Effect of Liposome Composition and Other Factors on the … · thin-layer chromatography....

(CANCER RESEARCH SO.6371-6378. October I. 1990]

Effect of Liposome Composition and Other Factors on the Targeting of Liposomesto Experimental Tumors: Biodistribution and Imaging Studies1

Alberto Gabizon,2 David C. Price, John Huberty, Robert S. Bresalier, and Demetrios PapahadjopoulosCancer Research Institute Â¡A.(j., I). P.] and Department of Radiology, [D. C. P., J. H.J, L'nirersity of California, San Francisco, California 9414}; (iastroinlestinalResearch Laboratories, I eteram Administration Medical Center, and Department of Medicine, I 'nirersity of California, San Francisco, California 94121 [R. S. B.J; and

Liposome Technology Inc., Mento Park, California 94025 Â¡A.CiJ

ABSTRACT

We have examined the distribution of radiolabeled liposomes in tumor-bearing mice after i.v. injection. Two mouse tumors (B16 melanoma,J6456 lymphoma) and a human tumor (LS174T colon carcinoma) inoculated i.m., S.C.,or in the hind footpad were used in these studies. Whenvarious liposome compositions with a mean vesicle diameter of ~ 100 nmwere compared using a radiolabel of gallium-67-deferoxamine, optimaltumor localization was obtained with liposomes containing a phosphati-dylcholine of high phase-transition temperature and a small molar fraction of monosialoganglioside or hydrogenated phosphatid)linositol (HPI).At 24 h after injection, average values of tumor uptake higher than 10%of the injected dose per g and liver-to-tumor ratios close to 1 werereproducibly obtained. Increasing the molar fraction of HPI from 9% to41% of the total phospholipid resulted in enhancement of liver uptakeand decrease of tumor uptake. Methodological aspects that influencevesicle size appear to affect significantly liposome localization in thetumor. However, varying the phospholipid dose within a 10-fold rangecaused only minor changes in the percent of injected dose recovered inthe tumor. A high uptake by tumors was also observed using otherradiolabels Â¡|'II|imiliuand indium-111-labeled bleomycin ("'In-Bleo)| inmonosialoganglioside- and HPI-containing liposomes. In the case of '"In-

Bleo, encapsulation in liposomes resulted in ~20- to 40-fold increase intumor accumulation of the radiolabel at 24 h after injection. The markedlocalization of liposomes in the mouse footpad inoculated with tumor asopposed to the contralateral mock-injected footpad was also documentedby imaging experiments with gallium-67-deferoxamine and '"In-Bleo-labeled liposomes. These results support the contention that some gly-colipid-containing liposomes previously shown to have long circulatinghalf-lives accumulate significantly in a variety of tumors and are promising tools for the delivery of anti-tumor agents.

INTRODUCTION

Liposomes have been used as carriers of cytotoxic drugs witha strategy based on reduction of toxicity and/or passive deliveryto liver-infiltrating tumors (1, 2). The fast and dominant uptakeof liposomes by the RES' (3, 4) has prevented so far the

adoption of a more direct strategy based on selective homingof liposomes to tumors. Targeting of liposomes to tumorsrequires, most importantly, a prolonged circulation half-life ofintact liposomes which, in turn, depends on a reduction of therate of clearance by the RES. In addition, there is a need tominimize the leakage of liposome contents during their prolonged stay in the blood stream. Recent developments haveshown that the inclusion of some glycolipids in the liposomebilayer coupled with phospholipids of high phase-transition

Received 3/5/90: accepted 6/15/90.The costs of publication of this article were defrayed in part by the payment

of page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1Work supported until November 1987 by a grant from the National CancerInstitute (CA-35340) and subsequently by Liposome Technology. Inc.. and theAmerican Cancer Society.

1To whom requests for reprints should be addressed, at Department of

Oncology. Hadassah Medical Center. P.O. Box 12.000. Jerusalem. Israel 91120.*The abbreviations used are: Bleo. bleomycin; Ch. cholesterol; DF, deferox-

amine: DSPC. distearoylphosphatidylcholine: GM|. monosialoganglioside; HPI,hydrogenated phosphatidylinositol: PC. phosphatidylcholine: PC. phosphatidyl-glycerol; RES. reticulocndothelial system; c'cID/g. percentage of injected dose per

g.

temperature, cholesterol, and careful size control result in inhibition of RES uptake with concomitant enhancement oftumor uptake (5).

In this report, we describe tissue distribution and imagingstudies with transplantable mouse and human tumor modelsusing 3 different radiolabels of liposomes. The findings hereindicate that the concentration of liposome-encapsulated radio-labels in tumors is well above that of most other tissues andapproximates the values obtained in the liver. The evidencegathered strongly supports the validity of a direct tumor targeting approach with liposomes.

MATERIALS AND METHODS

Preparation of Liposomes. The sources of materials used in this study-

are the same as reported previously (5). The main fatty acid componentsof the hydrogenated phospholipids used were 66% stÃ©arateand 33%palmitate for phosphatidylinositol, and 85% stÃ©arateand 15% palmitatefor phosphatidylcholine. Liposomes were prepared by thin lipid filmhydration followed by repeated extrusions through polycarbonate membranes of defined pore size (0.05 urn) as described previously (5). Forhydration, we used isotonic solutions of saline (pH range, 6.0-7.0) witheither DF (20-25 mivi), Bleo (5-15 units/ml), or ['Hjinulin (250 MCi/3

ml). The mean vesicle size obtained with this methodology was in therange of 70 to 120 nm with a SD not larger than 30% of the Gaussianmean as measured by dynamic laser scattering. In some cases, whereindicated, liposomes were prepared by solvent injection (6) in thefollowing way: lipids were weighed and dissolved in a 7:3 mixture ofethanohdimethyl sulfoxide: the lipid solution was warmed to 60Â°Cand

injected through a 21- to 23-gauge needle, at a rate of =10 ml/min,into the bulk water phase (consisting of a isotonic solution of NaCl and20 mM DF), which was kept at 60Â°Cand constantly stirred up; the

organic solvents were diluted to 15% of the total volume, thus enablingliposome formation. Ethanol, dimethyl sulfoxide, and unencapsulatedDF were removed by repeated cycles of cartride dialysis (Diaflo hollow-fiber cartridge: Amicon, Danvers, MA) against isotonic saline or glucose 5%. Liposomes were concentrated =5-fold during the last dialysisstep. Liposomes were used either without any further manipulations orafter additional extrusion through 0.05 ^m-pore polycarbonate membranes and Sephadex G-75 gel filtration to remove any DF released bythe extrusion procedure.

Phospholipid concentration was determined by a phosphate assay(7, 8). Liposomes were stored either under argon or in vacuum-sealedtubes at 5Â°Cand tested within 1 month after preparation.

Radiolabeling of Liposomes. The method of labeling preformed liposomes containing DF with 67Ga citrate has been described in detail

(9). Labeling preformed liposomes containing Bleo (Bristol-Myers,Syracuse, NY) with '"InCI (New England Nuclear, Boston. MA) was

done following a similar technique. Briefly, 20 n\ of a 1 mg/ml solutionof oxine in ethanol or oxine hemisulfate in saline was added to analiquot of 100 to 200 jiCi of '"InCI and incubated at 60T for 15 min.Bleo-containing liposome suspensions (5-15 (<mnl phospholipid/ml)were incubated for 1 h at 60Â°Cwith 2 /Â¿Ci'"In-oxine per Â¿irnolphospholipid. Removal of unencapsulated '"In-oxine was achieved bypassing the liposome suspension through an anion-exchange resin(AG 1X2, acetate form, 200-400 mesh size; Bio-Rad, Emeryville, CA)in the same way as done for 67Ga-oxine (9). This method resulted in>80% encapsulation of '"In in Bleo-containing liposomes. Previouswork had shown that '"In is translocated from the '"In-oxine complex

6371

Association for Cancer Research. by guest on August 23, 2020. Copyright 1990 Americanhttps://bloodcancerdiscov.aacrjournals.orgDownloaded from

https://bloodcancerdiscov.aacrjournals.org

LIPOSOME TARGETING TO TUMORS

to Bleo with an efficiency close to 100% as assessed by separation onthin-layer chromatography.

Tritium-labeled inulin (New England Nuclear) was encapsulated in

liposomes by adding it to the hydration buffer (physiological saline).The rest of the procedure was as described above for the thin lipid film/extrusion method (5); 2.8% of the initial amount of [3H]inulin wasentrapped in GMi-DSPC-Ch liposomes.

Animals and Tumors. Age-matched C57BL/6 and BALB/c inbredfemale mice from Simonsen Laboratories (Gilroy, CA) and NCR-NUathymic nude outbred female mice from the National Cancer Institute(Frederick, Ml)) were used in these studies. The mouse B16 melanomawas inoculated either i.m. into the hind leg (IO5 cells) or into the hindfootpad (2.5 x 10" cells) of syngeneic C57BL/6 mice. The mouse J6456lymphoma (10) was inoculated into the hind leg (10'' cells) of syngeneic

BALB/c mice. The human LS174T colon carcinoma (11) was transplanted by s.c. injection of IO7cells into the flank of nude mice. Cellswere injected in serum-free RPMI-1640 medium (GIBCO, New York,NY).

Biodistribution Studies. Radiolabeled liposomes were injected i.v.into the tail vein of tumor-bearing mice. Tumor weight was in the rangeof 0.5 to 2.0 g for mice bearing i.m. or s.c. tumors, and 0.1 to 0.4 g formice bearing footpad tumors. Retro-orbital bleeding (=1 ml) underether anesthesia followed by complete dissection were done in all micegiven injections of "Ga- or "'In-labeled liposomes. Organs were

weighed and their radioactivity quantitated in a gamma counter usingintegral counting above 10 keV. Blood volume and correction factorsfor the blood content of various tissues were determined in age- andsex-matched mice with '"In oxine-labeled syngeneic erythrocytes as

described previously (5). The results are expressed as %ID/g. Statisticalsignificance of the differences between the %ID/g means was analyzedusing the 2-sided Student's / test.

In the case of ['H]inulin-labeled liposomes, blood, liver, and tumorwere examined for radioactivity content in a beta counter after solubil-ization with Protosol (New England Nuclear). Quenching correctionfactors were obtained by measuring the radioactivity of blank tissuesfrom noninjected mice spiked with known amounts of | 'I Ijimilin.

Imaging Studies. Imaging was performed with a Pho-Gamma IVSiemens scintillation camera with pinhole collimator at 13-inch distance from the animal lying in anterior view. Mice were anesthesizedby i.p. injection of ketamine before each picture. The picture exposuretime was 15 min. Images were digitized and stored in a DEC 11/34computer using Gamma-II software for region of interest quantitativeanalysis.

RESULTS

Effect of Liposome Composition on the Biodistribution of 67Ga-

labeled Liposomes in Mice Bearing B16 Melanoma

The first step of this study was to examine in tumor-bearingmice the tissue distribution of various liposome preparations,different in composition but similar in size. For this purpose, 5types of liposomes were tested in mice bearing B16 melanomausing the 67Ga-DF radiolabel. Mice were sacrificed 24 h after

injection on the basis of previous experience indicating thatmaximal tumor accumulation occurs around this time point(5). The optimal formulations for further studies on tumortargeting were selected on the following bases: (a) highestabsolute values of tumor uptake as reflected in the %ID/gtumor; (b) lowest liver/blood and liver/tumor uptake ratios;and (c) highest tumor/carcass uptake ratio. As seen in Table 1,the most favorable results were obtained with HPI- and GM,-containing formulations when considering both the absolutetumor uptake and the tissue uptake ratios. Injection of the free67Ga-DF label resulted in only 0.1% ID/g tumor (see Table 1,

footnote c). As expected, liposomes with fully saturated fattyacyl chains showed a markedly improved distribution patternover those with unsaturated fatty acyl chains despite having the

same phospholipid head group (dipalmitoylphosphatidylgly-cerol-DSPC versus PG-PC). Liposomes containing phospho-lipids of widely different phase transition temperature (PG andDSPC) resulted in low tissue recoveries probably as a result ofphase separation and leakage of the label (5). Of note, skinuptake was several-fold higher in mice given injections of liposomes with extended circulation time and high tumor uptake.No such increase was observed in kidneys, intestine, and carcass. The highest uptake values per g tissue were seen in thespleen. However, given its small weight, the absolute uptake ofspleen was still below that of liver and skin for most formulations tested.

Biodistribution of (J\I,-containing Liposomes in Nude MiceBearing the LS174T Human Colon Carcinoma

It was important to examine whether a high tumor uptakewould also be observed in a human tumor model. For thispurpose, we tested one of the optimal formulations (GM,-DSPC-Ch) in nude mice bearing a human colon carcinoma,LS174T. The results shown in Table 2 point at marked tumorand skin uptake of liposomes, with an average value 5.3-foldhigher than carcass uptake, exceeded only by liver and spleen.

Biodistribution of |3H]Inulin-labeled GM|-containing Liposomesin Tumor-bearing Mice

The results presented in Tables 1 and 2 were obtained withthe 67Ga-DF label. To confirm these findings with an alternative

labeling method, we injected GM^DSPC-Ch liposomes withencapsulated [3H]inulin in C57BL/6 mice bearing the B16

melanoma and in nude mice bearing the LS174T tumor andexamined the label distribution in blood, liver, and tumor. Asseen in Table 3, the values obtained are of similar magnitudeto those observed when the same liposomes were labeled with67Ga-DF. Liver-to-tumor ratios as low as 2 to 3 were obtained.

Tissue Distribution Studies with HPI-containing Liposomes inMice Bearing the B16 Melanoma

To assess the contribution of HPI to the liposome distribution pattern in tumor-bearing mice, we tested 3 preparationscontaining increasing molar fractions HPI and 1 preparationcontaining no HPI, while the size and the other liposomecomponents were kept the same. The results, presented in Fig.1, indicate that the most favorable pattern of distribution wasobtained with liposomes containing a phospholipid fraction ofHPI of 9 to 23%. Liposomes with a higher HPI content showeda remarkable enhancement of liver uptake and minimal uptakeby the tumor. The effect appeared to be liver-specific sinceuptake by the spleen as well as other tissues was markedlyreduced. In the case of DSPC-Ch liposomes lacking HPI, liverand spleen uptake were both similarly increased as comparedwith low HPI-containing preparations, although tumor uptakewas not significantly affected. Thus, there appears to be different mechanisms by which the lack or the excess of the negativelycharged lipid component, HPI, alter the pattern of tissue distribution.

The effect of method of preparation and particle size distribution on the pharmacological behavior of liposomes was studied using a preparation made by solvent injection with orwithout extrusion. Solvent injection resulted in unilamellarvesicles of small diameter (<150 nm) (6) without any need forextrusion. Extrusion through 0.05 /im-pore membranes furtherreduced the average diameter and SD from 138 Â±56 to 97 Â±22 nm. Larger vesicles were not examined since prior work in

6372




Table 1 Biodistribution of'Ga-labeled liposomes in C57BL/6 mice bearing the BI6 melanoma"

Liposomecompositionc%ID/g

tissue(SD)Blood''Liver*SpleenKidneysIntestineSkinCarcass'Tumor''Tissue

uptakeratiosLiver/bloodLiver/tumorTumor/carcassPG-PC-Ch0.3(0.1)35.7(1.9)61.7(16.6)2.3

(0.3)0.5(0.0)0.3(0.0)0.4(0.1)1.8(1.0)119.019.84.5PG-DSPC-Ch1.4(0.4)9.8

(0.6)6.0(0.2)1.8(0.1)0.2

(0.0)0.7(0.5)0.2(0.0)1.5(0.4)7.06.57.5DPPG*-DSPC-Ch10.8(1.5)20.3

(2.4)23.2(2.5)3.2

(0.3)0.8(0.1)1.8(0.4)0.9(0.2)4.5

(0.6)1.94.55.0HPI-DSPC-Ch8.4

(2.0)14.1(0.3)17.1(2.9)3.3(0.6)1.1(0.1)2.4(0.2)0.5(0.0)5.3

(0.4)1.72.710.6GM,-DSPC-Ch7.8(1.7)22.0

(0.8)37.8(3.6)2.9(0.9)0.9(0.1)1.9(0.9)0.7(0.1)8.4

(0.3)2.82.612.0

Â°Tumor cells inoculated i.m. in the hind leg; dose, 1 fimol phospholipid/mouse i.V.; mice sacrificed 24 h after injection; results are the mean of 3 to 6 animals per

experimental group.* DPPG. dipalmitoylphosphatidylglycerol.' Molar ratio, 1-10-5. When the liposome radiolabel, 67Ga-DF, was injected in free form, tissue uptake was <0.2%ID/g, except for kidneys, which showed an

uptake of 1.7%ID/g.d Blood, liver, and tumor values of dipalmitoylphosphatidylglycerol, HPI, and GM]-containing liposomes significantly different from those of PG-containing

liposomes (IP < 0.05).' Carcass consists of all skeletal muscles, bones, and appendages.

Table 2 Biodistribution of*'Ga-labeled CM,-containing liposomes in nude micebearing the LSI 74T human colon carcinoma0

%ID/g tissue (SD)

50.OA

BloodLiverSpleenKidneysLungsHeartIntestineSkinCarcassTumorExperiment

13.6

(0.4)12.5(1.6)25.9(3.3)3.6

(0.6)1.1(0.2)1.0(0.2)1.0(0.0)5.2

(0.6)0.9(0.1)5.1

(0.4)Experiment

24.2(0.1)19.3(2.6)38.3

(4.5)3.3(0.3)0.8(0.6)1.3(0.2)1.1

(0.2)4.3(0.3)0.8(0.0)4.4

(0.5)" Tumor cells injected s.c. in the flank; each mouse received i.v. 1 fimol

phospholipid of GM|-DSPC-Ch (1-10-5) liposomes; mice were sacrificed 24 hafter liposome injection; the results are the average of 3 mice/experiment; adifferent liposome batch was tested in each experiment.

Table 3 Biodistribution off'H/inulin-labeled GM'Â¡-containingliposomes intumor-bearing mice"

B16 tumor LS174T tumor

%ID/g tissue(SD)BloodLiverTumorTissue

uptakeratiosLiver/bloodLiver/tumor4.9(0.1)21.9(0.4)10.6(0.2)4.52.19.0

(0.8)17.9(2.7)6.3

(2.9)2.02.8

" Tumor cells inoculated i.m. in the leg (B16) or s.c. in the flank (LS174T);dose. 1 f/mol phospholipid/mouse of GMi-DSPC-Ch (1-10-5) i.v.; mice sacrificed24 h after injection; results are the mean of 5 mice.

tumor-free mice had shown that the circulation time of HPI-and GMi-containing liposomes was markedly shortened whenvesicles of 200-nm size or larger were used (data not shown).As seen in Table 4, the highest tumor uptake (13.2% ID/g) andlowest liver-to-tumor ratio (0.9) were obtained with the extruded liposomes. These results suggest that size optimizationis a critical factor even within a narrow range of discrimination.

In order to study the effect of lipid dose, we used the liposomepreparation made by solvent injection and extrusion. As shownin Fig. 2, tumor uptake was not significantly affected by increasing the lipid dose over a 10-fold range (0.3 pmo\ to 3 /Â¿molphospholipid) reaching values between 10 and 15% ID/g inagreement with the results obtained in Table 4. Liver-to-tumorratios were close to 1 (1.2 to 1.3) for all 3 dose levels tested.Similar results were obtained when 67Ga-DF-labeled GM,-DSPC-Ch liposomes were tested in mice bearing the J6456

oQ

10 20 30 40

HPI Content (molar % of total phospholipid)

Fig. 1. Effect of liposome HPI content on liposome biodistribution in tumor-bearing mice. 67Ga-DF-labeled DSPC-Ch liposomes and HPI-DSPC-Ch lipo

somes containing increasing molar fractions of HPI (9,23, and 41 %) Â»ereinjectedi.v. in C57BL/6 mice bearing i.m. leg implants of the B16 melanoma. Dose, 1fimol phospholipid per mouse. Mice were sacrificed 24 h after liposome injection.Each liposome formulation was tested in 3 mice. Liver and spleen values forDSPC-Ch (HPI = 0) were significantly higher (2P < 0.05) than those for HPI-DSPC-Ch (HPI, 9-23%). Blood, spleen, and tumor values for HPI-DSPC-Ch(HPI = 41%) were lower (IP < 0.05) than those for other formulations. Oâ€”O,Blood; â€¢â€”â€¢,liver; A A, spleen; A A, tumor.

Table 4 Biodistribution of"JGa-labeled HPl-HPC-Ch liposomes in mice bearingthe BI6 melanoma: effect of size reduction"

No extrusion Extrusion

Mean size in nm(SD)%ID/gtissue(SD)Blood"Liver*SpleenKidneysIntestineSkinCarcassTumor*Tissue

uptakeratiosLiver/bloodLiver/tumorTumor/carcass138(56)4.2(1.9)19.3(1.7)18.3(0.0)4.3

(0.2)1.6(0.2)1.0(0.1)0.8(0.1)7.7

(2.3)4.62.59.697

(22)3.7(2.1)11.8(2.8)9.5(1.6)5.7(1.1)1.9(0.2)1.4(0.1)1.3(0.3)13.2(2.3)3.20.910.2

" Tumor cells inoculated in the

mouse i.v.; liposome composition,injection with or without extrusion;results are the average of 3 mice.

* Blood values of nonextruded

different. Liver and tumor valuessignificantly different (2P = 0.017 a

hind footpad; dose, 1 ,,nml phospholipid/HPI-HPC-Ch (1-10-5) prepared by solventmice sacrificed 24 h after liposome injection;

versus extruded liposomes not significantlyof nonextruded versus extruded liposomes

Â¡nd0.043 respectively).

6373




20.0

Â£ 10.0-

OQ

5.0

2.0

u

Z 1.CH

0.50.0 0.5 1.0 1.5 2.0 2.5 3.0

Phospholipid dose (umoles / mouse)

Fig. 2. Effect of phospholipid dose on the biodistribution of liposomes intumor-bearing mice. "Ga-DF labeled HPI-HPC-Ch liposomes were injected i.v.

in C57BL/6 mice bearing footpad implants of the B16 melanoma. Mice weresacrificed 24 h after liposome injection. Each dose level was tested in 3 mice. O,Blood; â€¢,liver; A. carcass; A. tumor.

lymphoma in the dose range of 0.25 to 5 /jmol phospholipidper mouse, namely no significant change in the relative tumoruptake (data not shown).Tissue Distribution Studies with '"In-Bleo-labeled Liposomes

To evaluate the potential significance of liposome uptake bytumors with regard to a cytotoxic drug, we examined thebiodistribution of Bleo labeled with ' "In, either as free complexor encapsulated in GMi- and HPI-containing liposomes. Theseexperiments were done with 2 mouse tumor models: the B16melanoma and the J6456 lymphoma. The results are summarized in Table 5 and Fig. 3. Liposome encapsulation increasedBleo retention in all body tissues, probably by diminishing renalclearance. The tumor uptake of liposome-encapsulated Bleowas relatively high as compared with other tissues, with theexception of liver and spleen (Table 5). These factors resultedin 2 favorable parameters of tissue distribution pattern: (a) asubstantial increase of the label concentration in tumors of micetreated with liposome-encapsulated Bleo as opposed to freeBleo (=20- to 40-fold); and (b) an increased uptake by tumoras opposed to carcass in mice treated with liposome-encapsulated Bleo (~7- to 16-fold). As found previously using a 67Ga-

DF liposome label (5), peak increases in tumor concentrationwere observed at 24 h after injection (Fig. 3).Imaging Studies with Free and Liposome-encapsulated 67Ga-DFand '"In-Bleo

The use of -y-ray-emitting radiolabels enabled us to carryimaging studies to complement the information obtained from

tissue distribution experiments. C57BL/6 mice bearing B16melanoma implants in the right hind footpad were used as theanimal tumor model. The left hind footpad was mock-injectedwith cell-free medium. In the first experiment, HPI-HPC-Ch(mean size, 100 Â±39 nm) and PG-PC-Ch (mean size, 74 Â±15nm) liposomes labeled with 67Ga-DF were injected into 3 miceeach, and free ft7Ga-DF was injected into 1 mouse. Images were

obtained within l h (blood pool image), and at 24, 42, and 70h after injection. Figs. 4 (A to />), 5 (A to D), and 6 (A and B)show successive images obtained with HPI-HPC-Ch, PG-PC-Ch, and free label, respectively. The label was more rapidlyexcreted from animals receiving PG-PC-Ch than from animalsreceiving HPI-HPC-Ch. The average number of counts obtained at 70 h was 20.3% of the initial blood pool image countsfor the former, and 30.3% for the latter. Free 67Ga-DF was

rapidly excreted into the urine, and only 10% of the blood poolimage counts were recovered at 42 h injection. With both typesof liposomes, there was marked label concentration in the upperabdomen (liver and spleen). In addition, label concentration inthe tumor area was clearly enhanced in mice receiving HPI-HPC-Ch as compared with mice receiving PG-PC-Ch. Whenthe mice were sacrificed and the dissected tissues checked in a-y-counter at 70 h post-injection, the %ID/g tumor and liver-to-tumor ratios were 1.8 (SD = 0.3) and 3.8 for PG-PC-Ch,and 7.2 (SD = 1.0) and 1.9 for HPI-HPC-Ch.

An additional imaging experiment was carried out with '"In-Bleo-labeled HPI-HPC-Ch liposomes and free '"In-Bleo. Two

mice given injections of liposomes and 1 mouse given aninjection of the free drug were imaged within l h (blood poolimage), and at 18, 44, and 67 h after injection. When free '"In-

Bleo was injected, the label was concentrated in the urinarybladder, and only 10% of the blood pool image counts wererecorded at 18 h, thus pointing at the fast renal excretion rate.No label uptake was detected in the tumor-injected footpad. Inthe case of "'In-Bleo encapsulated in HPI-HPC-Ch liposomes

(Fig. 7, A to D), 50% of the blood pool image counts were stillrecovered at 67 h. The label was concentrated in the upperabdomen (liver and spleen) and in the tumor area. Computerized quantitation of regions of interest showed a 5- to 11-foldincreased uptake by the tumor as compared with the normalmock-injected contralateral leg. Upon sacrifice and 7-countingof the dissected organs, the %ID/g tumor and carcass were 7.9Â±0.3 and 1.2 Â±0.0, respectively (tumorcarcass ratio, 6.6).

DISCUSSION

The results of this study provide strong evidence on the abilityof at least 2 liposome formulations to accumulate in transplant-

Table 5 Biodistribution of free '"In-Bleo and "'In-Bleo encapsulated in GMÂ¡-and HPI-containing liposomes in tumor-bearing mice"

%ID/g tissue(SD)*BloodLiverSpleenKidneysLungsHeartIntestineSkinCarcassTumorFreeO.I

(0.0)0.4(0.0)0.3(0.0)3.0(1.0)0.2(0.1)0.4

(0.0)0.1(0.0)0.2(0.0)0.1(0.0)0.3

(0.0)B16

tumorGMl-DSPC-Ch7.4

(0.9)23.0(1.1)37.8

(2.7)4.5(1.2)1.5(0.4)1.2(0.1)1.2

(0.6)2.9(0.9)1.1(0.2)9.2(1.9)HPI-HPC-Ch2.2

(0.3)26.7(2.4)54.7(5.0)12.1(1.1)2.6(1.0)2.0

(0.2)1.9(0.1)1.9(0.2)1.0(0.1)6.8

(2.3)Free0.6

(0.3)0.5(0.4)0.8(0.1)1.9(0.6)1.0(0.3)0.8

(0.3)0.2(0.1)0.3

(0.2)0.2(0.1)0.4

(0.3)J6456

tumorGMl-DSPCCh2.5

(0.9)44.8(0.6)80.9(2.1)3.5

(0.2)1.9(0.6)2.2(1.1)1.8(0.1)1.4(0.1)0.6

(0.0)9.6(1.6)HPI-DSPC-Ch3.8

(0.0)43.2(1.1)45.4

(0.6)6.6(0.6)2.8(0.4)3.3(0.5)3.4(0.1)2.9(0.1)1.1

(0.2)17.3(7.7)

* Tumor cells inoculated i.m in the hind leg; phospholipid dose. 1 to 1.4 ^mol/mouse; Bleo dose. 5 mU/mouse. except for the B16 tumor HPI-HPC-Ch group,

which received 50 mU/mouse; mice sacrificed 24 h after injection: results are the average of 3 mice.* Except for kidneys, lungs, and heart, all other tissue values were significantly higher for "'In-Bleo in liposomes as compared with free '"In-Bleo (2P < 0.05).

6374




50.0Â«jD

Â« 20.0'+J

Â§ 10-Â°-

-T 5.0

<no

Q

el?

2.0

1.0-

0.5

0.25 10 15 20 25 30 35 40 45 50

Hours after liposome injectionFig. 3. Biodistribution of "'ln-Bleo-labeled liposomes in tumor-bearing mice.

"'In-Bleo-labeled HPI-DSPC-Ch liposomes were injected i.v. in BALB/c micebearing i.m. implants of the J6456 lymphoma. Dose, 1 timo\ phospholipid and 5mU Bleo per mouse. Each time point was tested in 3 mice. O, Blood; â€¢,liver; A.carcass; A, tumor.

able mouse and human tumors. This phenomenon appears tobe nonspecific since: (a) there is no known tumor-specific ligandon the liposome surface; (b) it has been observed with severaltumors of different histology and origin (mouse B16 and J6456,human LS174T); (c) it is not dependent on the site of tumorinoculation (i.m., s.c., footpad). Given the inherent lack oftumor specificity, this approach could have the advantage ofgeneral applicability, although the heterogeneous spectrum ofhuman tumors precludes so far any extrapolations to the clinicalsituation. Another important feature of this phenomenon is therelative degree of selectivity of tumor uptake versus uptake by

most normal tissues. This is exemplified by the high tumor-to-

carcass ratios. In most cases, tumor uptake was still below theuptake by liver and spleen, organs with fenestrated capillariesand rich in cells of the RES (12). Nevertheless, we were able toobtain very low liver-to-tumor ratios in the range of 1 to 2. Thissubstantial improvement in the pattern of liposome biodistribution was the combined result of an absolute decrease of liveruptake and an absolute increase in tumor uptake. The significance of these changes is likely to be a reduced risk of toxicityto the RES and an enhanced anti-tumor effect of liposome-delivered drugs. Of note, an additional organ in which we haveobserved an increased uptake of liposomes is the skin. This wasspecially conspicuous in nude mice in which the lack of furresults in a relative increase of the fraction of skin weightrepresented by the vascularized dermis. A significant uptake ofsmall unilamellar liposomes by skin and intestine has beenreported by Hwang et al. (13). It is as yet unclear whether acommon mechanism is responsible for liposome localization inskin and in tumors.

As reported previously (5), there is an inverse relationshipbetween liposome clearance by the RES and a prolonged circulation time of liposomes. In turn, there appears to be a directcorrelation between prolonged circulation time and liposomelocalization in tumors (5). This is underscored by the results ofthis study in which low liver-to-blood ratios were generallycorrelated with low liver-to-tumor ratios.

This and previous studies (5, 14-17) have identified severalimportant characteristics of liposome formulation, conferringoptimal pharmacological properties: a rigid bilayer matrix consisting of distearoyl or fully hydrogenated PC and Ch, a smallfraction of a negatively charged glycolipid such as GMi or fully

Fig. 4. Imaging of tumor-bearing mousegiven an injection of "Ga-DF-labeled HPI-HPC-Ch liposomes. C57BL/6 mice bearingright hind footpad implants of the B16 melanoma received i.v. 1.3 /.mol phospholipidand 29.5 nCi of the label in liposome-en-trapped form. Anterior views were obtained1 (A), 24 (B), 42 (C), and 70 (D) h afterinjection. Note the higher uptake of label inthe tumor-injected footpad as compared withthe normal mock-injected footpad.

6375




Fig. 5. Imaging of tumor-bearing mousegiven an injection of "Ga-DF labeled PG-PC-

Ch liposomes. C57BL/6 mice bearing right hindfootpad implants of the B16 melanoma receivedi.V. 1.3 Â¿imuÃphospholipid and 46.5 ÃŸC'iof the

label in liposome-entrapped form. Anteriorviews were obtained 1 (A), 24 (fi), 42 (C), and70 (O) h after injection. In contrast to Fig. 4,the difference in uptake between the tumor-injected footpad and the normal mock-injectedfootpad is barely noticeable. Note also that theurinary bladder is visualized in the l h bloodpool image.

Fig. 6. Imaging of tumor-bearing mousegiven an injection of free 67Ga-DF. C57BL/6

mice bearing right hind footpad implants ofthe B16 melanoma received i.v. 33 nC\ of thefree label. Anterior views were obtained 1 (A)and 42 (B) h after injection. In contrast toFigs. 4 and 5, the label is rapidly excreted, thusresulting in accumulation of the label in theurinary bladder. No label uptake occurs in thetumor-injected footpad.

hydrogenated phosphatidylinositol, and a relatively small vesicle diameter (=100 nm). The differences in tumor uptake between unextruded and extruded solvent injection-prepared liposomes (Table 4) suggest that even small changes in size havean important pharmacological impact.

The observations of Proffitt et al. (18, 19) and Ogihara et al.(20, 21) in tumor-bearing mice and subsequent reports onimaging studies in cancer patients (22, 23) with small DSPC-Ch neutral vesicles would appear to indicate that significanttumor targeting is achievable without any need for negativelycharged glycolipids. These observations (18-23) were madewith weak chelation complexes ('"In-nitriloacetic acid, 67Ga-

nitriloacetic acid), which result in high tissue backgrounds asopposed to 67Ga-DF (9). Using such neutral vesicles, we alsofound a high uptake by the B16 melanoma, although hepato-

splenic uptake was relatively high, thus resulting in an overallsuboptimal biodistribution pattern when compared with liposome preparations containing a low molar ratio of HPI (Fig.1). As we proposed previously (5), the effect of HPI and CM,may be related to prevention of vesicle aggregation in additionto decrease of the recognition of the liposome surface by plasmaopsonins. Interestingly, a high molar ratio of HPI overrides thefavorable effect of a low molar ratio by shifting liposomedistribution to the liver and concomitantly reducing the amountreaching other organs. The increased liver uptake of thesehighly negatively charged vesicles is likely to be the result ofopsonization and receptor-mediated endocytosis (24).

Accumulation of liposomes in tumors may result from extravasation into the interstitial tumor space or endocytosis bycapillary endothelial cells of the tumor microvasculature. Al-

6376




Fig. 7. Imaging of tumor-bearing mousegiven an injection of "'In-Bleo-labeled HPI-HPC-Ch liposomes. C57BL/6 mice bearingright hind footpad implants of the B16 melanoma received i.v. 5 /*mol phospholipid and23.5 fiCi of the label (=0.25 units Bleo) inliposome-entrapped form. Anterior views wereobtained I (A), 18 (A), 44 (C), and 67 (D) hafter injection. Note the high uptake of labelin the tumor-injected footpad as comparedwith the normal footpad.

though no morphological evidence is available yet to supporteither possibility, the apparent lack of significant saturation oftumor uptake when the liposome dose is increased supports anextravasation mechanism mediated by convective transport.The probability of nonspecific localization of a macromoleculeor small particle at the tumor site may be affected by a varietyof factors including mean residence time in circulation (5),tumor microvascular permeability, particle size, and rate ofconvective transport (25-27).

An important consequence of drug encapsulation in liposomes is reduction of clearance rate (17, 28) and increased bodyexposure to liposome-encapsulated drug. This effect is clearly

shown in biodistribution and imaging experiments comparingfree Bleo and liposome-encapsulated Bleo. Increased drug dep

osition in sensitive tissues could result in side effects, whichwill partially counteract the therapeutic gain achievable byenhanced drug delivery to tumors. In addition, the bioavailabil-ity of liposome-encapsulated drug at the tissue microenviron-mental level may be affected positively or negatively by a varietyof factors besides extravasation. Among those, movement ofliposomes across the interstitial fluid, uptake of liposomes bycells, and rate of drug release in situ. Thus, although the presentfindings point at liposomes as a valuable tool in pharmacological manipulation, their implications on the therapeutic indexof liposome-delivered drugs are still uncertain.

ACKNOWLEDGMENTS

We wish to thank Ninfa Lopez, RenÃ©eShiota, and Ann Strubbe fortechnical assistance.

REFERENCES1. Perez-Soler, R. Liposomes as carriers of antitumor agents: toward a clinical

reality. Cancer Treat. Rev., 16: 67-82. 1989.2. Gabizon, A. Liposomes as a drug delivery system in cancer chemotherapy.

In: F. H. Roerdink and A. M. Kroon (eds.). Drug Carrier Systems, Horizonsin Biochemistry and Biophysics Series, Vol. 9, pp. 185-211. Chichester:John Wiley and Sons, 1989.

3. Poste, G. Liposome targeting in vivo: problems and opportunities. Biol. Cell,47: 19-38, 1983.

4. Weinstein, J. N. Liposomes as drug carriers in cancer chemotherapy. CancerTreat. Rep., 68: 127-135, 1984.

5. Gabizon, A., and Papahadjopoulos, D. Liposome formulations with prolonged circulation time in blood and enhanced uptake by tumors. Proc. Nail.Acad. Sci. USA, 85: 6949-6953. 1988.

6. Lichtenberg, D.. and Barenholz, Y. Liposomes: preparation, characterization,and preservation. In: D. Glick (ed.). Methods in Biochemical Analysis, Vol.33, pp. 337-462. New York: John Wiley and Sons, 1988.

7. Bartlett, G. R. Phosphorus assay in column chromatography. J. Biol. Chem.,254:466-468, 1959.

8. Bligh, E. G., and Dyer, W. J. A rapid method of total lipid extraction andpurification. Can. J. Biochem. Physiol., 37:911-917. 1959.

9. Gabizon. A.. Huberty, J., Straubinger, R. M., Price, D. C., and Papahadjopoulos. D. An improved method for /'// vivo tracing and imaging liposomesusing a "Gallium-deferoxamine complex. J. Liposome Res., /: 123-135,

1988.10. Gabizon, A., and Trainin, N. Enhancement of growth of a radiation-induced

lymphoma by T cells from normal mice. Br. J. Cancer, 42: 551-558, 1980.11. Bresalier. R. S.. Raper, S. E., Hujanen, E. S., and Kim, Y. S. A new animal

model for human colon cancer mÃ©tastases.Int. J. Cancer, 39:625-630, 1987.12. Mclntyre, N. Hepatic function in health and disease: implications in drug

carrier use. In: G. Gregoriadis, J. Senior, and G. Poste (eds.). Targeting ofDrugs with Synthetic Systems, Life Sciences Series. Vol. 113, pp. 87-96.New York: Plenum Press. 1986.

13. Hwang. K. J., Padki. M. M., Chow, D. D.. Essien, H. E.. Lai. J. Y.. andBeaumier, P. L. Uptake of small liposomes by nonreticuloendothelial tissues.Biochim. Biophys. Acta, 907: 88-96, 1987.

14. Hwang, K. J. Liposome pharmacokinetics. In: M. J. Ostro (ed.). Liposomes:From Biophysics to Therapeutics, pp. 109-155. New York: Marcel Dekker.1987.

15. Gregoriadis, G. Fate of injected liposomes: observations on entrapped soluteretention, vesicle clearance, and tissue distribution in vivo. In: G Gregoriadis(Ã©d.).Liposomes as Drug Carriers: Recent Trends and Progress, pp. 3-18.Chichester: John Wiley and Sons, 1988.

6377




16. Allen, T. M., and Chonn, A. Large unilamellar liposomes with low uptakeinto the reticuloendothelial system. FEBS Lett., 223: 42-46, 1987.

17. Gabizon, A., Shiota, R., and Papahadjopoulos. D. Pharmacokinetics andtissue distribution of doxorubicin encapsulated in stable liposomes with longcirculation times. J. Nati. Cancer Inst., 81: 1484-1488, 1989.

18. Proffitt. R. T., Williams. L. E., Presant, C. A., Tin, G. W., Uliana, J. A..Gamble, G. C., and Baldeschwieler, J. D. Liposomal blockade of the reticuloendothelial system: improved tumor imaging with unilamellar vesicles.Science (Wash. DC), 220.-502-505, 1983.

19. Proffitt, R. T., Williams, L. E., Presant, C. A., Tin, G. W., Uliana, J. A.,Gamble, G. C., and Baldeschwieler, J. D. Tumor imaging potential ofliposomes loaded with ln-111-NTA: biodistribution in mice. J. NucÃ.Med.,24:45-51, 1983.

20. Ogihara, I., Kojima, S., and Jay. M. Tumor uptake of "Ga-carrying liposomes. Eur. J. NucÃ.Med., / /: 405-411,1986.

21. Ogihara-Umeda, I., and Kojima, S. Increased delivery of Gallium-67 totumors using serum-stable liposomes. J. NucÃ.Med., 29: 516-523, 1988.

22. Turner, A. F., Presant, C. A., Proffitt, R. T., Williams, L. E., Winsor, D.

W., and Werner, J. L. ln-111-labeled liposomes: dosimetry and tumordepiction. Radiology, 166: 761-765. 1988.

23. Presant, C. A., Proffitt, R. T., Turner, A. F., Williams, L. E., Winsor, D.,Werner, J. L., Kennedy. P.. Wiseman. C.. Gala, K., McKenna, R. J.. DouglasSmith. J., Bouzaglou, S. A., Callahan, R. A., Baldeschwieler, J., and Crossley,R. J. Successful imaging of human cancer with Indium 111-labeled phospho-lipid vesicles. Cancer (Phila.), 62: 905-911,1988.

24. Dijkstra, J., Van Galen, W. J. M., Hulstaert. C. E., Kalicharan, D.. Roerdink,F. H.. and Scherphof. G. L. Interaction of liposomes with Kupffer cells invitro. Exp. Cell Res., 750: 161-176, 1984.

25. Jain, R. K., and Gerlowski, L. E. Extravascular transport in normal andtumor tissues. CRC Crii. Rev. Oncol. Hematol., 5: 115-170. 1986.

26. Peterson, H. 1. (ed.) Tumor Blood Circulation: Angiogenesis, Vascular Morphology, and Blood Flow of Experimental and Human Tumors. Boca Raton:CRC Press. 1979.

27. Jain. R. K. Delivery of novel therapeutic agents in tumors: physiologicalbarriers and strategies. J. Nati. Cancer Inst., */: 570-576, 1989.

28. Juliano, R. L., and Stamp, D. Pharmacokinetics of liposome-encapsulatedanti-tumor drugs. Biochem. Pharmacol., 27: 21-27, 1978.

6378



Effect of Liposome Composition and Other Factors on the … · thin-layer chromatography....

Documents

Transcript of Effect of Liposome Composition and Other Factors on the … · thin-layer chromatography....