Enhanced Sensitivity to 1-@-D ... - Cancer Research

3 downloads 117 Views 1MB Size Report
Vitro Antineoplastic Drug Screen to assess whether sensitivity to any of the ..... II inhibitors showed the highest correlation (data not shown). This list contained ...
[CANCER RESEARCH56. 5211-5216. November IS, 19961

Enhanced Sensitivity to 1-@-D-Arabinofuranosylcytosine and Topoisomerase II Inhibitors in Tumor Cell Lines Harboring Activated ras Oncogenes' Han-Mo Koo, Anne Monks, Andrei Mikheev,2 Larry V. Rubinstein, Marcia Gray-Goodrich, Mary Jane McWilliams, W. Gregory Alvord, Herbert K. Oie, Adi F. Gazdar, Kenneth D. Paull, Helmut Zarbl,2 and George F. Vande Woude3 ABL-Basic Research Program, National Cancer Institute-Frederick Cancer Research and Development Center, Frederick. Maryland 21702-1201 IH-M. K., M. J. M., G. F. V. WI; Science Applications International Corporation-Frederick. National Cancer Institute-Frederick Cancer Research and Development Center, Frederick. Maryland 2! 702 (A. M.. M. G-G.J; Massachusetts Institute of Technology. Cambridge. Massachusetts 02139 [A. M.. H. ii: Division of Cancer Treatment. National Cancer Institute, NIH, Bethesda, Maryland 20892 IL V. R., H. K. 0., K. D. P.!: Data Management Services, Inc., National Cancer Institute-Frederick Cancer Research and Development Center. Frederick, Maryland 21 702 [W. G. A.J: and Simmons Cancer Center and Department of Pathology, Universiti of Texas Southwestern Medical Center at Dallas, Dallas, Texas 75235 [A. F. G./

ABSTRACT We used human tumor cell lines from the National Cancer Institute's In Vitro Antineoplastic Drug Screen to assess whether sensitivity to any of the —45,000compounds tested previously correlated with the presence of a ma oncogene.

Among

these cell lines, the mutations

in Ki-ras2

clustered

in

non-small cell lung and colon carcinoma subpanels, and five of the six leukemia lines contained mutations in either N-ras or Ki-ras2. These analyses revealed a striking correlation with i-(l-D-arabinofuranosylcy. tosine (Ara-C) and 2,2'-O-cyclocytidine sensitivity in the cell lines harbor ing ras mutations compared to the tumor lines with wild-type ras alleles. Strong correlations were also found with topoisomerase (topo) IJ inhibi tors, especially 3'-hydroxydaunorubicin and an olivacine derivative. These

differential

sensitivities

persisted

in an additional

22 non-small

cell

The NCI-ADS database presently contains the differential sensitiv ity patterns of the 60 tumor cell lines to over 45,000 compounds. The sensitivity

of each cell line to a given compound

can be displayed

as

a mean graph, based on the drug concentration that gives a 50% inhibition of cell growth (G150) in 2-day assay (20, 21). Individual patterns can be compared to the entire database using the COMPARE pattern recognition program (21). COMPARE searches for similarities in mean graph patterns and ranks the similarity between patterns by a PCC (2 1). COMPARE analyses of the NCI-ADS database has iden tified compounds with similar chemical structures and/or related biochemical

mechanisms

of action

(21—29) and, recently,

substrates

and inhibitors of the P-glycoprotein, which mediates MDR, have been identified

(30, 31).

lung carcinoma lines (ras mutations, n 12 and wild-type ras, n 10). Thus, the association with Ara-C sensitivity was greatest while topo II inhibitors showed a lower, but significant, correlation. These results sug

We used COMPARE to assess whether compounds present in the database displayed selective growth inhibition against the tumor cell

gest that the ras oncogene

lines with ras oncogenes.

may play a determinant

role in rendering

tumor

The compounds

identified

by COMPARE

were further tested against an independent set of tumor cell lines. Our

cells sensitive to deoxycytidine analogues and topo II inhibitors.

results

revealed

that the presence

of a ras oncogene

in the tumor cell

lines correlated with enhanced sensitivity to deoxycytidine analogues including Ara-C, 2,2'-O-cyclocytidine, and gemcitabine. Low but

INTRODUCTION Tumor progression is a multistep process that is driven by genetic instability (1, 2). Defects in specific genes like DNA repair genes or the frequently mutated tumor suppressor genes and activated onco genes have been implicated in the genetic instability of tumor cells (reviewed in Refs. 3—9).In addition, profound changes occur in checkpoint function as well as in the regulation of apoptosis (reviewed in Refs. 5—12).Cancer cells in general display decreased sensitivity to a wide variety of therapeutic agents (13—16),but we and others (7—10) have proposed that through compromising cell cycle checkpoint func tions, the oncogene and tumor suppressor gene of cancer cells may be responsible for their preferential sensitivity to certain antineoplastic agents. To begin to test this hypothesis, we determined whether the presence of mutations in ras alleles in the 60 human tumor cell lines of the NCI-ADS4 (17) correlated with enhanced sensitivity to any of the —45,000natural and synthetic compounds that have been tested in the screen. We chose to examine the ras alleles since ras oncogene activation is the most frequently occurring gain-of-function mutation detected in human tumors (18, 19).

significant correlation 3'-hydroxydaunorubicin

MATERIALS

was also shown to topo II inhibitors, and an olivacine derivative.

especially

AND METHODS

Compounds and Cell Lines. Compoundsused in this studywereobtained through

the Drug Synthesis

and Chemistry

Branch,

NC!. The human tumor cell

lines of NCI-ADS and culture conditions has been described previously (20). The tumor cell lines derived from NSCL carcinoma (32) were maintained and assayed

for drug sensitivity

in RPM!

supplemented

with 10% FBS and 2 mM

L-glutamine.

Sequence Determination of ras Mutations. GenomicDNAs wereisolated from the cell lines of NCI-ADS using a Cell Culture DNA kit (Qiagen). The exons

I and 2 of each

ras gene

were amplified

by PCR from the genomic

DNAs using the primer pairs specific for each exon/intron junction. The PCR amplifications Elmer)

were

performed

and were all preceded

in a GeneAmp

PCR

System

with one cycle of denaturation

9600

(Perkin

at 94°C for 3 mm

and finished with the incubation at 72°Cfor 3 mm. The PCR conditions used were 40 cycles of: 94°C(15 s), 48°C(15 s), 72°C(15 s with 1 s extension per cycle) for Ki-ras2

exon I and N-ras exon I ; 94°C ( 15 s), 52°C ( I 5 s), 72°C (15

Received 7/1/96; accepted 9/I 9/96.

5 with

The costs of publication of this article were defrayed in part by the payment of page

(15 s), 48°C(15 s), 70°C(25 s with I s extension per cycle) for N-ras exon 2; and 94°C(15 s), 56°C(15 s), 72°C(15 s with I s extension per cycle) for Ha-rasl exon 1. The products were then purified from a 6% polyacrylamide

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

Supported

in

part

by

the

National

Cancer

Institute,

Department

in accordance with of

Health

and

Human

Servies, under contract with ABL. 2 Present

address:

Fred

gel. DNA sequence

Hutchinson

Cancer

Research

Center,

I 124

Columbia

3 To whom

requests

for reprints

4 The

abbreviations

used

are:

should

be addressed,

at ABL-Basic

Research

per

cycle)

for

was determined

Ki-ras2

exon

2 and

for both strands

Ha-rasl

directly

exon

2; 94°C

from the purified

PCR products using the same primers as those for PCR amplification.

Street,

Cl-OlS. Seattle. Washington 98104. National Cancer Institute-Frederick B, Frederick, MD 21702-1201.

1 s extension

sequencing

Program,

Cancer Research and Development Center, P. 0. Box

reactions

were

carried

out

with

the

PRISM

Ready

DNA

Reaction

DyeDeoxy Terminator cycle sequencing kit (ABI), and the products were analyzed on an ABI 373A DNA sequencer. The DNA sequences of primers

NCI-ADS,

National

Cancer

Institute

In

Vitro

Antineo

used

plastic Drug Screen; G130 and Gl@0,50 and 30% inhibition of cell growth, respectively;

PCC, Pearson correlation coefficient; MDR, multidrug resistance; Ara-C, l-@3-o-arabino furanosylcytosine; topo, topoisomerase; NSCL, non-small cell lung; FBS, fetal bovine serum; DGGE. denaturing gradient gel electrophoresis; AML, acute myeloid leukemia.

for

PCR

amplification

and

sequencing

were:

KR15,

GGCCTGCT

GAAAATGACTGA and KRI3, GTCCTGCACCAGTAATATGCfor Ki-ras2 exon 1; KR25, CCAGACTGTG11TCTCCC'ITC and KR23, CACAAA GAAAGCCCTCCCCA for Ki-ras2 exon 2; NR15, GACTGAGTACAAACT

5211

ras AND SENSITIVITY TO Ara-C AND TOPO 11 INHIBITORS

GGTGGandNR13, GGGCCTCACCTCTATGGTG for N-ras exon 1; NR25, GGTGAAACCTG1TrGTTGGA

and

NR23,

ATACACAGAGGAAGCCT

Table 1 Activating mutations andcolon of ras genes in the cell lines of leukemia. NSCL cancer subpanels of the NCI-ADSMutation

TCG for N-ras exon 2; HR15, CAGGCCCCTGAGGAGCGATG and HR13, UCGTCCACAAAATGGTTCT

CAGGArFCCTACCGG

for Ha-rasl

exon

1; and HR25,

TCCTG

and HR23, GGTFCACCTGTACTGGTGGA

Cellline

Subpanel

for

Ha-rasl exon 2. DGGE. The PCR amplification of each exon from genomic DNA, purifi cation and end-labeling of the PCR products, heteroduplex formation, and DNA fragments by DGGE, one primer from each pair used for the direct DNA was linked

at its 5-end

to a 54-bp

long,

thermostable,

OC-rich

clamp sequence (33). Drug Sensitivity Assays, Data Calculation, and COMPARE Algorithm. The assay methodology, sensitivity data calculation, and the COMPARE algorithm have been described in detail previously (20, 21). Briefly, cellular response to a compound was evaluated using a sulforhodamine-B assay after 48 h continuous exposure of cells with a range of the compound concentration, and the sensitivity of each cell line was calculated for the compound concen tration yielding a GI50 (20). The mean graph pattern generated from the

NSCL cancer

Colon cancer

by exact distribution

methods

using STATXACT

software

Software Co.).

a The

allelic

genomic

DNA.

status

The

of

exon

I 2: 12:

A549/ATCC EKVX

mutKl2: AGT(Ser)WIWtKI:

HOP-62 HOP-92

mut>wtKl2: TGT(Cys)WIWtWIWtKl:

mut>WtKl2: TGT(Cys)WtWtK2: mutK61: CAT(His)wtWI

COLO-205 HCC2998 HCT-l 16

(CYTEL

I 2:

mut/wtN12: TGT(Cys)K I: mutlwtK GCT(Ala)N I : mutlwtN TGT(Cys)KI:

NCI-H226 NCI-H23 NCI-H322M NCI-H460 NCI-H522

relative sensitivities of cell lines tested was used by the COMPARE pattern recognition program to calculate PCCs for all compounds in NCI-ADS data base (21). Statistical Analysis. The 1-sided Ps by Wilcoxon rank sum test were calculated

1: mutlwtK GAT(Asp)Nl: mut/wtNl2: GTF(Val)N2: mutN61: CTA(Leu)wtwtNl:

HL-60(TB) K562 MOLT-4 RPMI-8226 SR

DGGE procedure were described elsewhere (33). To facilitate separation of the sequencing

Dc.iciE°Sequence―K

CCRF-CEM

Leukemia

wt WI Kl: mut/Wt

WI Wt K13: GAC(Asp)

HCT- I5

K 1: mut/WI

K I 3: GAC(Asp)

HT-29

WI

wt

KM-12

WI

wt

SW620

Kl: mut

ras

genes

number

determined

containing

by

Kl2: G1T(Val)

DGGE

a mutation

is also

analysis

of

indicated:

PCR-amplified

K, Ki-ras2;

N,

N-ras; mut, allele With a mutation; Wt, wild-type allele; >, elevated DGGE signal for the

mutant allele compared to the wild-type signal.

RESULTS

b The

Determination of Activating Mutations of ras Genes in Tumor

Cell Lines of NCI-AD5. The mutation status of ras alleles in each

sequence

of

a mutation

independently

determined

by

direct

sequencing

of

PCR-amplified genomic DNA. The codon number harboring a mutation is indicated With the sequence and amino acid encoded. C A

silent

mutation

at

59th

codon

of

Ha-rasl

(GCC—+GCT)

was

also

found.

tumor cell line of NCI-ADS was determined from PCR-amplified

genomic DNA prepared from each cell line. Each PCR-amplified DNA sample was then analyzed for the presence of an activating mutation in the 12th, 13th and 61st codons of the Ki-, N- and Ha-ras alleles (1 8, 19) by direct DNA sequencing and DGGE. Among the 60 cell lines analyzed, 17 cell lines were found to contain activating mutations in at least one of the ras alleles. Most of the cell lines with activating mutations in ras genes clustered in three of the nine cancer-subpanels of the NCI-ADS; mutations in the Ki-ras2 allele were found in four cell lines of the NSCL cancer subpanel and in three cell lines of the colon cancer subpanel, whereas five of six leukemia lines contained mutations in either the N-ras or Ki-ras2 allele (Table 1). The additional mutant ras alleles were detected in cell lines of the remaining six cancer subpanels: N..rasArg6l in SK-MEL-2 (melano ma); K-ras@12 in OVCAR-5 (ovarian carcinoma); K@rasAst@3in SN12C (renal carcinoma); and K@rasA@@1@l3 in MDA-MB-231 and H@rasAI@@l2 in HS578T (breast carcinomas). The frequency of activat ing mutations in ras genes among the NCI-ADS cell lines were

comparablewith that found previouslyin other human tumors and tumor cell lines (18, 19). Compounds That Are Preferentially Active against Tumor Cell Lines with Activated ras Oncogenes. We used COMPARE to assess whether any compounds in the database of NCI-ADS displayed se lective growth inhibition of the tumor cell lines in the NSCL and colon carcinoma subpanels containing activated ras oncogenes. We initially focused on the 16 cell lines among which the distribution of mutant and wild-type ras allele-containing cell lines was approxi mately equal (Table 1), allowing an optimal analysis of the database using COMPARE. To establish the ras mean graph for this analysis, we assigned the cell lines with activated ras oncogenes a positive value to equate with preferential drug sensitivity, whereas those with wild-type ras alleles were given a negative value. The COMPARE analysis was performed to identify compounds in the database that: (a) were tested at least three times against all sixteen cell lines; (b) gave a PCC of >0.75 with the ras mean graph pattern (under these 5212

conditions, i.e., n = 16 cell lines, the probability of an observing correlation is 1 @-@-

carboquone

(PCC = 0.76). However, these compounds displayed poor correla 5213

1•@

T + $2 .1.@t!I53

(I, C

oncogenes and enhanced sensitivity to the topo II inhibitors, 3'hydroxydaunorubicin (47), and an olivacine derivative, NSC 659687 (Ref. 48; PCC = 0.82 and 0.8 1, respectively; Fig. 2). The range of response for these compounds was, however, narrower than for the deoxycytidine analogues (compare Fig. I with Fig. 2), but as with the deoxycytidine compounds, the same four ras oncogene-containing leukemia lines were more sensitive to the topo II inhibitors than wild-type K562 (Table 2). Positive correlations with the ras mutant NSCL and colon carci noma cell lines were also observed for a 3-deazauridine derivative compound (PCC = 0.77), the lipophilic antifolate trimetrexate (PCC

were

cells (20), and the

presence of oncogenic ras mutations in five of six cell lines in the subpanel (Table 1) introduced a bias to the mean graph. These leukemia lines, therefore, were not a suitable data set to generate a statistically relevant differential pattern for COMPARE analysis. Nev ertheless, consistent with the findings in NSCL and colon carcinoma cell lines, four of the five ras oncogene-containing leukemia lines (Table 1) were highly sensitive to the deoxycytidine analogues corn pared to K562 line with wild-type ras alleles (Table 2). In addition to the deoxycytidine analogues, the COMPARE analy sis also revealed

tions with leukemia lines in which the K562 line with wild-type ras alleles was generally more sensitive to the compounds than ras oncogene-containing lines (data not shown). The RPMI-8226 cells,

> .

a,

Cl)

.

io@—

(I, C a,

5$7

Cl)

•1

4 i@-@

10@@—

Mut

Wt

Mut

Wt

Fig. 2. Sensitivity of the NSCL and colon carcinoma cell lines of NCI-ADS to topo II inhibitors 3'-hydroxydaunorubicin (a) and olivacine derivative NSC 659687 (b). Symbols and cell line number are as described in the legend to Fig. 1.

ras AND SENSITIVITYTO Am-C AND TOPOII INHIBITORS Table 2 Sensitivities of leukemia lines of NCI-ADS to deoxycytidine analogues and topo Ii inhibitors Deoxycytidine M)Ara-CCCRF-CEM― analogueCell

line―Sensitivity

K562 Itt―2,2'O-cyclocytidineCCRF-CEM―RPMI-8226―5.01

l0@GemcitabineHL-6O(TB)―

lines

were

ordered

by

the

b Cell

lines

with

ras oncogenes.

level

RPMI-8226―

1.29 X l0―

K5628.91

1.41 x

CCRF-CEM― SR― HL-60(TB)― RPMI-8226―

.00 1.51 1.82 2.19 9.33

K5621

I .20 X

K562

1.00 X l0@NSC

CCRF-CEM―

X I0' 4.37 X ltt'

MOLT-4―

8.71x 10'

SR― RPMI-8226―

1.48 X l0@@ 2.51 X l0-@

K5622.51 a Cell

1.91 X 10' 3.47 X l0@3'-Hydroxy

RPMI-8226―1

HL-60(TB)―

of

sensitivity

(GI@0 X 10' 1.51 X l0@ 2.40 X l0@ 2.51 x l0@

X X X X X

10.8 10.8 l0@ l0@

.20 1.58 2.00 5.62 1.91

SR― MOLT@4b

line―Sensitivity MOLT-4― HL-6O(TB)― CCRF-CEM―

X 6.17 X 2.19 X 2.29 X

MOLT@4b HL-60(TB)― SRb

II inhibitorCell

(Gl,0 M)Topo

daunorubicinSRb

l0@ i07 l0@ iO-7 l0'

659687MOLT-4―

X X X x x

10' Itt' 10' 10' Itt'

2.75 X l0@ to

each

compound.

inhibitors compared to other ras oncogene-containing leukemia lines (Table 2), also displayed the relative resistance to antifolates and alkylating agents (data not shown). Tests of Am-C and Topo II Inhibitors on Additional NSCL

lc). The basis of correlation between ras oncogenes and enhanced sensitivity to deoxycytidine analogues is not yet understood. How ever, with an equimolar saturation dose of Ara-C (i.e., 10 nM), we observed more efficient apoptotic cell death in the ras oncogene

Tumor Cell Lines. To extend our findings with the cell lines of

containing

NCI-ADS, we tested an additional 22 NSCL cell lines for which the

Despite a large variety of well-characterized topo II inhibitors tested in NCI-ADS, significant correlations with the ras mutation pattern was only observed with 3'-hydroxydaunorubicin and the oh vacine derivative, NSC 659687. Further characterization of these topo II inhibitors revealed that 3'-hydroxydaunorubicin and the ohivacine derivative were superiorly capable of overcoming MDR.―These re

profileof ras mutationshad been determinedpreviously(32). These cell lines, 10 of which contained wild-type ras alleles, were treated with Ma-C, 3'-hydroxydaunorubicin, and the olivacine derivative, NSC 659687, as well as trimetrexate. The same assay conditions as in

NCI-ADS were used, except that these cell lines were assayed in growth medium with 10% FBS instead of the 5% FBS growth con ditions used in the NCI-ADS (20). The sensitivity to Ara-C was evaluated at a drug concentration that gave a 30% inhibition of cell growth (G130), since 50% growth inhibition was not achieved, even at 1 mM Ara-C for most of the cell lines (data not shown). The differ entials in drug sensitivity observed in these cell lines were less than in NCI-ADS cell lines (compare Figs. 1 and 2 with 3). However, the NSCL cell lines with mutant ras alleles again strongly associated with greater sensitivity to Ara-C (1-sided P = 0.01 by Wilcoxon rank sum test); the mean GI30 Ara-C concentrations for mutant and wild-type ras-containing cell lines were 1.35 X i07 M and 1.95 X lO_6 M, respectively (Fig. 3a). As in the NCI-ADS test, the differential re sponse to 3'-hydroxydaunorubicin and the olivacine derivative com pound was lower than with Ara-C and also showed a lower correlation (P = 0.04 for both compounds; Fig. 3b and c). DISCUSSION Our search for chemotherapeutic agents in the NCI-ADS database that are preferentially more effective at inhibiting the growth of tumor cells harboring activated ras oncogenes identified two different classes

of compounds,

deoxycytidine

analogues

and topo II inhibitors

(Figs. 1 and 2, and Table 2). The preferential activities of the com pounds against ras oncogene-containing tumor cells were further

tested using additional tumor cell lines derived from NSCL carcino

cells than in cells with wild-type

sults suggested

that the activation

ras alleles.5

of ras oncogenes

in tumor

cells

specifically enhanced the intrinsic sensitivity to topo II inhibitors, but this ras oncogene-associated

sensitization

could be unveiled

only with

MDR-escaping topo II inhibitors or in the absence of MDR in the target cells.6 We also performed COMPARE analysis using all 60 cell lines and found that, again, the deoxycytidine analogues and the topo II inhibitors showed the highest correlation (data not shown). This list contained many more topo II inhibitors than deoxycytidine analogues with the highest PCC of —0.55(data not shown). Although most of the tumor cell lines with ras oncogenes displayed highly enhanced sensitivity to deoxycytidine analogues, SW620 and HCT- 15 derived from colon carcinomas and one NSCL carcinoma line were less sensitive to deoxycytidine analogues (Figs. 1 and 3). However, these cell lines were quite sensitive to the topo II inhibitors (Figs. 2 and 3). The sensitivity

of these cells to the topo II inhibitors

may be related to the “enhancing― effect of topo II inhibitors when combined with a deoxycytidine analogue, as in the standard Ara-C/ daunorubicin combination regimen used in the treatment of acute leukemias (49, 50). Does the ras oncogene play a significant role in rendering tumor cells sensitive to deoxycytidine analogues and topo II inhibitors? In a study of 99 patients with AML, Neubauer et a!. (5 1) observed a relationship between the presence of a ras oncogene in the heukemic cells and improved overall survival of the patients in response to treatment primarily comprised of Ara-C and a topo II inhibitor. It was speculated that the presence of the ras oncogene accounted for the prolonged survival of patients, either by its association with a lower

mas (Fig. 3). Although Ara-C and related compounds have been the single most effective group of therapeutics for the treatment of acute leukemias (49), these compounds have been essentially inactive against solid tumors. Gemcitabine, however, a deoxycytidine ana 5 H-M. Koo, unpublished observation. logue showing some activity against solid tumors (40—46),displayed 6 H.-M. Koo, M. Gray-Goodrich, A. Vaigro-Wolff, M. J. McWilliams, preferential activity against the mutant ras cell lines like Ara-C (Fig. Monks, and G. F. Vande Woude, manuscript in preparation. 5214

K. D. Paull, A.

@

;@ ras AND SENSITIVITY TO Ma-C AND TOPO II INHIBITORS

a

b

_____

io-7— I

•

1.. .@

S

ii +

S

>.

S

io@

S S

10.8.

1 @-@-

C

@@-8

•

.T

10.6@

U, C

I

a,

Cl)

10.8_

S

@,

@:

@‘ ..

@5S

U) C

.@ C

a,

.1

a,

.

Cl)

S

Cl)

io@—

S

.

10@-

S S

.

T .Ã '+

S

.

10-s—

. .

10.6_

. S -—-3

10@-

Mut

Wt

Mut

Wt

Mut

Wt

Fig. 3. Sensitivity of additional NSCL carcinoma cell lines to: a, Ara-C: GI30 = 1.35 X l0@ M for Mut (•)and 1.95 X l0_6 M for Wt ( •); b, 3'-hydroxydaunorubicin:

GI50 9.77 x l0—@ MforMutand 1.86X l0_6 si for Wt;andc, olivacinederivativeNSC659687:GI50= 3.47X 10' MforMutand 1.26X l0@ Mfor Wt.Thecompoundswere tested at ten 10-fold dilutions. Mean drug concentrations conferring a 30% inhibition of cell growth for Ara-C and a 50% inhibition for the topo II inhibitors for each cell line were determined from three independent experiments. Mean sensitivities (+) for each group of cell lines are shown; bars, SE.

heukemic burden or by rendering the leukemic cells more sensitive to the chemotherapeutic agents used (5 1). In addition, consistent with a ras oncogene affecting the chemosensitivity of AML, 6 of 10 initially ras oncogene-positive AML patients, who had been treated with Ara-C and a topo II inhibitor, had only wild-type ras alleles in their leukemic cells at relapse, indicating that the ras oncogene-positive leukemic cells had been selectively eliminated (52). Furthermore, a cohort of ras oncogene-positive AML patients exhibited a significantly longer duration of complete remission when treated with progressively higher doses of Ara-C as consol idation therapy, whereas the postremission Ara-C dose intensifi cation did not significantly impact on complete remission duration of AML patients with wild-type ras alleles (53). These clinical observations (51—53) are consistent with our results illustrating that the tumor cell hines derived from heukemias and NSCL and colon carcinomas with activated ras oncogenes were preferentially more sensitive to deoxycytidine analogues and topo II inhibitors. In addition, these results suggest that the combination therapy used in AML treatment (51) may be more effective at killing leukemic cells containing ras oncogenes, through either additive or syner gistic mechanisms (49, 50). Our results combined with the AML clinical data reveal a signifi cant correlation between the ras oncogene status of tumor cells and their sensitivity to clinically relevant drugs. One application of these findings may be to determine the therapeutic efficacy of the deoxy cytidine analogues and topo II inhibitors on other tumors bearing frequent ras mutations, such as NSCL, colon and, notably, pancreatic carcinomas (18, 19); it should be noted that gemcitabine has been approved recently for the treatment of locally advanced pancreatic 5215

carcinoma (54), in which a large percentage of the tumors contain activated ras oncogenes (18, 19). If the relationship between cancer specific

genotypes

and antineophastic

drugs

provides

an explanation

for the preferential sensitivity of tumor cells to successful chemother apy protocols, it is hoped that such approaches would direct chemo therapy away from purely empirical approaches to strategies based on cancer-specific

genotypes.

ACKNOWLEDGMENTS We thank B. Chabner, M. Grever, G. Johnson, and E. Sausville for helpful discussion

and continued

support.

We thank M. Strobel,

K. Fukasawa,

I. Daar,

E. Harlow, and C. D. Bloomfield for critical reading of the manuscript. We also thank C. Rhoderick and M. Reed for preparation of the manuscript.

REFERENCES I . Nowell, P. C. The clonal evolution of tumor cell populations. Science (Washington

DC), 194: 23—28,1976. 2. Nicolson, G. L. Tumor cell instability, diversification, and progression to the meta static phenotype: from oncogene to oncofetal expression. Cancer Res., 47: 1473— 1487,1987. 3. Modrich, P. Mismatch repair, genetic stability, and cancer. Science (Washington DC), 266: 1959—1960,1994.

4. MacPhee, D. G. Mismatch repair, somatic mutations, and the origins of cancer. Cancer Res., 55: 5489—5492, 1995. 5. Kaufmann, W. K. Cell cycle checkpoints and DNA repair preserve the stability of the human genome. Cancer Metastasis Rev.. 14: 31—41,1995.

6. HarIwell, L. Defects in a cell cycle checkpoint may be responsible for the genomic instability of cancer cells. Cell, 71: 543—546,1992. 7. Weinert, T., and Lydall, D. Cell cycle checkpoints, genetic instability and cancer. Cancer Biol., 4: 129—140,1993. 8. Vande Woude, G. F., Schulz, N., Thou, R., Paules. R. S., Daar, I., Yew, N., and Oskarsson, M. Cell cycle regulation. oncogenes, and antineoplastic drugs. In: General

ras AND SENSITIVITY TO Ara-C AND TOPO II INHIBITORS Motors Cancer Research Foundation: 1990 Views of Cancer Research, pp. 128—143. Philadelphia, PA: J. B. Lippincott Co., 1990.

9. Murakami,M. S., Strobel,M. C., and VandeWoude,G. F. Cell cycleregulation, oncogenes, and antineoplastic drugs. In: P. M. Howley, M. A. Israel, and L. A. Liotta (eds.), The Molecular Basis of Cancer, pp. 3—17.Philadelphia: W. B Saunders, Co.,

1994.

10. Hartwell,L. H., and Kastan,M. B. Cellcyclecontroland cancer.Science(Wash ington DC), 266: 1821—1828,1994. 11. Thompson, C. B. Apoptosis in the pathogenesis and treatment of disease. Science (Washington DC), 267: 1456—1462, 1995. 12. Kastan, M. B., Canman, C. E., and Leonard, C. J. p53, cell cycle control and apoptosis: implications for cancer. Cancer Metastasis Rev., 14: 3—15,1995. 13. Lowe, S. W., Bodis, S., McClatchey, A., Remington, L., Ruley. H. E., Fisher, D. E.,

Bates, S. E. Rhodamine efflux patterns predict P-glycoprotein substrates in the National Cancer Institute Drug Screen. Mol. Pharmacol., 46: 627—638, 1994. 32. Mitsudomi, T., Viallel, J., Mulshine, J. L., Linnoila, R. I., Minna, J. D., and Gazdar,

A. F. Mutations of ras genes distinguish a subset of non-small-cell lung cancer cell lines from small-cell lung cancer cell lines. Oncogene, 6: 1353—1362,1991. 33. Mikheev, A., Cha, R. S., and Zarbl, H. Detection of point mutations in Ras in tumor cell lines by denaturant gradient gel electrophoresis. Methods Enzymol., 255: 442—

451,1995.

16. Morgan,W. F., and Murnane,J. P. A role for genomicinstabilityin cellular

34. Chabner, B. A. Anticancer Drugs. In: V. T. DeVita, Jr., S. HeIlman, and S. A. Rosenberg (eds.), Cancer: Principles & Practice of Oncology, Ed. 4, pp. 325—417. Philadelphia: J. B. Lippincott Co., 1993. 35. Hoshi, A., KanzaWa, F., Kubetani, K., Saneyoshi, M., and Arai, Y. 2,2'-O-Cyclocy tidine: antitumor cytidine analog resistant to cytidine deaminase. Gann. 62: 145—146, I971. 36. Novotny, L., Farhall, H., Janku, I., and Beranek, J. Comparative study of cyclocyti dine and arabinosylcytosine disposition in rats. Neoplasma. 35: 707—714,1988. 37. Heinemann, V., Hertel, L. W., Grindey, G. B., and Plunkett, W. Comparison of the cellular pharmacokinetics and toxicity of 2',2'-difluorodeoxycytidine and 1-@3-oarabinofuranosylcytosine. Cancer Res., 48: 4024—4031, 1988. 38. van Haperen, V. W. T. R., and Peters, G. J. New targets for pyrimidine antimetabo lites for the treatment of solid tumours. Pharmacy World Sci., JO: 104—112, 1994. 39. Grunewald, R., Kantarjian, H., Du, M., Faucher, K., Tarassoff, P., and Plunkett, W. Gemcitabine in leukemia: a phase I clinical, plasma, and cellular pharmacology study.

radioresistance? Cancer Metastasis Rev., 14: 49—58,1995. 17. Grever, M. R., Schepartz, S. A., and Chabner, B. A. The National Cancer Institute.

40. Abbruzzese, J. L., Grunewald, R., Weeks, E. A., Gravel, D.. Adams, T., NOWak,B..

Housman, D. E., and Jacks, T. p53 status and the efficacy of cancer therapy in vivo. Science (Washington DC), 266: 807—810, 1994. 14. Schaefer, D. I., Livanos, E. M., White, A. E., and Tlsty, T. D. Multiple mechanisms of N-(phosphonoacetyl)-L-asparlate drug resistance in SV4O-infected precrisis human fibroblasts. Cancer Res., 53: 4946—4951, 1993. 15. Fan, S., El-Deity, W. S., Bae, I., Freeman, J., Jondle, D., Bhatia, K., Fomace, A. J., Jr., Magrath, I., Kohn, K. W., and O'Connor, P. M. p53 gene mutations are associated with decreased sensitivity of human lymphoma cells to DNA damaging agents. Cancer Res., 54: 5824—5830, 1994.

J. Clin.Oncol.,10:406—413, 1992.

Cancer Drug and Discovery and Development Program. Scm. Oncol., 19: 622—638, 1992. 18. Bos, J. L. ras oncogenes in human cancer: a review. Cancer Res., 49: 4682—4689, 1989.

19. 20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

Mineishi, S., Tarassoff. P., Satterlee, W., Raber, M. N., and Plunkeu, W. A phase I clinical, plasma, and cellular pharmacology study of gemcitabine. J. Clin. Oncol., 9:

491—498, 1991.

41. Lund, B., Kristjansen, P. E. G., and Hansen, H. H. Clinical and preclinical activity of 2',2'-difluorodeoxycytidine (gemcitabine). Cancer Treatment Rev., 19: 45—55,1993. Barbacid, M. ras oncogenes: their role in neoplasia. Eur. J. Clin. Invest., 20: 225—235, 42. Casper, E. S., Green, M. R., Kelsen, D. P., Heelen, R. T., Brown, T. D., Flombaum, 1990. C. D., Trochanowski, B., and Tarassoff, P. G. Phase H trial of gemcitabine (2,2'Monks, A., Scudiero, D., Skehan, P., Shoemaker, R., Paull, K., Vistica, D., Hose, C., difluorodeoxycytidine) in patients with adenocarcinoma of the pancreas. Invest. New Langley, J., Cronise, P., Vaigro-Wolff, A., Gray-Goodrich, M., Campbell, H., Mayo, J., and Boyd,M. Feasibilityof a high-fluxanticancerdrug screenusinga diverse Drugs, 12: 29—34,1994. 43. Carmichael, J., Fink, U., Russell, R. C. G., Spittle, M. F., Harris, A. L., Spiessi, G., panel of cultured human tumor cell lines. J. Nail. Cancer. Inst., 83: 757—766,1991. and Blalter, J. Phase II study of gemcitabine in patients with advanced pancreatic Paull, K. D., Hamel, E., and Malspeis, L. Prediction of biochemical mechanism of cancer. Br. J. Cancer, 73: 101—105,1996. action from the in vitro antitumor screen of the National Cancer Institute. In: W. 0. 44. Shepherd, F. A., Gatzemeier, U., Gotfried, M., Weynants, P., Coltier, B., Groen, H., Foye (ed.) Cancer Chemotherapeutic Agents, pp. 9—45.Washington, DC: American Rosso, R., Mattson, K., Cornes-Funes, H., Tonato, M., Hatty, S., and Voi, M. An Chemical Society, 1995. extended phase II study or gemcitabine in non-small cell lung cancer. Proc. Am. Soc. Gupta, M., Abdel-Megeed, M., Hold, Y., Kohihagen, G., Paull, K., and Pommier, Y. Eukaryotic DNA topoisomerases mediated DNA cleavage induced by a new inhibitor: Clin. Oncol., 12: 330, 1993. 45. Abratt, R. P., Bezwoda, W. R., Falkson, G., Goedhals, L., Hacking, D., and Rugg, NSC 665517. Mol. Pharmacol., 48: 658—665,1995. T. A. Efficacy and safety profile of gemcitabine in non-small-cell lung cancer: a phase Solary, E., Leteurtre, F., Paull, K., Scudiero, 0., Hamel, E., and Pommier, Y. Dual II study.J. Clin.Oncol.,12: 1535—1540, 1994. inhibition of topoisomerase II and tubulin polymerization by azatoxin, a novel 46. Anderson, H., Lund, B., Bach, F., Thatcher, N., Walling, J., and Hansen, H. H. cytotoxic agent. Biochem. Pharmacol., 45: 2449—2456, 1993. Single-agent activity of weekly gemcitabine in advanced non-small-cell lung cancer: Leteurtre, F., Sackeu, D. L., Madalengoitia, J., Kohihagen, G., MacDonald, T., a phase II study. J. Clin. Oncol., 19: 1821—1826,1994. Hansel, E., Paull, K. D., and Pommier, Y. Azatoxin derivatives with potent and 47. Fuchs, E.-F., Horton, D., Weckerle, W., and Winter-Mihaly, E. Synthesis and anti selective action on topoisomerase II. Biochem. Pharmacol., 49: 1283—1290,1995. tumor activity of sugar-ring hydroxyl analogues of daunorubicin. J. Med. Chem., 22: Leteurtre, F., Kohlhagen, G., Paull, K. D., and Pommier, Y. Topoisomerase II 406—411, 1979. inhibition and cytotoxicily of the anthrapyrazoles DuP 937 and DuP 941 (Losox 48. Jasztold-Howorko, R., Landras, C., Pierre, A., Alassi, G., Guilbaud, N., Kraus antrone) in the National Cancer Institute preclinical antitumor drug discovery screen. Berthier, L., Leonce, S., Rolland, Y., Prost, J-F., and Bisagni, E. Synthesis and J. Nat!.CancerInst.,86: 1239—1244, 1994. evaluation of 9-hydroxy-5-methyl-(and 5,6-dimethyl-6H-pyrido[4,3-b]carbazole1Cleaveland, E. S., Monks, A., Vaigro-Wolfe, A., Zaharevitz, D. W., Paull, K., N-[(dialkylamino)alkyl]carboxamides, a new promising series of antitumor olivacine Ardalan, K., Cooney, D. A., and Ford, H. J. Site of action of Iwo novel pyrimidine derivatives. J. Med. Chem., 37: 2445—2452, 1994. biosynthesis inhibitors accurately predicted by the COMPARE program. Biochem. 49. Keating, M. J., Estey, E., and Kantarjian, H. Acute leukemia. In: V. T. DeVita, Jr., S. Pharmacol., 49: 947—954,1995. Hellman, and S. A. Rosenberg (eds.), Cancer: Principles & Practice of Oncology, Ed. Bai, R., Paull, K. D., Herald, C. L., Malspeis, L., Pettit, G. R., and Hansel, E. 4, pp. 1938—1964.Philadelphia: J. B. Lippincott Co., 1993. Halichondrin B and homohalichondrin B: marine natural products binding in the 50. Rohatiner, A. Z. S., and Lister, T. A. The treatment of acute myelogenous leukemia. Vinca domain of tubulin. J. Biol. Chem., 266: 15882—15889,1991. In: E. S. Henderson and T. A. Lister (eds.), Leukemia, Ed. 5, pp. 485—514.Phila Paull, K. D., Lin, C. M., Malspeis, L., and Hamel, E. Identification of novel delphia: W. B. Saunders Co., 1990. antimitotic agents acting at the tubulin level by computer-assisted evaluation of 51. Neubauer, A., Dodge, R. K., George, S. L., Davey, F. R., Silver, R. T., Schiffer, C. A., differential cytotoxicity data. Cancer Res., 52: 3892—3900, 1992. Mayer, R. J., Ball, E. D., Wurster-Hill, D., Bloomfield, C. D., and Liu, E. T. Kuo, S-C., Lee, H-Z., Juang, J-P., Lin, Y-T., Wu, T-S., Chang, J-J., Lednicer, D., Prognostic importance in the rex proto-oncogenes in de novo acute myeloid leukemia. Paul), K. D., Lin, C. M., Hamel, E., and Lee, K-H. Synthesis and cytotoxicily of Blood,83: 1603—161 1,1994. 1,6,7,8-substituted 2-(4'-substituted phenyl)-4-quinolones and related compounds: 52. Liu, E. T. The role of ras gene mutations in myeloproliferalive disorders. Clinics Lab. identification as antimitotic agents interacting with tubulin. J. Med. Chem., 36: Med., 10: 797—807, 1990. 1146—1156, 1993. 53. Liu, E. T., Dodge, R., Meyer, A., Thang, X. X., Schiffer, C., Mayer, R., and Alvarez, M., Paull, K., Monks, A., Hose, C., Lee, J-S., Weinstein, J., Grever, M., Bloomfield, C. 0. De novo acute myeloid leukemias harboring ma mutations exhibit Bates, S., and Fojo, T. Generation of a drug resistance profile by quantitation of preferential sensitivity to dose intensive cytarabine as consolidation therapy. Blood, mdr-1/P-glycoprotein in the cell lines of the National Cancer Institute Anticancer 86: (Suppl. 1): 2380, 1995. Drug Screen. J. Clin. Invest., 95: 2205—2214,1995. 54. ODAC okays gemcitabine for pancreatic cancer. Cancer LeIt., 21: 6—7,1995. Lee, J-S., Paul!, K., Alvarez, M., Hose, C., Monks, A., Grever, M., Fojo, A. T., and

5216