tumor-targeted intracellular delivery of anticancer drugs through the...

Tumor-targeted intracellular delivery of anticancer drugsthrough the mannose-6-phosphate/insulin-like growthfactor II receptor

Jai Prakash1,2, Leonie Beljaars1, Akshay K. Harapanahalli1, Mieke Zeinstra-Smith1,2, Alie de Jager-Krikken1,

Martin Hessing2, Herman Steen2 and Klaas Poelstra1,2

1 Department of Pharmacokinetics, Toxicology and Targeting, Groningen Research Institute for Pharmacy, University of Groningen,

The Netherlands2 BiOrion Technologies BV, Groningen, The Netherlands

Tumor-targeting of anticancer drugs is an interesting approach for the treatment of cancer since chemotherapies possess

several adverse effects. In the present study, we propose a novel strategy to deliver anticancer drugs to the tumor cells

through the mannose-6-phosphate/insulin-like growth factor receptor (M6P/IGF-IIR) which are abundantly expressed in several

human tumors. We developed a drug carrier against M6P/IGF-II receptor by modifying human serum albumin (HSA) with M6P

moieties. M6P-HSA specifically bound and internalized into M6P/IGF-IIR-expressing B16 melanoma cells as demonstrated with

radioactive studies and anti-HSA immunostaining. In vivo, M6P-HSA rapidly accumulated in subcutaneous tumors in tumor and

stromal components after an intravenous injection. To demonstrate the application of M6P-HSA as a drug carrier, we coupled

doxorubicin to it. Dox-HSA-M6P conjugate could release doxorubicin at lysosomal pH and showed M6P-specific binding and

uptake in tumor cells. In vitro, a short exposure with Dox-HSA-M6P induced killing of tumor cells, which could be blocked by

excess M6P-HSA. In vivo, Dox-HSA-M6P distributed to tumors and some other organs while free doxorubicin distributed to all

organs but slightly to tumors. In B16 tumor-bearing mice, Dox-HSA-M6P significantly inhibited the tumor growth whereas an

equimolar dose of free doxorubicin did not show any anti-tumor effect. In addition, targeted doxorubicin did not show any

side-effects on liver and kidney function tests, body weight and blood cell counts. In conclusion, M6P-HSA is a suitable

carrier for delivery of anticancer drugs to tumors through M6P/IGF-IIR. Improved antitumor effects of the targeted doxorubicin

by M6P-HSA suggest that this novel approach may be applied to improve the therapeutic efficacy of anticancer drugs.

Tumor-specific delivery of anticancer drugs has been claimedas a potential approach for the treatment of cancer1,2 andmany efforts have been made to deliver anticancer drugs totumors through tumor cell-specific receptor-mediated drugdelivery approaches.3–5 However, low and nonselectiveexpression in tumors and poor ligand-uptake capacity of thetarget receptors are common problems in these approaches.Therefore, a receptor which is highly expressed on tumorcells compared to the normal cells and has a high internaliza-tion capacity might be a suitable target receptor for drugdelivery purposes.

Mannose-6-phosphate/insulin-like growth factor (M6P/IGF-II) receptor, also known as cation-independent M6P re-ceptor, is a multifunctional receptor, involved in the trans-port of cellular proteins from the cell surface or trans-Golginetwork to lysosomes.6,7 There is also another type of recep-tor so-called cation-dependent M6P receptor that also bindsto M6P-bearing proteins. However, it is not present on cellsurface and only involved in the transport of some lysosomalenzymes from trans-Golgi network to lysosomes.8 M6P/IGF-IIR has 2 types of ligands; (i) M6P-containing ligands e.g.latent transforming growth factor-beta and lysosomalenzymes and (ii) non-M6P-containing ligands e.g. IGF-II andretinoic acid. M6P/IGF-IIR has a large extracellular domainwith 4 distinct binding sites especially for M6P-containingligands, which potentiates its internalization capacity forM6P-containing proteins.6 After binding to M6P/IGF-IIR,the ligand-receptor complex is internalized into lysosomesand degraded by the lysosomal enzymes. M6P/IGF-IIR ishighly expressed in fetal and neonatal tissues but it declinespostnatally.9 However, in adults M6P/IGF-IIR has beenreported to be induced in several human carcinoma cellssuch as breast cancer,10 pancreatic cancer,11 gastric cancer,12

melanoma13 and hepatocellular carcinoma.14 In addition,there is accumulating evidence that M6P/IGF-IIR plays an

Key words: insulin-like growth factor receptor, mannose-6-phosphate,

intracellular delivery, stromal targeting

Grant sponsor: STW Valorisation Grant, The Netherlands

DOI: 10.1002/ijc.24914

History: Received 6 Jul 2009; Accepted 8 Sep 2009; Online 30 Sep

2009

Correspondence to: Jai Prakash, Department of Pharmacokinetics,

Toxicology and Targeting, Groningen Research Institute for

Pharmacy, University of Groningen, Antonius Deusinglaan 1,

9713 AV, Groningen, The Netherlands, Tel: þ31-50-3636414, Fax:

+31-50-3633247, E-mail: [email protected]

Can

cerTherapy

Int. J. Cancer: 126, 1966–1981 (2010) VC 2009 UICC

International Journal of Cancer

IJC

important role in the regulation of tumor growth and metas-tasis.15,16 M6P/IGF-IIR elicits these actions by degradingIGF-II, a putative mitogen for tumor growth, and by traffick-ing lysosomal enzymes, which facilitate tissue invasion andmetastasis, to lysosomes.17 M6P/IGF-IIR has been foundto be frequently mutated during oncogenesis within theliver,18–21 prostrate,22 lung23 and breast,24 although thismostly involves the IGF-II binding site rather than the sitesfor M6P-ligand.25

Since M6P/IGF-IIR is highly expressed in many tumorsand has a high ligand internalization capacity, we hypothe-sized that this receptor might be a potential target to deliveranticancer drugs to tumors. In earlier studies, we developed aM6P/IGF-IIR–specific carrier by modifying human serum al-bumin with M6P sugar moieties (M6P28-HSA) to deliverantifibrotic drugs to hepatic stellate cells for the treatment ofliver fibrosis.26,27 In the present study, we investigatedwhether M6P-HSA can serve as a carrier to deliver drugs totumor cells expressing M6P/IGF-IIR. To achieve this, we firstscreened for M6P/IGF-IIR–expressing tumor cell lines andstudied the binding and uptake of M6P-HSA in these cellsin vitro. Thereafter, we determined the distribution and accu-mulation of M6P-HSA in vivo in tumor-bearing mice. Fur-thermore, we conjugated doxorubicin to M6P-HSA andexamined its efficacy in vitro for cell-viability and in vivo forits effects on tumor growth.

Material and MethodsMaterials

Mouse melanoma cells (B16-F10) were obtained from Ameri-can Type Culture Collection (ATCC, Rockville, MD), andmouse colon carcinoma cells (C26) were kindly provided byProf. Molema (Medical Biology, University Medical CentreGroningen, The Netherlands). Polyclonal goat-anti-M6P/IGFII receptor IgG was purchased from Santa Cruz Biotech-nologies, Santa Cruz, CA, and monoclonal rat-anti-mousePECAM-1 (CD31) from BD PharMingen. Rabbit-anti-humanserum albumin (HSA) was obtained from ICN Biomedics,Eschwege, Germany, and goat-anti-HSA and mouse-anti-a-smooth muscle actin antibody from Sigma. Rat anti-F4/80antibody was purchased from AbD serotec, Oxford, UK.Human serum albumin (HSA, CealbVR ) was purchased fromSanquin, Amsterdam, Netherlands.

Western blot analyses

M6P/IGF-II receptor expression in tumor cells, tumors anddifferent organs was determined using Western blot analyses.Then 70–80% confluent cells in 12-well plate were isolatedwith RIPA buffer (250 mM Tris-HCl, 150 mM NaCl, 0.1%Igepal in 0.5% sodium deoxycholate, 0.5% SDS with 1 prote-ase inhibitor cocktail tablet (Roche Diagnostics, Mannheim,Germany) in 10 ml). Tissues were lysed in RIPA buffer witha homogenizer. Twenty micrograms of protein from eachsample was applied on the 10% SDS-PAGE gel and the pro-teins were transferred to polyvinylidene fluoride membrane

electrophoretically. Membranes were blocked with 5% nonfatmilk in Tris-buffered saline containing 0.5% Tween-20 andthen incubated with anti-M6P/IGF-II receptor or anti-actinantibody (Millipore, Temecula, CA) at 4�C overnight. Subse-quently, after 3 times washing horseradish peroxidase-conju-gated secondary antibody was applied for 1 hr. Protein bandswere developed with ECL detection reagent (Perkin-ElmerLife Sciences, Boston, MA).

Immunohistochemistry and immunofluorescence

Immunohistochemical stainings were performed on 4-lmthick frozen sections made by cryostat (Leica). Sections werefixed in acetone for 20 min and then incubated with primaryantibody of interest for 1 hr. After 3 times washing withPBS, endogenous peroxidase activity was blocked by incubat-ing with 0.05% hydrogen peroxide for 20 min. Thereafter,sections were washed 3 times with PBS and incubated withhorseradish peroxidase labeled secondary antibody for30 min. In case of fluorescent staining, no hydrogen peroxi-dase treatment was performed, and secondary antibodies la-beled with either FITC (green) or TRITC (red) were added.Subsequently, sections were washed and incubated with hem-atoxyllin or DAPI (for fluorescent staining) for nuclear stain-ing. After this, sections were mounted with glycerol/kiesel-guhr or glycerol/PBS (for fluorescent staining) solution afterwashing. Sections were visualized under a light microscope orfluorescent microscope.

Synthesis of M6P-HSA and Dox-HSA-M6P

M6P-HSA was synthesized by chemically modifying humanserum albumin (HSA) with mannose-6-phosphate (M6P) asreported earlier.26,28 To couple doxorubicin to M6P-HSAthrough hydrazone linkage, doxorubicin was first modified toprepare doxorubicin-maleimide as described elsewhere29 andthen conjugated to M6P-HSA. In brief, doxorubicin (17.2lmol) was reacted with 3,30-N-[e-maleimidocaproic acid]hy-drazide (51.2 lmol, Pierce) in methanol for 24 hr and thendoxorubicin-maleimide product was isolated through crystal-lization. To couple doxorubicin-maleimide to M6P-HSA,thiol (–SH) groups were introduced in M6P-HSA (1.3 lmol)using N-succinimidyl-S-acetylthioacetate (28.5 lmol, Sigma)cross-linker. Then, doxorubicin-maleimide (18.2 lmol) wasreacted with M6P-HSA–SH (1.3 lmol) for 3 hr to obtaindoxorubicin-HSA-M6P conjugate. The final product was dia-lyzed and purified using gel filtration chromatography to col-lect the monomeric fraction of the conjugates.

Characterization of M6P-HSA and Dox-HSA-M6P

The monomeric form of M6P-HSA was characterized bydetermining the molar ratio of M6P to HSA using anion-exchange chromatography (MonoQ column, Amersham Bio-sciences, Uppsala, Sweden) and a phosphate assay asdescribed earlier.26 The drug/protein coupling ratio in Dox-HSA-M6P was subsequently determined by analyzing doxor-ubicin with high-performance liquid chromatography (HPLC,

Can

cerTherapy

Prakash et al. 1967


Waters, Milford, MA) and by analyzing protein concentra-tion using Lowry’s assay (Bio-Rad). Dox-HSA-M6P was fur-ther characterized to determine change in net-negative chargeusing anion-exchange chromatography and to determine thepercentage monomeric form on gel filtration chromatography(Amersham Biosciences).

Drug-release studies with Dox-HSA-M6P

Since doxorubicin was conjugated to the carrier through anacid-sensitive hydrazone linkage, we first determined thedrug releasing property of Dox-HSA-M6P conjugate at differ-ent pH levels. The conjugate (equivalent to 5 lg/ml doxoru-bicin) was incubated at 37�C in phosphate buffers of pH 7.0,5.0 and 3.0. At specific time points, 50 ll sample was col-lected and analyzed with HPLC. HPLC system comprised ofC18 reversed-phase column (SunFireTM, Waters) with a fluo-rescent detector (Waters; Ex: k450 nm and Em: k550 nm).Samples were eluted with the mobile phase (phosphate bufferpH 7.0: acetonitrile; 86:14) at a flow rate of 1 ml/min. Doxor-ubicin peaks were quantified with EmpowerTM Software(Waters).

Cell experiments

B16, C26 and 3T3 cells were maintained on Dulbecco’smodified Eagle’s medium (DMEM, BioWhittaker, Verviers,Belgium) supplemented with 10 % fetal calf serum and anti-biotics (penicillin, 50 units/ml plus streptomycin, 50 ng/mlfor B16 and 3T3 and gentamicin for C26 cells, 10 lg/ml) at37�C in a humidified incubator containing 5% CO2. Mediumfor 3T3 cells was added with 2 mM L-glutamine.

Cell viability studies

We determined the effect of doxorubicin and Dox-HSA-M6Pon the cell viability of B16, C26 and 3T3 cells. Cells wereseeded into the 96-well plate as 1 � 104 cells/well and after24 hr they were incubated with different compounds for 2 hrand then washed twice and cultured for another 46 hr. Ala-mar Blue assay (Serotec, Oxford, UK) was performed todetermine cell viability by incubating cells for 4 hr at the endof the experiment.

In vitro binding and uptake of radiolabeled conjugates

To examine the uptake of M6P-HSA and Dox-HSA-M6Pthrough M6P-binding receptors on cancer cells, we radiola-beled these conjugates with 125I using Chloramine-T methodand studied their uptake in B16-F10 cells. About 1 � 105

cells/well were seeded in 12-well plate and allowed to attachovernight. Cells were incubated with 1% BSA for 15 min andthen with different competitors in 0.2% BSA in DMEM for15 min and subsequently incubated with 1 � 105 cpm/wellof 125I-M6P-HSA or 125I-Dox-HSA-M6P conjugate at 37�C.At specific time-points, medium was removed and the cellswere washed 3 times with cold medium. Cells were lysed andisolated with 1 M sodium hydroxide and cell-associated

radioactivity was measured using a c-counter (Riastar; Pack-ard Instruments, Palo Alto, CA).

Cellular uptake of Dox-HSA-M6P

B16-F10 cells (1 � 105 cells/well) were grown in 6-well platesup to 90% confluency and then incubated at 37�C with Dox-HSA-M6P (equivalent to 5 lM doxorubicin). At differenttime points, cells were washed 3 times with PBS and isolatedby trypsinization. The cell membranes were ruptured byultrasonification for 15 sec and then treated with 0.5 N HClto release doxorubicin from the conjugate. The samples werecentrifuged and supernatants were injected into HPLC todetermine cell-associated doxorubicin concentrations.

To visualize the uptake of the Dox-HSA-M6P conjugate,B16-F10 cells (2 � 104 per well were seeded in 8-well glassplate (Lab-Tek, Nunc Inc., Naperville, IL). Cells were incu-bated with Dox-HSA-M6P (equivalent to 10 lg/ml doxorubi-cin) at 37�C in the incubator. After 15, 60 and 120 min, cellswere washed thrice with PBS and fixed with 4% formalin for15 min at 4�C. Doxorubicin fluorescence was visualized incells with a fluorescent microscope.

Subcutaneous tumor model in mice

Normal male C57BL/6 and Balb/c mice (20–25g) wereobtained from Harlan (Zeist, The Netherlands) and kept at a12:12 hr light/dark cycle and received ad libitum normal diet.All experimental protocols for animal studies were approvedby the Animal Ethics Committee of the University of Gro-ningen. To induce subcutaneous tumors, B16-F10 cells andC26 cells suspended in 100 ll of PBS were injected subcuta-neously in the flank of C57BL/6 and Balb/c mice, respectively(1 � 106 cells per mice). Tumor growth was followed bymeasuring tumor size using a digital Vernier Capilar and tu-mor volume was established using the following formula (a� b2/2) where a denotes to the tumor length and b denotesto the tumor width.

Tumor uptake of M6P-HSA

In the B16 and C26 subcutaneous tumors with tumor volumeof approximately 1,500 mm3, M6P-HSA (1.5 mg/mouse) wasinjected intravenously under anesthesia through the penilevein of mice. After 2 hr, tumors were isolated and frozen inisopentane. Anti-HSA immunostaining was performed oncryosections as described above, in order to localize M6P-HSA in these tumors.

Organ and tumor distribution of Dox-M6P-HSA

We examined the whole body distribution of 125I-Dox-HSA-M6P in B16 tumor-bearing mice. A tracer dose of 125I-Dox-HSA-M6P was injected in tumor-bearing mice (n ¼ 3) andafter 30 min animals were sacrificed. Subsequently, tumorsand several organs were collected in tube and analyzed forthe radioactivity using a c-counter. The percentage of theinjected dose was calculated by determining the total organweight.

Can

cerTherapy

1968 Tumor-targeting of anticancer drugs


To compare the distribution of Dox-HSA-M6P and freedoxorubicin, a single dose of Dox-HSA-M6P (1.0 mg equiva-lent to 50 lg doxorubicin per mouse) or doxorubicin (50 lg/mouse) was injected intravenously under anesthesia throughthe penile vein of mice. After 2 hr, tumors and many organs(lungs, kidneys, liver, spleen and heart) were isolated and fro-zen in isopentane. Cryosections of tumors and normal tissueswere analyzed under fluorescent microscope for localizingdoxorubicin staining in Dox-HSA-M6P– and doxorubicin–treated mice. Anti-HSA immunostaining was performed oncryosections as described above, in order to localize Dox-HSA-M6P in these tumors.

Effect of Dox-HSA-M6P on tumor growth

B16 tumors were induced in C57BL/6 mice as mentionedabove. Animals were injected intravenously with either vehi-cle (PBS), M6P-HSA, doxorubicin or Dox-HSA-M6P whenthe tumor size of 6100 mm3 was attained. This tumor sizehas been shown to be suitable for the start of the treatment.30

Tumor size was measured at the time of dosing under anes-thesia. All animals were sacrificed when one of the tumors inany group exceeded the tumor volume of �2,000 mm3.

Statistical Analyses

Data are presented as mean 6 standard error mean (SEM).The statistical analyses were performed using Student’s t-test

Figure 1. Expression of M6P/IGF-II receptors in cultured tumor cells, subcutaneous tumors and normal organs in mice. Western Blot

showing the protein bands for M6P/IGF-IIR detected at the molecular weight of approximately 300 KDa and the reference protein actin at

42 KDa determined by molecular weight markers in B16 and C26 cells (a) and in subcutaneous tumors and normal organs (b). Lanes 1–6

indicate the tumor, kidneys, spleen, liver, lungs and heart, respectively in C26 tumors-bearing balb/c mice (a0) and in B16 tumor-bearing

C57/BL6 mice (b0). (c) and (d) showing the representative microscopic pictures of the immunohistochemical staining for M6P/IGF-IIR in

tumors (a0) and normal organs such as kidneys (b0), spleen (c0), liver (d0), lungs (e0) and heart (f0) in C26 tumors-bearing balb/c mice and in

B16 tumor-bearing C57/BL6 mice, respectively.

Can

cerTherapy

Prakash et al. 1969


with p < 0.05 as the minimal level of significance. Cell-viabil-ity data were fitted for sigmoidal dose–response curve to cal-culate IC50 using GraphPad Prism 4 software (La Jolla, CA).

ResultsExpression of M6P/IGF-II receptor in tumors and

normal tissues

We compared the expression of M6P/IGF-IIR in tumors andnormal organs using Western Blot and immunohistochemicalstaining. We found that M6P/IGF-IIR was highly expressedin both B16 melanoma and C26 colon carcinoma cells and intheir subcutaneous tumors (Fig. 1). In contrast, the expres-sion levels of M6P/IGFII-R were much lower in normal

organs such as kidneys, spleen, liver, lungs and heart asshown by Western Blot and immunostaining for M6P/IGFII-R.

We furthermore determined the localization of the recep-tor in tumors using immunostaining. We showed that M6P/IGF-IIR was highly expressed on tumor and stromal cells ofthese tumors as illustrated in Figure 2a. In stromal compart-ment of both tumor types, most of the cells were positive fora-smooth muscle actin (a-SMA), a marker for fibroblastsand pericytes (surrounding blood vessels) as shown in Figure2a. From these data, we demonstrate that M6P/IGF-IIR wasmost likely present on fibroblast-like stromal cells but not onendothelial cells as there was no colocalization of endothelial

Figure 2. Localization of the expression of M6P/IGF-II receptors in subcutaneous tumors. (a) Representative microscopic pictures showing

the staining for M6P/IGF-IIR and alpha-smooth muscle actin (a-SMA) in tumors. S, stroma; T, tumor. M6P/IGF-IIR staining was distributed

throughout the cells in B16 tumor but was more localized in C26 tumors. Arrows indicate the perinuclear localization of the receptor in

both tumor types. (b) Immunofluorescent pictures illustrating that M6P/IGF-IIR (green color) are not present on endothelial cells (CD31) as

there is no colocalization. The negative controls show the background for M6P/IGF-IIR and CD31 immunostaining without adding first

antibodies but secondary antibodies in tumor sections.

Can

cerTherapy



cells (CD31 staining) with M6P/IGF-IIR (Fig. 2b). In addi-tion, the CD31 staining shows that stroma has higher vascu-larisation compared with the tumor tissue.

Synthesis of M6P-HSA

Since M6P-containing ligands bind to M6P/IGF-IIR,6 weconjugated M6P sugar moieties to HSA to develop a carrierto target M6P/IGF-IIR expressed on tumor cells. M6P-HSAcontained approximately 28 M6P molecules per molecule ofalbumin as determined by the phosphate assay and anion-exchange chromatography analyses.27 This is above thethreshold of 21 sugar moieties which allows binding of con-structs to this target receptor.26,27

Cellular binding and uptake of M6P-HSA

Figure 3a illustrates the binding and uptake of radiolabeled(125I)-M6P-HSA in B16 cells. At 4�C, binding of 125I-M6P-HSA became constant after 1 hr, whereas the radioactivityat 37�C was continuously increased with time until 6 hr.The increase in the radioactivity at 37�C indicates thatM6P-HSA was internalized into the cells. M6P-specificbinding and uptake was shown with competition studies(Figs. 3b and 3c); both binding and uptake of 125I-M6P-HSA were inhibited by an excess of unlabeled M6P-HSAbut not by the excess of unmodified HSA or insulin-likegrowth factor-II (IGF-II), the non-M6P containing ligand ofM6P/IGF-IIR. These data also suggest that binding of M6P-

Figure 3. In vitro binding and uptake of M6P-HSA in B16-F10 cells. (a) The cell-associated radioactivity after incubating 125I-M6P-HSA with

B16 cells at 4�C (open circle) and at 37�C (closed circle). At 4�C, cells were detached from the plate at 24 hr so no measurement could be

performed at this time point. The experiments were performed in triplicate. (b) and (c) demonstrate the binding and uptake specificity of 125I-

M6P-HSA in B16 cells at 4�C and at 37�C, respectively. M6P–HSA and HSA concentrations were 1 mg/ml and IGF-II was 0.5 lg/ml. N ¼ 3. *p

< 0.05 and **p < 0.01 versus the control (125I-M6P-HSA). (d) Representative anti-HSA immunofluorescent staining (red fluorescence) showing

the binding of M6P-HSA in B16 cells. B16 cells (2 � 104 cells/well) were incubated with 100 lg/mL of M6P-HSA at 4�C for 2 hr. To

demonstrate the M6P/IGF-IIR–specific binding, cells were preincubated with goat anti-M6P/IGF-IIR antibody or goat anti-PDGF-bR (irrelevant

antibody as a control) for 1 hr at 4�C and then added with M6P-HSA. Anti-HSA staining was performed using rabbit anti-HSA antibody and

nuclear staining is DAPI. (a0) control, (b0) M6P-HSA, (c0) M6P-HSA plus anti-M6P/IGF-IIR, (d0) M6P-HSA plus anti-PDGF-bR. Pictures show that

binding of M6P-HSA to B16 cells was blocked with antibody against M6P/IGF-IIR but not with an irrelevant antibody, indicating the receptor-

specific binding. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Can

cerTherapy

Prakash et al. 1971


HSA should not interfere with the binding of IGF-II to thisreceptor.

We furthermore investigated the M6P/IGF-IIR–specific bind-ing of M6P-HSA. We blocked the receptor with anti-M6P/IGF-IIR antibody by pre-incubating for 1 hr and found that bindingof M6P-HSA was completely inhibited as shown with anti-HSAimmunofluorescent staining (Fig. 3d). Moreover, preincubationwith an irrelevant antibody (anti-PDGFb receptor) did not in-hibit the binding of M6P-HSA. These data suggest that M6P-HSA binds to B16 cells specifically through M6P/IGF-IIR.

We also examined the intracellular uptake of M6P-HSAin stromal cells like 3T3 fibroblasts and in B16 tumor cells.The intracellular immunofluorescent staining showed that

M6P-HSA was internalized by both 3T3 and B16 cells within2 hr (Fig. 4). The pattern of the accumulated M6P-HSA cor-responded with the perinuclear staining for M6P/IGF-IIR. Incontrast, unmodified HSA did not even bind to these cells.Similar results were obtained with C26 cells, shown by anti-HSA immunostaining (data not shown). These results suggestthat M6P-HSA is specifically taken up by M6P/IGF-IIR–expressing fibroblast-like stromal cells and tumor cells via areceptor-mediated process.

Tumor distribution of M6P28-HSA

Since tumor cells are able to internalize M6P-HSA in vitro, wedetermined the distribution of M6P-HSA in subcutaneous

Figure 4. In vitro uptake of M6P-HSA in 3T3 fibroblasts and B16-F10 cells. Representative anti-HSA immunofluorescent staining

(green fluorescence) showing the internalization of M6P-HSA in 3T3 fibroblasts (as a model for stromal cells) and in B16 cells. B16 or 3T3

cells (2 � 104 cells/ well) were incubated with 100 lg/mL of either M6P-HSA or HSA at 37�C for 2 hr. Cells were washed 3 times with

PBS, fixed, and anti-HSA (goat anti-HSA antibody) and anti-M6P/IGF-IIR stainings were performed. Arrows in the enlarged pictures illustrate

the accumulation of M6P-HSA into the cells which closely corresponds to the pattern of the receptor staining. No uptake of HSA was found

in both cell types. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Can

cerTherapy



tumors of B16 and C26 cells in mice. We found that M6P-HSA was highly distributed to the B16 and C26 tumor 2 hr af-ter an intravenous injection especially to the endothelial layer,as indicated by strong anti-HSA staining in tumor tissue(Fig. 5). The staining was also found to be associated with thestromal cells as shown in high magnification pictures. Thestaining in both tumors was stronger in stromal cells comparedwith tumor cells which might be related to the higher vascular-isation and the higher receptor expression in stromal cells com-

pared to tumor cells (see Fig. 2). From these results we con-cluded that M6P-HSA rapidly distributed to the tumors andaccumulated in both tumor cells and stromal cells.

Synthesis and characterization of Dox-HSA-M6P

Next, we conjugated doxorubicin to M6P-HSA to explore theapplicability of M6P-HSA as a drug carrier to tumor cells.Dox-HSA-M6P was successfully synthesized with a drug toprotein ratio of 6:1 (Fig. 6a). Modification of the protein may

Figure 5. In vivo targeting of M6P-HSA in B16-F10 and C26 subcutaneous tumors. M6P-HSA was administered intravenously in B16 or

C26 tumor-bearing mice and tumors were isolated 2 hr after the injection. Anti-HSA immunostaining (red color) was performed to localize M6P-

HSA in tumor cryostat sections. In both tumor types, M6P-HSA was strongly present in the endothelial layer and also distributed throughout the

tumor (panel a and b). The high magnification pictures show the accumulation of the carrier in stromal cells indicated by arrows for the

intracellular localization of the carrier (panel c and d). Control pictures show the anti-HSA immunostaining in untreated tumors (panel e and f).Can

cerTherapy

Prakash et al. 1973


form oligomers or polymers which are generally engulfed bymacrophages. We therefore purified Dox-HSA-M6P conju-gate using gel filtration chromatography to obtain monomericform (96% in the final product; Fig. 6b). Furthermore, wefound that modification of HSA with M6P provided a nega-tive charge to the molecule which led to a shift of the M6P-HSA peak to the right-side in the FPLC-chromatogramwhereas coupling of doxorubicin to M6P-HSA only slightlyaffected the net negative charge (Fig. 6c).

Doxorubicin was coupled to M6P-HSA through the acid-sensitive hydrazone linkage to induce the release of doxorubi-cin at lower pH inside the lysosomes after receptor internal-ization. Indeed, we found that doxorubicin was rapidlyreleased at lysosomal pH 3.0 (86%) and 5.0 (61%) butremained rather stable at physiological pH 7.0 with a releaseof only 12% (Fig. 6d). These results suggest that Dox-HSA-M6P can release doxorubicin in lysosomes while remainingstable at extracelluar pH levels.

Cellular uptake and in vitro effects of Dox-HSA-M6P

We examined the binding and uptake of Dox-HSA-M6P inB16 cells to ensure that doxorubicin modification did notaffect the binding properties of M6P-HSA. We found that125I-labeled Dox-HSA-M6P bound and internalized into thecells indicated by the enhanced cell-associated radioactivity at37�C than 4�C (p < 0.05, Fig. 7a). In addition, the cell-asso-ciated radioactivity at 4�C and at 37�C was clearly blockedby an excess of M6P-HSA but not by HSA indicating itsspecificity for the M6P-recognizing receptor. Similar resultswere obtained when these experiments were performed inC26 colon carcinoma cells (data not shown).

Next, we compared the rate and amount of doxorubicindelivered by the Dox-HSA-M6P conjugate into the cells (Fig.7b). We found a continuous increase of doxorubicin levelswithin the cells at 37�C which indicates that the conjugateeffectively delivered doxorubicin at a constant rate at leastuntil t ¼ 6 hr. To visualize the uptake of Dox-HSA-M6P, we

Figure 6. Synthesis and characterization of Dox-HSA-M6P. (a) Chemical structure of Dox-HSA-M6P with 6 molecules of doxorubicin coupled

through the hydrazone linkage and 28 molecules of M6P-HSA covalently bound to one molecule of albumin (HSA). (b) The chromatogram

of size-exclusion chromatography illustrates that purified Dox-HSA-M6P conjugate contained 96% monomers. (c) The overlay of the anion-

exchange column chromatogram for HSA, M6P-HSA and Dox-HSA-M6P. (d) pH-dependent release of doxorubicin from Dox-HSA-M6P.

The conjugate released doxorubicin at lysosomal pH (3.0 and 5.0) but remained rather stable at pH 7.0. The experiments were performed

in triplicate representing each symbol as mean 6 SEM.

Can

cerTherapy



incubated the construct with B16 cells and examined intra-cellular doxorubicin accumulation with fluorescent micro-scope at different time points. We found that Dox-HSA-M6P was increasingly accumulated into the cells as indicatedby the increasing intensity of the doxorubicin fluorescence(Fig. 7c).

Since Dox-HSA-M6P was efficiently taken up by the cells,we allowed the conjugate to be internalized only for 2 hr andthen washed out the unbound conjugate. Thereafter cell via-bility was determined at t ¼ 46 hr. We found that Dox-HSA-M6P killed tumor cells effectively with an IC50 value of0.43 lM in B16 cells and 1.1 lM in C26 cells (Figs. 8a and8b). In 3T3 cells only the cell growth was significantly inhib-ited (Fig. 8c). The efficacy of free doxorubicin (IC50 ¼0.15 lM in B16 cells and 1.0 lM in C26 cells) was similar tothe conjugate in tumor cells but higher in 3T3 cells (IC50 ¼6.57 lM). Again, these effects in B16 cells could be blockedby an excess of M6P-HSA and not by HSA (Fig. 8d), indicat-

ing that the effects of Dox-M6P-HSA were mediated via aM6P-recognizing receptor.

In vivo distribution of Dox-HSA-M6P

We determined the whole body distribution of 125I-Dox-HSA-M6P in B16 tumor-bearing mice after 30 min of theinjection. Data showed that the radiolabeled construct wasdistributed to the tumors and other organs such as liver,lungs, spleen and kidneys (Fig. 9). The accumulation instomach and small intestine might be due to the released io-dine label that normally accumulates in these organs.

Furthermore, we compared the accumulation of Dox-HSA-M6P and free doxorubicin in tumors and variousorgans at 2 hr after injections by doxorubicin staining andanti-HSA immunostaining. Doxorubicin fluorescent picturesdemonstrated that Dox-HSA-M6P accumulated into tumorin intact-form whereas there was very weak staining in freedoxorubicin-treated tumors. Free doxorubicin was detected in

Figure 7. In vitro uptake of the Dox-HSA-M6P conjugate. (a) Binding and uptake of the radiolabeled (125I)-Dox-HSA-M6P in B16 cells after

incubating for 2 hr at 4�C (dark grey bars) or 37�C (light grey bars). M6P-HSA and HSA concentrations were 1 mg/mL. Mean 6 SEM, n ¼ 3.

*p < 0.05 and **p < 0.01. (b) Uptake efficiency of Dox-HSA-M6P in B16 cells. The cell-associated doxorubicin levels indicate to the bound

plus free drug as determined with HPLC, triplicate samples. (c) Microscopic pictures show the doxorubicin fluorescence in B16 cells at

different time points after incubating with Dox-HSA-M6P conjugate (equivalent to 10 lg/ml doxorubicin) at 37�C. [Color figure can be

viewed in the online issue, which is available at www.interscience.wiley.com.]

Can

cerTherapy

Prakash et al. 1975


all analyzed organs but Dox-HSA-M6P was accumulated inliver and spleen apart from tumors. The data was confirmedwith anti-HSA immunostaining in Dox-HSA-M6P-treatedanimals (Fig. 10). The anti-HSA staining was found to belocalized in non-parenchymal cells of the liver and red-pulpof the spleen.

In vivo effects of Dox-HSA-M6P on tumor growth

For the proof-of-principle, whether targeting of doxorubi-cin through M6P-HSA would improve the efficacy of dox-orubicin, we investigated the effects of Dox-HSA-M6Pconjugate in B16-F10 subcutaneous tumors in mice. Treat-ment with Dox-HSA-M6P significantly inhibited the tu-mor growth compared to the vehicle–treated tumors asindicated by the reduced tumor size in the conjugate–treated group (Figs. 11a and 11b). In contrast, equimolardoses of free doxorubicin did not reduce tumor growth,even with higher number of doses. We also investigatedwhether the M6P-HSA and the doxorubicin construct

Figure 8. Effects of Dox-HSA-M6P on the growth of tumor cells and fibroblasts. Cells were incubated with Dox-HSA-M6P conjugate (filled

triangle) or free doxorubicin (filled square) only for 2 hr, washed and then cell viability was determined after 46 hr. N ¼ 3. (d) M6P/IGF-IIR-

specific activity of Dox-HSA-M6P in B16 cells. The effect of Dox-HSA-M6P (DHM) on target cells was significantly blocked by a 5-fold higher

concentration of M6P-HSA but not by a 5-fold higher concentration of unmodified HSA. **p < 0.01.

Figure 9. Organ-distribution of 125I-Dox-HSA-M6P in vivo. B16

tumor-bearing mice were injected with a tracer dose of 125I-Dox-HSA-

M6P, and after 30 min tumors and many organs were collected in

tubes and analyzed with c-counter for determining the radioactivity in

these samples. Data represent the average of n¼ 3 mice.

Can

cerTherapy



Figure 10. Localization of Dox-HSA-M6P and doxorubicin in B16 tumors and normal organs. Representative microscopic pictures of the

doxorubicin staining (red fluorescence) and anti-HSA immunohistochemical staining (red color) in Dox-HSA-M6P and doxorubicin treated

B16 tumor-bearing mice to localize the conjugate in tumor, kidneys, liver, heart, spleen and lungs. In B16 tumor bearing mice, a single

dose of Dox-HSA-M6P and doxorubicin was administered and after 2 hr different organs were isolated and examined for fluorescent

staining and for anti-HSA staining.

Can

cerTherapy

Prakash et al. 1977


displayed side effects with the applied doses. We foundthat any of the treatments (PBS, M6P-HSA, doxorubicinor Dox-HSA-M6P) did not show any side effect as deter-mined by the measurement of body weights, blood cellcount (white blood cells, red blood cells and platelet) andparameters reflecting liver function (ALT, AST and alka-line phosphatase plasma levels; Table 1). Since we foundthe accumulation of our construct in nonparenchymalkupffer cells in liver and in spleen, we performed the im-munostaining for F4/80 antigen (marker for macrophagese.g. kupffer cells and macrophage cells in red pulp ofspleen). We found that there was no toxicity on thesecells as there was no difference in the cellular stainingand intensity in the Dox-HSA-M6P treated mice (Fig.11c). There was also no mortality in these groups.

Altogether, our in vivo data show that Dox-HSA-M6Paccumulated into tumors which is associated with anenhanced therapeutic efficacy of Dox-HSA-M6P compared tofree doxorubicin.

DiscussionThe current study is the first example of a targeted therapyto tumors through M6P/IGF-IIR. Although several app-roaches have been proposed for tumor-specific delivery inparticular with antibodies and peptides,5,31,32 further develop-ments are required in this field as these approaches are lim-ited by the absence of target receptors in certain tumor cellsand low drug internalization due to limited receptor-medi-ated endocytosis. The present study offers a novel strategy todeliver anticancer drugs to the tumor cells through the insu-lin-like growth factor-II receptor using M6P-HSA carrier.

M6P/IGF-IIR have been reported to be expressed in sev-eral human tumors10–14; however, only a few studies showedits expression and distribution in animal models.33,34 Thecurrent study showed that M6P/IGF-IIR was abundantlyexpressed on tumor cells and fibroblast-like stromal cells inB16 and C26 tumors. This may allow the access of M6P/IGF-IIR-specific carrier to both tumor and stromal cells. Fur-thermore, low expression of this receptor in major organslike kidney, liver, lung heart and spleen (Fig. 1) indicates itspreference for tumors. However, there are evidence showingthat M6P/IGF-IIR expression is inversely related to the tumorgrowth and invasiveness in many tumors.17 Therefore, thepresent drug targeting approach might be limited to thetumors that possess the M6P/IGF-IIR.

In the present study, efficient uptake of M6P-HSA carrierin many M6P/IGF-IIR-expressing cells indicates its high cel-lular uptake capacity. The rapid internalization may be a val-uable asset of our drug carrier, because it allows release ofdrugs after intercellular degradation of the biodegradable car-rier. Targeting to a specific receptor on tumor cells alwaysconfers the risk of receptor mutations but our receptor com-petition studies (Fig. 3c) show that the uptake is specific forM6P-containing moieties and the M6P-binding site is knownto be affected to a lesser extent.25 In line of our earlier

Figure 11. In vivo efficacy of Dox-HSA-M6P in B16-F10

subcutaneous tumor. B16 tumor-bearing mice were treated either

with PBS, Dox-HSA-M6P (290 lg/mouse/day equivalent to

16.5 lg/mouse/day doxorubicin), M6P-HSA (290 lg/mouse/day)

or free doxorubicin (16.5 lg/mouse/day) intravenously after

attaining the tumor size of 6100 mm3. Panel (a) shows the

representative pictures of the isolated B16 tumors at the end of

the experiment and panel (b) illustrates the tumor growth curve at

different days. Treatment with Dox-HSA-M6P significantly (*p <

0.05, **p < 0.01 vs. vehicle group) reduced the growth of the

tumor while free doxorubicin did not show any beneficial effects.

Two graphs indicate separate experiments and tumor pictures

represent both experiments; number of animals per group ¼ 6. (c)

Representative pictures showing the immunostaining for F4/80

antigen, a marker for macrophages, in liver and spleen of the

PBS– and Dox-HSA-M6P–treated mice. Pictures show no change in

the staining after the treatment with Dox-HSA-M6P indicating no

toxicity of the construct in Kupffer cells in liver and in

spleenocytes. [Color figure can be viewed in the online issue,

which is available at www.interscience.wiley.com.]

Can

cerTherapy



studies with M6P-HSA,26 we now demonstrate that M6P-HSA also accumulates into the fibroblasts-like stromal cellsin addition to tumor cells, which may be an additional bene-fit for tumor targeting. Delivery of anticancer drugs to stro-mal cells is gaining increased attention as a possible thera-peutic strategy for some types of tumors.35,36 The stromalcells contribute to the proliferation and invasion of tumorcells via secretion of several growth factors, cytokines andangiogenic factors, and consequently the delivery of drugsinto these particular cells may be quite beneficial.35

To directly demonstrate the application of M6P-HSA as adrug carrier, we conjugated doxorubicin to the core proteinHSA of the carrier through an acid-sensitive linkage. Asshown with different techniques, coupling of doxorubicin didnot affect the cell binding behavior of M6P-HSA. In addition,the Dox-HSA-M6P conjugate could be internalized by cellsafter a short exposure and the active drug was released intra-cellularly as indicated by the in vitro dose-effect studies (Fig.8). Although both free and targeted doxorubicin was equallyeffective, no benefits of the targeted drug can be expectedin vitro. Because free drug diffuses into the cells more rapidlycompared with the receptor-mediated process for the conju-gate (data not shown).

In vivo, Dox-HSA-M6P distributed to the tumors shortlyafter a single intravenous injection as shown with radiola-beled experiments and anti-HSA staining. Doxorubicin stain-ing in tumor tissues demonstrated that the Dox-HSA-M6Pconjugate was stable enough to reach tumors. In contrast,free doxorubicin was not detectable in tumors but depositedin most of the organs whereas targeted doxorubicin wasabsent in the organs (kidneys and heart) that are susceptibleto doxorubicin-induced toxicity. However, benefit of the tar-geted doxorubicin could be clearly demonstrated as Dox-HSA-M6P significantly inhibited the tumor growth of B16tumor compared with free drug. In addition, selective accu-mulation of Dox-HSA-M6P in tumors was shown with anti-HSA staining which might explain its higher efficacy.

Although we still have to examine the maximum tolerateddose for this construct, the applied doses were much lowerthan the doses used with many other published doxorubicindelivery systems such as polymers, liposomes, micelles andantibodies.37–40 Also, no signs of toxicity such as loss in bodyweight, change in blood cell counts, liver and kidneys func-tions, and number of kupffer cells and spleenocytes with thetargeted construct illustrate the safety of our approach.

In summary, the present study reveals a novel strategy todeliver anti-cancer drugs to tumor cells through the M6P/IGF-II receptor using the M6P28-HSA carrier. M6P-HSA wasefficiently taken up by both tumor and stromal cells in vitroand in vivo. M6P-HSA was successfully employed to deliverdoxorubicin to B16 tumors in mice which resulted in theenhanced efficacy of doxorubicin. Conclusively, our drug tar-geting approach using the M6P-HSA carrier offers newopportunities to deliver anticancer drugs to tumors for thetreatment of cancer.

AcknowledgementsCatharina Reker-Smit, Eduard Post and Annemiek M. van Loenen-Wee-maes are acknowledged for their excellent technical assistance. Authorsalso thank J.H. Pol from the Department of the Nuclear Medicine for theradiolabeling of the proteins. This study was supported by STW Valorisa-tion Grant, The Netherlands. The novelty and impact of the article is thatachieving therapeutic levels of anticancer drugs within tumors hasremained a challenge due to existence of several biological barriers, andpoor pharmacokinetic profile and narrow therapeutic window of thesedrugs. Receptor-mediated drug delivery to tumors is of great interest sinceit provides an intracellular delivery with high specificity. However, targetreceptors are usually present on either tumor, endothelial or stromal cellsof a tumor which limits the wide applicability of drug delivery approachesand therefore need combinational therapies. In the present study, we dem-onstrate a novel approach to deliver anticancer drugs to both tumor andstromal cells through mannose-6-phosphate/insulin-like growth factor-IIreceptor that is widely expressed in many human tumors. This is the firststudy showing tumor delivery of anticancer drugs through this receptor.Our novel drug targeting strategy may provide a new platform for antitu-mor therapies.

Table 1. Effect of different treatments on the liver function (plasma levels of alkaline phosphatase, AST, ALT), renal function (plasmacreatinine levels), blood cell counts (WBC, RBC and Platelets) and body weight of the tumor-bearing mice

Vehicle Dox-HSA-M6P M6P-HSA Doxorubicin

Alkaline phosphatase (U/l) 39.7 6 2.1 44.0 6 3.6 36.7 6 4.5 38 6 4.0

AST (U/l) 290 6 28.1 248 6 31.6 362.7 6 79.8 307 6 110

ALT (U/l) 31.7 6 1.76 54.2 6 8.1 35.7 6 3.14 61.5 6 20.5

Creatinine (lmol/l) 8.0 6 0.82 10.0 6 1.1 10.5 6 1.55 14.33 6 2.20

WBC count (�109/l) 11.0 6 0.65 9.2 6 1.9 13.2 6 0.95 10.4 6 3.0

RBC count (�1012/l) 7.8 6 0.40 8.4 6 0.41 7.54 6 0.38 7.13 6 0.68

Platelets count (�109/l) 1,243 6 46.6 1,376 6 60.7 1,245 6 117 1,226 6 170

Body weight (start of the treatment; g) 30.7 6 0.48 29.2 6 0.56 29.3 6 0.67 29.5 6 0.83

Body weight � tumor weight (end of the experiment; g) 30.9 6 0.55 29.1 6 0.54 29.9 6 0.57 30.1 6 0.35

Mean 6 SEM, n ¼ 5–6 mice per group.Abbreviations: AST, aspartate aminotransferase; ALT, alanine aminotransferase; WBC, white blood cells; RBC, red blood cells.

Can

cerTherapy

Prakash et al. 1979


References

1. Chari RV. Targeted cancer therapy:conferring specificity to cytotoxic drugs.Acc Chem Res 2008;41:98–107.

2. Shukla GS, Krag DN. Selective delivery oftherapeutic agents for the diagnosis andtreatment of cancer. Expert Opin Biol Ther2006;6:39–54.

3. Rihova B. Receptor-mediated targeted drugor toxin delivery. Adv Drug Deliv Rev1998;29:273–89.

4. Vyas SP, Singh A, Sihorkar V. Ligand-receptor-mediated drug delivery: anemerging paradigm in cellular drugtargeting. Crit Rev Ther Drug Carrier Syst2001;18:1–76.

5. Torchilin VP. Recent approaches tointracellular delivery of drugs and DNAand organelle targeting. Annu Rev BiomedEng 2006;8:343–75.

6. Ghosh P, Dahms NM, Kornfeld S.Mannose 6-phosphate receptors: new twistsin the tale. Nat Rev Mol Cell Biol 2003;4:202–12.

7. Hassan AB. Keys to the hidden treasuresof the mannose 6-phosphate/insulin-likegrowth factor 2 receptor. Am J Pathol2003;162:3–6.

8. Dahms NM, Olson LJ, Kim JJ. Strategiesfor carbohydrate recognition by themannose 6-phosphate receptors.Glycobiology 2008;18:664–78.

9. Funk B, Kessler U, Eisenmenger W,Hansmann A, Kolb HJ, Kiess W.Expression of the insulin-like growthfactor-II/mannose-6-phosphate receptor inmultiple human tissues during fetal lifeand early infancy. J Clin Endocrinol Metab1992;75:424–31.

10. Berthe ML, Esslimani SM, Roger P, GleizesM, Lemamy GJ, Brouillet JP, Rochefort H.Mannose-6-phosphate/insulin-like growthfactor-II receptor expression levels duringthe progression from normal humanmammary tissue to invasive breastcarcinomas. Eur J Cancer 2003;39:635–42.

11. Ishiwata T, Bergmann U, Kornmann M,Lopez M, Beger HG, Korc M. Alteredexpression of insulin-like growth factor IIreceptor in human pancreatic cancer.Pancreas 1997;15:367–73.

12. Pavelic K, Kolak T, Kapitanovic S,Radosevic S, Spaventi S, Kruslin B, PavelicJ. Gastric cancer: the role of insulin-likegrowth factor 2 (IGF 2) and its receptors(IGF 1R and M6-P/IGF 2R). J Pathol 2003;201:430–8.

13. Laube F. Mannose-6-phosphate/insulin-likegrowth factor-II receptor in humanmelanoma cells: effect of ligands andantibodies on the receptor expression.Anticancer Res 2009;29:1383–8.

14. Cariani E, Lasserre C, Seurin D, HamelinB, Kemeny F, Franco D, Czech MP,Ullrich A, Brechot C. Differentialexpression of insulin-like growth factor IImRNA in human primary liver cancers,benign liver tumors, and liver cirrhosis.Cancer Res 1988;48:6844–9.

15. De Leon DD, Issa N, Nainani S, AsmeromY. Reversal of cathepsin D routingmodulation in MCF-7 breast cancer cellsexpressing antisense insulin-like growthfactor II (IGF-II). Horm Metab Res 1999;31:142–7.

16. Wise TL, Pravtcheva DD. Delayed onset ofIgf2-induced mammary tumors in Igf2rtransgenic mice. Cancer Res 2006;66:1327–36.

17. Hebert E. Mannose-6-phosphate/insulin-like growth factor II receptor expressionand tumor development. Biosci Rep 2006;26:7–17.

18. De Souza AT, Hankins GR, WashingtonMK, Orton TC, Jirtle RL. M6P/IGF2R geneis mutated in human hepatocellularcarcinomas with loss of heterozygosity.Nat Genet 1995;11:447–9.

19. De Souza AT, Yamada T, Mills JJ, JirtleRL. Imprinted genes in livercarcinogenesis. FASEB J 1997;11:60–7.

20. Jang HS, Kang KM, Choi BO, Chai GY,Hong SC, Ha WS, Jirtle RL. Clinicalsignificance of loss of heterozygosity forM6P/IGF2R in patients with primaryhepatocellular carcinoma. World JGastroenterol 2008;14:1394–8.

21. Yamada T, De Souza AT, Finkelstein S,Jirtle RL. Loss of the gene encodingmannose 6-phosphate/insulin-like growthfactor II receptor is an early event in livercarcinogenesis. Proc Natl Acad Sci USA1997;94:10351–5.

22. Hu CK, McCall S, Madden J, Huang H,Clough R, Jirtle RL, Anscher MS. Loss ofheterozygosity of M6P/IGF2R gene is anearly event in the development of prostatecancer. Prostate Cancer Prostatic Dis 2006;9:62–7.

23. Kong FM, Anscher MS, Washington MK,Killian JK, Jirtle RL. M6P/IGF2R ismutated in squamous cell carcinoma of thelung. Oncogene 2000;19:1572–8.

24. Hankins GR, De Souza AT, Bentley RC,Patel MR, Marks JR, Iglehart JD, Jirtle RL.M6P/IGF2 receptor: a candidate breasttumor suppressor gene. Oncogene 1996;12:2003–9.

25. Devi GR, De Souza AT, Byrd JC, Jirtle RL,MacDonald RG. Altered ligand binding byinsulin-like growth factor II/mannose6-phosphate receptors bearing missensemutations in human cancers.Cancer Res 1999;59:4314–19.

26. Beljaars L, Molema G, Weert B, BonnemaH, Olinga P, Groothuis GM, Meijer DK,Poelstra K. Albumin modified withmannose 6-phosphate: a potential carrierfor selective delivery of antifibrotic drugsto rat and human hepatic stellate cells.Hepatology 1999;29:1486–93.

27. Beljaars L, Olinga P, Molema G, de BP,Geerts A, Groothuis GM, Meijer DK,Poelstra K. Characteristics of the hepaticstellate cell-selective carrier mannose 6-phosphate modified albumin (M6P(28)-HSA). Liver 2001;21:320–8.

28. Sando GN, Karson EM. p-Isothiocyanatophenyl 6-phospho-alpha-D-mannopyranoside coupled to albumin. Amodel compound recognized by thefibroblast lysosomal enzyme uptake system.1. Chemical synthesis and characterization.Biochemistry 1980;19:3850–5.

29. Willner D, Trail PA, Hofstead SJ, KingHD, Lasch SJ, Braslawsky GR, GreenfieldRS, Kaneko T, Firestone RA.(6-Maleimidocaproyl)hydrazone ofdoxorubicin–a new derivative for thepreparation of immunoconjugates ofdoxorubicin. Bioconjug Chem 1993;4:521–7.

30. Sengupta S, Eavarone D, Capila I, Zhao G,Watson N, Kiziltepe T, Sasisekharan R.Temporal targeting of tumour cells andneovasculature with a nanoscale deliverysystem. Nature 2005;436:568–72.

31. Ricart AD, Tolcher AW. Technologyinsight: cytotoxic drug immunoconjugatesfor cancer therapy. Nat Clin Pract Oncol2007;4:245–55.

32. Brumlik MJ, Daniel BJ, Waehler R, CurielDT, Giles FJ, Curiel TJ. Trends inimmunoconjugate and ligand-receptorbased targeting development for cancertherapy. Expert Opin Drug Deliv 2008;5:87–103.

33. O’Gorman DB, Weiss J, Hettiaratchi A,Firth SM, Scott CD. Insulin-like growthfactor-II/mannose 6-phosphate receptoroverexpression reduces growth ofchoriocarcinoma cells in vitro and in vivo.Endocrinology 2002;143:4287–94.

34. Xie S, Kang JX. Differential expression ofthe mannose 6-phosphate/insulin-likegrowth factor-II receptor in human breastcancer cell lines of different invasivepotential. Med Sci Monit 2002;8:BR293–BR300.

35. Bouzin C, Feron O. Targeting tumorstroma and exploiting mature tumorvasculature to improve anti-cancer drugdelivery. Drug Resist Updat 2007;10:109–20.

36. Hofmeister V, Schrama D, Becker JC.Anti-cancer therapies targeting the tumor

Can

cerTherapy



stroma. Cancer Immunol Immunother2008;57:1–17.

37. Xiong XB, Huang Y, Lu WL, Zhang X,Zhang H, Nagai T, Zhang Q. Enhancedintracellular delivery and improvedantitumor efficacy of doxorubicin bysterically stabilized liposomes modifiedwith a synthetic RGD mimetic. J ControlRelease 2005;107:262–75.

38. Trail PA, Willner D, Bianchi AB,Henderson AJ, TrailSmith MD, Girit E,

Lasch S, Hellstrom I, Hellstrom KE.Enhanced antitumor activity ofpaclitaxel in combination with theanticarcinoma immunoconjugate BR96-doxorubicin. Clin Cancer Res 1999;5:3632–8.

39. Kamada H, Tsutsumi Y, Yoshioka Y,Yamamoto Y, Kodaira H, Tsunoda S,Okamoto T, Mukai Y, Shibata H,Nakagawa S, Mayumi T. Design of a pH-sensitive polymeric carrier for drug

release and its application in cancertherapy. Clin Cancer Res 2004;10:2545–50.

40. Ko J, Park K, Kim YS, Kim MS, Han JK,Kim K, Park RW, Kim IS, Song HK, LeeDS, Kwon IC. Tumoral acidic extracellularpH targeting of pH-responsive MPEG-poly(beta-amino ester) blockcopolymer micelles for cancer therapy.J Control Release 2007;123:109–15.

Can

cerTherapy

Prakash et al. 1981


tumor-targeted intracellular delivery of anticancer drugs through the...

Documents