P-Glycoprotein Expressing Multidrug Resistance and ...

Daunorubicin-resistant Chinese Hamster Ovary Cells Expressing Multidrug Resistance and a Cell-Surface P-Glycoprotein Norbert Kartner, Michael Shales, John R. Riordan, et al. Cancer Res 1983;43:4413-4419. Published online September 1, 1983.

Updated Version

Citing Articles

E-mail alerts Reprints and Subscriptions Permissions

Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/43/9/4413

This article has been cited by 7 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/43/9/4413#related-urls

Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected].

Downloaded from cancerres.aacrjournals.org on July 14, 2011 Copyright © 1983 American Association for Cancer Research

[CANCER RESEARCH 43, 4413-4419,

September 1983]

Daunorubicin-resistant Chinese Hamster Ovary Cells Expressing Multidrug Resistance and a Cell-Surface P-Glycoprotein1 Norbert Kartner,2 Michael Shales,3 John R. Riordan, and Victor Ling Ontario Cancer Institute, Princess Margaret Hospital and Department of Medical Biophysics, University of Toronto, Toronto M4X 1K9 [N. K., M. S., V. L.J, and Research Institute, The Hospital for Sick Children and Departments of Biochemistry and Clinical Biochemistry, University of Toronto, Toronto M5G 1X8 [J. R. R.Â¡,Ontario, Canada

ABSTRACT Independent lines of Chinese hamster ovary cells resistant to the antineoplastic drug, daunorubicin, were obtained by clonal isolation in increasing drug concentrations. A single daunorubicin-resistant phenotype typified by reduced cellular drug accu mulation was observed. These mutants displayed a complex phenotype of resistance to a variety of unrelated drugs. Such properties are similar to those of membrane-altered colchicineresistant lines (V. Ling and L. H. Thompson, J. Cell. Physiol., 83: 103-116,1974.). Analysis of the plasma membrane components of the daunorubicin-resistant clones by gel electrophoresis re vealed a prominent cell surface glycoprotein with a molecular weight of about 170,000. This component was immunologically cross-reactive with the cell surface P-glycoprotein of about the same molecular weight, previously identified in colchicine-resistant cells. Thus, it appears that the mechanism of resistance characterized by P-glycoprotein expression could be the basis of many drug-resistant phenotypes.

to the clinical situation, in that it would render tumor cells uniquely resistant to combination chemotherapy. In an effort to elucidate further the mechanism of drug resis tance, we describe here the isolation and characterization of CHO cells resistant to the clinically important, antineoplastic drug, daunorubicin. We were especially interested in determining whether CHO cells selected for resistance to daunorubicin could express a pleiotropic cross-resistance to unrelated drugs. We also asked whether, by analogy with colchicine-resistant CHO cells, such multidrug-resistant isolates would elaborate a cell surface glycoprotein alteration. MATERIALS AND METHODS

INTRODUCTION In the treatment of neoplastic disease, especially disseminated cancers, chemotherapy often fails after initial success and ap parent remission. Frequently, the recurring tumor is unresponsive to a number of apparently unrelated drugs to which the patient had not been exposed previously. Although the basis of this nonresponse is not well understood, one possible explanation is that drug-resistant subpopulations of tumor cells arise sponta neously, survive, and proliferate during the course of treatment (14). Model systems for the study of drug resistance have been developed both in vivo and in vitro (for reviews, see Refs. 5 and 14). It is significant that pleiotropic cross-resistance to numerous drugs, including many which are structurally dissimilar and have distinct cytotoxic targets, is frequently observed. This pleiotropic phenotype has been studied in detail in CHO4 cells in the colchicine-resistant (CHn) mutants originally isolated by Ling and Thompson (16). These cells have been shown to be crossresistant to a variety of antineoplastic drugs as a consequence of an alteration of the plasma membrane, which results in re1 Research was supported by the National Cancer Institute of Canada and by the Medical Research Council of Canada. 2 To whom requests for reprints should be addressed. 3 Present address: Banting and Best Department of Medical Research, University of Toronto, Toronto, Canada M5G 1L6. 4 The abbreviations used are: CHO, Chinese hamster ovary; PAGE, pdyacrylamide gel electrophoresis; SDS, sodium dodecyl sulfate; PCS, fetal calf serum; Â«MEM, Â«-modified Eagle's minimal essential medium; 1CÂ»,drug concentration at which cell growth in vitro is inhibited by 50%; PBS, phosphate-buffered saline; BSA, bovine serum albumin. Received December 16, 1982; accepted June 10, 1983.

SEPTEMBER

duced permeability to the drugs involved (2, 12, 16). A M, 170,000 plasma membrane glycoprotein (the P-glycoprotein) is expressed concomitantly with the colchicine-resistant phenotype (10, 11, 20), and it has been postulated that the P-glycoprotein modulates the cell surface membrane, by some mechanism not yet understood, to limit the permeation of drugs into the cell (12). This multidrug-resistant phenotype may be particularly relevant

Materials. Daunorubicin (Cerubidine) and rubidazone were obtained from Poulenc, Ltd. (Montreal, Canada); colchicine, puromycin, and eme tine were from Sigma Chemical Co. (St. Louis, Mo.); and vinblastine sulfate (Velbe) was from Eli Lilly (Columbus, Ohio). Polyethylene glycol 1000 was obtained from J. T. Baker Chemical Co. (Glen Ellyn, III.); ultrapure sucrose, from Schwarz/Mann (Spring Valley, N. Y.); fluorescamine (Fluram), from Roche Diagnostics (Hoffmann-La Roche Ltd., Vaudreuil, Quebec, Canada); and Staphylococcus aureus protein A, from Pharmacia (Don/al, Quebec, Canada). Reagents for PAGE were obtained from Bio-Rad Laboratories Ltd. (Mississauga, Ontario, Canada), except for SDS, specially purified for biochemical work, which was from Schwarz/Mann. Na125l(15 mCi//ig) was from Amersham Corp. (Oakville, Ontario, Canada), and [3H]colchicine (12.5 mCi/mg) was from New England Nuclear (Dorval, (pore size, 0.45 //m) was (Keene, N. H.). PCS was obtained from was obtained from Grand

Quebec, Canada). Nitrocellulose filter paper obtained from Schleicher and Schuell, Ltd. Flow Laboratories (Maclean, Va.). Calf serum Island Biological Co. (Grand Island, N. Y.); Â«-

MEM (22), containing asparagine monohydrate at 50 mg/liter, was ob tained as a commercially prepared powder (Grand Island Biological Co.) and was supplemented with streptomycin sulfate (100 mg/liter; Pfizer Canada, Inc., Kirkland, New Brunswick, Canada), potassium penicillin G (100 mg/liter; Ayerst Laboratories, Ltd., Montreal, Canada), and sodium bicarbonate (2.2 g/liter); Â«-special medium was formulated as Â«-MEM, but was lacking glycine, nucleosides, and deoxynucleosides (18). PBS consisted of 137 mM NaCI, 2.7 mw KCI, 8.1 HIM Na2HPO4, 1.3 rriM KH2PO4,0.9 mM CaCI2, and 0.3 mw MgCI2. Tris/0.9% NaCI consisted of 10 mw Tris-HCI, pH 7.4; 0.9% NaCI; and 0.01% NaN3. BSA/0.9% NaCI buffer consisted of 3% bovine serum albumin (Fraction V; Sigma) in Tris/0.9% NaCI buffer. Cells and Culture Conditions. The daunorubicin-resistant CHO cell lines DNR"5 and DNRR51 were selected from the drug-sensitive parent

4413

1983


N. Kartner et al. line AUXB1, by N. T. Bech-Hansen, in our laboratory. AUXB1 is a glycine,

PBS plus 15 rriM glucose at 8 x 106 cells/ml. For each time point, 0.5 ml

adenosine, thymidine, and proline auxotroph described by McBurney and Whitmore (18). The independently derived series of daunorubicinresistant CHO cell lines, DNRR1, DNRR15, and DNRR159, was selected

of suspension was combined with 1.0 ml of daunorubicin at 10 Â¿ig/mlin PBS plus 15 mW glucose, and incubated in a shaking water bath at 37Â°.

from the drug-sensitive

parent line AUXE29, an auxotroph which has an

adenosine and proline requirement complementary to that of AUXB1 (24). CHRC4 and CHRC5 are colchicine-resistant CHO cell lines inde pendently derived from AUXB1 as described previously by Ling and Thompson (16). All CHO cell lines were maintained in monolayer culture at 37Â° in humidified air containing 5% CO2. They were propagated in Â«-MEM supplemented with 10% FCS. Large-scale cultures for membrane prep aration were grown in liquid suspension culture in Â«-MEMsupplemented with 5% PCS and 5% calf serum. Drugs added to the media were first sterilized by filtration through a 0.45-/im pore membrane filter (Millipore Corp., Mississauga, Ontario, Canada), where necessary. Selection of Daunorubicin-resistant Clones. All daunorubicin-resistant clones were selected by plating 5 x 10s cells/100-mm Retri dish in growth medium containing the selective drug concentration. The drug concentration chosen was determined from a dose-response survival curve where variant populations were revealed (1, 24). After 8 to 10 days at 37Â°, the surviving colonies were picked by scraping and aspirating with a sterile Pasteur pipet and were subsequently cloned by plating in drug at limiting dilution. Successive clonal selections were performed at daunorubicin concentrations of 0.1, 0.5, and 1.0 ng/m\. Drug-resistant clones occurred at a frequency of approximately 10~6 at each stage. The daunorubicin resistance phenotype remained stable during growth in the absence of drug for at least 4 months. Stocks of each cell line were stored at -70Â° in Â«-MEM supplemented with 10% PCS and 10% dimethyl sulfoxide. Cell Hybridization. Intraspecific hybrids between drug-resistant and drug-sensitive lines were constructed as described previously (15). Briefly, monolayer cultures containing 106 cells of each line were treated with 50% polyethylene glycol 1000 in Â«-MEM for 1 min at 37Â°. Cells were then washed and incubated in complete medium for 24 hr before plating into selective medium, which consisted of Â«-special medium supplemented with 10% dialysed PCS. Some plates were simultaneously treated with different concentrations of daunorubicin in order to establish the degree of resistance of the total nascent hybrid population (15). Assay of Drug Resistance. Dose-response curves for presumptive daunorubicin-resistant clones were determined by plating in increasing concentrations of daunorubicin. After 10 days of incubation at 37Â°, colonies were stained with mÃ©thylÃ¨neblue and counted. For rapid assessment of resistance to a variety of drugs, suspension growth assays were performed as described previously (2). Cell densities were determined using a Model F electronic particle Coulter Counter (Coulter Electronics, Hialeah, Fla.) calibrated for CHO cells. The doubling times of parental and mutant lines were between 16 and 22 hr. ICso is defined as the concentration of drug inhibiting the growth of cells such that their rate of growth is reduced to one-half that observed under drug-free conditions, over a 48-hr period. The so-called relative resistance of a test cell line to a given drug is determined in growth assays as the ratio of ICso of the test cells to ICso of the drug-sensitive control cells [as adapted from the method of Bech-Hansen ef a/. (2)J. Assay of Drug Accumulation. To assay cellular accumulation of colchicine, cells were grown to near confluence (2 x 106 cells) on the bottoms of sterile glass scintillation vials. Cultures were washed free of medium and incubated with PBS containing 15 mwi glucose, which was replaced with |3H]colchicine (1 /Â¿M; 1 jiCi/ml) in PBS plus 15 mw glucose at different times, to obtain incubations ranging in duration from 0 to 30 min at 37Â°. The vials were then washed 3 times in ice-cold PBS. Cells were solubilized in 1 ml of 0.1 N NaOH and assayed for protein according to the method of Lowry ef al. (17). Radioactivity taken up by the cells was determined by liquid scintillation counting. To assay cellular accumulation of daunorubicin, cells were pelleted from exponentially growing suspensions, washed, and resuspended in

Cells were then pelleted and washed twice with cold PBS. Daunorubicin was extracted by sonication of the cells in 0.3 N HCI in 50% ethanol, followed by a 20-min centrifugation at 20,000 x g, and was quantitated by fluorometry on a Farrand Spectrofluorometer Mark I (Farrand Optical Co., Inc., Valhalla, N. Y.) at an excitation wavelength of 420 nm and an emission wavelength of 585 nm (10-nm bandpass). Blanks consisted of control samples to which no daunorubicin had been added. Daunorubicin concentration was interpolated from the linear portion of a standard curve. Plasma Membrane Purification. Plasma membrane was isolated from cell homogenates by isopycnic centrifugation on discontinuous sucrose gradients according to the method of Riordan and Ling (20). Membrane protein was assayed according to the method of Bohlen et al. (6) against BSA standards (Fraction V; Sigma). Electrophoretic Analysis of Membrane Protein. Plasma membranes were concentrated by centrifugation in a Beckman airfuge at 130,000 x g for 20 min, where necessary. PAGE was performed according to a slight modification of the method of Fairbanks eÃ-al. (8), as described previously (7). Protein profiles on SDS-polyacrylamide gels were replicaelectroblotted (Western-blotted) onto nitrocellulose filter paper which was then probed with antisera. The method used was essentially that of Towbin ef al. (25), as described previously (7). To improve further the specificity of P-glycoprotein detection, the antiserum was absorbed with surface proteins derived from drug-sensi tive CHO cells. These presumably compose the same complement of proteins as the drug-resistant membranes, with the notable exception of the P-glycoprotein (20). Previous attempts to prepare a specific antiserum by absorption with intact membrane vesicles, up to the equivalent of 10'Â° cells/ml, were not successful, as judged by Western blots. It is possible that the proteins bound to the nitrocellulose paper bear exposed antigenic determinants which are not normally accessible when the proteins are in a native conformation, imbedded in the lipid bilayer of the membrane. We reasoned, therefore, that proteins bound to nitrocellulose particles might be a more efficient absorption medium than intact mem branes. The details of this novel absorption method are described below. Nitrocellulose powder was prepared by cutting sheets of nitrocellulose filter paper (pore size, 0.45 ^m) into approximately 1-sq cm pieces with scissors. These were soaked with water and were further chopped to approximately 1-sq mm pieces with a razor blade. The wet mass was then ground using an unglazed porcelain mortar and pestle, and was sieved by scraping through an 80 mesh stainless steel screen. The sieved particles were elutriated in distilled water to remove fines and were stored in aliquots of 4Â°in distilled water containing 0.1% NaN3. After being washed in distilled water, these particles were coated with membrane protein as follows. About 4 mg of plasma membrane vesicles derived from drug-sensitive CHO cells were solubilized in the modified Fairbanks solubilizing buffer described previously (7). Only enough solubilizing buffer was used to achieve twice the saturating (w/w) SDS:protein ratio of 1.4:1 (at 2% SDS concentration). After cooling, solid urea was added to 9 M final concen tration and, after 10 min at room temperature, the solution was diluted to 5 ml with Tris/0.9% NaCI buffer. An aliquot of nitrocellulose powder representing 30 sq cm of nitrocellulose paper was added to the solution, and the suspension was dialysed against Western blotting electrolyte (25 mw Tris base; 192 mw glycine, pH 8.3; 20% methanol) overnight at room temperature with gentle end-over-end agitation of the dialysis bag. By this method, approximately 60% of the solubilized protein was bound to the nitrocellulose. Remaining binding sites were blocked with BSA/ 0.9% NaCI buffer at 37Â°for 1 hr. This solid-phase absorption medium was stored at 4Â°in BSA/0.9% NaCI buffer. Absorption of the antiserum was performed by adding the washed absorption medium to 5 ml of serum diluted 25-fold in BSA/0.9% NaCI buffer. This was incubated overnight at 4Â°in a 10-ml tube with end-over-

4414

CANCER


RESEARCH

VOL. 43

P-Glycoprotein in Daunorubicin-resistant end agitation.

The supernatant-absorbed

serum was collected

Cells

after

pelleting the nitrocellulose powder by centrifugation. The absorption medium is reusable after stripping the bound antibody with acidic buffer (three 20-min washes at 20Â° with 0.1 M glycine-HCI, pH 2.2; 20 rtiM magnesium acetate; and 50 mM KCI). The regeneration has been re peated at least 14 times without apparently depleting the capacity for absorption of the nitrocellulose-bound antigen.

RESULTS Isolation of Daunorubicin-resistant (DNRR) Clones. Dauno rubicin-resistant CHO cell clones were isolated in a series of single steps, without chemical mutagenesis, from the 2 drugsensitive parent lines, AUXB1 and AUXE29. The first step in the selection was accomplished in the presence of daunorubicin (0.1 Â¿Â¡g/ml). Clones growing in daunorubicin (0.5 ^g/ml) were then selected in a second step from the first-step isolates. A third selection was performed with daunorubicin at 1.0 Â¿Â¿g/ml, using the second-step isolates derived from AUXE29. A schematic representation of the selection procedures is shown in Chart 1. Daunorubicin-resistant clones were characterized subse quently for their degree of resistance to the selecting drug. Chart 2 shows survival curves of the series derived from AUXE29. Increased resistance to daunorubicin is seen with each succes sive step in the selection such that 9-, 18-, and 63-fold greater resistance is observed, relative to the drug-sensitive parent line. Similar results were obtained for the isolates derived from AUXB1 where, in the 2 successive steps, the clones were 13- and 41fold more resistant than the parent line (data not shown). An important question concerning resistance to antineoplastic drugs is whether such a phenotype could be expressed in fully diploid somatic cells or in polyploid tumor cells. The investigation of whether a drug-resistant phenotype is dominant or recessive is pertinent to this question (13). Such an investigation also serves to compare the daunorubicin resistance of clones se lected independently from AUXB1 and AUXE29. Using an ap proach described previously to examine the drug resistance of a population of nascent hybrid clones (15), we observed that somatic cell hybrids between daunorubicin-resistant lines and drug-sensitive complementary auxotrophs are resistant to dau

1.4 Chart 2. Dose-response of DNR" lines. Dose-response curves of the drugsensitive parent line AUXE29 and 3 daunorubicin-resistant lines (see Chart 1 for derivation). Curves, assays for colony-forming ability at increasing daunorubicin concentrations. Points, average of 3 assays for colony-forming ability at increasing daunorubicin concentrations.

norubicin (see Chart 3). This resistance, however, was not as great as that of the drug-resistant parent. Such codominant expression of drug resistance in somatic cell hybrids has been observed previously in colchicine-resistant CHO cells (15). Fur ther evidence of the hybrid nature of the clones shown in Chart 3 was provided by the analysis of metaphase chromosome spreads. Typically, the hybrid metaphase chromosome number ranged from 35 to 40 (data not shown). The metaphase chro mosome spreads of the resistant cell lines were similar to their parent lines, with a pseudodiploid modal chromosome number of 21. Thus, no ploidy change was associated with the dauno rubicin resistance phenotype. In order to characterize further the mechanism of resistance

AUXE29-5:!~DNRRl-2:5-DNRRl5-!:5-DNRRl59 AUXBI2:1-DNRR52:5-DNRR5I Chart 1. Derivation of DNR" lines. Lineage of the independently derived dau norubicin-resistant lines. The drug concentration used in each of the stepwise selections is given in tig/ml (numbers above arrows).

SEPTEMBER

0.01 Daunorubicin

(/Â¿g/ml)

Chart 3. DNRR expression in somatic hybrids. Dose-response curves of nascent intraspecific hybrids (DNRB51 x AUXE29 and DNRR159 x AUXB1 ) of daunorubicinresistant and -sensitive lines (see "Materials and Methods" for hybridization and dose-response assay conditions). The dose-response curves of the daunorubicinresistant line DNRR159 and a drug-sensitive control hybrid, AUXE29 x AUXB1, are shown for comparison. Points, average of 3 experiments.

4415

1983


N. Kartner et al. expressed Â¡nthe various daunorubicin-resistant isolates, we examined the rates of drug accumulation of these cells. Chart 4a shows that the rate at which daunorubicin is accumulated by the drug-resistant line DNAR51 is approximately 3-fold less than that observed in the drug-sensitive parent line AUXB1. Further more, as shown in Chart 4b, the rate of accumulation of colchicine is also significantly lower in these cells than in the parent cells. The colchicine-resistant line CHRC4, also derived from the sensitive parent line AUXB1, is shown for comparison in Chart 4b. Similar results were obtained for both daunorubicin and colchicine accumulation in DNRR159, relative to its drug-sensitive parent line AUXE29. Since daunorubicin and colchicine distinct intracellular cytotoxic targets, the reduced cellular mulation of both drugs most likely results from a plasma brane-associated mechanism that limits the net entry of

have accu mem these

diverse compounds (1, 16). Further evidence for such a mechanism was sought by deter mining the extent of pleiotropic cross-resistance of the dauno rubicin-resistant isolates to a variety of antineoplastic drugs. Pleiotropic Cross-Resistance Associated with DNRR. The results of growth assays for the series of isolates derived from AUXE29, in 5 different drugs, are summarized in Table 1. It is evident from these data that all of these daunorubicin-resistant clones are cross-resistant not only to other anthracyclines, such as Adriamycin and rubidazone, but also to unrelated drugs such as the Vinca alkaloid, vinblastine, and the antibiotic, puromycin. Indeed, the cross-resistance of these cells to vinblastine equals their primary resistance to the selective drug, daunorubicin. Furthermore, the selection for increased resistance to daunorub icin results in the concomitant increase in cross-resistance to all of the drugs listed in Table 1. This observation is strong evidence for a pleiotropic relationship among these various resistance

Table 1 Cross-resistance

of daunorubicin-resistant

lines

Relative resistance of cell lines was determined from 48-hr growth assays (see "Materials and Methods"). The relative resistance values represent the increase in ICso observed for a given drug-resistant line in a given drug, relative to the ICÂ«, observed for the control, drug-sensitive parent line AUXE29. ICM values were determined in triplicate experiments and generally vary less than 10% from the given mean value. Relative resistance of cell lines DrugDaunorubicin Adriamycin Rubidazone Vinblastine PuromycinAUXE2911

1 1 1DNR"14

2 2 3 2DNRR1512

15 26 11 30 11 3924 16DNR"15937

phenotypes, presumably as a result of a single mechanism for multidrug resistance. Having established the pleiotropy of cross-resistance in the daunorubicin-resistant isolates, and their reduced accumulation of both daunorubicin and colchicine, it seems very likely that these cells are resistant as a result of a plasma membrane alteration. It is of interest to compare the cross-resistance patterns of the 2 independently derived isolates that are most resistant to daunorubicin with those of the 2 independently derived colchi cine-resistant lines described previously (2, 16). Table 2 sum marizes these comparative data. It is seen that the cross-resis tance patterns of the daunorubicin-resistant clones are different from those of the colchicine-resistant clones. In particular, the daunorubicin-resistant clones are about 2-fold more resistant to daunorubicin than to colchicine, whereas the colchicine-resistant clones are at least 2-fold more resistant to colchicine than to daunorubicin. For clones selected with the same drug, their cross-resistance patterns are similar, and only minor differences are observed. Thus, it appears that the drug of selection deter mines to a large extent the specificity of the membrane-associ ated pleiotropic phenotype expressed by the resistant cells. Comparison of DNRHand CHRPlasma Membranes. Although the specificity of cross-resistance suggests that the daunorubicin and colchicine resistance phenotypes are different, the associ ated membrane alterations must be generally similar, in that both cell types exclude diverse drugs that diffuse passively across the plasma membrane. For this reason, it is of interest to deter mine whether the membrane alteration associated with dauno rubicin resistance is of the same nature as that characterized in the colchicine-resistant clones. The only molecular alteration known to be consistently associated with the colchicine-resistant cells is the expression of the P-glycoprotein. It has been pro

20 0 Time (min)

5

Chart 4. Reduced drug accumulation in DNR" lines, a, accumulation of dauno rubicin in intact, viable cells. The drug-resistant line DNR"51 and its drug-sensitive parent line AUXB1 are compared (see "Materials and Methods* for assay condi tions) 6. Accumulation of colchicine in intact, viable cells. The daunorubicinresistant line DNRB51 is compared with the colchicine-resistant line CHRC4 and the drug-sensitive parent line AUXB1 (see "Materials and Methods" for assay condi tions). Total protein per assay is approximately experiments.

0.4 mg. Points, average of 3

posed previously that this cell surface glycoprotein may be the effector molecule responsible for altered membrane permeability (12). To test whether such a protein is expressed concomitantly with the daunorubicin resistance phenotype, we examined iso lated plasma membrane fractions from the various drug-resistant isolates and their drug-sensitive parent lines using SDS-PAGE. The results of such an analysis are shown in Fig. 1. It is clear from Fig. 1 that an intensely staining protein band of about the same molecular weight as the P-glycoprotein of colchicine-re sistant CHO cells is expressed in the daunorubicin-resistant cells. We were able to characterize further the anomalous, dauno rubicin resistance-associated band by probing replica electro-

4416

CANCER


RESEARCH

VOL. 43

P-Glycoprotein in Daunorubicin-resistant

Cross-resistance

Table 2 of daunorubicin-resistant lines differs with colchicine-resistant

lines ICÂ» values were determined from 48-hr growth assays (see "Materials and Methods"). The relative resistance values given for DNRR51, CHRC4, and CHBC5 represent the fold increase in ICÂ»for a given drug, relative to the \CX observed for the control, drug-sensitive parent line AUXB1. The drug-sensitive parent line AUXE29 was used as a control to compute the relative resistance values for DNR"159. ICÂ»values were determined as averages of 3 experiments. The greatest variation from the computed relative resistance values doe? not exceed 15%. Relative resistance of cell lines DrugDaunorubicm Colchicine Vinblastine Puromycin EmetineDNRR5141

2139 25 22 24CHRC43288 38 11DNRR15937

160 29 58 105 29 15CHRC576 29

blots (Western blots) of the SDS-polyacrylamide gels with a rabbit antiserum prepared against colchicine-resistant CHO cells. It is seen in Fig. 2A that this antiserum clearly detects the Pglycoprotein of the latter cells, among many other cell surface antigens. Cross-absorption of the antiserum with plasma mem brane proteins of drug-sensitive cells, as described under "Ma terials and Methods," has greatly improved the specificity of the serum for detection of the P-glycoprotein. The results of such an absorption are seen in Fig. 28. It is apparent that the anom alous protein band found in the daunorubicin-resistant cells not only is of similar apparent molecular weight as the P-glycoprotein of colchicine-resistant cells, but is also cross-reactive with an antiserum of high specificity for the P-glycoprotein. Metabolic labeling studies with radiolabeled glucosamine and amino acids have indicated that the daunorubicin resistance-associated pro tein is a glycoprotein synthesized by the mutant cells (data not shown). DISCUSSION In this study, we describe the step-wise, clonal isolation of daunorubicin-resistant CHO cell lines. The resistant mutants can be obtained readily at a frequency of about 1 in 106 cells plated, in the absence of prior mutagenesis. This frequency is typical for the selection of genetic mutants resistant to specific drugs in a variety of mammalian systems (14). We have found that, in all the clones isolated in the 2 independent series of selections, a multidrug-resistant phenotype was expressed in addition to the primary resistance to daunorubicin. The expression of this phe notype, together with the observation of reduced drug accumu lation and the expression of a new high molecular weight cell surface glycoprotein in the daunorubicin-resistant lines, provides strong evidence that these isolates are analogous to the mem brane-altered colchicine-resistant mutants characterized previ ously (2,10,11,16, 20, 21). Thus, it appears that this mutation is relatively common, even in isolates derived by using selective agents unrelated to colchicine. This conclusion has broad signif icance in the context that multidrug-resistant phenotypes are observed in clones selected for resistance to daunorubicin, actinomycin D, vincristine, vinblastine, and Adriamycin in a variety of cell systems (for reviews, see Refs. 1, 5,13, and 14). In many of these cases, cell surface glycoprotein changes have been reported (4, 5, 7, 9,10,19). The possibility of a general correlation between multidrug

SEPTEMBER

Cells

resistance and P-glycoprotein expression raises the question of how P-glycoprotein is functionally involved. Any model where Pglycoprotein is presumed to participate directly in the mainte nance of reduced drug accumulation must take into account the differences between the cross-resistance patterns of daunorub icin and colchicine-resistant CHO cells. In this context, we spec ulate that the variation in specificity of resistance to a variety of drugs could result from at least 2 different mechanisms. One is the alteration of the membrane microenvironment, in which the P-glycoprotein is presumed to function. This could involve intrin sic variation in, for example, phospholipid, glycolipid, or ancillary membrane (or cytoplasmic) protein components. Another possi bility is that the structure of the P-glycoprotein itself is variable, either in its carbohydrate or in its peptide moiety. Microheterogeneity in a cell surface glycoprotein is usually attributed to variation in its carbohydrate moiety rather than in its peptide sequence. It is unlikely, however, that cell surface carbohydrate is functionally involved in drug resistance. We have found that colchicine-resistant cells, which are selected addition ally for phytohemagglutinin resistance, are not altered in their degree of resistance to colchicine, or in their pattern of crossresistance to other drugs. The precise pattern of multidrug resistance remains unaltered, despite a substantial reduction in the amount of cell surface carbohydrate in general and in the amount of carbohydrate associated with the P-glycoprotein spe cifically.5 Likewise, Beck and Cirtain (3) have shown, in human vinblastine-resistant cells, that neither removal of cell surface carbohydrate by Pronase digestion nor inhibition of glycosylation with tunicamycin has any effect on the expression of the resis tance phenotype. We conclude, therefore, that the carbohydrate moiety does not play a dominant role in determining the multidrug-resistant phenotype. On the other hand, the diffuse nature of the P-glycoprotein band in SDS-PAGE indicates a molecule that may be heteroge neous in molecular weight. Likewise, we have found that there is a slight, although reproducible, difference in molecular weight of the P-glycoprotein detected in colchicine- and daunorubicinresistant cells, the daunorubicin resistance-associated protein being lower by approximately M, 5000. This is consistent with the notion that variation in the structure of the P-glycoprotein might account for variation in the specificity of the barrier to accumulation of various drugs. Such variability could be ac counted for if the P-glycoprotein were to present a class of similar proteins with different specificities, the appropriate member(s) of the putative gene family being expressed or amplified under a given selective pressure. The possibility of minor differ ences in the peptide sequence of P-glycoprotein in daunorubicinand colchicine-resistant lines is of interest in this regard and will require further study. A major conclusion of our present work is that the expression of homologous P-glycoproteins may be a factor that is common to various multidrug-resistance phenotypes, regardless of the precise nature of their cross-resistance patterns. This fact may be exploitable for the detection and control of therapy-resistant neoplastic cells in patients. It will be important to determine whether P-glycoprotein-like molecules are expressed in associ ation with the drug resistance in spontaneous human tumor cells of patients who are unresponsive to combination chemotherapy. 5 T. Sudo, V. Ling, and L. Siminovich, unpublished observations.

4417

1983


N. Kartner et al. ACKNOWLEDGMENTS The authors wish to acknowledge the excellent technical assistance of N. Aton, S. Fahim, and M. Naik.

12. 13.

REFERENCES 14. 1. Baker, R. M., and Ling, V. Membrane mutants of mammalian cells in culture. Methods Membrane Biol., 9. 337-384, 1978. 2. Bech-Hansen, N. T., Till, J. E., and Ling, V. Pleiotropic phenotype of colchicineresistant CHO cells: cross-resistance and collateral sensitivity. J. Cell. Physiol., 88. 23-31. 1976. 3. Beck, W. T., and Cirtain, M. C Continued expression of Vinca alkaloid resistance by CCRF-CEM cells after treatment with tunicamycin or Pronase. Cancer Res., 42: 184-189, 1982 4. Beck, W. T., Mueller, T. J., and Tanzer, L. R. Altered surface membrane glycoproteins in Vinca alkaloid-resistant human leukemic lymphoblasts. Cancer Res.. 39. 2070-2076,1979. 5. Siedler, J. L., and Peterson, R. H. F. Altered plasma membrane glycoconjugates of Chinese hamster cells with acquired resistance to actinomycin D, daunorubicin. and vincristine. In: A. C. Salterelli, J. S. Lazo, and J. R. Berlino (eds), Molecular Action and Targets for Cancer Chemotherapeutic Agents, pp. 453-482. New York: Academic Press, Inc., 1981. 6. Bohlen, P., Stein, S., Dairman. W., and Udenfriend, S. Fluorometric assay of protein in the nanogram range. Arch. Biochem. Biophys., 755: 213-220,1973. 7. Debenham, P. G., Kartner, N.. Siminovitch, L., Riordan. J. R., and Ling, V. DNA-mediated transfer of multiple drug resistance and plasma membrane glycoprotein expression. Mol. Cell. Biol., 2. 881-889, 1982. 8. Fairbanks, G., Steck, T. L., and Wallach, D. H. F. Electrophoretic analysis of the major polypeptides of the human erythrocyte membrane. Biochemistry, 70.2606-2617,1971. 9. Garman, D., and Center. M. S. Alterations in cell surface membranes in Chinese hamster lung cells resistant to Adriamycin. Biochem. Biophys. Res. Commun., 705: 157-163, 1982. 10. Juliano, R. L., and Ling, V. A surface glycoprotein modulating drug permeability in Chinese hamster ovary cell mutants. Biochim. Biophys. Acta, 455: 152162,1976. 11. Juliano, R. L., Ling, V., and Graves, J. Drug-resistant mutants of Chinese

15. 16. 17.

18

19.

20.

21.

22. 23.

24. 25.

hamster ovary cells possess an altered cell surface carbohydrate component. J. Supramol. Struct., 4: 521-526,1976. Ling, V. Drug resistance and membrane alteration in mutants of mammalian cells. Can. J. Genet. Cytol.. 77. 503-515, 1975. Ling, V. Genetic aspects of drug resistance in somatic cells. In: R. Schabel (ed.). Fundamentals in Cancer Chemotherapyâ€”Antibiotics Chemotherapy, Vol. 23. pp. 191-199. Basel: S. Karger AG. 1978. Ling. V. Genetic basis of drug resistance in mammalian cells. In: N. Bruchovsky and J. H. Goldie (eds.). Drug and Hormone Resistance in Neoplasia, Vol. 1, pp. 1-19. Boca Raton, Fla.: CRC Press, Inc., 1982. Ling, V., and Baker, R. M. Dominance of colchicine resistance in hybrid CHO cells. Somatic Cell Genet.. 4: 193-200, 1978. Ling, V., and Thompson, L. H. Reduced permeability in CHO cells as a mechanism of resistance to colchicine. J. Cell. Physiol., 83: 103-116, 1974. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. Protein measurement with the folin phenol reagent. J. Biol. Chem., 793: 265-275, 1951. McBurney, M. W.. and Whitmore, G. F. Isolation and biochemical characteri zation of folate deficient mutants of Chinese hamster cells. Cell. 2: 173-182. 1974. Peterson, R. H. F., and Biedler, J. L. Plasma membrane proteins and glycopro teins from Chinese hamster cells sensitive and resistant to actinomycin D. J. Supramol. Struct., 9: 289-298,1978. Riordan, J. R., and Ling, V. Purification of P-glycoprotein from plasma mem brane vesicles of Chinese hamster ovary cell mutants with reduced colchicine permeability. J. Biol. Chem., 254: 12701-12705. 1979. See, Y. P., Carteen. S. A., Till, J. E., and Ling. V. Increased drug permeability in Chinese hamster ovary cells in the presence of cyanide. Biochim. Biophys. Acta, 373:242-252, 1974. Stanners, C. P.. Elicieri, G , and Green, H. Two types of ribosome in mousehamster hybrid cells. Nat. New Biol., 230: 52-54, 1971. Switzer, R. C., Merril, C. R., and Shifrin, S. A highly sensitive silver stain for detecting proteins and peptides in polyacrylamide gels. Anal. Biochem., 98: 231-237,1979. Thompson, L. H., and Baker, R. M. Isolation of mutants of cultured mammalian cells. Methods Cell Biol., 6: 209-281.1973. Towbin. H., Staehelin, T., and Gordon, J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some appli cations. Proc. Nati. Acad Sei. U. S. A.. 76: 4350-4354, 1979.

Fig. 1. P-glycoprotein detection by silver staining in DNR" lines. Plasma membranes were isolated as described previously (20). SDS-PAGE was performed to resolve proteins of isolated plasma membranes, and protein bands were visualized by silver staining according to the method of Switzer ef al. (23) (see "Materials and Methods'); 10 ng of membrane protein were loaded per Lane. Lane a. Sensitive parent line, AUXB1 ; Lane o, sensitive parent line AUXE29; Lane c, colchicine-resistant line CHRC5 (derived from AUXB1); Lane d, daunorubicin-resistant line DNR"159 (derived from AUXE29). p, region of the P-glycoprotein at M, 150,000 to 170,000. Daunorubicin resistance-associated P-glycoprotein-like bands (Lanes d and e) appear to be of slightly lower molecular weight than the P-glycoprotein (Lane c). Molecular weight markers (Bio-Rad) are shown in the unlabeled lane, farthest right. These consist of about 0.1 ^g each of myosin (M, 200,000), /i-galactosidase (M, 116,250), phosphorylase b (M, 92,500), BSA (M, 66,200), ovalbumin (M, 45,000), carbonic anhydrase (M, 31,000). soybean trypsin inhibitor (M, 21,500). and lysozyme (M, 14,400). Fig. 2. Immunochemical detection of P-glycoprotein in DNR" lines. SDS-PAGE was performed as in Fig. 1. but protein bands were immediately electroblotted onto nitrocellulose paper, rather than being silver stained. A and B. blots from 2 gels that were run simultaneously and treated identically except for the antibody overlay. In A and B, the Lanes are as in Fig. 1: Lane a, AUXB1; Lane b, AUXE29; Lane c, CHRC5; Lane d, DNRR51; Lane e, DNR"159; p, region of the P-glycoprotein. A, Blot was overlaid with antiserum (diluted 400-fold) prepared against isolated plasma membrane of the colchicine-resistant cell line CHRC5. The only reproducible difference seen by this method is the appearance of a heavy P-glycoprotein band in the drug-resistant cells. B, Duplicate of blot shown in A was overlaid with the same antiserum after preabsorbing with immobilized proteins that were detergent-solubilized from isolated plasma membranes of the drug-sensitive parent line AUXB1 (see "Materials and Methods" for details of absorption). The major difference observed between the drug-sensitive and -resistant lines is the P-glycoprotein band in the drug-resistant cells (Lanes c fo e). A variable band, which is routinely observed in both the drug-sensitive and drug-resistant cells, is seen at about M, 50,000. This may result from incomplete absorption of the antiserum and is probably not related to drug resistance or the P-glycoprotein.

CANCER

4418


RESEARCH

VOL. 43

P-Glycoprotein in Daunorubicin-resistant

a

b

c

d

Cells

e â€” 200 K â€” 116

â€”92 â€”66

â€”45 â€”31 â€” 21 â€” 14

abcde P-C

abcde â€”200 K

i

â€” 116

â€” 92 â€”66 â€”45 â€”31 â€” 21

â€” 14

A

SEPTEMBER

B

4419

1983
