
(approximately 2-5 X l06/ml).
Cryopreserved Tumor Samples. Tumor specimens
were obtained for diagnostic or therapeutic indications from
SCLC patients treated at the Kingston Regional Cancer Centre
from 1986 to 1994. With the patient’s informed consent, a
sample was sent for research purposes as part of a study ap-
proved by the Research Ethics Board of Queen’s University.
The source of the tumor cells and the treatment history of the
patients is indicated in Table 2. Effusion samples were centri-
fuged and washed in RPM! 1640 medium, and viable tumor
cells were separated from RBCs and non-viable cells on a
Ficoll-Hypaque density gradient. The proportion of tumor cells
present was determined by examination of cytospin prepara-
tions. Only samples containing more than 90% tumor cells were
cryopreserved. The cells were frozen in a viable state in 10%
DMSO in RPM! 1640 medium with 20% fetal bovine serum and
stored in liquid nitrogen. Cell lines were subsequently derived
from six of these tumor samples, as indicated in Table 2. In
addition, the cell line JN-M was established from a marrow
sample taken prior to treatment from the same patient as tumor
sample no. 4, which was a pleural effusion obtained at recur-
rence after chemotherapy treatment.
Drug Sensitivity Testing. The sensitivity of each cell
line was determined using a modified 3-(4,5-dimethylthiazol-2-
yl)-2,5-diphenyltetrazolium bromide assay (15). Four drugs, i.e.,
doxorubicin, vincristine, VP-I6, and cisplatin were tested, usu-
ally within 3 weeks of mRNA isolation. The dose-response
curves for the four drugs were summarized by calculating the
area under the curve, using the trapezoidal method as described
(15).
mRNA Isolation, cDNA Synthesis, and Oligonucleotide
Synthesis. Polyadenylated mRNA was isolated from approx-
imately i0� cells using the Micro-FastTrack mRNA isolation kit
(Invitrogen, San Diego, CA). The cDNA synthesis reaction
mixture consisted of 20 mr�i of each dinucleotide triphosphate
(Pharmacia, Biotech, Inc., Baie d’Urf#{233},Quebec, Canada), 10
mM Dli’, S ng/�i.1 random hexanucleotide primers (Pharmacia
Biotech, Inc.), I .35 units/p.l RNAguard (Pharmacia, Biotech,
Inc.), 1 unit/pA avian myeloblastosis virus reverse transcriptase
(Life Sciences, Inc., St. Petersburg, FL), and reverse tran-
scriptase buffer (40 mM MgCl2, 250 mrvi KC1, and 250 msi
Tris-HC1, pH 8.3, at 42#{176}C).mRNA (0.3 p.g) was added to each
20 �jJ cDNA synthesis reaction. To minimize variation between
samples, all cDNA synthesis reactions were carried out simul-
taneously under the same conditions. The cDNA synthesis re-
action was incubated at 37#{176}Cfor I h followed by enzyme
on May 9, 2018. © 1997 American Association for Cancer Research.clincancerres.aacrjournals.org Downloaded from
Clinical Cancer Research 117
Table 2 Clinical samples
The name of the corresponding cell line established from the same sample as in Table I is indicated. Cell line JN-M was established from a
marrow sample taken prior to treatment from the same patient as tumor sample no. 4. The results of RT-PCR screening for MRP and MDR I expression
are shown.
Sample no. Source” Prior treatment”
Subsequent
treatment’ Response”
Corresponding
cell line MDRI MRP
I Pleural effusion None None - HG-E - +
2 Pleural effusion None RT. CAV PR - - +
3 Pleural effusion VP/CP, CAV None - WL-E - +
4 Pleural effusion CAy, VP/CP None - JN�Me + +
5 Pleural effusion CAV VP/CP PD SV-E + +
6 Pleural effusion CAV None - - + -
7 Pleural effusion CAV, VP/CP None - - + +
8 Pericardial effusion CAV, VP/CP None - JS-E + -
9 Pleural effusion Oral VP None - LV-E + +
10 Pericardial effusion CAV. VP/CP. RT Oral VP PD GL-E - -
“ Tumor cells from which the cell line was derived.
h Therapy that the patient had received at the time that the cell line was established. CMC-VAP, cyclophosphamide. methotrexate. CCNU,
vincristine, Adriamycin, procarbazine: CAV. cyclophosphamide, Adriamycin. and vincristine; VP/CP. VP-l6 and cisplatin: RT. radiotherapy; Carb.
carboplatin; Mitox. mitoxantrone.
C Therapy that the patient received after the cell line was established.
�1 The response of the patient to further chemotherapy (if given) after the cell line was established. CR. complete response; PR. partial response:
PD, progressive disease; NA. not assessable; -, patient was not treated or treatment status is unknown.‘. The cell line JN-M was established from a bone marrow aspirate taken prior to therapy. A clinical sample from the same patient (no. 4) was
obtained at relapse following partial response to chemotherapy.
inactivation at 95#{176}Cfor 3 mm. The resulting cDNA was diluted
10-fold in sterile water.
The 25-mer primers specific for MRP, MDRI, and the
human TFRR were synthesized on a Biosearch 8750 DNA
synthesizer and purified by thin layer chromatography (Queen’s
University DNA Synthesis Laboratories, Kingston). The down-
stream (antisense) primers were biotinylated using biotin phos-
phoramidite (Prime Synthesis, Ashton, PA). The internal stand-
ard, TFRR, was used to control for variations in mRNA
extraction and cDNA synthesis. TFRR was selected as an inter-
nal standard because it is expressed ubiquitously at relatively
low levels (similar to that anticipated for MRP and MDRI), and
expression levels are not cell cycle dependent.
The sequences of the primers used for PCR are as
follows: MRP, upstream primer (5 ‘-AGTGACCTCTGGTC-
CTTAAACAAGG-3’) and downstream primer (5’-GAGG-
TAGAGAGCAAGGATGACTTGC-3 ‘); MDR 1 , upstream
primer (5 ‘-ACACCCGACTTACAGATGATGTCTC-3 ‘) and
downstream primer (5 ‘-CGAGATGGGTAACTGAAGT-
GAACAT-3’); TFRR, upstream primer (5’-GGATA-
I and downstream primer
(5 ‘-TGGAAGTAGCACGGAAGAAGTCTCC-3’).
RT-PCR and Q-PCR. Each PCR reaction contained
dinucleotide triphosphates ( 1 20 p.M; Pharmacia), Vent
NEBuffer (New England Biolabs, Mississauga, Ontario, Cana-
da), upstream (sense) primer (0.2 pM), downstream (antisense)
primer (0.2 pM), 0.005 unit/pA Vent DNA polymerase (New
England Biolabs), and cDNA (10 pi per 50 p.1 PCR reaction).
The reaction mixture was overlaid with light mineral oil. RT-
PCR screening of the cell lines and clinical samples was per-
formed independently with the primers specific for MRP,
MDRI, and TFRR.
The PCR was performed on a PHC-l Dri-Block thermo-
cycler (Techne, Cambridge, United Kingdom). Reactions were
allowed to proceed through one cDNA second strand synthesis
cycle of 95#{176}Cfor 1 .5 mm, 58#{176}Cfor I mm, and 72#{176}Cfor 3 mm,followed by 27 PCR cycles of 95#{176}Cfor 45 s, 58#{176}Cfor I mm,
and 72#{176}Cfor 1 mm, with a final PCR cycle of 95#{176}Cfor 45 s,
58#{176}Cfor I mm, and 72#{176}Cfor S mm. For reactions with the
MRP-specific primers, the annealing temperature was 52#{176}C.The
expression of MRP and TFRR was quantified by Q-PCR using
“mimic” standards in those cells lines which were shown to
express MRP by RT-PCR. The “gene-mimic” is a competitive
standard comprised of a heterologous Ba,nHIlEcoRI DNA frag-
ment of the human v-erbB gene (Clontech MIMIC Construction
kit, Mississauga, Ontario, Canada) that is flanked on either side
by the DNA sequence complementary to the gene-specific
primer of interest. This allows both the target cDNA and the
gene-mimic to be coamplified by the same primers. When the
target DNA and the gene-specific mimic DNA are present in
equal concentrations in the PCR reaction mixture, they compete
equally for the primers and result in equal quantities of ampli-
fication product.
For each cDNA sample, five PCR reactions were per-
formed using a 5-fold dilution series of the cDNA template. The
reaction components for each sample were assembled in a
master mixture, and a known amount of the gene-mimic stand-
ard was added, resulting in a series ofcompetitive reactions with
a constant concentration of the gene-mimic standard and van-
able concentrations of the target cDNA. For each sample. the
cDNA concentration at which the PCR product of the target
cDNA and the gene-mimic were present in equal amounts was
determined by interpolation. The level of MRP mRNA was
normalized to the level of TFRR expression in the same cell line
and then expressed as a percentage of the MRP level in H69AR
cells. The conditions for the Q-PCR were the same as for the
RT-PCR except that 36 cycles of amplification were carried out.
on May 9, 2018. © 1997 American Association for Cancer Research.clincancerres.aacrjournals.org Downloaded from
C-)
‘4:
0
I ‘)�
VP-16
C-)
‘4:
C.)
‘4:
7c
Cisplatin
1o0�
��nnnnHU- � #{149}
50
25
��HHnnnnnHHHHiffi�wwu�u�u�
C,)-)
C,)
Cell Line
I�I �
U) I
Cell Line
Fig. I Chemosensitivity was measured using a modified 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay. The dose-response
curves were summarized by calculating the AUC. The cell lines were ranked in order of AUC. Results for doxorubicin, vincristine, VP- 16, and
cisplatin are shown.
118 Multidrug Resistance in Small Cell Lung Cancer
50
25
Doxorubicin Vincristine
�
�thIc/)>.IO�,(coth>OL�O�Jc1: ,
�I< �I
. C,)Cell Line
Chemiluminescent Detection of PCR Products. Thedownstream primers contained a biotin group to allow for
chemiluminescent detection of the PCR products. PCR-ampli-
fled samples were separated on a 2% agarose gel, stained with
ethidium bromide, and photographed. The PCR products were
then transferred to a Zetaprobe membrane (Bio-Rad, Missis-
sauga, Ontario, Canada) by downward alkaline transfer (0.4 M
NaOH, 0.6 M NaCI); the membrane was UV-irradiated and then
agitated briefly in neutralizing buffer (1.5 M NaCl, 0.5 M Tris,
pH 8.0). The membrane was blocked in SDS buffer (5% w/v
SDS, 17 mM Na2HPO4, and 8 mr�i NaH2PO4) for 30 mm,
agitated in blocking buffer containing 4 units/iOO ml streptavi-
din-alkaline phosphatase conjugate (Boehringer Mannheim) for
10 mm, washed for 15 mm in a 1:10 dilution of blocking buffer,
and finally agitated in a wash buffer (1 m�i MgC12, 10 mM NaC1,
and 10 mM Tris-HC1, pH 9.5) for 30 mm. Lumi-Phos 530
(Boehringer Mannheim, Laval, Quebec, Canada) was applied,
and the membrane was placed within plastic page protectors and
allowed to incubate at 37#{176}Cfor 1.5-2.5 h. X-OMAT film
(Kodak) was exposed to the membrane for 30-60 s at room
temperature. The intensity of the bands on the film was deter-
mined on a laser scanning densitometer (Molecular Dynamics,
Sunnyvale, CA) using Image Quant 3.3 software.
Correlation Analysis. Statistical analyses of drug sensi-
tivity, levels of MRP expression, and chemotherapy treatment
histories were performed using the Systat software package
(version 5.0 for DOS). The cell lines were classified as treated
or untreated according to whether the patients from whom the
cell lines were derived had received chemotherapy at the time
the cell line was established (Table 1). The drug sensitivity data
were normally distributed. In contrast, the MRP mRNA levels
were highly skewed toward low values. Consequently, a loga-
rithmic transformation of the MRP values was performed to
approximate more closely a normal distribution. The cell line
H69AR was not included in the correlation analysis, because
unlike the other cell lines, it had undergone in vitro selection in
on May 9, 2018. © 1997 American Association for Cancer Research.clincancerres.aacrjournals.org Downloaded from
Clinical Cancer Research 119
Table 3 Pearson correlation analysis
Pearson correlation matrix is shown. The correlation coefficientsare indicated, and the P values indicating the statistical significance of
the correlations are shown in parentheses below the correlation coeffi-
cients.
Dox Vinc VP-l6 CisPt
Dox” 1.000
(0.000)Vinc” 0.739c
(<0.001)
1.000(0.000)
VP�l6a 0.360
(0.091)0.347
(0.104)1.000
(0.000)CisPt” 0.523k
(0.01 1)
0.624’
(0.001)0.539’
(0.008)1.000
(0.000)MRP” 0.422c
(0.045)0.336
(0.1 16)
0.205
(0.347)
0.080
(0.7 16)
a Correlation between sensitivity to doxorubicin (Dox), vincristine
(Vinc), VP-l6, and cisplatin (CisPt).b Correlation between drug sensitivity and MRP levels.
C Statistically significant correlations.
doxorubicin. Furthermore, this cell line is highly drug resistant
and has exceptionally high levels of MRP mRNA, resulting in
an undue influence on the correlation analysis. Pearson corre-
lation coefficients were calculated for all pairwise data combi-
nations, including sensitivity to doxorubicin, vincristine, VP-16,
and cisplatin, MRP mRNA levels, and treatment history.
RESULTS
Drug Sensitivity. The sensitivities of the 23 cell lines to
doxorubicin, vincristine, VP-16, and cisplatin are shown in Fig.
I . The dose-response curves were summarized by calculating
the AUC as described previously (15). The cell lines displayed
a spectrum of drug responsiveness. One of the cell lines, SV-E,
was highly resistant to the natural product drugs, doxorubicin,
vincristine, and VP-16, but had an intermediate level of sensi-
tivity to cisplatin. The cell line NCI-H69 (from which H69AR
was derived) was among the most sensitive to doxorubicin,
vincristine, and cisplatin but not to VP-l6. The results for
H69AR, which was the most resistant to doxorubicin and yin-
cristine, are shown for comparative purposes only.
The correlation between response to each of the four drugs
tested and response to the other three drugs is shown in Table 3.
Doxorubicin, vincristine, and VP-16 are natural products, and
resistance to these agents may be conferred by either MRP or
P-gp. On the other hand, neither MRP nor P-gp would be
expected to confer resistance to cisplatin. There was a highly
significant correlation between responsiveness to doxorubicin
and vincristine (r = 0.739; P = 0.0001). However, no signifi-
cant correlation was observed between VP-I 6 and either doxo-
rubicin (r = 0.360; P 0.091) or vincristine (r = 0.347; P =
0. 104). There was an unexpected correlation between respon-
siveness to cisplatin and the other drugs, including doxorubicin
(r 0.523; P = 0.010) and VP-16 (r 0.539; P = 0.008). The
association between cisplatin and vincristine was particularly
striking (r = 0.624; P 0.001).
Screening for MRP and MDR1 Expression in Cell Lines
and Clinical Samples. MRP expression was detected in 20 of
the 23 SCLC cell lines, whereas expression of MDRI was
detected in only 1 of 23 cell lines, i.e., SHP-77 (Table 1).
Because of the low mRNA yield from the clinical specimens, 33
cycles of amplification were required to obtain levels of the
control gene, TFRR, that were comparable to those obtained for
the cell lines. The results of screening the 10 cryopreserved
SCLC tumors for MRP and MDRI expression are shown in Fig.
2 and summarized in Table 2. Seven of the 10 samples ex-
pressed MRP, and 6 expressed MDR1. Of the 10 clinical sam-
ples, 6 were the original specimens from which cell lines were
derived. It is interesting to note that three of the clinical samples
with detectable MDRI expression (nos. 5, 8, and 9) cone-
sponded to cell lines that had no MDR1 expression (SV-E. JS-E.
and LV-E, respectively). Thus, it appears that the relative levels
of MDRI mRNA may have declined during in vitro culture.
Sample no. 4 was derived from the same patient as cell line
JN-M. The JN-M cell line (established from a bone marrow
aspirate obtained prior to chemotherapy) was negative for
MDRI but positive for MRP expression, whereas the clinical
sample (no. 4; a pleural effusion obtained from the same patient
at recurrence following partial response to chemotherapy) was
positive for both MRP and MDRI . Two of the clinical samples
(nos. 8 and 10) had no detectable MRP expression, whereas
MRP mRNA was detectable at very low levels in the cone-
sponding cell lines (JS-E and GL-E). Because only 10 clinical
samples were studied and only 3 of these samples were derived
from patients who subsequently received chemotherapy, it was
not possible to make any clinical correlations from this data.
Quantitation of MRP Expression in Cell Lines. Levels
of MRP mRNA expression were quantitated by Q-PCR in those
cell lines that were positive by RT-PCR. Since MDRI mRNA
was detectable in only one cell line (SHP-77), no attempt was
made to quantitate levels of expression of this gene. Because of
the limited amount of material available from the clinical spec-
imens, the levels of MRP and MDRI were not quantitated in
these samples.
An example of the Q-PCR for the cell line H69AR is
shown in Fig. 3. Serial 5-fold dilutions of the cDNA, along with
a constant amount of the “mimic” standard were amplifed using
TFRR- and MRP-specific primers. At high concentrations of
MRP cDNA, the mimic standard is not visible due to competi-
tion for primers, whereas at low concentrations of the MRP
cDNA, the “mimic” is able to compete for the primers and is,
therefore, amplified.
The relative levels of MRP expression in the cell lines are
shown in Fig. 4. The results are expressed as a percentage of the
levels in H69AR cells. Five of the 24 cell lines had no detectable
MRP expression, and in the remainder of the cell lines, the
levels ranged from 0. 1 to I 7.5% of the levels detected in
H69AR.
Correlation of MRP mRNA Expression and Drug Re-
sistance. The results of the Pearson correlation analysis are
shown in Table 3. The correlation between resistance to doxo-
rubicin and MRP expression by Q-PCR (r = 0.422) was signif-
icant (P = 0.045). Although there was a significant correlation
between response to doxorubicin and to vincristine (r = 0.739;
p � 0.001), the correlation between MRP mRNA levels and
vincristine resistance, although positive, was not statistically
significant (r = 0.336; P = 0.116). No correlation was found
between MRP levels and sensitivity to either VP-I6 or cisplatin.
on May 9, 2018. © 1997 American Association for Cancer Research.clincancerres.aacrjournals.org Downloaded from
123456 7 8 9 10 Sample#
�. *‘� 1r[10* 1* � 4pj� 4� 1� #{149}esa
MRP
MDR1
TFRR
cDNA dilution
.)n
0� � � - -
‘�
00
-J<0)
LI.
EO
120 Multidrug Resistance in Small Cell Lung Cancer
Fig. 2 Screening for MDRI
and MRP expression in cryo-
preserved SCLC samples. Tencryopreserved tumor samples
were screened for the expres-
sion of MDR1 and MRP usingRT-PCR as described in “Mate-
rials and Methods.” TFRR was
used as an internal control.
‘b ‘�‘.
(vcY�#{234}#{234}-_ 0. O� 0. 0.
TFRR
TFRR-MIMIC
MRP
MAP-MIMIC
Fig. 3 Quantitation of MRP expression using mimic standards. Serial5-fold dilutions of cDNA derived from cell line H69AR along with aconstant amount of the gene-mimic standard were amplified with prim-ers specific for TFRR and MRP.
There was no correlation between the chemotherapy treatment
histories of the patients from whom the cell lines were derived
and either MRP levels or response to doxorubicin, vincristine,
VP-l6, or cisplatin.
DISCUSSION
Ever since combination chemotherapy became the standard
treatment for SCLC nearly two decades ago, there have been no
major improvements in results of therapy of this disease. The
acquisition of resistance to multiple chemotherapeutic agents
continues to be the major impediment to cure (16). Despite the
efforts of many groups, the molecular basis of clinically ac-
quired resistance is not well understood. To identify mecha-
nisms of resistance, we and others have studied cell lines that
have been selected in vitro for drug resistance. Although a
variety of alterations associated with resistance have been iden-
tified, it remains to be determined whether such alterations are
present in patients with drug-resistant tumors and whether they
are responsible for the drug resistance phenotype. In this study,
we examined cell lines and clinical samples obtained directly
from patients with a spectrum of clinically drug-sensitive and
drug-resistant tumors (Tables 1 and 2). Such investigations may
give a better indication of mechanisms that are involved in
clinical drug resistance.
In the 23 SCLC cell lines examined in this study, there was
a close correlation between response to doxorubicin and to
vincristine. This finding was not unexpected, because doxoru-
bicin and vincristine are both natural product compounds that
1
10
0
Cell Line
Fig. 4 Quantitation of MRP expression SCLC cell lines by Q-PCR.The levels of MRP mRNA are expressed as a percentage of the levels in
H69AR. The cell lines have been ranked in order of MRP levels.
are included in the spectrum of cross-resistance that character-
izes the multidrug resistance phenotype. VP-16 is also included
in the muitidrug resistance phenotype, and thus it is somewhat
surprising that we detected no significant correlation between
response to VP-16 and response to either doxorubicin or yin-
cristine. On the other hand, there was also a close correlation
between response to cisplatin and to doxorubicin, vincristine,
and VP-16. Because cisplatin is not included in the multidrug
resistance phenotype, this finding suggests the presence of re-
sistance mechanisms other than MRP or P-gp in these uns-
elected SCLC cell lines.
Only one of the cell lines, i.e., SHP-77, was positive for
MDR1 expression by RT-PCR. SHP-77 also expressed MRP and
was one of the most resistant cell lines to both doxorubicin and
vincristine but was not highly resistant to VP-16 or cisplatin
(Fig. 1). In contrast to the cell lines, 6 of 10 clinical samples of
SCLC expressed MDR1 (Fig. 2).
P-gp expression does not occur frequently in multidrug-
resistant SCLC cell lines, although it has been detected (17). Lai
et a!. (4) measured expression of MDR1 in lung cancers of all
major histological types as well as corresponding normal lung
tissues and tumor cell lines. In most of these tumors, including
the SCLC samples, the expression of MDR1 mRNA was low or
undetectable. However, it is interesting to note that in three of
the four SCLC samples in which MDR1 levels were measured in
both tumor samples and corresponding cell lines, there was a
decline in MDR1 expression in the derived cell line.
In contrast, two reports suggest that there may be a relation-
on May 9, 2018. © 1997 American Association for Cancer Research.clincancerres.aacrjournals.org Downloaded from
Clinical Cancer Research 121
ship between clinical drug resistance and MDRI levels in SCLC.
Holzmayer et a!. (18) found that the presence of even very low
levels of MDR1 expression, as detected by PCR, correlated with
lack of response to chemotherapy in seven tumor samples from
SCLC patients. Poupon et aL (19) used Northern blot analysis to
detect MDR1 mRNA in xenografts derived from seven SCLC
patients. Expression of this gene was detected in all but two of the
xenografts, and these two samples corresponded to the two patients
in the series who were long-term survivors. They also noted that the
levels of MDR1 mRNA were higher in tumor samples obtained
directly from patients compared with those which had been pas-
saged in nude mice. Although the results are suggestive, larger
studies are required before firm conclusions can be drawn about the
clinical relevance of MDR1 expression in SCLC.
In our study, MDR1 expression was detected more fre-
quently in clinical samples than in cell lines established from
SCLC patients. Furthermore, in some cases, MDR1 expression
was detectable in the clinical sample, whereas the cell line
established from the same material was negative. In human
tumor cells, increased levels of P-gp are most often due to
increased gene expression rather than gene amplification (16).
This may explain the lack of persistence of P-gp overexpression
when the tumor cells are propagated in vitro. Another possibility
is that the cell types that become established as permanent cell
lines may not reflect the heterogeneity of tumor cells present in
the original sample. Furthermore, small numbers of contaminat-
ing nonmalignant cells in the clinical samples could lead to
false-positive results. The frequent detection of P-gp in clinical
samples of SCLC rather than established cell lines suggests that
this transport protein could play a more significant role in
clinically acquired multidrug resistance in SCLC than previ-
ously thought.
On initial screening, we detected MRP expression in 20 of
our 23 cell lines (83%). It is interesting to note that the cell line
with the highest level of MRP mRNA (SV-E) was also among
the most resistant to doxorubicin, vincristine, and VP-16 but did
not express MDR1 mRNA. However, despite the fact that the
relative resistance of the SV-E cell line to these agents was
comparable to that of H69AR, the level of MRP mRNA expres-
sion in this line was only 17.5% of that of H69AR, suggesting
that MRP expression may not be the only factor accounting for
the resistance of this cell line. In the panel of cell lines, there
was a significant correlation between MRP mRNA levels and
doxorubicin resistance. Although vincristine resistance cone-
lated strongly with doxorubicin resistance, the correlation be-
tween vincristine resistance and MRP mRNA levels, although
positive, was not statistically significant. Furthermore, there was
no apparent correlation between levels of MRP mRNA and
response to VP-16. Although VP-16 is one of the drugs to which
MRP confers resistance, it is possible that other resistance
mechanisms, such as altered topoisomerase II, may obscure
significant correlations with MRP levels. MRP does not confer
resistance to cisplatin (9, 10), and the lack of correlation be-
tween MRP mRNA levels and cisplatin resistance was not
surprising.
It is possible that the relationship between drug response
and MRP expression may be stronger than the correlation anal-
ysis would seem to indicate. As noted, there are other drug
resistance mechanisms that may be involved in SCLC, and these
could obscure significant relationships with drug response. For
example, our analyses showed strong correlations between cis-
platin resistance and resistance to doxorubicin, vincristine, and
VP-16, suggesting the presence of other mechanisms that confer
resistance to all four of these drugs. The correlation coefficients
were similar when only those cell lines derived from untreated
patients were included in the analysis, indicating that these
relationships do not result from prior treatment with these four
drugs. Because cisplatin resistance is not conferred by MRP (9,
10) (and does not correlate with MRP levels in this study), the
correlations of doxorubicin, vincristine, and VP-16 with MRP
mRNA levels could be obscured.
Although increased MRP and MDR1 expression have been
clearly implicated in the resistance of certain tumors, they are
not the only factors that may result in multidrug resistance. For
example, resistance to multiple chemotherapeutic agents has
been associated with increased drug detoxification by glutathi-
one and its associated enzymes (20). Although alterations of
glutathione and related enzymes have been detected in multi-
drug-resistant SCLC cell lines, the functional significance of
these changes remains to be determined (21, 22). In a study
using many of these same unselected SCLC cell lines, we found
no significant correlation between drug response and levels of
GSH and associated enzymes (12).
Increasing evidence has indicated that reduced levels or
function of topoisomerase II are important factors in drug re-
sistance in SCLC as well as other cancers. It is now recognized
that topoisomerase II is the common intracellular target for
several of the natural product drugs that are also part of the
classical “multidrug resistance phenotype.” In this collection of
SCLC cell lines, we have shown that reduced levels of topoi-
somerase IIcz correlate with resistance to a variety of agents,
including some drugs that are not known to exert their cytotox-
icity through this target (23). Other studies using unselected
lung cancer cell lines have also shown an inverse correlation
between topoisomerase II levels and drug resistance (24, 25).
In summary, the data presented here indicate that drug resist-
ance in SCLC is complex and unlikely to be explained by a single
resistance mechanism. Our results further emphasize the impor-
tance of examining clinical samples directly to identify clinically
important resistance mechanisms. Results obtained with estab-
lished cell lines may be misleading, in view of the discordance that
we observed in MDRI expression in tumor samples and cell lines
established from these samples. However, the analysis of clinical
samples is technically demanding and results may be difficult to
interpret. Often the diagnosis of SCLC is made on very small
amounts of tumor tissue, which may be partially necrotic. Further-
more, tumor heterogeneity and infiltration with nonmalignant cells
may pose significant problems.
The ideal method for measuring levels of drug resistance
genes or proteins in clinical samples should be sensitive, repro-
ducible, applicable to small samples, and capable of distinguish-
ing between nonmalignant infiltrating cells and tumor cells. The
quantitative PCR technique used in this study fulfills some of
these criteria but requires a homogeneous population of tumor
cells. Normal bronchial epithelium is known to express MRP,
and this could complicate the analysis of MRP expression in
lung tumor samples. In some cases, measurement of protein
levels may be more informative than gene expression levels.
on May 9, 2018. © 1997 American Association for Cancer Research.clincancerres.aacrjournals.org Downloaded from
122 Multidrug Resistance in Small Cell Lung Cancer
4 Unpublished observations.
However, the quantitation of MRP protein may be complicated
by its relatively high sensitivity to proteolytic degradation.4 The
use of immunohistochemistry for detecting multidrug resistance
proteins at the cellular level has a number of advantages but may
not necessarily reflect levels of functional protein.
Despite the infrequent expression of P-gp in SCLC cell
lines, this transport protein appears to be more frequently de-
tectable in clinical samples of these tumors. Thus, it is possible
that P-gp may play a significant role in clinical drug resistance
in SCLC. The multidrug resistance protein MRP is detectable in
a majority of cell lines as well as clinical samples of SCLC and
correlates with resistance to doxorubicin in SCLC cell lines. The
difference in expression of these multidrug resistance genes
between cell lines and clinical samples emphasizes the impor-
tance of studying clinical material and correlating results with
drug responsiveness and clinical outcome. An understanding of
clinically significant mechanisms of resistance in SCLC may
lead to effective strategies to overcome this important clinical
problem.
ACKNOWLEDGMENTS
We thank Monika Vasa, Iva Kosatka, and Heather Baker for
excellent technical assistance and Drs. Nidhi Jam and Shelagh Mirski
for helpful suggestions.
REFERENCES
1 . Ihde, D. C., Pass, H. I., and Glatstein, E. J. Small cell lung cancer. in:
V. T. DeVita, Jr., S. Hellman, and S. A. Rosenberg (eds.), Cancer:Principles and Practice of Oncology, pp. 723-758. Philadelphia: J. B.
Lippincott Co., 1993.
2. Bradley, G.. Juranka, P., and Ling, V. Mechanism of multidrug
resistance. Biochim. Biophys. Acta, 948: 87-128, 1988.
3. Gerlach, J. H. Structure and function of P-glycoprotein. In: R. F.Ozols (ed), Drug Resistance in Cancer Therapy, pp. 37-52. Boston:
Kluwer Academic Publishers, 1989.
4. Lai, S. L., Goldstein, L. J., Gottesman, M. M., Pastan, I., Tsai, C. M.,Johnson, B. E., Mulshine, J. L., and Ihde, D. C. MDRJ gene expression
in lung cancer. J. NatI. Cancer Inst., 81: 1 144-1150, 1989.
5. Goldstein, L. J., Galski, H., Fojo, A., Willingham, M., Lai, S. L.,Gazdar, A., Pirker, R., and Green, A. Expression of a multidrug resist-ance gene in human cancers. J. Nail Cancer Inst., 81: 1 16-124, 1989.
6. Noonan, K. E., Beck, C., Holzmayer, T. A., Chin, J. E., Wunder,
J. S., Andrulis, I. L., Gazdar, A. F., and Willman, C. L. Quantitative
analysis of MDRJ (multidrug resistance) gene expression in human
tumors by polymerase chain reaction. Proc. Natl. Acad. Sci. USA, 87:
7160-7164, 1990.
7. Milroy, R., Plumb, J. A., Batstone, P., Maclay. A., Wishart, G. C.,
Hay, F. G., Candlish, W., and Adamson, R. Lack of expression ofP-glycoprotein in 7 small cell lung cancer cell lines established bothfrom untreated and from treated patients. Anticancer Res., 12: 193-200,1992.
8. Cole. S. P. C., Bhardwaj, G., Gerlach, J. H., Mackie, J. E., Grant,C. E., Almquist, K. C., Stewart, A. J., Kurz, E. U., Duncan, A. M. V.,and Deeley, R. G. Overexpression of a transporter gene in a multidrug-
resistant human lung cancer cell line. Science (Washington DC), 258:
1650-1654, 1992.
9. Grant, C. E., Valdimarsson, G., Hipfner, D. R., Almquist, K. C.,
Cole, S. P. C., and Deeley, R. G. Overexpression of multidrug resis-
tance-associated protein (MRP) increases resistance to natural product
drugs. Cancer Res., 54: 357-361, 1994.
10. Cole, S. P. C., Sparks, K. E., Fraser, K., Loe, D. W., Grant, C. E.,
Wilson, G. M., and Deeley, R. G. Pharmacological characterization of
multidrug resistant MRP-transfected human tumor cells. Cancer Res.,
54: 5902-5910, 1994.
11. Campling, B. G., Haworth, A. C., Baker, H. M., Greer, D. L.,
Holden, J. J. A., Bradley, W. E. C., Pym, J., and Dexter, D. F. Estab-lishment and characterization of a panel of human lung cancer cell lines.
Cancer (Phila.), 69: 2064-2074, 1992.
12. Campling, B. G., Baer, K., Baker, H. M., Lam, Y. M., and Cole,S. P. C. Do glutathione and related enzymes play a role in drugresistance in small cell lung cancer cell lines? Br. J. Cancer, 68:
327-335, 1993.
13. Dalton, W. S., Durie, B. G. M., Alberta, D. S., Gerlach, J. H., andCress, A. E. Characterization of a new drug-resistant human myeloma
cell line that expresses P-glycoprotein. Cancer Res., 46: 5125-5130,
1986.
14. Mirski, S. E. L., Gerlach, J. H., and Cole, S. P. C. Multidrug
resistance in a human small cell lung cancer cell line selected in
Adriamycin. Cancer Res., 47: 2594-2598, 1987.
15. Campling, B. 0., Pym, J., Baker, H. M., Cole, S. P. C., and Lam,Y. M. Chemosensitivity testing of small cell lung cancer using the MUassay. Br. J. Cancer, 63: 75-83, 1991.
16. Cole, S. P. C. Multidrug resistance in human lung cancer. in: H. I.Pass, J. Mitchell, D. H. Johnson, and A. T. Tumsi (eds.), Lung Cancer.Principles and Practice, pp. 169-204, Philadelphia: J. B. LippincouCompany, 1996.
17. Reeve, J. G., Rabbitts, P. H., and Twentyman, P. R. Amplificationand expression of mdrl gene in a multidrug resistant variant of smallcell lung cancer cell line NCI-H69. Br. J. Cancer, 60: 339-342, 1989.
18. Holzmayer, T. A., Hilsenbeck, S., Von Hoff, D. D., and Roninson,I. B. Clinical correlates of MDRI (P-glycoprotein) gene expression inovarian and small-cell lung carcinomas. J. Natl. Cancer Inst., 84: 1486-1491, 1992.
19. Poupon, M. F., Arvelo, F., Goguel, A. F., Bourgeois, Y., Jacrot, M.,Hanania, N., Arriagada, R., and La Chevalier, T. Response of small-celllung cancer xenografts to chemotherapy: multidrug resistance and direct
clinical correlates. J. Nail Cancer Inst., 85: 2023-2029, 1993.
20. Tew, K. D. Glutathione-associated enzymes in anticancer drugresistance. Cancer Res., 54: 4313-4320, 1994.
21. Meijer, C., Mulder, N. H., Timmer-Bosscha, H., Zijlstra, J. G., andDe Vries, E. G. E. Role of free radicals in an Adriamycin-resistanthuman small cell lung cancer cell line. Cancer Res., 47: 4613-4617,1987.
22. Cole, S. P. C., Downes, H. F., Mirski, S. E. L., and Clements, D. J.
Alterations in glutathione and glutathione-related enzymes in a multi-drug resistant small cell lung cancer cell line. Mol. Pharmacol., 37:
192-197, 1990.
23. Campling. B. G., Baer, K. A., Lam, Y-M., Gerlach, J. H., Cole.
S. P. C., and Mirski, S. E. L. Correlation of topoisomerase II levels and
drug response in small cell lung cancer. Proc. Am. Soc. Cancer Res., 35:
453, 1994.
24. Giaccone, G., Gazdar, A. F., Beck, H., Zunino, F., and Capranico,G. Multidrug sensitivity phenotype of human lung cancer cells associ-ated with topoisomerase II expression. Cancer Res., 52: 1666-1674,
1992.
25. Kasahara, K., Fujiwara, Y., Sugimoto, Y., Nishio, K., Tamura, T.,
Matsuda, T., and Saijo. N. Determinants of response to the DNAtopoisomerase II inhibitors doxorubicin and etoposide in human lungcancer cell lines. J. Natl. Cancer Inst., 84: 1 13-1 18, 1992.
on May 9, 2018. © 1997 American Association for Cancer Research.clincancerres.aacrjournals.org Downloaded from
1997;3:115-122. Clin Cancer Res B G Campling, L C Young, K A Baer, et al. small cell lung cancer.Expression of the MRP and MDR1 multidrug resistance genes in
Updated version
http://clincancerres.aacrjournals.org/content/3/1/115
Access the most recent version of this article at:
E-mail alerts related to this article or journal.Sign up to receive free email-alerts
Subscriptions
Reprints and
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Permissions
Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)
.http://clincancerres.aacrjournals.org/content/3/1/115To request permission to re-use all or part of this article, use this link
on May 9, 2018. © 1997 American Association for Cancer Research.clincancerres.aacrjournals.org Downloaded from