OVERCOMING MULTI-DRUG RESISTANCE USING AN INTRACELLULARANTI-MDR1 SFVYuji HEIKE

1–3, Keizo KASONO1, Chikara KUNISAKI

1, Seiji HAMA3, Nagahiro SAIJO

2, Takashi TSURUO4, Douglas A. KUNTZ

5,David R. ROSE

5 and David T. CURIEL1*

1UAB Gene Therapy Center, Wallace Tumor Institute, Birmingham, AL, USA2Pharmacology Division, National Cancer Center Research Institute, Chuo-ku, Tokyo, Japan3Department of Medical Oncology and Clinical Research, National Shikoku Cancer Center, Matsuyama, Japan4Institute of Molecular and Cellular Biosciences, University of Tokyo, Bunkyo-ku, Tokyo, Japan5Ontario Cancer Institute and Department of Medical Biophysics, University of Toronto, Toronto, Ontario, Canada

We made an intracellular single-chain variable fragment(sFv) from the C219 monoclonal antibody that recognizedthe intracellular domain of the multidrug resistance (MDR)gene product, P-glycoprotein (P-gp). Immuno-cytochemistryusing the FITC conjugated anti-C-myc tag antibody showedthat the sFv protein was expressed in the cytoplasm of thecells. Although transfection of the sFv did not result in thedown-regulation of P-gp expression in P-gp positive MDRcells as determined by flow cytometry analysis, Adriamycin(ADM) uptake and Rhodamine123 (Rh123) retention wereincreased by the C219 intra-cellular sFv transfection. Thetransfected cells exhibited a higher sensitivity to ADM usinga 10-day colony formation assay. The conventional 3-dayMTT assay showed the drug resistant tendency in C219 sFvtransfected cell we tested. The growth rate of C219 sFvtransfected cells was delayed in all non-MDR and MDR cellsthat might be the reason why C219 transfected cells exhib-ited the drug resistant tendency in the MTT assay. Despitethis unexpected effect of C219 sFv on growth rate, our datasuggest that the intra-cellular sFv technique could knockoutMDR functionally and may offer a means of increasing theeffectiveness of tumor chemotherapy.© 2001 Wiley-Liss, Inc.

Key words: multidrug resistance; single-chain antibody; functionalknockout; Adriamycin

Multiple drug resistance (MDR) has been studied extensivelybecause it is one of major problems in cancer chemotherapy. AMDR geneproduct, P-glycoprotein (P-gp), on thesurfaceof MDRcells was found to play an important role in the mechanism ofMDR.1–3 P-gp is now believed to be an energy dependent drugefflux pump, that reduces the accumulation of drugs in MDRcells.4,5 As P-gp is known to be expressed by both acquired andintrinsic MDR cells in human cancers,1–3 therapeutic approachesthat target it should be useful in overcoming multidrug resistance.Various chemical compounds (including many dihydropyridinetype calcium channel blockers such as verapamil, R-verapamil,diltiazem, quinidine and trifluoperazine) have been found to re-verse multidrug resistance in vitro,6,7 and in a few instances invivo.8–10 Unfortunately, most of these agents produce undesirableside effects in vivo, that cause cardiovascular disorders at theconcentrations needed to reverse multidrug resistance in vitro.11,12

Recent advances in molecular technology made it possible tointroduce genetic modification of the cell to the field of treatment.Reports have already described that treatment of anti-sense oligo-nucleotide13,14 against P-gp and anti-P-gp ribozyme15 overcameMDR and enhanced the chemosensitivity of the treated cells.Single chain antibodies have recently emerged as highly specifictools for the intracellular modulation of target proteins, and assuchhave led to a variety of genetic therapies. We have already madeseveral intracellular sFvs and reported the use of these on themolecular knockout of the gene product or on the functionalknockout of the gene product.16–20

Several monoclonal antibodies are known to react to P-gp.Some, such as MRK16, recognize the extracellular glycosylatedportion5,21–23whereas others, including C219, recognize the intra-cellular domain.24,25In the present study, we report our use of sFv

designed from C219 that binds to regions near the ATP-bindingcassette of the cytoplasmic domain of P-gp, and the effects ofintracellular expression of C219 sFv on MDR. Furthermore, wediscuss the possible utility and problems of our strategy in thetreatment of MDR cancers.

MATERIAL AND METHODS

C219 sFv and control sFv

C219 Mab was developed by Kartner et al.24 The epitope ofC219 has been identified25 and the sFv derived from it had alsobeen characterized.26,27 The sFv fragment of C219 sFv was am-plified by polymerase chain reaction (PCR) methods using thefollowing primers: 59-CCG GCC CAG CCG GCC GAC ATTGTGATG ACA CAG TC-39 for an upper primer and 59-AATGCG GCC GCT GAG GGG ACG GTG ACC GTG G-39 for alower primer. PCR product was ligated into the Sfi I/Not I sites ofeukaryotic expression vector pcDNA3 (Invitrogen, Carlsbad, CA)modified in our laboratory.17 As shown in Figure 1, the C219 sFvvector contains the C219 sFv gene fused with c-myc tag (EQKLI-SEEDLN). Thecontrol plasmid consisted of an expression plasmidfor the sFv against LMP-1.20

Cancer cells and cultures

Human ovarian tumor A2780 and its ADM-resistant variants,2780-AD, were kindly supplied by Drs. R.F. Ozols and T.C.Hamilton, National Cancer Institute, Bethesda, MD. Human my-elogenous leukemia K562 cells resistant to Adriamycin (ADM)(K562/ADM) were established in Dr. Tsuruo’s laboratory as de-scribed previously.28 The characteristics of these cell lines havealready been summarized in detail.29 These cell lines and their sFvtransfected cells were cultured in RPMI1640 tissue culture media(Mediatech Inc., Herndon, VA) supplemented with L-glutamine (2mM), penicillin (100 IU/ml), streptomycin (25 mg/ml) and 10%heat-inactivated fetal calf serum (FCS) at 37°C in ahumidified 5%CO2 atmosphere.

Abbreviations: MDR, multiple-drug resistance; ADM, Adriamycin.

Grant sponsor: Ministry of Health and Welfare, Japan; Grant sponsor:Ministry of Education, Science, Culture and Sports, Japan; Grant sponsor:National Institutes of Health; Grant number: R01 CA72532, R01CA68245.

*Correspondence to: UAB Gene Therapy Center, 620 Wallace TumorInstitute, 1824 Sixth Avenue South, Birmingham, AL 35294-3300. Fax:205-975-7476. E-mail: David.Curiel@ccc.uab.edu

Received 15 May 2000; Revised 16 October 2000; Accepted 18 October

Published online 9 February 2001

Int. J. Cancer: 92, 115–122 (2001)© 2001 Wiley-Liss, Inc.

Publication of the International Union Against Cancer

ReagentsFetal calf serum (FCS) was purchased from GIBCO (Grand

Island, NY). ADM was obtained from Kyowa Hakko Kogyo(Tokyo, Japan). Geneticin (G418) and Rhodamine123 (Rh123)were obtained from Sigma (St. Louis, MO).

Transfection of sFv plasmid and establishmentof the sFv transfected cells

DH5a transformed by pcDNA3 plasmid containing sFvs andwere cultured in LB medium with 500mg/ml ampicillin. Thegrowing phase bacteria were collected and plasmids were purifiedby Maxi kit according to the manufacturer’s manual (Promega,Madison, WI).

sFv gene transfection into the A2780 and 2780-AD was per-formed by the Lipofectamine method (Promega) according themanufacturer’s manual. sFv transfection into the K562 and K562/ADM was performed by an electroporation. The transfected cellswere cultured in RPMI 1640 medium containing 500mg/ml ofG418 for 7 days and further cultured in the medium containing 300

mg/ml of G418 for 2 weeks. The established cell lines werecultured in a G418 free condition for at least 1 week before use toavoid any effects from G418.

Expression of C219 sFv protein in eukaryotic cellsTo detect the expression and localization of the C219 sFvin situ,

A2780, 2780-AD, K562 and K562/ADM cells transfected with thesFvs were plated on glass coverslips. After 48 hr, the cells werewashed and fixed by PBS containing 4% paraformaldehyde. Thenthe cell membranes were permeabilized with PBS containing 0.2%Triton X-100 (Sigma). An anti-c-Myc rabbit polyclonal antibody(Santa Cruz Biotechnology, Santa Cruz, CA) at 1:400 dilution wasadded and incubated for 1 hr at room temperature, followed byincubation with fluorescein isothiocyanate (FITC)-conjugatedpolyclonal goat anti-rabbit IgG antibody (Calbiochem, La Jolla,CA ) at 1:400 dilution for an additional 1 hr. After washing withPBS(2) twice, the stained cells were then evaluated using immu-nofluorescence microscopy.

Analysis of P-gp expression by flow cytometryA2780, 2780-AD, K562 and K562/ADM cells were harvested

and suspended in RPMI 1640 medium at 13 106/tube. Aftercentrifugation for 5 min, the cells were resuspended in 50ml ofPBS containing MRK16 (10mg/ml) (Kyowa Hakkou Co. Tokyo,Japan) for 30 min at 4°C. After washing twice with PBS, the cellswere resuspended in 50ml of PBS containing the second-stageantibody, F(ab9)2 of fluorescein-isothiocyanate (FITC)-conjugatedgoat anti-mouse IgG (H1L) (Immunotech, Marseille, France), at adilution of 1:50 for 30 min at 4°C. They were then washed twice,resuspended in PBS and stored at 4°C in the dark until needed.Cells incubated with isotype control as a first antibody were used

FIGURE 1 – Schema of the eukaryotic expression vectors of sFvs.The open reading frame of all sFvs are fused with a c-myc epitope toallow easy detection by Immunocytochemistry and were ligated intopcDNA3 plasmid. The sFvs were driven by a CMV promoter.

FIGURE 2 – Immunofluorescent staining of cytoplasmic targeted C219 sFv. (a) 2780-AD cells transfected with C219 sFv, (b) 2780-AD cellstransfected with control sFv, (c) 2780-AD cells transfected without sFvs. (d) K562/ADM cells transfected with C219 sFv, (e) K562/ADM cellstransfected with control sFv, (f) K562/ADM cells transfected without sFvs.

116 HEIKE ET AL.

to measure the background. Flow cytometry was performed in aFACSCaliver system (Becton Dickinson, Lincoln Park, NJ).

Effects of intra-cellular sFvs on cell growthTransfected cells growing in log phase were replated at 53 103

cells per well in 24-well plates (Falcon, Becton Dickinson, Frank-lin Lakes, NJ). The viable cell numbers were counted periodicallyusing the Trypan-blue-exclusion test over a 7-day period. Tripli-cate wells were used for each treatment, and the average of eachgroup was calculated. These experiments were performed 3 timesindependently.

Uptake of fluorescent ADMTo estimate the functional activity of the transfected C219 sFv,

the uptake of the fluorescent drug ADM was determined by flowcytometry. Briefly, cells were cultured in phenol red-free RPMI1640 medium (GIBCO, Rockville, MD) supplemented with 10%FCS for 48 hr and were then washed with phenol red-free 5% FCSRPMI 1640 medium. Cells were aliquoted and incubated at 37°Cin phenol red-free 5% FCS RPMI 1640 medium containing 50mMADM for 3 hr. Thereafter, cells were washed twice with mediumand kept on ice. The fluorescence intensity of 13 104 control cellsand sFv transfected cells was determined by flow cytometry andanalyzed using the Cell Quest software program (Becton Dickinson).

Rhodamine 123 (Rh123) efflux studiesSamples containing 13 106 cells/ml were incubated with

Rh123 at a final concentration of 0.5 mg/ml at 37°C for 30 min foruptake. After washing, Rh123-stained cells were incubated in adye-free RPMI medium for additional 90 min for efflux. Flowcytometry was performed using FACS Caliver system (BectonDickinson) for Rh123-associated fluorescence. We decided the cutoff value as the 80% non-transfected cells treated Rh123 are inRh123 negative and residual 20% cells were in positive area ineach cell individually. Data was represented as the percentage ofRh123 positive cell in each group.

Drug sensitivity assay using a 3-day MTT testDrug sensitivity using the conventional MTT assay was mea-

sured as described previously.30 Briefly, cells (13 103/well) werecultured at 37°C for 5 hr in 50ml of RPMI 1640 medium inflat-bottomed 96-well microtiter plates (Falcon, Becton Dickin-son). Then the cells were incubated without or with graded con-centrations of ADM for 72 hr. Drug sensitivity was measuredusing the MTT assay as described previously.31

Drug sensitivity of sFv transfected cells using a 10-daycolony formation assay

Transfected cells growing in log phase were replated at 53 102

cells/2 ml of pre-warmed 0.3% agarose in culture medium with 0,10 or 20 ng/ml of ADM. After placing the plates at 4°C for 30 minto fix the gel, 1 ml of the culture medium containing an identicalconcentration of ADM was added to each well. The plates wereincubated for 10-days at 37°C in a humidified 5% CO2 atmo-sphere. The medium including ADM was replaced with newmedium every 3 days to maintain ADM concentration and gelwetness. The viable colony number, in which more than 50 viablecells were contained, in each well was counted on Day 10 using aphase-contrast microscope. Triplicate wells were used for eachtreatment, and the average of each group was calculated. Theseexperiments were performed 3 times independently.

Statistical analysisThe statistical significance of differences between groups was

analyzed by Student’st-test (2-tailed).

RESULTS

Plasmid constructionFigure 1 shows the construction of C219 sFv and control sFv

expression vectors. Both sFvs were cloned into the Sfi I/Not I siteof pcDNA3 plasmid and include a c-myc detection tag.

Expression of sFvs in the cellsFigure 2 shows the expression and localization of C219 sFv and

control sFv in 2780-AD and K562/ADM cells. These plasmidshave c-myc tags at the 39 end of the sFvs and localization wasdetected using the anti-c-myc tag antibody. C219 sFv (a,d) and thecontrol sFv (b,e) were expressed in the cytoplasmic area of thecells. The brightness of cells was almost the same between C219

FIGURE 3 – Growth curves of MDR and their parental cell lines withor without sFv transfection. Cell growth is shown as the ratio of thecell number of the indicated day to that of Day 0. (a) —●—: A2780,—Œ—: 2780-AD,- - -●- - -: A2780-control sFv,- - -Œ- - -: 2780-AD-control sFv,—C—: A2780-C219 sFv,—‚—: 2780-AD-C219sFv, (b)—●—: K562, —Œ—: K562/ADM, - - -●- - -: K562-control sFv,- - -Œ- - -: K562/ADM-control sFv,—C—: K562-C219 sFv,—‚—:K562/ADM-C219 sFv.

117OVERCOMING MDR BY C219 SFV

sFv transfected cells and control sFv transfected cells. We checkedthe sFv expression in A2780 and K562 cell lines, and determinedthat the sFv expressed well in the cytoplasmic area of the cells(data not shown).

Growth retardation of cells by C219 sFv gene transfectionThe growth of A2780, 2780-AD, K562 and K562/ADM cells

with or without sFv gene transfection are shown in Figure 3. Thedoubling time of A2780, 2780-AD, control sFv transfected A2780,control sFv transfected 2780-AD, C219 sFv transfected A2780 andC219 sFv transfected 2780-AD were 21 hr, 19.5 hr, 20.5 hr, 17.5hr, 26.0 hr and 24.5 hr, respectively. The doubling time of K562,K562/ADM, control sFv transfected K562, control sFv transfectedK562/ADM, C219 sFv transfected K562 and C219 sFv transfectedK562/ADM were 22.5 hr, 20.0 hr, 20.5 hr, 18.5 hr, 29.0 hr and25.5 hr, respectively. It was obvious that the growth rate of allC219 sFv transfected cells was delayed in both MDR positive andnegative cell lines.

Effect of C219 sFv gene-transfection on P-gp expressionTo evaluate sFv-mediated modulation of P-gp expression, trans-

fected cells were labeled with the P-gp-specific antibody MRK16,incubated with a fluorescein-conjugated secondary antibody, andanalyzed by flow cytometry. As shown in Figure 4, the expressionlevel of P-gp was the same between A2780 and C219 sFv trans-fected A2780, between 2780-AD and C219 sFv transfected 2780-AD, between K562 and C219 sFv transfected K562 and betweenK562/ADM and C219 sFv transfected K562/ADM, respectively.

ADM uptake and Rh123 retention in C219 sFv transfected cellsFigure 5 summarizes the data determined as the mean fluores-

cence values per cell (MF/cell) after 3 hr of ADM incubation incontrol and sFv gene-transfected cells. The ADM uptake in non-transfected and control sFv transfected A2780 cells was 120.5MF/cell and 116.8 MF/cell, respectively. The ADM uptake ofC219 sFv transfected A2780 cells was 115.2 MF/cell. No statisti-cal difference was found between these groups. ADM uptake innon-transfected, control sFv and C219 sFv transfected 2780-ADcells was 28.5 MF/cell, 32.3 MF/cell and 54.1 MF/cell, respec-tively. The accumulation of ADM was 1.9-fold higher in C219 sFvtransfected 2780-AD cells when compared with non-transfected,control sFv transfected 2780-AD cells. ADM concentration innon-transfected, control sFv transfected and C219 sFv transfectedK562 cells was 66.7 MF/cell, 65.1 MF/cell and 73.8 MF/cell,respectively. No statistical difference was found between theseresults. The ADM concentration in non-transfected, control sFvtransfected and C219 sFv transfected K562/ADM cells was 16.6MF/cell, 15.0 MF/cell and 50.0 MF/cell, respectively. Thus, thedrug accumulation was 3-fold higher in C219 sFv transfected cellswhen compared with non-transfected cells.

Figure 5 also showed Rh123 retardation in C219 sFv transfectedMDR cells. In the assay, we decided the cut off value as the 80%non-transfected cells treated Rh123 are in Rh123 negative andresidual 20% cells were in positive area in each cell individually.In C219 sFv transfected 2780-AD cells, 34% of total cells wereRh123 positive, although 20.0 % and 17.9% of 2780-AD and

FIGURE 4 – Representative FACS analyses of P-gp expression of (a) A2780, (b) 2780-AD, (c) K562 and (d) K562/ADM cells transfected withor without control and C219 sFvs. The cells were incubated with anti-P-glycoprotein MAb or mouse immunoglobulin isotype control.

118 HEIKE ET AL.

control sFv transfected 2780-AD cells were Rh123 positive re-spectively. In C219 sFv transfected K562/ADM cells, 36.1% oftotal cells were Rh123 positive, although 20.0% and 20.7% ofK562/ADM and control sFv transfected K562/ADM cells wereRh123 positive respectively.

Drug sensitivity assay using 3-day MTT assayTo determine if C219 sFv transfection caused a sensitization of

MDR cells to ADM, cells were treated with ADM for 3 days at aconcentration range of 0.3–1,000 ng/ml (Fig. 6). Cytotoxicity wasexpressed as percentage growth inhibition when compared withuntreated control cells. Cytotoxicity was reduced in all cells trans-fected with C219 sFv. IC50s of non-transfected, control sFv orC219 sFv transfected A2780 were 4.0 ng/ml, 5.4 ng/ml and 21.2ng/ml, and those of non-transfected, control sFv transfected andC219 sFv transfected 2780-AD were 20.8 ng/ml, 15.8 ng/ml and90.1 ng/ml, respectively. IC50s of non-transfected, control sFvand C219 sFv transfected K562 cells were 3.2 ng/ml, 2.9 ng/mland 21.0 ng/ml, and those of non-transfected K562/ADM, controlsFv transfected K562/ADM and C219 sFv transfected K562/ADMwere 30.7 ng/ml, 21.5 ng/ml and 79.8 ng/ml, respectively.

Colony formation assayChemosensitivity was assayed using 10-day colony formation

assay. Table I shows the colony numbers of indicated cells cul-tured in 0.3 % agar medium containing 0, 10 or 30 ng/ml ADM.Under the 30 ng/ml ADM condition, colony numbers of K562/ADM, of control sFv transfected K562/ADM and C219 sFv trans-fected K562/ADM were 61, 44 and 13, respectively. Those of2780-AD, control sFv transfected 2780-AD and C219 sFv trans-fected 2780-AD were 38, 37 and 12, respectively. Colony sizediffered between C219 sFv transfected cells and others. The col-ony size of C219 sFv transfected A2780, 2780-AD, K562 andK562/ADM were markedly smaller than their parental and controlsFv transfected cells (data not shown).

DISCUSSION

One of the most serious problems in current cancer chemother-apy is intrinsic or acquired drug resistance. We and numerousother researchers have studied the mechanisms of drug resistanceto develop therapies to overcome this problem. P-gp is a mem-

brane ATPase that serves as an efflux pump for multiple anticanceragents.4,5 This protein has 12 transmembrane domains divided intotwo homologous halves, each of which includes an ATP-bindingcassette domain that catalyzes ATP hydrolysis.32,33 Many reportssuggest that the calcium antagonists and analogs can overcomedrug resistance by competition with anti-cancer drugs. Many ofthese compounds, however, produced severe side effectsin vivo atsufficient concentrations. Consequently, effective treatments uti-lizing such strategies are not currently available.

Given the recent progress in the field of molecular biology, thestrategy of knocking out of an activated oncogene or a generesponsible for drug resistance may offer the best hope for a noveltreatment for cancer. Based on this concept, new methods includ-ing anti-sense DNA/RNA, ribozyme and chemicals have beendeveloped and tested. Intracellular sFvs provide a novel method toknockout gene products. A sFv is made from the VH and VLdomains of a monoclonal antibody and may contain localizationsignals to target it to a particular sub-cellular location such as theendoplasmic reticulum (ER) or nucleus. We have already estab-lished several intracellular sFvs such as the erbB2 sFv and testedthese in cancer therapy. The data suggested that some of thesecould knockout the production of the gene product itself and othersshowed knockout of its function.16–20,34

In the present study, we elected to use the C219 monoclonalantibody that binds to regions near the ATP-binding cassette in theC- and N-terminal regions of P-gp25 because of the followingresearch findings: C219 is known to inhibit photolabeling of P-gpwith 8-azido-[a-32P]ATP and [3H] azidopine25 and cyclosporin Amodulates the C219-P-gp binding;35 Kokubo et al.36 also reportedthat C219 inhibits the ATPase activity of P-gp by inhibiting ATPbinding; the sFv derived from C219 has been developed andstructurally analyzed.26,27

An important point to consider when using the sFv strategy isthe protein synthesis pathway. The extracellular portion of a newlysynthesized molecule is exposed on the inside of the ER and thecytoplasmic domain of the molecule is in the cytoplasm. Aftermoving to the Golgi body, sugars are added to the extracellularportion of what will become glycoproteins.37 The epitope of theC219 monoclonal antibody is located near the ATP binding do-main of the cytoplasmic portion of P-gp and this portion should be

FIGURE 5 – ADM accumulation and Rh123 retention in sFv transfected cells. (a) ADM fluorescence intensity was measured by a FACS Caliverflow cytometer as mean fluorescence of 13 104 cells. Each value represents the average of triplicate experiments. (b) The cut off value wasset as the 80% non-transfected cells treated with Rh123 are in Rh123 negative and residual 20% cells were in positive area in each cellindividually. Percentage of Rh123 positive cells are shown in each group.

119OVERCOMING MDR BY C219 SFV

present in the cytoplasm not in the ER. Therefore C219 sFvexpressed in the cytoplasm should bind to P-gp. Based on thisconcept, we constructed and tested a cytoplasmically targeted sFvderived from C219.

Transfected C219 sFv localized in the cytoplasm in a diffusemanner (Fig. 2) and inhibited the function of P-gp that resulted ina high concentration of intracellular ADM and Rh123 (Fig. 5).Unexpectedly, the C219 sFv transfection induced growth retarda-tion in both the MDR and non-MDR cells we tested (Fig. 3).Current reports suggest a cross-reactivity of the C219 antibodywith p815c-erbB2 as well as other proteins.38–40 As p815c-erbB2 isthought to have multifunctional activity including an involvementin the cell cycle, this cross reactivity might be a causal factor in theunexpected growth retardation of C219 sFv transfected cells. P-gpis a member of the ATP-binding cassette superfamily, a family thatincludes numerous proteins with diverse functions. The slow cellgrowth observed in all C219 sFv transfected cells might be causedby the cross reactivity of C219 sFv with other ATP-bindingcassette transporters. Alternately the growth retardation may bedue to inhibition of the intrinsic role of P-gp in normal cellmetabolism where it is suggested to have a role in phospholipidtransport. In any event, this growth retardation effect is undesirable

TABLE I 1

Cell lines ADM(2) ADM10 ng/ml

ADM30 ng/ml

K562 3186 21 126 4 06 0K562-Control sFv 2946 16 226 3 16 1K562-c219 sFv 2996 33 186 5 16 0K562/ADM 2786 8 1146 8 616 5K562/ADM-Control sFv 2776 11 1056 12 446 5K562/ADM-c219 sFv 2846 15 536 7 136 4A2780 1856 7 366 6 36 2A2780-Control sFv 1766 14 406 4 36 1A2780-c219 sFv 1816 19 346 8 26 1A2780/AD 1596 13 756 9 386 6A2780/AD-Control sFv 1496 15 806 9 376 3A2780/AD-c219sFv 1626 10 516 11 126 31sFv transfected cells growing in log phase were replated at 53 102

cells/2 ml of pre-warmed 0.3% agarose in culture medium with 0, 10and 30 ng/ml of ADM. The viable colony number of the wells werecounted on Day 10 using a phase-contrast microscope. Triplicate wellswere used for each treatment, and the average of each group wascalculated.

FIGURE 6 – Effects of sFv gene transfection on sensitivity of the cells to ADM. Cells were incubated for 72 hr in RPMI 1640 with variousconcentrations of ADM, and viable cells were then measured by MTT assay. (a) A2780, (b) 2780-AD, (c) K562, (d) K562/ADM. —●— sFv(-),—C— Control sFv,—‚— C219 sFv.

120 HEIKE ET AL.

for cancer chemotherapy, because in many cases slower growingcells show higher resistance to anti-cancer drugs. In accordancewith this, the 3-day MTT assay showed higher resistance to ADMin all C219 sFv transfected cells than non-transfected and controlsFv transfected cells (Fig. 6). We believe that this undesirableresult was caused by the unexpected growth retardation of C219sFv transfected cells.

When using the 10-day colony formation assay, however, wewere able to show that C219 sFv transfected MDR cells were moresensitive to ADM when compared with non-transfected and con-trol sFv transfected MDR cells (Table I). Another finding from thecolony formation assay showed that the colonies of C219 sFvtransfected cells were markedly smaller than those of non-trans-fected and control sFv transfected cells that might be the result ofthe slow growth of C219 sFv transfected cells.

In C219 sFv transfected MDR cells, although expression of P-gpwas not reduced (Fig. 4), intracellular accumulation of ADM wasenhanced. Other studies of intracellular cytoplasmic type sFvsfound that cytoplasmic sFvs appear not to knockout protein ex-pression but to knockout the molecules functionally. In contrastmany ER directed sFvs showed reduced levels of the protein itselfin the cell.16–20,34

Given these findings, we conclude that C219 sFv transfectionfunctionally knocked out P-gp allowing increased intracellulardrug accumulation and diminution of the MDR phenotype whenassessed using the 10-day colony formation assay. The sFv alsoinduced growth retardation, however, that lead to drug resistancewhen assessed using the 3-day MTT assay. In practice it is ex-tremely difficult, if not impossible, to maintain high, long-termADM concentration in patients because of its toxicity. This factmight indicate that the C219 sFv strategy will not be the mosteffective clinical method to overcome drug resistance. We havebegun efforts to develop an sFv that recognizes the MDR sequencespecifically and selectively inhibits only MDR expression or func-tion to develop the ideal anti-MDR therapy.

ACKNOWLEDGEMENTS

This work was supported in part by a grant-in-aid for CancerResearch for the Second Term of the Comprehensive 10-YearStrategy for Cancer Control from the Ministry of Health andWelfare, Japan, by a grant-in-aid from the Ministry of Education,Science, Culture and Sports, Japan and by the National Institutesof Health (R01 CA72532 and R01 CA68245).

REFERENCES

1. Bell DR, Gerlach JH, Kartner N, Buich RN, Ling V. Detection ofP-glycoprotein in ovarian cancer: a molecular marker associated withmultidrug resistance. J Clin Oncol 1985;3:311–5.

2. Fojo AT, Ueda K, Slamon DJ, Poplack DG, Gottesman MM, PastanI. Expression of a multidrug-resistance gene in human tumors andtissues. Proc Natl Acad Sci USA 1987;84:265–9.

3. Tsuruo T, Sugimoto Y, Hamada H, Roninson I, Okumura M, AdachiK, et al. Detection of multidrug resistance markers, P-glycoproteinand mdr1 mRNA, in human leukemia cells. Jpn J Cancer Res 1987;78:1415–9.

4. Hamada H, Tsuruo T. Characterization of the ATPase activity of the170- to 180-kDa membrane glycoprotein (P-glycoprotein) associatedwith multidrug resistance. Cancer Res 1988;48:4926–32.

5. Hamada H, Tsuruo T. Purification of the 170- to 180-kDa membraneglycoprotein associated with multidrug resistance. 170- to 180-kDamembrane glycoprotein is an ATPase. J Biol Chem 1988;263:1454–8.

6. Hamilton TC, Young RC, Ozols RF. Experimental model systems ofovarian cancer: applications to the design and evaluation of newtreatment approaches. Semin Oncol 1984;11:285–98.

7. Rogan AM, Hamilton TC, Young RC, Klecker RW Jr, Ozols RF.Reversal of Adriamycin resistance by verapamil in human ovariancancer. Science 1984;224:994–6.

8. Ozols RF, Cunnian RE, Klecker RW, Hamilton TC, Ostchega Y,Parrillo JE, et al. Verapamil and Adriamycin in the treatment ofdrug-resistant ovarian cancer patients. J Clin Oncol 1987;5:641–7.

9. Tolcher AW, Cowan KH, Solomon D, Ognibene D, Goldspiel B,Change R, et al. Phase I crossover study of pacritaxel with r-verapamilin patients with metastatic breast cancer. J Clin Oncol 1996;14:1173–84.

10. Milroy R, Connory L, Banham F, Hutcheon D, MacIntyre D, Stack B.Randomized trial of verapamil in addition to chemotherapy for smallcell lung cancer. Thorax 1987;42:209.

11. Gottesman MM, Pastan I. Clinical trials of agents that reverse multi-drug-resistance. J Clin Oncol 1989;7:409–11.

12. Mickisch GH, K̈ossig J, Keilhauer G, Schlick E, Tschada RK, AlkenPM. Effects of calcium antagonists in multidrug resistant primaryhuman renal cell carcinomas. Cancer Res 1990;50:3670–4.

13. Motomura S, Motoji T, Takahashi M, Wang YH, Shiozaki H, Sug-awara I, et al. Inhibition of P-glycoprotein and recovery of drugsensitivity of human acute leukemic blast cells by multidrug resis-tance gene (mdr1) antisense oligonucleotides. Blood 1998;91:3163–71.

14. Cucco C, Calabrett B.In vitro and in vivo reversal of multidrugresistance in a human leukemia-resistant cell line by mdr1 antisenseoligodeoxynucleotides. Cancer Res 1996;56:4332–7.

15. Kobayashi H, Takemura Y, Wang FS, Oka T, Takao O. Retrovirusmediated transfer of anti-MDR1 hammerhead ribozymes into multi-drug-resistant human leukemia cells: screening for effective targetsites. Int J Cancer 1999;81:944–50.

16. Stackhouse MA, Buchsbaum DJ, Grizzle WE, Bright SJ, Olsen CC,Kancharla S, et al. Radiosensitization mediated by a transfectedanti-erbB-2 single-chain antibodyin vitro and in vivo. Int J RadiatOncol Biol Phys 1998;42:817–22.

17. Kasono K, Piche A, Xiang J, Kim HG, Bilbao G, Johanning F, et al.Functional knock-out of c-myb by an intracellular anti-c-Myb single-chain antibody. Biochem Biophys Res Comm 1998;251:124–30.

18. Grim JE, Deshane J, Siegal GP, Alvarez RD, DiFiore P, Curiel DT.The level of erbB2 expression predicts sensitivity to the cytotoxiceffects of an intracellular anti-erbB2 sFv. J Mol Med 1998;76:451–8.

19. Grim JE, Siegal GP, Alvarez RD, Curiel DT. Intracellular expressionof the anti-erbB-2 sFv N29 fails to accomplish efficient target mod-ulation. Biochem Biophys Res Comm 1998;250:699–703.

20. Piche A, Grim J, Rancourt C, Gomez-Navarro J, Reed JC, Curiel DT.Modulation of Bcl-2 protein levels by an intracellular anti-Bcl-2single-chain antibody increases drug-induced cytotoxicity in thebreast cancer cell line MCF-7. Cancer Res 1998;58:2134–40.

21. Georges E, Tsuruo T, Ling V. Topology of P-glycoprotein as deter-mined by epitope mapping of MRK-16 monoclonal antibody. J BiolChem 1993;268:1792–8.

22. Scheper RJ, Bulte JW, Brakkee JG, Quak JJ, van der Schoot JJ, BalmAJM, et al. Monoclonal antibody JSB-1 detects a highly conservedepitope on the P-glycoprotein associated with multidrug-resistance.Int J Cancer 1988;42:389–94.

23. Mechetner EB, Roninson IB. Efficient inhibition of P-glycoprotein-mediated multidrug resistance with a monoclonal antibody. Proc NatlAcad Sci USA 1992;89:5824–8.

24. Kartner N, Evernden-Porelle D, Bradley G, Ling V. Detection ofP-glycoprotein in multidrug-resistant cell lines by monoclonal anti-bodies. Nature 1985;316:820–3.

25. Georges E, Bradley G, Gariepy J, Ling V. Detection of P-glycoproteinisoforms by gene-specific monoclonal antibodies. Proc Natl Acad SciUSA 1990;87;152–6.

26. Hoedemaeker FJ, Signorelli T, Johns K, Kuntz DA, Rose DR. Asingle chain Fv fragment of P-glycoprotein-specific monoclonal anti-body C219. Design, expression, and crystal structure at 2.4 A reso-lution. J Biol Chem 1997;272:29784–9.

27. van den Elsen JM, Kunts KA, Hoedemaeker FJ, Rose DR. AntibodyC219 recognizes an alpha-helical epitope on P-glycoprotein. Proc NatlAcad Sci USA 1999;96:13679–84.

28. Tsuruo T, Iida-Saito H, Kawabata H, Oh-Hara T, Hamada M, UtakojiT. Characteristics of resistance to Adriamycin in human myelogenousleukemia K562 resistant Adriamycin and in isolated clones. Jpn JCancer Res 1986;77:682–92.

29. Hamada H, Tsuruo T. Growth inhibition of multidrug-resistant cellsby monoclonal antibodies against P-glycoprotein. In: Roninson IB.Molecular and cellular biology of multidrug resistance in tumor cells.New York: Plenum Publishing Corp., 1991;373–91.

30. Heike Y, Hamada H, Inamura N, Sone S, Ogura T, Tsuruo T.Monoclonal antibody-dependent killing of multidrug-resistant tumorcells by human mononuclear cells. Jpn J Cancer Res 1990;81:1155–61.

31. Vugrin D. Systematic therapy of metastatic renal cell carcinoma.Semin Nephrol 1987;7:152–62.

32. Cornwell MM, Tsuruo T, Gottesman MM, Pastan I. ATP-bindingproperties of P-glycoprotein from multidrug-resistant KB cells.FASEB J 1987;1:51–4.

121OVERCOMING MDR BY C219 SFV

33. Schurr E, Raynord M, Bell JC, Gros P. Characterization for themultidrug resistance protein expressed in cell clones stably transfectedwith the mouse MDR1 cDNA. Cancer Res 1989;162:2729–33.

34. Wang-Johanning F, Gillespie GY, Grim J, Rancourt C, Alvarez RD,Siegal GP, et al. Intracellular expression of a single-chain antibodydirected against human papillomavirus type 16 E7 oncoproteinachieves targeted antineoplastic effects. Cancer Res 1998;58:1893–1900.

35. Georges E, Zhang JT, Ling V. Modulation of ATP and drug bindingby monoclonal antibodies against P-glycoprotein. J Cell Physiol 1991;148:479–84.

36. Kokubo N, Cohen D, Watanabe T. Functional modulation of ATPaseof P-glycoprotein by C219, a monoclonal antibody against P-glyco-protein. Biochem Biophys Res Comm 1997;230:398–401.

37. Jones AD, Marasco WA. Antibodies for targeted gene therapy: ex-tracellular a gene targeting and intracellular expression. Adv DrugDeliv Rev 1998;31:153–70.

38. Liu B, Sun D, Xia W, Hung MC, Yu D. Cross-reactivity ofC219 anti-p170mdr-1 antibody with p185c-erbB2 in breast cancercells: cautions of evaluating p170mdr-1. J Natl Cancer Inst 1997;89:1524–9.

39. Beaulieu E, Demeule M, Pouliot JF, Averill-Bates DA, Murphy GF,Beliveau R. P-glycoprotein of blood brain barrier: cross-reactivity ofMab C219 with a 190 kDa protein in bovine and rat isolated braincapillaries. Biochim Biophys Acta 1995;1233:27–32.

40. Chan HS, Ling V. Anti-P-glycoprotein antibody C219 cross-reactivitywith c-erbB2 protein: diagnostic and clinical implications. J NatlCancer Inst 1997;89:1473–6.

122 HEIKE ET AL.