the detection of rhodamine 123 efflux at low levels of drug resistance

The detection of rhodamine 123 effluxat low levels of drug resistance

MALCOLM WEBB, CARY L. RAPHAEL,* HEINKE ASBAHR,* WENDY N. ERBER* AND BRIAN F. MEYER*Department of Haematology, Woden Valley Hospital, Canberra, and*Department of Haematology, Royal Perth Hospital, Perth, Western Australia

Received 9 October 1995; accepted for publication 1 February 1996

Summary. Although many cell models of multidrug resistance(MDR) have been developed, most have been high-resistancemodels which generate up to 100-fold increases in drugresistance. However, the drug concentrations required toachieve these levels of resistance are much higher than thosefound in vivo. In this paper we describe the development of acell model that reflects the resistance levels that are likely tobe found clinically. We then investigated the methods used todetect MDR1 expression at these low levels of drug resistance.We demonstrated that the immunological and PCR-basedmethods are unable to detect increased MDR1 expressionin cells with a <5.2-and 6.5-fold increase in vinblastineresistance, respectively, in our drug-resistant sublines.The rhodamine 123 (Rh 123) efflux assay was able to

discriminate the vinblastine-sensitive parent cells from all thevinblastine-resistant sublines, including cells with a 1.7-foldincrease in resistance. However, this assay is non-specific tothe MDR1 gene and may detect the activity of other drugefflux proteins such as the MDR-associated protein (MRP).Our results show that the Rh 123 efflux assay is able to detectthe activity of drug efflux proteins such as the P-glycoprotein,MRP or other efflux systems at the low levels of drugresistance that are likely to be attained in vivo.

Keywords: P-glycoprotein, multidrug resistance, monoclonalantibody, rhodamine, PCR.

One of the major obstacles to the successful treatment ofmany malignancies is the development of drug resistance tocommonly used cytotoxic compounds. Multidrug resistance(MDR) is a phenomenon where the malignant cell developscross resistance to a variety of unrelated cytotoxic drugs.Expression of the MDR1 gene produces a membrane protein,termed the P-glycoprotein, which actively pumps thecytotoxic agent from the cell (Kartner et al, 1985). Over-expression of the MDR1 gene has been associated with thedevelopment of drug resistance in vitro and a correlationbetween MDR1 over-expression and a poor response tochemotherapy has been established (Dalton et al, 1990;Salmon et al, 1989; Epstein et al, 1989; Grogan et al, 1993).

Although mammalian cell models expressing the P-glycoprotein have been developed, most of these haveconcentrated on the higher levels of drug resistance (e.g.100-fold increase in resistance). To achieve this level ofresistance, cells are required to be continuously exposed to a

P-glycoprotein transportable drug such as vinblastine atconcentrations of at least 200 ng/ml over a period of months.However, in clinical practice, vinblastine is usually adminis-tered as a 10 mg intravenous bolus injection �3/week andonly transiently rises above a serum level of 100 ng/ml beforefalling, within minutes, to a level below 20 ng/ml (Nelson etal, 1980). These serum levels bear little resemblance to theconcentrations used in the high-level resistance cell models.

The methods used to detect MDR1 expression areimmunological, functional and molecular, with the immuno-logical methods being the most widely applied. The mono-clonal antibodies most commonly used are the C219 and JSB-1 which are directed against internal epitopes and theMRK16 which is directed against an external epitope (Groganet al, 1990). Functional assays that utilize the P-glycopro-tein’s ability to pump dyes such as rhodamine 123 fromthecell have been used (Neyfakh, 1988), along with RNA-based assays such as RNA slot-blot analysis (Gottesman et al,1989), RNAase protection assay and the reverse transcrip-tion-polymerase chain reaction (RT-PCR) (Noonan et al,1990). These methods have been extensively investigated in

British Journal of Haematology, 1996, 93, 650–655

650 # 1996 Blackwell Science Ltd

Correspondence: Dr M. Webb, Haematology Department, WodenValley Hospital, P.O. Box 11, Woden, ACT 2606, Australia.

651Detection of Rhodamine 123 Efflux

# 1996 Blackwell Science Ltd, British Journal of Haematology 93: 650–655

high-resistance models, but this is not the case at low levels ofresistance.

We believe a cell model for MDR should be established toreflect the resistance levels that are likely to be foundclinically. In this paper we describe the development of sucha cell model that covers a wide range of resistance levels, andhave investigated the methods used to detect MDR1 expres-sion at these different resistance levels. We demonstrated thatthe immunological and PCR-based methods are unable todetect increased MDR1 expression in cells with a <5.2- and6.5-fold increase in vinblastine resistance, respectively, in ourdrug-resistant sublines. The Rh 123 efflux assay is able todetect significant differences in cells with a 1.7-fold increasein vinblastine resistance as compared to the drug-sensitiveparent cell line.

MATERIALS AND METHODS

Cell lines. A drug-sensitive T-lymphoblastic cell line(CCRF-CEM) was cultured in media consisting of RPMI1640 (Gibco BRL, Gaithersgurg, Md., U.S.A.), 20% fetal calfserum (CSL, Parkville, Victoria, Australia) and maintainedin a 5% CO2 incubator. Resistance was induced by seriallyculturing the cells in increasing concentrations of vinblas-tine (DBL, Musgrove, Victoria, Australia). Sublines wereproduced from the parent cell line after remaining viable inculture with a steady concentration of vinblastine over 3months and named according to that concentration ofvinblastine.

Degree of resistance. The degree of resistance was assessed byplacing 105 cells from the sublines into a range ofconcentrations of vinblastine. These were cultured for 2 dand then counted. The IC50 of each subline was determinedand the degree of resistance was calculated by dividing thesubline IC50 by the drug-sensitive parent cell line IC50 (Beck etal, 1979).

Immunofluorescence (flow cytometry). 2� 106 cells fromeach subline were incubated with and without 2.5 units/sample of neuraminidase for 30 min at 378C. The cells werewashed with phosphate-buffered saline (PBS) + 1% bovineserum albumin (BSA) (CSL, Parkville, Victoria, Australia) andthen incubated with 2 �g of the monoclonal antibody MRK16(courtesy of Dr Tsuruo) at 48C for 20 min. Cells were washedthree times with PBS (1000 rpm, 5 min, 48C) and incubatedwith fluorescein isothiocyanate conjugated F(ab0)2 (Silenus,Hawthorn, Victoria, Australia) for 20 min at 48C. After twowashes with PBS, cells were fixed with 1% paraformaldehyde(Sigma, St Louis, U.S.A.). The samples were analysed on aCoulter Epics-Profile II flow cytometer. The results wereexpressed as a fluorescence index which was calculated bymultiplying the mean channel fluorescence with the percen-tage of positive cells. The experiment was performed intriplicate on three different occasions and the Student t-testwas performed on these results.

Immunocytochemistry (alkaline phosphatase anti-alkalinephosphatase) (APAAP). Cytospins of 105 cells from all sublineswere made, air dried and fixed with acetone for 5 min. TheMDR1 specific murine monoclonal antibody JSB-1 (Sanbio,Amsterdam, The Netherlands) was used as the primary

antibody at a 1/10 dilution for 30 min at room temperature.The cytospins were washed and incubated with a rabbit anti-mouse immunoglobulin (Dakopatts, Carpinteria, Calif.,U.S.A.) at a dilution of 1/50 for 30 min at room temperature.The cytospins were washed and the APAAP complex (8 mgalkaline phosphatase (Sigma, St Louis, U.S.A.) per ml of anti-alkaline phosphatase antibody (AP-7:Oxford) 1/10) wasapplied for 30 min at room temperature. The above incuba-tions were repeated for 10 min at room temperature. Thesubstrate solution, 10 mg of naphthanol AS-MX phosphate(Sigma, St Louis, U.S.A.) per ml of N,N-dimethylformide with0.5 ml of 1 M of levamisole (Sigma, St Louis, U.S.A.) and fastred added, was applied in a humid chamber for 20 min atroom temperature. The cytospins were washed andcounterstained with haematoxylin for 1 min. The sampleswere graded from negative (ÿ) through varying grades ofpositive (++++).

RNA extraction, reverse transcription and polymerase chainreaction. Total cellular RNA was extracted from all CCRF-CEMsublines using RNAzol (Biotecx, Friendswood, Texas, U.S.A.)according to the supplier’s instructions. cDNA was producedby reverse transcription of total RNA (1�g) using AMVreverse transcriptase (15 units, Promega, Woods Hollow,Madison, Wis., U.S.A.) with oligo (dT) as the primer in a 20�lvolume as described by the manufacturer. PCR amplificationof the 307 bp cDNA was performed as per manufacturer’srecommendations (Perkin Elmer Cetus, Norwalk, Conn.,U.S.A.) by using 5�l of the reverse transcriptionmix producedabove and 0.2�g of MDR or actin-specific primers in a finalvolume of 100�l. The MDR1 cDNA was amplified usingoligonucleotides with the following sequences: primer 1, 50-AGGTCGGGATGGATCTTGAAGGG GA-30 and primer 2, 50-ATCATTGATATCACTTCTATTAGTG-30. Actin cDNA wasamplified using oligonucleotides with these sequences:primer 1, 50-TCACTCATGAAGATCCTC-30 and primer 2, 50-TTCGTGGATGCCACAGGAC-30. PCR parameters of 40 s at948C, 30 s at 558C and 2 min at 728C for 30 cycles were used.20�l aliquots of each PCR product were electrophoresedthrough a 1% agarose gel containing ethidium bromide. ThePCR product was visualized on a UV transilluminator andtransferred to a Hybond N (Amersham, Bucks., U.K.)hybridization membrane. The blot was hybridized with afluorescein-dUTP-labelled MDR1 internal oligonucleotide 50-TGGCTGCCATCATCCATGGG-30 (30 oligolabelling and detec-tion system, Amersham, Bucks., U.K.) under conditionsrecommended by the supplier.

Rhodamine 123 efflux assay. 5� 105 cells from eachsubline were incubated with 150 ng/ml of Rh 123 (Sigma,St Louis, U.S.A.) for 30 min at 378C in RPMI medium. Thecells were washed twice in PBS and resuspended in RPMIand allowed to efflux for 3 h at 378C. The cells were thenwashed twice and analysed on a Becton Dickinson FASCscanFlow cytometer. Cells from each subline that had not beenexposed to rhodamine were used as controls. The resultswere expressed as the fluorescence index calculated bymultiplying the mean channel fluorescence with thepercentage of positive cells. The experiment was performedin triplicate on three separate occasions and the Student ttest was applied to the results.

RESULTS

Resistance of sublinesFrom the vinblastine-sensitive CCRF-CEM parent cell line, 11vinblastine-resistant sublines were developed and designatedby the maximum concentration of vinblastine in which theywere cultured during establishment. The concentration ofvinblastine used in selection ranged from 1 to 200 ng/ml. Thedegree of resistance ranged from 1.0 to 106.8 (Table I).Differences in the degree of resistance of individualsublines ascompared to the drug-sensitive parent cell line were highlysignificant in all cases (P < 0:001).

Expression of the P-glycoprotein in sublinesFig 1 shows the relationshipbetween the degree of vinblastine

resistance in the sublines and P-glycoprotein expression usingthe MRK16 monoclonal antibody. No significant difference(P > 0:1) in P-glycoprotein expression was seen in sublinesresistant to <15 ng/ml of vinblastine (i.e. with degrees ofresistance <5.2) relative to the drug-sensitive parent cell line.Sublines resistant to 515 ng/ml of vinblastine (i.e. withdegrees of resistance 55.2) had significant increases(P < 0:01) in P-glycoprotein expression as detected usingthe MRK16 monoclonal antibody. Neuraminidase treatmenthad no effect on the binding of MRK16 to the drug-sensitiveparent cell line or the vinblastine-resistant sublines (data notshown), which is consistent with other investigators’ findingsin the CEM cell line (Cumber et al, 1990). Immunocytochem-ical staining with the monoclonal antibody JSB-1 showed no


652 Malcolm Webb et al

Fig 2. RT-PCR amplification of MDR1 mRNA inthe vinblastine-resistant sublines. Lanes 2–12represent ethidium bromide stained PCRproducts from sublines with degrees ofresistance 1.0, 3.3, 4.3, 5.2, 6.5, 10.8, 14.5,38.2, 38.2, 78.2 and 106.8 respectively. Lane1: marker. �-actin mRNA used as a control forRNA integrity.

Fig 1. P-glycoprotein expression relative to thedegree of resistance of the vinblastine-resistantsublines. Immunophenotypic analysis by flowcytometry was performed using the MRK16antibody on cells from the vinblastine-resistantsublines. The sublines were represented by theirdegrees of resistance relative to the drug-sensitive parent cell line. Each experiment wasperformed in triplicate and error barsrepresenting one standard deviation from themean are shown. (A) All sublines.(B) The sublines resistant to vinblastineconcentrations (<20 ng/ml) attained in vivo.



detectable difference in MDR1 gene expression between thedrug-sensitive cell line and sublines resistant to415 ng/ml ofvinblastine (i.e. with degrees of resistance45.2) (Table II).

PCR analysis of MDR1 mRNAFollowing RT and 30 cycles of PCR, MDR1 mRNA productwas detected in sublines resistant to520 ng/ml of vinblastine(i.e. with degrees of resistance56.5). The assay was unableto detect MDR1 expression in sublines resistant to <20 ng/ml(i.e. with degrees of resistance <6.5). As may be seen from the

ethidium bromide stained gel (Fig 2), the product intensityincreased with the degree of resistance (upper panel).Sensitivity could be enhanced by increasing the cycles ofPCR and/or Southern blotting and probing of the PCRproduct. However, the level of expression could not bediscriminated between the degrees of resistance of 1 and5.2 (data not shown). RNA integrity and quantity loaded wasequivalent for all sublines as indicated by RT-PCR amplifica-tion of actin mRNA (Fig 2; lower panel).

Rhodamine efflux studiesThere was a significant change in the efflux patterns of Rh123 in all sublines as compared to the drug-sensitive parentcell line (Fig 3). The fluorescence index for the parent cell linewas 28 693 but this fell to 3793 in the subline with a degreeof resistance of 1.7 (P value <0.05). The fluorescence indexvalues for the more resistant sublines were even lower(P values <0.05).

Fig 3. Functional P-glycoprotein expression relative to the degree ofresistance of the vinblastine-resistant sublines. Cells from eachsubline were incubated with Rh123 and analysed by flow cytometry.The sublines are represented by their degree of resistance relative tothe drug-sensitive parent cell line. Each experiment was performed intriplicate and error bars representing one standard deviation from themean are shown. (A) All sublines. (B) The sublines resistant tovinblastine concentrations (<20 ng/ml) attained in vivo.

Table II. Immunocytochemical detection of P-glycoprotein expression in the vinblastine-resistant sublines. The JSB-1 antibody was usedin an APAAP technique on cytospins of cellsfrom each vinblastine-resistant subline. Anindependent observer graded the level ofP-glycoprotein expression of each sample fromnegative (ÿ) to four plus (++++).

Cell line Staining intensity

CEM �=ÿ

CEMVb5 �=ÿ

CEMVb10 �=ÿ

CEMVb15 �=ÿ

CEMVb20 �

CEMVb30 �

CEMVb40 �

CEMVb50 ��

CEMVb75 ��

CEMVb100 ��

CEMVb200 ��

Table I. The degree of vinblastine resistance ofeach subline relative to the drug-sensitive parentcell line (CEM). The sublines were designated bythe maximum concentration of vinblastine inwhich they were cultured during establishment.The degree of resistance was calculated bydividing the subline IC50 by the drug-sensitiveparent cell line IC50.

Cell line Degree of resistance

CEM 1CEMVb1 1.7CEMVb5 3.3CEMVb10 4.3CEMVb15 5.2CEMVb20 6.5CEMVb30 10.8CEMVb40 14.5CEMVb50 38.2CEMVb75 38.2CEMVb100 78.2CEMVb200 106.8

DISCUSSION

In this study we developed 11 vinblastine-resistant sublinesfrom the drug-sensitive T-lymphoblastic cell line, CCRF-CEM.These sublines exhibited resistance levels ranging from 1.7- to106.8-fold increase in vinblastine resistance as compared tothe parent cell line. The sublines at the lower range (<6.5-foldincrease in the degree of resistance) were resistant tovinblastine concentrations (<20 ng/ml) that can be attainedin vivo. This enabled us to test the methods for detection ofMDR1 expression at low and high levels of drug resistance.

The MRK16 antibody was able to differentiate cells thathad not been exposed to vinblastine from cells resistant to515 ng/ml of vinblastine (i.e. with degrees of resistance55.2). However, it was unable to discriminate betweensublines resistant to <15 ng/ml of vinblastine (i.e. withdegrees of resistance <5.2). The JSB-1 antibody was even lessdiscriminating, because it was unable to detect any differ-ences between drug-sensitive cells and sublines resistant to415 ng/ml (i.e. with degrees of resistance 45.2). RT-PCRwas able to discriminate between sublines resistant to520 ng/ml of vinblastine (i.e. with degrees of resistance56.5) and the drug-sensitive parent cell line (Fig 2). Theseresults demonstrate that the two monoclonal antibodies andRT-PCR were unable to detect increases in MDR1 expressionin cells with a <5.2- and 6.5-fold increase in vinblastineresistance, respectively, in our drug-resistant sublines.

The Rh 123 efflux assay was able to discriminate drug-sensitive cells from all the vinblastine-resistant sublines. Itwas able to detect significant differences (P value <0.05)between drug-sensitive parent cells and cells resistant to 1 ng/ml of vinblastine (i.e. with degrees of resistance of 1.7).However, unlike the other methods, the Rh 123 efflux assay isnon-specific to the MDR1 gene and may detect the activity ofother drug efflux proteins such as the MDR-associated protein(Cole et al, 1992). It is possible that a system such as the MRPis responsible for the vinblastine resistance at the lower levelsof resistance in this cell line (Slapak et al, 1994). This wouldexplain why the Rh 123 efflux assay is able to detect the lowlevels of vinblastine resistance whereas the MDR1-specificassays are not.

Since the vinblastine levels found in vivo are mainly<20 ng/ml, it is likely the lower levels of resistance (i.e.resistance generated by <20 ng/ml of vinblastine) are asimportant clinically as the higher resistance levels. It wouldbe important to be able to detect these low levels of drugresistance because they may reflect an early, more frequentand possibly more reversible phase in the development ofclinical drug resistance. Our results show that the Rh 123efflux assay is able to detect the activity of drug efflux proteinssuch as the P-glycoprotein, MRP, or other efflux systems atthese low levels of resistance (i.e. cells with a 1.7-fold increasein vinblastine resistance) in our resistant sublines.

Most studies report the correlation between MDR1 over-expression, as detected by immunological and molecularmeans, with a poor response to chemotherapy (Guerci et al,1995; Wood et al, 1994). These studies also report asignificant proportion of patients who have no evidence ofMDR1 over-expression but develop clinical drug resistance,

and it is currently believed that these patients become drugresistant due to other mechanisms. Our results suggest thatthe Rh 123 efflux assay may be helpful in identifyingsubpopulations of these patients in whom the activity ofdrug efflux proteins contribute to clinical drug resistance.

ACKNOWLEDGMENTS

The authors thank Professor P. Herdson, Professor P. Boardand Dr O. Chisholm for critical reading and helpful comments.This study was supported by a grant from the Royal PerthHospital Medical Research Foundation.

REFERENCES

Beck, W., Muellar, T. & Tanzer, L. (1979) Altered surface membraneglycoproteins in vincristine alkaloids-resistant human leukaemiclymphoblasts. Cancer Research, 39, 2070–2076.

Cole, S., Bhardwaj, G., Gerlach, J., Mackie, J., Grant, C., Almquist, K.,Stewart, A., Kurz, E., Duncan, A. & Deeley, R. (1992) Over-expression of a transporter gene in a multidrug resistant humanlung cancer cell line. Science, 258, 1650–1654.

Cumber, P., Jacobs, A., Hoy, T., Fischer, J., Whittaker, J., Tsuruo, T. &Padua, R. (1990) Expression of the multiple drug resistance gene(mdr-1) and epitope masking in chronic lymphatic leukaemia.British Journal of Haematology, 76, 226–230.

Dalton, W.S., Grogan, T.M., Meltzer, P.S., Scheper, J.J., Durie, B.G.,Taylor, C.W., Miller, T.P. & Salmon, S.E. (1989) Drug resistance inmultiple myeloma and non-Hodgkin’s lymphoma: detection of P-glycoprotein and potential circumvention by addition of verapamilto chemotherapy. Journal of Clinical Oncology, 7, 415–423.

Epstein, J., Xiao, H.Q. & Oba, B.K. (1989) P-glycoprotein expression inplasma cell myeloma is associated with resistance to VAD. Blood,74, 913–917.

Gottesman, M., Goldstein, I., Bruggeran, E., Currier, S., Galski, H.,Cardarelli, C., Thiebaut, F., Willgham, M. & Pastan, J. (1989)Molecular diagnosis of MDR. Cancer Cells, 7, 75–80.

Grogan, T., Dalton, W., Rybski, C., Meltser, P., Richter, L., Gleason, M.,Pindur, J., Cline, A., Schelper, R., Tsuruo, T. & Salmon, S. (1990)Optimisation of immunocytochemical P-glycoprotein: assessmentof MDR plasma cell line myeloma using three antibodies.LaboratoryInvestigation, 63, 815–823.

Grogan, T., Spier, C., Salmon, S., Matzner, M., Rybski, J., Weinstein, R.,Scheper, R. & Dalton, W. (1993) P-glycoprotein expression inhuman plasma cell myeloma: correlation with prior chemotherapy.Blood, 81, 490–495.

Guerci, A., Merlin, J., Missoum, N., Feldman, L., Marchal, S., Witz, F.,Rose, C. & Guerci, O. (1995) Predictive value for treatment in acutemyeloid leukemia of cellular daunorubicin accumulation and P-glycoprotein expression simultaneously determined by flowcytometry. Blood, 85, 2147–2153.

Kartner, N., Evernden, K., Porelle, D., Bradley, G. & Ling, V. (1985)Detection of P-glycoprotein in MDR cell lines by monoclonalantibodies. Nature, 316, 820–823.

Nelson, R., Dyke, R. & Root, M. (1980) Comparative pharmoco-kinetics of vindesine, vincristine, and vinblastine in patients withcancer. Cancer Treatment Reviews, 7, 17–24.

Neyfakh, A. (1988) Use of fluorescent dyes as molecular probes forthe study of multidrug resistance. Experimental Cell Research, 174,168.

Noonan, K., Beck, C., Holzmayer, T., Chin, J., Wunder, J., Andrulis, I.,


654 Malcolm Webb et al



Gazdar, A., Willian, C., Griffith, B., Von Hoff, D. & Roninson, I. (1990)Quantitation analysis of MDR1 gene expression in human tumorsby polymerase chain reaction. Proceedings of the National Academyof Sciences of the United States of America, 87, 7160–7164.

Salmon, S.E., Grogan, T.M., Miller, T., Scheper, R. & Dalton, W.S.(1989) Prediction of doxorubicin resistance in vitro in myeloma,lymphoma, and breast cancer by P-glycoprotein staining. Journal ofthe National Cancer Institute, 81, 696.

Slapack, C.A., Mizunuma, N. & Kupe, D.W. (1994) Expression of themultidrug resistance associated protein and P-glycoprotein indoxorubicin-selected human myeloid leukemia cells. Blood, 84,3113–3119.

Wood, P., Burgess, R., MacGregor, A. & Liu Yin, J. (1994) P-glycoprotein expression on acute myeloid leukaemia blast cells atdiagnosis predicts response to chemotherapy and survival. BritishJournal of Haematology, 87, 509–514.

the detection of rhodamine 123 efflux at low levels of drug resistance

Documents