nuclear delivery of doxorubicin via folate-targeted ...enable bypassing pgp-mediated efflux. here,...

Nuclear Delivery of Doxorubicin via Folate-targeted Liposomes withBypass of Multidrug-resistance Efflux Pump1

Dorit Goren, Aviva T. Horowitz, Dina Tzemach,Mark Tarshish, Samuel Zalipsky, andAlberto Gabizon2

Sharet Institute of Oncology [D. G., A. T. H., D. T., A. G.] andInterdepartmental Unit [M. T.], Hadassah Hebrew University MedicalCenter, Jerusalem 91120, Israel, and ALZA Corporation, MountainView, California 94039 [S. Z.]

ABSTRACTFolic acid, attached to polyethyleneglycol-derivatized,

distearoyl-phosphatidylethanolamine, was used to targetinvitro liposomes to folate receptor (FR)-overexpressing tumorcells. Confocal fluorescence microscopic observations dem-onstrated binding and subsequent internalization of rho-damine-labeled liposomes by a high FR-expressing, murinelung carcinoma line (M109-HiFR cells), with inhibition byfree folic acid. Additional experiments tracking doxorubicin(DOX) fluorescence with DOX-loaded, folate-targeted lipo-somes (FTLs) indicate that liposomal DOX is rapidly inter-nalized, released in the cytoplasmic compartment, and,shortly thereafter, detected in the nucleus, the entire processlasting 1–2 h. FR-mediated cell uptake of targeted liposomalDOX into a multidrug-resistant subline of M109-HiFR cells(M109R-HiFR) was unaffected by P-glycoprotein-mediateddrug efflux, in sharp contrast to uptake of free DOX, basedon verapamil-blockade experiments with quantitation ofcell-associated DOX and flow cytometry analysis. Deliveryof DOX by FTLs to M109R-HiFR cells increased continu-ously with time of exposure, reaching higher drug concen-trations in whole cells and nuclei compared with exposure tofree DOX. The in vitro cytotoxic activity obtained with DOX-loaded FTLs was 10-fold greater than that of the nontar-geted liposome formulation, but was not improved over thatof free DOX despite the higher cellular drug levels obtainedwith the targeted liposomes in M109R-HiFR cells. However,if M109R-HiFR cells were exposed to drugsin vitro andtested in an in vivo adoptive assay for tumor growth insyngeneic mice along a 5-week time span, FTL DOX wassignificantly more tumor inhibitory than free DOX. It issuggested that the biological activity of liposomal DOX re-

leased inside the cellular compartment is reducedin vitrodue to the aggregated state of DOX, resulting from theliposome drug-loading process, and requires a long period oftime and/or an in vivo environment for full expression.

INTRODUCTIONFR3, a GPI membrane-anchored glycoprotein of 38 kDa

(1), with extremely high affinity for folate (kDa' 10210M for

a-isoform and kDa' 1029M for b-isoform; Ref. 2), is over-

expressed in a wide variety of epithelial tumors (2–5). The FRparticipates in the cellular accumulation of folates in a numberof epithelial cells through a process of endocytosis (6). In thisprocess, ligand-bound receptor is internalized and released fromthe receptor through intravesicular reduction in pH. Ligand-freereceptor is then recycled to the cell surface (6, 7). The receptor-mediated uptake of folic acid has been proposed as a potentiallyuseful target in cancer treatment (7, 8) and as a route to promoteentry of attached macromolecules or liposomes into cells (9).

Liposomes with folate residues conjugated through a lipo-some-grafted PEG spacer are taken up avidly by KB cells(human nasopharyngeal cancer cell line; Refs. 10 and 11). Thebinding of these FTLs to KB cells is mediated by cell-surfaceFR, as demonstrated by competitive inhibition with excess freefolate or with antiserum against the FR (10, 11). By providing adifferent pathway of tumor cell drug uptake, the use of targetedliposomes for chemotherapeutic delivery may conceivably cir-cumvent the MDR drug efflux mechanism, leading to resistance(12, 13). The Pgp, located in the plasma membrane, is an activeefflux pump of cytotoxic agents conferring multidrug resistanceto cancer cells (12, 13). Intracellular entry of drug-loaded lipo-somes via endocytosis, followed by release of entrapped agentin cytoplasm (11), is an alternative route of drug entry that mayenable bypassing Pgp-mediated efflux.

Here, we investigate the mechanism of drug delivery andactivity of DOX-loaded FTLs as compared with free drug anddrug encapsulated in nontargeted liposomes, and we report ondrug uptake studies and tumor cell growth assays with DOX-sensitive and -resistant, high-FR-expressing murine tumor cells.

MATERIALS AND METHODSLiposome Preparation. Preparation of liposomes and

the sources of liposome components, including DSPE-PEG(3350)-Folate, were as described previously (14). DPPE-rhodamine was obtained from Avanti Polar Lipids (Birming-ham, AL). The following formulations were prepared: (a)

Received 11/12/99; revised 2/11/00; accepted 2/14/00.The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisementin accordance with 18 U.S.C. Section 1734 solely toindicate this fact.1 Supported by the Israel Science Foundation (Jerusalem, Israel) and byALZA Corporation (Mountain View, CA).2 To whom requests for reprints should be addressed, at Sharet Instituteof Oncology, Hadassah Medical Center, Kiryat Hadassah, P.O. Box12000, Jerusalem 91120, Israel. Fax: 972-2-6430622; E-mail: [email protected].

3 The abbreviations used are: FR, folate receptor; GPI, glycosyl-phos-phatidylinositol; PEG, polyethyleneglycol; FTL, folate-targeted lipo-somes; Pgp, P-170 glycoprotein; MDR, multidrug-resistance; DOX,doxorubicin; DSPE, distearoyl-phosphatidylethanolamine; DPPE, di-palmitoyl-PE; HiFR, high expression of FR in M109 cells.

1949Vol. 6, 1949–1957, May 2000 Clinical Cancer Research

Research. on March 7, 2020. © 2000 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

http://clincancerres.aacrjournals.org/

hydrogenated soybean phosphatidylcholine/cholesterol/DSPE-PEG-Folate/DPPE-rhodamine (molar ratio, 99.4:70:0.5:0.1); (b)same lipid composition (see Ref. 1), excluding DPPE-rhoda-mine, and loaded with DOX in the liposome water phase usingan ammonium sulfate gradient (250 mM), as described previ-ously (15).

In addition, we also tested a nontargeted pegylated liposomalformulation of DOX (Doxil) provided by Sequus Pharmaceuticals(Menlo Park, CA), with a lipid composition as described previously(16). Unless otherwise indicated, the final DOX:phospholipid ratiowas;150mg/mmol. The mean vesicle size of all liposome formu-lations was in the range of 70–100 nm.

Cell Lines. The high-FR M109 cell lines, includingDOX-sensitive and -resistant variants, have been described pre-viously (14). The cells were cultured in folate-depleted (;2 nM,versus2.26 mM folate under normal culture conditions) RPMI1640 (Biological Industries, Ltd., Beyt Haemek, Israel) with10% FCS (Life Technologies, Inc., Grand Island, NY). The twoselected sublines adapted to grow under low folate conditionsare referred to as M109-HiFR, for the sensitive line, andM109R-HiFR for the resistant line.

Confocal Microscopy. Tumor cells were plated, 24 hbefore each experiment, on 22-mm coverslips inserted into35-mm culture dishes. Exposure times of the cells to liposomesor DOX are indicated for each experiment. Phospholipid con-centration and DOX concentration in these experiments were inthe range of 30–150mM and 10–40mM, respectively. Aftermedium removal, cells were washed with PBS and fixed withPBS-buffered (pH 7) 4% formaldehyde/1.5% methanol solution(Bio-Lab, Jerusalem, Israel) at 4°C for 15 min, then washedthree times with PBS (Life Technologies, Inc.). Next, the cov-erslips were put on slides coated with buffered mounting me-dium consisting of 90%glycerol/10% PBS with 0.1% NaN3 and3% DABCO (triethylenediamine; Sigma Chemical Co., St.Louis, MO), to prevent fading. To rule out any artifact caused bythe processing of microscopic samples, such as liposomal drugleakage, the DOX tracking experiments were done with fixedcells and then repeated with live cells to confirm the observa-tions. Microscopic visualization of live (nonfixed) cells wasdone in PBS containing 2 mM MgSO4/1 mM HEPES (pH 7.5;Sigma Chemical Co.). Examination was done with invertedZeiss confocal laser scanning microscope (LSM410; Carl Zeiss,Jena, Germany). For rhodamine, maximum excitation was per-formed by a 543-nm line of the internal He-Neon laser, andfluorescence emission was observed above 570 nm with long-pass barrier filter LP-570. For DOX, maximum excitation wasperformed by a 488-nm line of internal Argon laser, and fluo-rescence emission was observed above 515 nm with long-passbarrier filter LP-515. A water immersion objective, C-Apochro-mat 633 1.2 W corr. (Zeiss), was used. Images were convertedto TIFF format, and the contrast level and brightness of theimages were adjusted by using the Zeiss LSM410 program.

Verapamil Blockade of Drug Efflux. M109R-HiFRcells in monolayer were exposed to 5mM DOX as free drug orin FTLs for 1 h, in the presence or absence of 10 mM verapamil(Teva, Netanya, Israel). Cells were washed to remove nonasso-ciated drug and further incubated in the presence or absence ofverapamil for 2 h. Cells were released from tissue culture disheswith 0.05% trypsin/0.02% EDTA (Life Technologies, Inc.),

followed by PBS washing (centrifugation, 7 min, 5003 g),suspended and split into two fractions: one for cellular DOXdetermination and the other one for flow cytometry assay.

Measurement of Cell-associated DOX. After cells werecounted and pelleted by centrifugation, DOX was extracted bytreatment with 0.075N HCl in 90% isopropyl alcohol at 4°Covernight. Following centrifugation, the supernatant was col-lected and used for fluorescent DOX determination at Ex 470nm/Em 590 nm in a Kontron SFM25 spectrofluorimeter (Kon-tron, Zurich, Switzerland). Extracts from untreated cells wereused as blank. DOX concentration was expressed as ng DOX-equivalents per 106 cells. In some experiments, cell-associateddrug was also measured by high-performance liquid chromatog-raphy and fluorescence detection, as reported previously (17).

Flow Cytometry. Suspended M109R-HiFR cells, as de-scribed above, were analyzed with a FACS-star plus (BectonDickinson, Mountain View, CA) flow cytometer. Cells werepassed at a rate of;1000 cells/s through a 70-mm nozzle, usingsaline as the sheath fluid. A 488-nm argon laser beam at 250mW served as the light source for excitation. Fluorescenceemission was measured using a 575-nm DF 26 band-pass filter.

Cellular and Nuclear DOX Quantitation. M109R-HiFR cells were exposed to free DOX or FTL-DOX for 1 h and4 h. M109R-HiFR cells, released by typsinization as describedabove, were suspended at a concentration of 53 106 cells/ml for10 min at 4°C in a 100 mM NaCl solution with 1 mM EDTA, 1%Triton X-100 (Sigma Chemical Co.), and 10 mM Tris buffer (pH7.4; Sigma Chemical Co.). The suspension was then centrifuged(15 min, 8003 g), and the resulting precipitate of cell nucleiwas separated from the supernatant cell cytosol. DOX extractionfrom both fractions followed, as described before.

Cytotoxicity Studies. M109HiFR and M109R HiFRcells in folate-depleted RPMI 1640 were seeded in 96-wellplates at a density of 103 cells/well (six replicates). Twenty-fourh later, cells were exposed for 1 h to free DOX, FTL-DOX, orDoxil. The cells were washed twice and incubated further for120 h in drug-free medium. In some experiments, cells wereexposed continuously to drugs for 72 h. Cell growth was as-sayed colorimetrically using 2.5% glutaraldehyde as fixative,followed by methylene blue staining and absorbance measure-ments at 620 nm on an automated plate reader (18). Growthrates were calculated as reported previously (18).

In Vivo Adoptive Tumor Growth Assay. FemaleBALB/c mice, 10 weeks of age, were obtained from the HebrewUniversity Animal Breeding House (Jerusalem, Israel) andmaintained in a specific pathogen-free facility at Hadassah Med-ical Center. The experiments were done under ethical approvalfrom our Institutional Animal Care and Use Committee.M109R-HiFR cells inin vitro suspension (107 cells/ml) wereexposed to 10mM DOX either as free drug, Doxil, or FTL for2 h, washed with PBS to remove non-cell-associated drug, andresuspended at a concentration of 23 107 cells/ml. Healthy,syngeneic BALB/c mice received injections into the right hindfootpad with 50ml (106 cells) from the above cell suspension.The footpad thickness was measured with calipers once or twicea week for 5 weeks. After 35 days, mice were sacrificed, thefinal number of tumors was recorded, and the control andtumor-inoculated footpads were sectioned at the ankle level andweighed. Tumor weight was estimated as the difference between

1950DOX Delivery by FTLs



the weight of the normal and tumor-bearing footpad. The sta-tistical significance of differences in the final incidence oftumors per group was analyzed by contingency tables and theFisher’s exact test.

RESULTSFTL Binding to M109-HiFR Cells Is Mediated by the

FR. Association of rhodamine-labeled FTLs to M109-HiFRcells in folate-free RPMI 1640 was observed within 30 min ofexposure (Fig. 1a). At 50 min, subsequent liposome internal-

ization and accumulation in the cell cytosol was observed, asshown in Fig. 1b. However, under competitive free folate con-centrations (2 mM, equivalent to;1000-fold the concentrationof liposomal folate), liposome binding was prevented com-pletely, indicating involvement of FR in the liposome-targetingprocess at both 30- and 50-min time points (Fig. 1,c and d).Further evidence of the involvement of the GPI-anchored FR inthe interaction of FTLs with M109 cells comes from experi-ments with phosphatidylinositol-phospholipase C-treated cells:exposure of phosphatidylinositol-phospholipase C-pretreated

Fig. 1 Confocal microscopy after 30 min (aandc) and 50 min (bandd) of incubation of rhodamine-labeled FTL with M109-HiFR cells at 37°C,in the absence (aandb) or presence of 2 mM free folic acid, a 1000-fold excess over FTL folate (candd).

1951Clinical Cancer Research



M109-HiFR cells to rhodamine-labeled FTLs (1 h at 4°C),resulted in no detectable binding, whereas the same liposomeswere effectively bound by nontreated cells with no apparenttemperature interference (data not shown).

DOX-loaded FTLs Bind to and Are Internalized intoM109R-HiFR Cells. In this experiment, we followed free andliposomal DOX access to M109R-HiFR cells by tracking theintense fluorescence of DOX molecules. M109R-HiFR cellexposure to DOX as free drug or loaded in FTLs revealed thatin both cases the drug gains access to the cell nucleus, as shownin Figs. 2 and 3, but the kinetics and apparently the route ofuptake differ. The influx of free DOX through cell membranewas very rapid, as indicated by the bright cytoplasmic stainingwithin 7 min of exposure (Fig. 2a). At 30 min, the free drug wasalready completely localized in the nucleus (Fig. 2b). The ki-netics of cellular interaction with DOX-loaded FTL was remark-ably different. Liposome attachment to the cell membrane wasobserved within 30 min (Fig. 3a). By 60 min, internalization hastaken place and liposomal DOX was detected in the cytosol, andin a few cells, the drug began to appear in the nucleus (Fig. 3b).After 90 min, liposome-delivered DOX has reached the nucleusin most of the cells, whereas the cytoplasmic drug fluorescencehas mostly disappeared (Fig. 3c). In contrast to FTL, a formu-lation of nontargeted liposomes coated with PEG (known com-mercially as Doxil) showed absolutely no association withM109R cells (Fig. 3d), even after 4 h of incubation. In this case,consistent with previous observations, the drug has no access tothe cells unless leakage into extracellular medium occurs (18).This type of experiment was repeated with fresh and fixed cellswith essentially similar results.

To examine drug efflux, M109R-HiFR cells were exposedto FTL-DOX or free DOX for 1 h and then further incubated in

drug-free medium for 24 h. DOX fluorescence in cells treatedwith free drug disappeared almost completely in contrast to amarked residual fluorescence in cells treated with FTL-DOX(data not shown).

Intracellular Delivery of DOX via FTLs OvercomesPgp-mediated Drug Efflux. The activity of the Pgp effluxpump in M109RHiFR cells was examined by flow cytometry inthe rhodamine-123 efflux assay, an indicator of Pgp-mediatedresistance (19, 20), and found to be highly effective and sensi-tive to verapamil blockade (data not shown). In the next seriesof experiments, we tested the efflux of free DOX and FTL DOXand its sensitivity to verapamil by flow cytometry. In agreementwith the rhodamine studies, flow cytometry analysis showed ashift in the curve, indicating a clear increase in cell fluorescencein M109R-HiFR cells after 1 h exposure to free DOX in thepresence of verapamil (Fig. 4A). In contrast, the cellular level offluorescence in M109R-HiFR cells after a 1-h exposure to drugin FTL appears identical in presence or absence of verapamil(Fig. 4B).

The flow cytometry observations were confirmed by quan-titative fluorometry of DOX from cell extracts, a method that isrelatively unaffected by quenching artifacts. Cell retention offree DOX was;4-fold higher in the presence of verapamil,whereas identical drug levels accumulated in cells exposed toliposome-targeted drug in the presence or absence of verapamil(Table 1). Furthermore, when the cellular DOX levels in theabsence of verapamil are compared, a 4–6-fold advantage fortargeted liposomal DOX over free DOX is noticeable (Table 1).These results indicate that free drug diffusing into the cells ispumped out by Pgp and/or other proteins sensitive to verapamilblockade, whereas drug entry via liposome internalization

Fig. 2 Confocal microscopy of free DOX uptake by M109R-HiFR cells cultured in folate-depleted RPMI 1640 at 37°C. Incubation times: 7 min (a)and 30 min (b).




provides an alternative delivery pathway not controlled by theMDR efflux machinery.

Additional experiments quantifying cell-associated drug indrug-sensitive and drug- resistant M109 cells (Table 2) confirm theenhancement of drug delivery with targeted liposomes to the re-sistant cells at 1 and 4 h of incubation. In the case of drug-sensitivecells (M109-HiFR), there was no advantage of FTL over free DOX(Table 2). Drug uptake was negligible when a nontargeted, liposo-mal DOX formulation (Doxil) was used (Table 2). The dependenceof drug uptake on the number of FR-bound and internalized lipo-

somes is supported by an experiment presented in Fig. 5 examiningtwo formulations of FTL that differ only in their drug:lipid ratio bya factor of;10. Cellular levels of DOX were consistently higherby a similar factor of;10, in favor of the high drug-to-lipidformulation. The progressive accumulation of targeted liposomaldrug by tumor cells along a 24-h incubation period, without evi-dence of efflux or plateau level, is also apparent from the curves inFig. 5.

Nuclear Delivery of FTL DOX. To investigate quanti-tatively the delivery of targeted liposomal DOX to the nucleus,

Fig. 3 Confocal microscopy of liposomal DOX uptake by M109R-HiFR cells cultured in folate-depleted RPMI 1640 at 37°C.a, b, andc, binding,internalization, and nuclear localization of DOX delivered by FTL after a 30-min, 60-min, and 90-min incubation, respectively.d, cells exposed for4 h to nontargeted PEG-coated liposomes encapsulating DOX (Doxil).




cell fractionation experiments were done. As seen in Fig. 6,most of the drug is found in the nuclear fraction with both freeDOX and FTL DOX already after 1 h of incubation. The nucleardrug concentration obtained with FTL DOX clearly surpassesthe concentration obtained with free DOX, especially after 4 hof incubation. In fact, the drug concentration in cells treatedwith free DOX did not increase at all when incubation wasprolonged from 1 h to 4 h. In contrast, in the case of cells treatedwith FTL-DOX, there was a 2-fold increase during the sameperiod. No significant amounts of metabolites were detected

after a 1–4-h exposure of tumor cells to either free DOX orFTL-DOX by high-performance liquid chromatography analysisof cell-associated drug (data not shown), indicating that FTLdelivers intact drug to the nucleus.

In Vitro Cytotoxic Activity of DOX Delivered by FTL,Nontargeted Liposomes, or as Free Drug. We exploredwhether delivery of DOX via FTL would increase the drug

Fig. 4 Flow cytometry of M109R-HiFR cells exposed for 1 h to 10 mM

DOX as free drug (A) or encapsulated in FTL (B) in the presence orabsence of 10 mM verapamil. X axis, fluorescence intensity; Y axis, cellnumber; black line, no verapamil;gray line, verapamil added. In thecase of free DOX (A), the curve shifts to the right, pointing to anincrease in the amount of drug retained in the cells in the presence ofverapamil, whereas in the case of FTL-DOX (B) no such shift occurs.

Fig. 5 Drug accumulation by M109R-HiFR cells exposed to high andlow drug:lipid ratio FTL. DOX concentration, 10mM. The DOX:phos-pholipid ratios of the high and low ratio preparations were, respectively,137.6mg/mmol and 11.3mg/mmol.

Table 1 DOX levels in M109R-HiFR cells (ng/106 cells) in thepresence or absence of verapamila

Exposure time (min) 30 60Free-DOX 166 1 286 3Free-DOX1 verapamil 606 4 1126 3FTL-DOX 1046 4 1336 12FTL-DOX 1 verapamil 1016 10 1246 6a M109R-HiFR cells in monolayer (24-multiwell plates) were ex-

posed for 30 min or 1 h at 37°C to 5mM (2.9 mg/ml) free DOX orFTL-DOX in the presence or absence of 10 mM verapamil. Thereafter,cells were washed and incubated further for 2 h in the presence orabsence of 10 mM verapamil.

Table 2 DOX accumulation in cells (ng/106 cells) exposed to freeDOX, nontargeted liposomal DOX (Doxil), and targeted liposomal

DOX (FTL-DOX)a

Cell tested M109-HiFR M109R-HiFR M109R-HiFR

Exposure time 1 h 1 h 4 hFree DOX 1536 3 816 8 1106 11Doxil 5.26 0.3 1.36 0.1 Not doneFTL-DOX 1426 2 1806 10 2856 8

a M109R-HiFR and M109-HiFR cells in monolayer (24-multiwellplates) were exposed to 10mM DOX at 37°C. The drug concentration inM109R-HiFR cells treated with FTL-DOX was significantly greaterthan with free DOX (P, 0.0001,t test).




cytotoxicity against FR-overexpressing cells. M109-HiFR cellswere exposed to free or liposomal DOX for 1 h, then washedand further incubated for 120 h in fresh medium. As seen in Fig.7A, the growth inhibition curve of DOX in FTL was similar tothat of free DOX, but clearly superior (10-fold drop in IC50) tothat of DOX in nontargeted liposomes, stressing the key role ofliposome binding and internalization in enhancement of cyto-toxic activity.

A similar cytotoxicity assay was done using the MDRsubline M109R-HiFR (Fig. 7B). Once more, a clear enhance-ment of cytotoxicity was obtained when the folate-targetedformulation is compared with the nontargeted formulation.However, despite higher drug levels accumulating in cells ex-posed to the targeted liposomal drug (Tables 1 and 2), this typeof assay did not reveal increased cytotoxic activity of the tar-geted preparation compared with free DOX. Cytotoxicity assaysusing continuous exposure to drug (72 h) pointed to similarresults in terms of the relative activity of the different forms(targeted, nontargeted, free) of DOX delivery.

Superior Tumor-inhibitory Activity of DOX Deliveredby FTL in an in Vivo Adoptive Tumor Growth Assay. Toexamine the biological activity of drug delivered by FTL inanother model, we exposed M109R-HiFR cellsin vitro to thetest drug and thereafter inoculated them into the mouse footpad.In this way, the growth of cells is tracked along a much longertime span than inin vitro experiments, and the influence ofinvivomicro-environmental factors is brought into play. However,unlike therapeutic experiments, this type ofin vivo adoptiveassay is unaffected by pharmacokinetic and biodistribution fac-tors that would have complicated the interpretation of results.The results (Table 3) point to a statistically significant decreaseof the number of tumor takes in mice injected with tumor cellsexposed to FTL, as compared with free DOX, Doxil, and con-trol, after 5 weeks follow-up. Tumor weights were also smallerfor the FTL group (Table 3). The kinetics of tumor growth, asestimated by the mean footpad tumor thickness, shows a clearlyslower growth rate for the FTL group from the second weekafter inoculation over the other treatment groups (Fig. 8).

DISCUSSIONRecent reports have indicated the feasibility of using fo-

late-conjugated liposomes to target drugs to cancer cells and to

augment theirin vitro efficacy (10, 11). Folate-targeted systemspossess tumor cell specificity due to the frequent overexpressionof the FR in human carcinomas. Folates and its conjugates entercancer cells by FR-mediated endocytosis (6, 7). The normalpermeability barriers that limit drug entry into cells are by-passed, allowing even macromolecules or liposomes to enter

Fig. 6 Accumulation of DOX in nuclei and cytosol of M109R-HiFRcells after exposure to free and FTL DOX. M109R-HiFR cells wereexposed to 10mM DOX for 1 and 4 h.

Fig. 7 Cytotoxicity of DOX delivered as free drug, in nontargetedliposomes (Doxil) or in FTL against M109-HiFR cells (A) andM109R-HiFR cells (B) exposed for 1 h to the drug, washed, andfurther incubated for 120 h in drug-free medium.E, free DOX-treated;M, FTL-DOX-treated;Œ, Doxil-treated. Note the significantshift to the right (i.e.,decreased toxicity) of all of the growth curvesin B as compared withA, as a result of the greater resistance ofM109R-HiFR cells to DOX. Each point consists of six replicates;SDs do not exceed6 15%.

Table 3 In vivoadoptive tumor growth assay in mice inoculatedwith M109R-HiFR tumor cells exposedin vitro to free DOX, FTL-

DOX, and Doxila

Celltreatment

Final tumorincidenceb (%)

Tumor weight (mg)median (range)

Untreated 13/20 (65%) 381 (48–825)Free DOX 8/19 (42%) 239 (32–683)Doxil 10/19 (53%) 397 (13–512)FTL-DOX 2/20 (10%) 57 (27–87)a Results of two experiments. DOX concentrationin vitro, 10 mM.b Fisher’s exact test: FTL-DOXversusuntreated,P 5 0.0008;

FTL-DOX versusDoxil, P 5 0.0057; FTL-DOXversusfree DOX,P 50.0310. All other comparisons, not significant.




FR-bearing cells readily (7). Folate conjugates such as proteinsand nucleic acids have been observed to remain intact for hoursfollowing uptake by cancer cells (9, 21). Apparently, vitamin-mediated delivery may constitute a more protected pathway forintracellular delivery.

We have examined the likelihood that targeted liposomeswill be taken up by carcinoma cells with overexpressed FR in anin vitro system. In this study and a previous study (14), thefolate residue, anchored to the liposome bilayer by itsg-car-boxyl group through a PEG-DSPE linker, was found to functionas a specific and efficient targeting agent. The involvement ofcell surface FR in the binding and internalization of FTL intoM109-HiFR was validated by inhibition of association anduptake of these liposomes in the presence of excess free folateand, by the loss of liposome binding due to enzymatic cleavageof GPI, the FR cell membrane anchor (1). Although solublePEG-folate seems to have less affinity than free folate for theFR, liposomal PEG-folate actually has an increased bindingaffinity due to the multivalency of liposomal binding (14).

In contrast to rhodamine-PE, a liposome bilayer componentthat localizes in the cytoplasm, FTL DOX readily gains accessto the cell nucleus after internalization. The small size of thepores of the nuclear membrane makes this compartment inac-cessible to liposomes. Therefore, this observation suggests thatduring the process of vesicular docking on cell surface and/orsubsequent internalization path, destabilization of the liposomalcarrier occurs, resulting in release of entrapped DOX and accu-mulation of the drug in the nucleus. It is unclear what triggersdrug release from endocytosed liposomes. The acidic milieu ofthe lysosomal or endosomal vesicles (6, 22) is unlikely to havean effect because the liposomes used here and the gradientretaining the drug are stable at pH 4–5 within the relevant timeframe (23). However, loss of cholesterol or bilayer damage byenzymatic activity of phospholipases will destroy the protongradient of liposomes and lead to DOX leakage.

Pgp is known to reduce cellular drug accumulation byacting as an ATP-dependent efflux pump of a great manystructurally distinct hydrophobic compounds (12). The mannerin which Pgp recognizes these different substrates is unknown.

Pgp has been found in the membrane, Golgi apparatus, andnucleus (24, 25). In the M109R-HiFR cell line, confocal mi-croscopy studies, together with fluorometric measurements ofcell-associated drug, indicate that DOX readily enters the celland concentrates in the nucleus, but is subsequently excretedvery effectively with no observable nuclear fluorescence left at24 h. The efflux of drug is blocked by verapamil, indicating thatit is Pgp mediated (19). Removal of DOX from the nucleus ispresumably mediated by nuclear Pgp (24) or other proteinsinvolved in the nuclear-cytoplasmic trafficking and compart-mentalization of drugs (26). A notable finding of this study isthe failure of the efflux mechanism to pump out and reduce therate of accumulation of DOX when delivered by FTLs. Themechanism for this is unclear because the liposome-targeteddrug is not retained in endosomal vesicles, but rather concen-trates effectively in the nucleus, albeit with a slower kineticsthan free drug. It is likely that part of the liposomal drug is in adifferent physical form than free drug. There is evidence ofintraliposomal precipitation of DOX in liposomes loaded via anammonium sulfate-generated gradient (27, 28). Aggregation ormolecular stacking due to self-association of the intracellularliposomal drug is a possibility, because DOX dimerization hasbeen observed for concentrations greater than 10mM (28–30).This would explain the inability of cells to pump out liposomalDOX, and, at the same time, the lack of a cytotoxic advantagefor folate-targeted liposomal DOX over free DOX in thein vitrotests, despite higher drug levels accumulating in cells exposed tothe former. Aggregation and precipitation of DOX after lipo-some encapsulation using the ammonium sulfate method is afully reversible process, with complete resolubilization and res-toration of drug activity on collapse of the gradient in extracel-lular medium (18, 28). However, the high intracellular concen-tration of targeted liposomal DOX and short time scale ofinvitro cytotoxicity experiments may not enable this process to becompleted for full biological expression of the drug activity. Incontrast, the results of thein vivo adoptive assay of tumorgrowth give a clear indication that targeted liposomal DOX hasa greater tumor-inhibitory potential than free DOX in line withthe levels of drug delivered to tumor cells. One factor that may

Fig. 8 In vivo adoptive assay of tumor growth.M109R-HiFR cells (106/mouse) were inoculatedinto the right hind mouse footpad after priorinvitro exposure to 10mM free DOX, Doxil, orFTL-DOX for 2 h. The mean size of tumor-inoc-ulated footpad along time of observation is shown.




account for the relatively higher biological activity of FTLinvivo is the lag period of growth commonly observed afterin vivotumor implantation and the long time span of the experiment(.30 days), which may enable disaggregation and full bioavail-ability of intracellular drug before rapid cell growth occurs.

In conclusion, our results support the proposition that FTLoffers an attractive means of delivering DOX into tumor cells,which is insensitive to Pgp-mediated drug efflux and moreeffective than free DOX and nontargeted liposomal DOX. FR-mediated drug delivery has the potential to circumvent multi-drug resistance and may be especially useful if it can be coupledwith the in vivo pharmacological advantages of long-circulatingliposomal delivery systems, such as stable drug retention andtumor accumulation.

ACKNOWLEDGMENTSWe thank D. Rund, A. Taraboulos, and Y. Barenholz for fruitful

discussion, and O. Drize for technical assistance.

REFERENCES1. Rothberg, K. G., Ying Y. S., Kolhouse, F. J., Kamen, B. A., andAnderson, R. G. W. The glycophospholipid-linked folate receptor in-ternalizes folate without entering the clathrin-coated pit endocytoticpathway. J. Cell Biol.,110: 637–649, 1990.2. Gates, S. B., Mendelsohn, L. G., Shackelford, K. A., Habeck, L. L.,Kursar, J. D., Gossett, L. S., Worzalla, J. F., Shih, C., and Grindey, G. B.Characterization of folate receptor from normal and neoplastic murinetissue: influence of dietary folate on folate receptor expression. Clin.Cancer Res.,2: 1135–1141, 1996.3. Weitman, S. D., Lark, R. H., Coney, L. R., Fort, D. W., Frasca, V.,Zurawski, V. R., Jr., and Kamne, B. A. Distribution of the folatereceptor GP38 in normal and malignant cell lines and tissues. CancerRes.,52: 3396–3401, 1992.4. Ross, J. F., Chaudhuri, P. K., and Ratnam, M. Differential regulationof folate receptor isoforms in normal and malignant tissues in vivo andin established cell lines. Cancer (Phila.),73: 2432–2443, 1994.5. Mantovani, L. T., Miotti, S., and Menard, S. Folate binding proteindistribution in normal tissues and biological fluids from ovarian carci-noma patients as detected by the monoclonal antibodies MOv 18 andMOv19. Eur. J. Cancer,30A: 363–369, 1994.6. Anderson, R. G. W., Kamen, B. A., Rothberg, K. G., and Lacy, S. W.Potocytosis sequestration and transport of small molecules. Science(Washington DC),255: 410–411, 1992.7. Wang, S., and Low, P. S. Folate mediated targeting of anti-neoplasticdrugs, imaging agents, and nucleic acids to cancer cells. J. ControlledRelease,53: 39–48, 1998.8. Kranz, D. M., Patrick, T. A., Brigle, K. E., Sprinella, M. J., and Roy,E. J. Conjugates of folates and anti-T-cell receptor antibodies specifi-cally target folate-receptor-positive tumor cells for lysis. Proc. Natl.Acad. Sci. USA,92: 9057–9061, 1995.9. Leamon, P. C., and Low, S. P. Delivery of macromolecules intoliving cells: a method that exploits folate receptor endocytosis. Proc.Natl. Acad. Sci. USA,88: 5527–5576, 1991.10. Lee, R. J., and Low, P. S. Delivery of liposomes into cultured KBcells via folate receptor-mediated endocytosis. J. Biol. Chem.,269:3198–3204, 1994.11. Lee, R. J., and Low, P. S. Folate-mediated tumor cell targeting ofliposome-entrapped doxorubicinin vitro. Biochim. Biophys. Acta,1233: 134–144, 1995.12. Gottesman, M. M., Pastan, I., and Ambudkar, S. V. P-glycoproteinand multi drug resistance. Curr. Opin. Genet. Dev.,6: 610–617, 1996.13. Dey, S., Ramachandra, M., Pastan, I., Gottesman, M. M., andAmbudkar, S. V. Evidence of two nonidentical drug resistance sites in

the human P-glycoprotein. Proc. Natl. Acad. Sci. USA,97: 10594–10599, 1997.

14. Gabizon, A., Horowitz, A. T., Goren, D., Tzemach, D., Mandel-baum-Shavit, F., Qazen, M. M., and Zalipsky, S. Targeting folatereceptor with folate linked to the extremities of poly(ethylene glycol)-grafted liposomes:in vitro studies. Bioconjugate Chem.,10: 289–298,1999.

15. Haran, G., Cohen, R., Bar L. K., and Barenholz, Y. Transmembraneammonium sulfate gradients in liposomes produce efficient and stableentrapment of amphipathic weak bases. Biochim. Biophys. Acta,1151:201–215, 1993.

16. Working, P. K., and Dayan, A. D. Pharmacological-toxicologicalexpert report. Caelyx. Hum. Exp. Toxicol.,15: 751–785, 1996.

17. Goren, D., Horowitz, A. T., Zalipsky, S., Woodle, M. C., Yarden,Y., and Gabizon, A. Targeting of stealth liposomes to erbB-2 (Her/2)receptor:in vitro and in vivo studies. Br. J. Cancer,74: 1749–1756,1996.

18. Horowitz, A. T., Barenholz, Y., and Gabizon, A.In vitro cytotox-icity of liposome-encapsulated doxorubicin: dependence on liposomecomposition and drug release. Biochim. Biophys. Acta,1109:203–209,1992.

19. Lamy, T., Drenou, B., Grulois, I., Fardel, O., Jacquelinet, C.,Goasguen, J., Dauriac, C., Amiot, L., Bernard, M., and Fauchet, R.Multi-drug resistance (MDR) activity in acute leukemia determined byrhodamine 123 efflux assay. Leukemia (Baltimore),9: 1549–1555,1995.

20. Webb, M., Raphael, C. L., Asbahr, H., Erber, W. N., and Meyer,B. F. The detection of rhodamine 123 efflux at low levels of drugresistance. Br. J. Haematol,93: 650–655, 1996.

21. Wang, S., Lee, R. J., Cauchon, G., Gorenstein, D. G., and Low, P. S.Delivery of antisense oligonucleotides against the human epidermalgrowth factor receptor into cultured KB cells with liposomes conjugatedto folate via polyethyleneglycol. Proc. Natl. Acad. Sci. USA,92: 3318–3322, 1995.

22. Lee, R. J., Wang, S., and Low, P. S. Measurement of endosome pHfollowing folate receptor-mediated endocytosis. Biochim. Biophys.Acta, 1312: 237–242, 1996.

23. Barenholz, Y., Amselem, S., Cohen, R., Gelvan, D., Samuni, A.,Golden, E. B., and Gabizon, A. Stability of liposomal doxorubicinformulations: problems and prospects. Med. Res. Rev.,13: 449–491,1993.

24. Baldini, N., Scotlandi, K., Serra, M., Shikita, T., Zini, N., Ognibene,A., Santi, S., Ferracini, R., and Maraldi, N. M. Nuclear immunolocal-ization of P-glycoprotein in multidrug-resistant cell lines showing sim-ilar mechanism of doxorubicin distribution. Eur. J. Cell Biol.,68:226–239, 1995.

25. Molinari, A., Calcabrini, A., Meschini, S., Stringaro, A., Del Bu-falo, D., Cianfriglia, M., and Arancia, G. Detection of P-glycoprotein inthe Golgi apparatus of drug-untreated human melanoma cells. Int. J.Cancer,75: 885–893, 1998.

26. Bradshaw, D. M., and Arceci, R. J. Clinical relevance of transmem-brane drug efflux as a mechanism of multidrug resistance. J. Clin.Oncol.,16: 3674–3690, 1998.

27. Lasic, D. D., Fredrik, P. M., Stuart, M. C. A., Barenholz, Y., andMcIntosh, T. J. Gelation of liposome interior: a novel method for drugencapsulation. FEBS Lett.,312: 255–258, 1992.

28. Li, X., Hirsh, D. J., Cabral-Lilly, D., Zirkel, A., Gruner, S. M.,Janoff, A. S., and Perkins, W. R. Doxorubicin physical state in solutionand inside liposomes loaded via a pH gradient. Biochim. Biophys. Acta,1415: 23–40, 1998.

29. Eksborg, S. Extraction of daunorubicin and doxorubicin and theirhydroxyl metabolites: self-association in aqueous solution. J. Pharm.Sci., 67: 782–785, 1978.

30. Mennozi, M., Valentini, L., Vannini E., and Arcamone, F. Self-association of doxorubicin and related compounds in aqueous solution.J. Pharm. Sci.,73: 766–770, 1984.




2000;6:1949-1957. Clin Cancer Res Dorit Goren, Aviva T. Horowitz, Dina Tzemach, et al. with Bypass of Multidrug-resistance Efflux PumpNuclear Delivery of Doxorubicin via Folate-targeted Liposomes

Updated version

http://clincancerres.aacrjournals.org/content/6/5/1949

Access the most recent version of this article at:

Cited articles

http://clincancerres.aacrjournals.org/content/6/5/1949.full#ref-list-1

This article cites 28 articles, 8 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/6/5/1949.full#related-urls

This article has been cited by 21 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://clincancerres.aacrjournals.org/content/6/5/1949To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/content/6/5/1949.full#ref-list-1

http://clincancerres.aacrjournals.org/content/6/5/1949.full#related-urls

http://clincancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



nuclear delivery of doxorubicin via folate-targeted ...enable bypassing pgp-mediated efflux. here,...

Documents