Amplification of the ATP-Binding Cassette 2 Transporter Gene Is Functionally Linked with Enhanced Efflux of Estramustine in Ovarian Carcinoma Cells1

[CANCER RESEARCH 58, 1332-1337. April 1. 1998] Advances in Brief

Naomi M. Laing, Martin G. Belinsky, Gary D. Kruh, Daphne W. Bell, Jonathan T. Boyd, Linda Barone,2 Joseph R. Testa, and Kenneth D. Tew3

Departments of Pharmacology IN. M. L.. J. T. B.. L B.. K. D. T.I and Medical Oncology Â¡M.C. B.. G. D. K.. D. W. B.. J. R. T.]. Fax Chase Cancer Center, Philadelphia, Pennsylvania 19111

Abstract affinity for EM (6) suggests a mechanism by which its overexpression contributes to EM resistance. We have also detected enhanced efflux An estramustine-resistant human ovarian carcinoma cell line, SKIM, of EM in an EM-resistant cell line, suggesting that increased expres was generated to explore resistance mechanisms associated with this sion of a transporter could also contribute to EM resistance. agent. Cytogenetic analysis revealed that SKEM cells have a homoge Although EM can bind to pgp and modulate its transport activity (7, neously staining region (hsr) at chromosome 9q34. Microdissection of the 8), pgp-overexpressing cells do not display cross-resistance to EM. In hsr, followed by fluorescence in situ hybridization to SKEM and normal addition, overexpression of pgp is not detected in any EM-resistant metaphase spreads, confirmed that the amplified region was derived from sequences from 9q34. In situ hybridization with a probe specific lor .I/if?, cell lines studied to date (9). These observations suggest that a a gene located at 9q34 that encodes an ATP-binding cassette 2 (ABC2) transporter distinct from pgp may contribute to EM resistance. transporter, indicated that this gene is amplified (Â»-foldinthe estramus To explore further EM resistance mechanisms, we characterized tine-resistant cells. Southern analysis confirmed that ABC2 was amplified SKEM in this study, an EM-resistant human ovarian carcinoma cell in SKEM, and Northern analysis indicated that the ABC2 transcript was line (SKOV3). Cytogenetic analysis revealed an hsr in this cell line overexpressed ~5-fold. The ABC1 gene located at 9q22â€”31was not am located at chromosome 9q34, the location of ABC2, which encodes a plified in the resistant cells, and niRNA levels of several other ABC partially characterized ABC transporter (10). Further analysis re transporter genes were unaltered. Consistent with the concept that in vealed that ABC2 is amplified and overexpressed in SKEM, that creased ABC2 expression contributes to the resistant phenotype, we ob antisense-mediated down-regulation of ABC2 expression sensitized served that the rate of efflux of dansylated estramustine was increased in SKEM compared with control cells. In addition, antisense treatment the cell line to EM, and that SKEM displays enhanced efflux of EM. directed toward ABC2 mRNA sensitized the resistant cells to estramus This study thus suggests that overexpression of ABC2 contributes to tine. Together, these results suggest that amplification and overexpression EM resistance by functioning as an efflux pump of the drug. These of ABC2 contributes to estramustine resistance and provides the first data provide the first evidence for a critical cellular role for this new indication of a potential cellular function for this product. ABC transporter family member.

Introduction Materials and Methods

EM,4 a conjugate of nor-nitrogen mustard and estradici, is an Development of Estramustine-resistant Cell Lines. SKOV3 ovarian car effective chemotherapeutic agent that is presently used to treat met- cinoma cells were maintained in a-MEM medium and cloned, and then one astatic, hormone-refractory prostate cancer (1). The cytotoxic activity clone was used for all subsequent experiments. Resistant cells were developed of EM relates to its ability to induce metaphase arrest by interfering by exposure to EM at concentrations starting at 1 /J.Mwith a gradual increase with microtubule structure and function (2). Biochemical studies have to 15 /Â¿Mover a 6-month period. The resulting SKEM cells were maintained shown that EM binds to microtubule-associated proteins and tubulin in 15 /UMEM. (3, 4), the consequence of which is depolymerization of microtubules. Chromosome Microdissection and DNA Amplification. Metaphase tu mor cells of SKEM were harvested by overnight exposure to Colcemid (0.01 To understand mechanisms associated with resistance to this agent, /Â¿g/ml).Metaphase spreads were prepared on 24 x 60-mm coverslips, air- we have previously characterized EM-resistant tumor cell lines and dried, and banded with trypsin-Giemsa. A single copy of the hsr was dissected studied the effects of EM on MDR cell lines. These studies demon under a Zeiss Axiovert inverted microscope with UV light-sterilized glass strated that alterations in the expression of the ÃŸ-subunitof the target needles controlled by an Eppendorf Micromanipulator 5171, as described in protein tubulin are associated with resistance to EM and other anti- detail previously (11). Briefly, needle tips containing dissected chromosomal microtubule agents. In addition, increased expression of the ÃŸn, material were broken off directly into a 0.5-ml thin-walled Eppendorf tube. isotype of tubulin was observed in EM-resistant prostate carcinoma Microdissected DNA was PCR-amplified by the sequence-independent ampli cells (5). The subsequent observation that this isotype has reduced fication method of Bohlander et al. (12). As a positive control, 20 pg of human placenta! DNA was amplified simultaneously. A negative control without DNA template was also included. A 10-ju.l sample of each reaction was Received 12/22/97: accepted 2/16/98. visualized by ethidium bromide staining following electrophoresis through a The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 2.0% agarose gel. 18 U.S.C. Section 1734 solely to indicate this fact. Microdissection-FISH. Metaphase spreads from SKEM cells and phyto- 1This work was supported in part by N1H Grants CA06927 and RR05539. NIH Grant hemagglutin-stimulated lymphocytes from human peripheral blood were pre CA53893 (to K. D. T.), and by an appropriation from the Commonwealth of Pennsylva nia. pared according to the method of Fan et al. (13). Cultures were synchronized 2 Present address: Viro-Pharma. Malvem, PA 19355. by treatment with 5-bromodeoxyuridine (0.18 fig/ml; Sigma Chemical Co.) for 1To whom requests for reprints should be addressed, at Department of Pharmacology, 16 h, followed by release from the block by incubation in fresh medium Fox Chase Cancer Center, 7701 Burholme Avenue, Philadelphia. PA !91I1. Phone: containing thymidine (2.5 jug/ml) for 6 h. Metaphase cells were harvested, and (215)728-3137; Fax: (215)728-4333. 4 The abbreviations used are: EM. estramustine; MDR. multidrug resistance: pgp, chromosome spreads were prepared according to standard procedures. P-glycoprotein: ABC, ATP-binding cassette: hsr, homogeneously staining region; FISH, A 1-jnl aliquot of the primary PCR was labeled with biotin-16-dUTP (Enzo fluorescence in xilu hybridization; DAPI. diamidino-2-phenylindole. Diagnostics) in a secondary PCR reaction (14), and 200-400 ng (5-10 Â¿tl)of 1332

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1998 American Association for Cancer Research. ABC2 GENE AMPLIFICATION AND DRUG RESISTANCE this DNA were hybridized to metaphase spreads from both SKEM cells and microscope, and the images were analyzed by ISEE analysis software (Inno- normal blood lymphocytes. vision). Excitation wavelength was 360 nm, and the emission was recorded The ABC2 probe was a 1.75-kb purified PCR product of human ABC2 with a broad spectrum emission filter. cDNA. The primers used to generate the fragment were chosen from the Daunorubicin Fluorescence. Cells on coverslips were treatedwith 10 UM sequence of an expressed sequence tag representing part of the human ABC2 daunorubicin (Sigma) for 2 h. Cells were washed and imaged after various time cDNA (EST0600) and from a 500-bp sequence from the PCR analysis of intervals using the 568-nm excitation line of a Bio-Rad MRC-600 laser ABC2 message levels. The 1.75-kB fragment was gel purified, cloned into the scanning confocal microscope. TA vector (Clontech), and sequenced. The sequence exhibited nearly >90% homology with the murine ABC2 sequence. The purified probe was labeled by Results nick translation with biotin-16-dUTP (Boehringer Mannheim) and mapped to ABC2 Is Amplified in the EM-resistant Cell Line. EM-resistant normal human chromosomes. FISH and detection of immunofluorescence were carried out essentially as ovarian carcinoma cells were generated by exposing sensitive cells to described previously (15). Biotinylated probes were denatured, preannealed gradually increasing concentrations of the drug. The resulting cells with excess Cot-1 DNA, and hybridized overnight at 37Â°Cto metaphase (designated SKEM) exhibited a 5-6-fold level of resistance to EM spreads on a 24 x 60-mm slide. Hybridization sites were detected with compared with the SKOV3 parental cells. Because drug-resistant cells fluorescein-labeled avidin (Oncor) and amplified by the addition of anti-avidin often display newly acquired chromosomal abnormalities, karyotypic antibody (Oncor) and a second layer of fluorescein-labeled avidin. The chro analysis of parental SKOV3 and SKEM cells was performed. The mosome preparations were counterstained with DAPI and observed with a Giemsa-banded metaphase chromosomes from SKEM cells showed Zeiss Axiophot epifluorescence microscope equipped with a cooled charge the presence of an abnormal chromosome 9 not present in the parental coupled device camera (Photometries, Tucson AZ) operated by a Macintosh cells. This aberrant chromosome 9 contained an hsr located at band computer work station. Digitized images of DAPI staining and fluorescein signals were captured, pseudo-colored, and merged using Oncor version 1.6 9q34 (Fig. la). We used chromosome microdissection techniques to software. characterize the amplicon of the hsr. Several copies of the hsr from Southern Analysis. DNA was isolated with a Puregenekit (CentraSys SKEM cells were microdissected (Fig. \b), pooled, and PCR ampli tems) and treated with RNase (l h at 37Â°C).Aliquotsof DNA (25 ^ig) were fied. To confirm that the amplified DNA came from within the hsr. the then cut with one of four restriction enzymes (12 h at 37Â°C),separatedon an PCR products were hybridized back to SKEM metaphase spreads. As agarose gel, and transferred to a nylon membrane. Hybridization of the anticipated, the resulting fluorescence was localized to the hsr region 32P-labeled 1.75-kb ABC2 probe to the blot was performed with RapidHyb in chromosome 9q34 (Fig. If, arrow, for discussion of Fig. \d, see (Amersham) according to the manufacturer's instructions. below). Hybridization to a normal copy of chromosome 9 in SKEM Northern Analysis. RNA was isolatedwiththeRNeasykit(Quiagen).The was also observed (Fig. \c, arrowhead). FISH analysis of the micro- RNA was then treated with DNase and purified with the RNeasy kit. RNA samples were separated by electrophoresis through a 1% agarose-2.2 Mform aldehyde gel and transferred to Magna NT membrane (Micron Separations, Westborough, MA). The membranes were hybridized with 32P-labeledprobes * 1 and imaged as described for the Southern analysis. For the analysis of both ABC1 and ABC2 mRNA levels, primers were chosen from the published murine sequences and were used to amplify the corresponding sequences from human cDNA. The resulting PCR products were sequenced, found to exhibit 93% (ABC1)and 90% (ABC2)homology with their murine counterparts, and then labeled with "P. RT-PCR Analysis. cDNA synthesis and RT-PCRwere performedas de scribed (16). Primers for RT-PCR analysis were derived from the published coding sequences as follows: MDRl, nucleotides 1325-1347, 1502-1523; MRP, 4208-4229. 4414-4435; murine ABC2 (EMBL accession no. X75927), 2503-2522, 3004-3025; ABC C (EMBL accession no. X97187), 21-39,459- 479; and cMOAT(GenBank accession no. U63970), 1-21, 516-536. Primers for 18S rRNA were as described (16).The predicted target size for each of the primer pairs was 198bp (MDRl), 229 bp (MRP), 523 bp (ABC2),459 bp (ABC C), 536 bp (cMOAT), and 311 bp (18S rRNA). PCR cycling conditions were 30 s at 94Â°C,30s at 60Â°C,and30 s at 72Â°C. Antisense Phosphorothioate Oligonucleotide Treatment and Cytotoxic- ity Assay. SKEMcells were platedin 24-well plates and incubatedfor 24 h at 37Â°C.Thecells were exposed to Superfect-oligonucleotide complexes [1 /ng of oligonucleotide and 5 ftl of Superfect (Qiagen) in 400 Â¡Ãºserum-containing media] for 2 h, washed, and then incubated with media containing serum and 5 JIM oligonucleotide for 48 h. For the cytotoxicity experiments, various concentrations of EM were then added to the wells. After 48 h, cell viability was determined with the MTS (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxy- methoxyphenyl)-2-(4-sulfophenyl)-2//-tetrazolium) assay. For the analysis of mRNA levels, the cells were harvested after the oligonucleotide treatment, and Cytogenetic changes in SKEM partial metaphase spread of normal the RNA was isolated with an RNeasy kit. Two phosphorothioate oligonu- Giemsa-banded chromosomes from SKEM cells. Ins t, normal chromosome 9 (left) and an aberrant copy of chromosome 9 containing an h: (hrucken. h. the same metaphase cleotides were used: 638 (GTCTGCACCAGTGAGTGCCG) and 613 ,â€¢).c.partial metaphaM- spread of (CAGTTTGTCAGCGGCTGCAAC). spread as in a after microdissection of the hsr (am SKEM cells illustrating chromosomal in aim suppres ion nybridi/ation of PCR products Dansylated EM Fluorescence. Cells wereplatedontoglass coverslipsand from the microdissected hsr. Arrow, specific hybrid /.ation of the microdissected DNA grown to 50% confluency. Coverslips were incubated at 37Â°Cfor 1 h in over the entire hsr; arrowhead, the native location of the microdissecled material at a-MEM with 10 Â¿tMdansylatedEM (Pharmacia, Helsingborg, Sweden). Cov of the nonna I chromosome 9 homologue contained within this cell line. t/. chromosomal erslips were then washed, placed in fresh a-MEM, and returned to the incu mapping of ARC2 to SKEM by FISH. Partial human metaphase spread showing specific hybridization signals over the derivative chromosome 9 bearing the hsr in this cell line. In bator. At various times following removal of the drug, cells were imaged by a both c and d, the photographs represent computer-enhanced, merged images of fluorescein Quantix cooled charge coupled device camera coupled to a Nikon TE300 signal* (green) and DAPI-stained chromosomes (pseudo-colored in red}. 1333

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1998 American Association for Cancer Research. ABCl GENE AMPLIFICATION AND DRUG RESISTANCE dissected DNA probe to normal lymphocyte metaphase spreads also indicated that the hsr sequences were from 9q34 (data not shown). In a addition, strong hybridization was also observed over the telomeric region of an unidentified marker chromosome present in SKEM cells but not parental SKOV3. A recently identified ABC cassette transporter gene, ABC2, was previously localized to this 9q34 (14). To determine whether the ABC2 gene was amplified within the hsr of the EM-resistant cells, a human ABC2 probe was synthesized based upon the published murine sequence. FISH analysis of human metaphase chromosomes with a murine probe suggested that the human homologue of murine ABC2 was localized to 9q34 (10). To confirm that the human homologue of ABC2 is located at 9q34, human ABC2 probes were generated. Al though the sequence of the human ABC2 cDNA is not known, primers based on the published murine ABC2 sequence produced a 523-bp PCR product using human cDNA as template. From this PCR product, a second human ABC2 cDNA fragment, 1750 bp in length, was made by performing PCR using primers from within the 523-bp fragment and from a human expressed sequence tag (EST0600). The 1750-bp fragment was sequenced and found to be >90% homologous to the mouse sequence and was used in FISH analysis of chromosomes from the EM-resistant cells. Hybridization of this probe to SKEM met aphase spreads revealed multiple signals distributed throughout the hsr present in this cell line (Fig. Id), confirming that the human ABC2 1 was present and amplified with the hsr. Hybridization signals were also observed at 9q34 of the normal homologue within SKEM. In addition, a marker chromosome that showed strong hybridization to the PCR products of the microdissected hsr also hybridized to the human ABC2 probe. To estimate the number of copies of the ABC2 locus present in the hsr, additional FISH analysis of SKEM cells were performed using the ABC2 cDNA probe. Fluorescent signals were detected over the hsr in all 19 spreads examined. Hybridization to the normal chromosome 9 1 2 34 5 6 78 9 10 111213 homologue was observed in 12 of 18 metaphases containing this Fig. 2. Analysis of the ABCl and ABC2 genes and the transcripl levels for various chromosome. In a single metaphase, a normal homologue of chromo transporters, a. Southern blots for the ABCl and ABC2 genes. Lanes 1 and 3, SKOV3 DNA; Lanes 2 and 4, SKEM DNA. Lanes I ana 2 were probed with an ABC2 probe, and some 9 was not present. A total of 107 signals was observed over the Lanes 3 and 4 with an ABCl probe, b. Northern blots for ABC2 and ABCl. Lanes 1 and 19 hsrs scored. On average, 5.6 signals were distributed over each hsr 3, SKOV3; Lanes 2 and 4, SKEM. Lanes I and 2, ABC2; Lanes 3 and 4. ABCl. Increased (median, 6; range, 2-12). The mean number of signals on the normal expression (5-10-fold) of ABC2 but not ABCl message was observed in the EM-resistant cells, c, analysis of the message levels of various ABC cassette transporter genes by chromosome 9 was 0.9 (median, 1; range, 0-2). On this basis, it is reverse transcription-PCR. Lane 1, markers; Lanes 2 and 3, ABC2; Lanes 4 and 5, a estimated that the hsr contains approximately six copies of the ABC2 second set of ABC2 primers; Lanes 6 and 7. ABC C; Lanes 8 and 9. MRP; Lanes 10 and //. cMOAT: Lanes 12 and 13. 18S control. Lanes 2, 4. 6. 8. 10. and 12 represent SKOV3, gene. Additional copies of ABC2 were also present in the unidentified whereas Lanes 3, 5. 7, 9, II. and 13 represent SKEM cells. The message for ABC2 was marker chromosome, but this was not quantitated because the chro the only one increased in the SKEM EM-resistant cell line, pgp message levels were at mosome could not be reliably scored in all cells. undetectabie levels in both cell lines. Amplification of the ABC2 gene was confirmed by Southern anal ysis of DNA isolated from parental SKOV3 cells and SKEM cells (Fig. 2a, Lanes I and 2). Quantitation of the ABC2 signals indicated in these cells. The RT-PCR results for ABC2 confirm that the message ~8-fold amplification of the ABC2 gene in resistant cells, a value in for this transporter is overexpressed in the resistant cells (Fig. 2c, approximate agreement with the FISH data. The ABC! gene encodes Lanes 2-5). The mRNA levels were detectable and unchanged for a protein closely related to ABC2 and is located on chromosome ABC-C, MRP, and cMOAT (Fig. 2c, Lanes 6-11). Analysis of pgp 9q22-31. ABCÃŒwas not amplified in the SKEM cells, as shown in mRNA levels by RT-PCR revealed undetectabie levels in both cell Fig. IM (Lanes 3 and 4). These results indicate that the amplicon of the lines. The lack of pgp mRNA expression corresponds with the cyto- hsr is derived from the 9q34 region and contains the ABC2 gene but genetic observation of the apparent loss of both copies of the long arm not ABC1. of 7q, where the MDR1 gene is localized, in SKOV3 cells (data not ABC2 Transcript Is Overexpressed in SKEM Cells. To deter shown). mine if ABC2 transcript is overexpressed in SKEM cells, RNA blot SKEM Cells Display Increased Efflux of EM. The observation analysis was performed using a 523-bp human ABC2 probe. As shown that ABC2 is overexpressed in SKEM raised the possibility that it in Fig. 1b (Lanes 1 and 2), ABC2 transcript levels in resistant cells are might contribute to EM resistance by functioning as an efflux pump increased ~5-fold compared with parental cells. Northern analysis of for EM. To assess this possibility, the cellular kinetics of EM were the levels of ABCl mRNA was also performed and showed no characterized using dansyl-EM, a fluorescent analogue of EM ( 18). differences in the two cell lines (Fig. 2b, Lanes 3 and 4). Equivalent Both wild-type and resistant cells were loaded with dansyl-EM, ÃŸ-actinmRNA levels indicated equivalent RNA loading (data not washed, and imaged at various times afterward. In both cell lines, shown). RT-PCR analysis was performed to assess the possibility that dansyl-EM was localized primarily in the cytoplasm, with little flu other ABC transporters may contribute to the EM-resistant phenotype orescence observed in the nucleus. The amount of dansyl-EM in the 1334

SKOV3 SKEM

Omin.

30 min.

60 min.

Omin. 30 min. 60 min. 120 min.

SKOV3

SKEM

Fig. 3. a, dansyl-EM fluorescence in SKOV3 and SKEM cells. Cells were incubated in media with 25 /UMdansyl-EM for 2 h. washed, and inured at various lime pomis The SKEM cells efflux dansyl-EM at an increased rate compared with the sensitive SKOV3 cells. Dansyl-EM appears mainly in the otopluMN. with little observed in (In1nucleus, b. daunomycin fluorescence in SKOV3 and SKEM cells. No differences in the rate of daunomycin efflux were observed. Over the lime OHII^ of the experiment, daunomycin became localized to the perinuclear region. Little efflux occurred in either of the cell lines over the 2-h period.

1335

crodissection-FISH to identify the chromosomal location of genes ^ 30n contributing to drug resistance. The microdissection of an hsr acquired during the development of resistance to EM combined with the mapping of the native location of the hsr sequence provided an entry point for further molecular analyses using a positional candidate gene approach. A similar strategy has been used by others to identify amplified putative oncogenes associated with tumorigenesis (18). In contradistinction to the high levels of resistance associated with the pgp-mediated MDR phenotype, our previous experience suggests that EM resistance ratios of 6-8-fold are maximally attained (5, 9). Although EM photoaffinity analogues will bind pgp, this protein does not provide protection against the cytotoxic effects of the drug. As a consequence, EM resistance is distinct from the MDR phenotype, and pgp-overexpressing cells are sensitive to EM (9). The relatively low CON +AS SKOV3 resistance ratio in the SKEM cells is quantitatively consistent with the levels of ABC2 overexpression. It is also feasible that the affinity of SKEM EM for ABC2 is low and that this is also a contributory factor to the low-level resistance. Fig. 4. The effecls of various ABC2 antisense phosphorothioale oligonucleotides on the cytotoxicity of EM in SKEM cells. The IC50s for EM in SKEM cells treated with the sense EM binds to tubulin and microtubule-associated proteins (3, 4, 6), (+5) and antisense (+.45) oligonucleotides and for untreated control cells. The SKOV3 causes aberrant spindle formation in dividing cells (2), and alters the parental cell line was included for comparison. Inset, Northern analysis of ABC2 mRNA kinetics of tubule cycling (19). Each, or all, of these contribute to the levels after 48 h of antisense treatment. CON, control. antimitotic cytotoxic properties of the drug. It has been shown that EM has different binding affinities for individual ÃŸ-tubulinisotypes (6). Thus, DU 145 cells that overexpress ÃŸmtubulin express resistance sensitive and resistant cells after loading was similar (Fig. 3a, O min). to EM and some other antimicrotubule agents (5). In addition, we However, after 30-60 min, the concentration of retained drug was have found that the SKEM cells overexpress the ÃŸ,vtubulin isotype significantly lower in the resistant cells. To determine whether en compared to wild-type, and this may also be a contributory factor in hanced efflux is specific for EM, the cells were also imaged in the the overall resistant phenotype. In light of the important selective presence of daunomycin, a natural product drug with intrinsic fluo advantage of survival in response to toxic drug stress, the presence of rescence. There was little export of daunomycin from either cell type multiple, complementary resistance mechanisms is not surprising. In over a 2-h period (Fig. 3h), consistent with the overexpression in the context of the present investigation, the relative importance of SKEM of a transporter specific for EM. altered tubulin and transporter expression is difficult to assess accu ABC2 Antisense Treatment Sensitizes the SKEM Cells to the rately. However, the ABC2 antisense data would support in quantita Cytotoxic Effects of EM. Further evidence to support a role for tive terms the more critical involvement of this transporter in deter ABC2 in EM resistance was obtained by using antisense phosphoro- mining resistance in SKEM cells. thioate oligonucleotides directed toward the ABC2 message. The Little is known about the normal function of the ABC2 transporter. effects of ABC2 antisense and control sense oligonucleotides on EM It is expressed in several murine tissues and in a wide variety of tumor cytotoxicity in SKEM were tested. Cells were treated with Superfect- cell lines at detectable levels (10). The fact that ABC2 is expressed at oligonucleotide complexes, washed, and then incubated with media high levels in the mouse brain suggests that, like pgp, it may play a containing serum and oligonucleotide for 48 h prior to the cytotoxicity role in the maintenance of the blood-brain barrier. ABC2 mRNA is assay. Treatment with an antisense construct decreased the ABC2 expressed in the murine uterus and is found at higher levels in the mRNA to a level similar to that observed in parental SKOV3 cells pregnant uterus, tissues that are highly responsive to estrogens. (Fig. 4, inset). The SKEM cells are ~5-fold resistant to EM, and this Although EM has an antimicrotubule mechanism of action, struc resistance is attenuated to ~3-fold after the antisense treatment (Fig. 4). turally the drug maintains the crystal structure of estradiol (20) and Conversely, sense oligonucleotides had little effect on the EM sensi has estrogenic properties when administered in vivo (21). EM is not a tivity of SKEM. These results support the idea that ABC2 mRNA natural product, and it is unlikely that ABC2 evolved in any natural levels influence the EM-resistant phenotype in the SKEM cells. setting to deal specifically with the toxic consequences of the drug. Discussion We are presently investigating the possible role of ABC2 in hormone transport. Because glucuronides of various steroids have been shown Recently, the murine ABC2 was isolated and sequenced. However, to be substrates for MRP (22) and recent studies have shown that the no specific cellular function has been ascribed to this new member of yeast ABC transporters PDR5 and SNQ2 can transport estradiol (23), the ABC transporter family (10). Here we present data showing that it remains feasible that ABC2 may be a mammalian counterpart, EM-resistant cells amplify and overexpress ABC2 at a level commen evolving as an endogenous transporter of steroid hormones and their surate with the degree of drug resistance. Of relevance, SKEM cells metabolites. passaged in the absence of drug revert to wild-type sensitivity, and Acknowledgments concomitant with this reversal is the disappearance of the hsr, leading to wild-type levels of ABC2 expression (data not shown). Moreover, We gratefully acknowledge Drs. Takahiro Taguchi and Ze Min Liu for antisense-mediated suppression of ABC2 achieves a significant sen- assistance with chromosome dissection and karyotype analysis. sitization of the resistant cells to subsequent EM treatment. Taken References together, these convergent experimental approaches indicate that ABC2 functions as a transporter of the estradiol-mustard conjugate, 1. Hudes. G. R.. Greenberg, R., Krigel, R. L., Fox, S., Scher, R., Litwin, S., Watts, P., Speicher. L.. Tew. K., and Comis, R. Phase II study of estramustine and vinblastine, acting in a manner conducive to enhancement of cellular protection. two microtubule inhibitors, in hormone-refractory prostate cancer. J. Clin. Oncol., 10: These studies demonstrate the value of combined chromosome mi- 1754-1761, 1992. 1336

2. Sheridan, V. R., Speicher, L. A., and Tew, K. D. The effects of estramustine on 14. Bohlander, S. K., Espinosa, R., Ill, Femald. A. A., Rowley, J. D., LeBeau, M. M., and mitotic progression in DU 145 human prostatic carcinoma cells. Eur. J. Cell Biol., 54: Diaz, M. O. Sequence-independent amplification and labeling of yeast artificial 268-276, 1991. chromosomes for fluorescence in situ hybridization. Cytogenet. Cell Genet., 65: 3. Speicher, L. A., Laing, N., Barone, L. R., Robbins, J. D., Seamon, K. B., and Tew, 108-110, 1994. K. D. Interaction of an estramustine photoaffmity analogue with cytoskeletal proteins 15. Bell, D. W., Taguchi, T., Jenkins, N. A., Gilbert, D. J., Copeland, N. G., Gilks, C. B., in prostate carcinoma cells. Mol. Pharmacol., 46: 866-872, 1994. and Zweidler-McKay, P. Chromosomal localization, in man and rodents, of a gene 4. Dahllof, B., Billstrom, A., Cabrai, F., and Hartley-Asp, B. Estramustine depolymer- (GF11) encoding a novel zinc finger protein reveals a new synlenic region between izes microtubules by binding to tubulin. Cancer Res., 53: 4573-4581, 1993. mouse and man. Cytogenet. Cell Genet., 70: 253-267, 1995. 5. Ranganathan, S., Dexter, D. W., Benetatos, C. A., Chapman, A. E., Tew, K. D., and 16. Clifford, S. C., Thomas, D. J., Neal, D. E., and Lunec. J. Increased mdrl gene Hudes, G. R. Increase of /3nl- and ÃŸ]Va-tubulin isotypes in human prostrate carci transcripts levels in high-grade carcinoma of the bladder determined by quantitative noma cells as a result of estramustine resistance. Cancer Res., 56: 2584-2589, 1996. PCR-based assay. Br. J. Cancer, 69: 680-686, 1994. 6. Laing, N., Dahllof, B., Hartley-Asp, B.. Ranganathan, S., and Tew, K. D. Interaction 17. Stearns, M. E.. Jenkins. D. P., and Tew, K. D. Dansylated estramustine, a fluorescent of estramustine with tubulin isotypes. Biochemistry, 36: 871-878, 1997. probe for studies of estramustine uptake and identification of intracellular targets. 7. Yang, C. P., Shen, H. J., and Horwitz, S. B. Modulation of the function of P- Proc. Nati. Acad. Sci. USA, 82: 8483-8487, 1985. glycoprotein by estramustine. J. Nati. Cancer Inst., 86: 723-725, 1994. 18. Guan, X. Y., Xu, J.. Anzick, S. L., Zhang, H.. Trent, J. M.. and Meltzer, P. S. Hybrid 8. Smith, C. D., Zilfou, J. T., Zhang, X., Hudes, G. R., and Tew, K. D. Modulation of P-glycoprotein activity by estramustine is limited by binding to plasma proteins. selection of transcribed sequences from microdissected DNA: isolation of genes within amplified region at 20qll-ql32 in breast cancer. Cancer Res., 56: 3446- Cancer (Phila.), 75: 2597-2604, 1995. 3450, 1996. 9. Speicher, L. A., Sheridan. V. R., Godwin, A. K., and Tew, K. D. Resistance to the 19. Panda, D., Miller, H. P., Islam, K., and Wilson, L. Stabilization of microtubule antimitotic drug estramustine is distinct from the multidrug resistant phenotype. Br. J. Cancer, 64: 267-273, 1991. dynamics by estramustine by binding lo a novel site in tubulin: a possible mechanistic basis for its antitumor action. Proc. Nati. Acad. Sci. USA, 94: 10560-10564, 1997. 10. Luciani, M. F., Denizot, F., Savary, S., Mattei, M. G., and Chinimi, G. Cloning of two novel ABC transporters mapping on human chromosome 9. Genomics, 21: 150-159, 20. Tew, K. D., Glusker, J. P., Hartley-Asp, B., Hudes, G., and Speicher, L. A. Preclinical 1994. and clinical perspectives on the use of estramustine as an antimitotic drug. Pharmacol. 11. Bell, D. W., Yeung, R. S., Bohlander. S. K., Cheng, J. Q., Jin, F., and Testa, J. R. A Ther., 56: 323-339, 1992. microdissection library of the rat renal carcinoma gene region. Cytogenet. Cell 21. Gunnarsson, P. O., and Forshell, G. P. Clinical pharmacokinetics of estramustine Genet., 70: 92-94, 1995. phosphate. Urology, 23: 22-27, 1984. 12. Bohlander, S. K., Espinosa, R., Ill, LeBeau, M. M., Rowley, J. D., and Diaz, M. O. 22. Jedlitschky, G., Leier, I., Bucholz, U., Barnouin, K., Kurz, G., and Keppler, D. A method for the rapid sequence-independent amplification of microdissected chro Transport of glutathione, glucuronate, and sulfate conjugates by the MRP gene- mosomal material. Genomics, 13: 1322-1324, 1992. encoded conjugate export pump. Cancer Res.. 56: 988-994, 1996. 13. Fan, Y-S., Davis, L. M., and Shows. T. B. Mapping small DNA sequences by 23. Mahe, Y., Lemoine, Y.. and Kuchler, K. The ATP binding cassette transporters Pdr5 fluorescence in situ hybridization directly on banded metaphase chromosomes. Proc. and Snq2 of Saccharomvces cerevisiae can mediate transport of steroids in vivo. Nati. Acad. Sci. USA, 87: 6223-6227. 1997. ]. Biol. Chem.. 271: 25167-25172, 1996.

1337

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1998 American Association for Cancer Research. Amplification of the ATP-Binding Cassette 2 Transporter Gene Is Functionally Linked with Enhanced Efflux of Estramustine in Ovarian Carcinoma Cells

Naomi M. Laing, Martin G. Belinsky, Gary D. Kruh, et al.

Cancer Res 1998;58:1332-1337.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/58/7/1332

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/58/7/1332. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.