lable at ScienceDirect

Biomaterials 34 (2013) 9142e9148

Contents lists avai

Biomaterials

journal homepage: www.elsevier .com/locate/biomateria ls

Choline transporter-targeting and co-delivery system for gliomatherapy

Jianfeng Li, Yubo Guo, Yuyang Kuang, Sai An, Haojun Ma, Chen Jiang*

Key Laboratory of Smart Drug Delivery (Fudan University), Ministry of Education, Department of Pharmaceutics, School of Pharmacy, Fudan University, 826Zhangheng Road, Shanghai 201203, China

a r t i c l e i n f o

Article history:Received 6 June 2013Accepted 12 August 2013Available online 28 August 2013

Keywords:Cancer therapyCholine transporterCo-deliveryDual targetingNanoparticles

* Corresponding author. Tel./fax: þ86 21 5198 0079E-mail address: jiangchen@shmu.edu.cn (C. Jiang)

0142-9612/$ e see front matter � 2013 Elsevier Ltd.http://dx.doi.org/10.1016/j.biomaterials.2013.08.030

a b s t r a c t

Combination of gene therapy and chemotherapy is a promising approach for glioma therapy. In thisstudy, a co-delivery system of plasmid encoding human tumor necrosis factor-related apoptosis-inducingligand (pORF-hTRAIL, Trail) and doxorubicin (DOX) has been simply constructed in two steps. Firstly, DOXwas intercalated into Trail to form a stable complex. Secondly, DOX-Trail complex was condensed byDendrigraft poly-L-lysine (DGL) to form a nanoscaled co-delivery system. Choline transporters are bothexpressed on bloodebrain barrier (BBB) and glioma, Herein, a choline derivate with high cholinetransporter affinity was chosen as BBB and glioma dual targeting ligand. Choline-derivate modified co-delivery system showed higher cellular uptake efficiency and cytotoxicity than unmodified co-deliverysystem in U87 MG cells. In comparison with single medication or unmodified delivery system,Choline-derivate modified co-delivery system induced more apoptosis both in vitro and in vivo. Thetherapeutic efficacy on U87 MG bearing xenografts further confirmed the predominance of this dualtargeting and co-delivery system.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Current treatments for glioblastoma multiforme (GBM) areinsufficient with a nearly universal recurrence after surgery, radi-ation therapy, and chemotherapy [1]. Poor chemotherapy outcomeafter surgery contributes to this high recurrence. The bloodebrainbarrier (BBB) existing at the early stage or the tumor margin ofglioma limits the penetration of most therapeutic agents [2]. Eventhough some chemotherapeutics could permeate across the BBB,single medication could not obtain optimal efficacy due to the drugresistance. Based on the above consideration, two strategiesincluding glioma specific targeting and combination therapy wereemployed for better outcomes.

For glioma targeting, two-order targeting strategy wascommonly employed in designing drug delivery or diagnosis sys-tems to overcome the BBB and accumulate into tumor. In brief, oneligand with binding affinity to the receptors or transportersexpressed on the vasculature endothelial cells could facilitate thetranscytosis across the BBB. Another ligand with high affinity to thereceptors overexpressed on glioma cells could enhance the accu-mulation into tumor [3e5]. Dual targeting to BBB and glioma with

..

All rights reserved.

one ligand was another choice. For example, Angiopep-2, one of thefamily of Kunitz domain-derived peptides, could target to the low-density lipoprotein receptor-related protein-1 (LRP1) expressed onbrain capillary endothelial cells (BCECs) and glial cells [6,7]. In ourprevious study, a choline derivate with high BBB choline trans-porter affinity was challenged to mediate the gene delivery intobrain [8]. Because of high malignancy, choline transporters areoverexpressed on glioma cells. This indicates that choline trans-porters may facilitate the glioma targeting [9]. In another work, wedeveloped a choline-derivate modified nanoprobe for glioma MRIimaging and gained notable signal contrast between tumor andnormal brain region [10]. Herein, choline derivate was selected as aBBB and glioma dual targeting ligand.

Combination therapy in glioma has attractedwide attention dueto the inefficiency of single medication. Attempts have been madeon the co-delivery of gene drug and chemotherapeutics. The syn-ergistic effect arisen from the co-delivery of human tumor necrosisfactor-related apoptosis-inducing ligand (pORF-hTRAIL, Trail) andpaclitaxel or doxorubicin (DOX) has been proved to gain enhancedanti-tumor effect [11,12]. In this study, Trail and DOX were chosenas model drugs.

A co-delivery system of DOX and nucleic acid was reported byBagalkot et al. [13]. DOX is known to intercalate within the doublehelix of DNA strand due to the presence of flat aromatic rings in thismolecule. At present, aptamer and plasmid DNAwere developed as

J. Li et al. / Biomaterials 34 (2013) 9142e9148 9143

vectors which could form stable complex upon DOX intercalation.Combined chemoimmunotherapy using plasmid-DOX complexachieved comparable therapeutic effect and reduced cardiotox-icities to DOX in two different tumor models. The therapeuticbenefits came from the improved pharmacokinetics of DOX byincreasing blood circulation. In turn, intercalation (complexation)can also protect the plasmid against nucleases [14]. While, a flawthat mars perfection may be the small size of complex resulting inrather fast renal clearance. To take the advantage of enhancedpermeability and retention (EPR) effect, plasmid-DOX complex wasfurther condensed by dendrimer to form nanoparticles (NPs). Thisnanosized co-delivery system preferentially accumulated in tumorand significantly inhibited its growth in a subcutaneous tumormodel [15].

In this study, a co-delivery system for glioma therapy wasestablished. Firstly, DOX was intercalated into Trail to form a stablecomplex. Trail served both as a DOX carrier and a gene drug.

Scheme 1. Construction of dual targeting and co-delivery system. DOX was intercalated inderivate modified DGL. This co-delivery system could accumulate into glioma cells by EPRtion therapy on glioma.

Secondly, DOX-Trail complex was condensed by Dendrigraft poly-L-lysine (DGL) to form a nanoscaled co-delivery system (Scheme1). Choline derivate was further conjugated to this system torealize the glioma specific drug delivery. Targeting efficiency andanti-tumor effect of the system were evaluated both in vitro andin vivo.

2. Materials and method

2.1. Materials

DGL generation 3 with 123 lysine groups were purchased from Colcom, France.a-Malemidyl-u-N-hydrox-ysuccinimidyl polyethyleneglycol (NHS-PEG-MAL, MW3500) was obtained from Jenkem Technology (Beijing, China). Doxorubicin hydro-chloride (DOX$HCl) was purchased from Beijing Huafeng United Technology Corp.The plasmid Trail and pGL2-control vector (InvivoGen, San Diego, CA, USA) werepurified using QIAGEN Plasmid Mega Kit (Qiagen GmbH, Hilden, Germany). AnnexinV-FITC TUNEL Apoptosis Assay Kit and TdT In Situ Apoptosis Detection Kit werepurchased from KeyGEN (Nanjing, China). Near-infrared probe (NIR783) was kindly

to Trail plasmid to form a stable complex which was further condensed by choline-and dual targeting effect. After cytoplasm releasing, Trail and DOX educe combina-

J. Li et al. / Biomaterials 34 (2013) 9142e91489144

provided by Prof. Cong Li (Molecular Imaging Institute, School of Pharmacy, FudanUniversity). Other reagents, if not specified, were purchased from SigmaeAldrich.

2.2. Cell lines and animals

U87 MG cells were obtained from American Type Culture Collection (ATCC,Rockville, MD, USA) and cultured at 37 �C in a humidified 5% CO2 atmosphere.Growth medium was Dulbecco’s modified Eagle’s Medium (DMEM) which supple-mented with 10% fetal bovine serum (FBS), 100 U/ml penicillin and 100 mg/mlstreptomycin. Male Balb/c nude mice of about 20 g body weight were purchasedfrom the Department of Experimental Animals, Fudan University and maintainedunder standard housing conditions. All animal experiments were carried out inaccordance with guidelines evaluated and approved by the ethics committee ofFudan University. For U87 MG xenograft model, nude mice were anaesthetized byintraperitoneal injection of 10% Chloral hydrate. U87 MG cells (1 � 105 in 5 ml PBS)were implanted into the right striatum (1.8 mm right lateral to the bregma and3 mm of depth) of the mice by using a stereotactic fixation device with mouseadaptor.

2.3. Preparation and characterization of DOX-Trail complex

A physical complex between Trail and DOX was made as described previously[15]. Briefly, DOX was dissolved in methanol at the concentration of 1 mg/ml as astock solution. Different picomoles of Trail were added in a fixed concentration ofDOX (3 mM) in PBS buffer, the mixture were vortexed for 1 min and the fluorescenceof DOX was monitored at excitation 480 nm and emission was recorded in the in-terval of 520e680 nm on a Perkine Elmer LS-55 spectrofluorometer. The florescencequenching was visualized by Cambridge Research & Instrumentation (CRi) in vivoimaging system (CRi, MA, USA).

2.4. Preparation and characterization of dendrimers/DOX-Trail NPs

DP (DGL-PEG) and DPC (DGL-PEG-10) were synthesized as described previously[8]. Dendrimers (DGL-PEG, and DGL-PEG-10) were diluted to appropriate concen-trations in PBS (pH 7.4). DOX-Trail complex or Trail solution (100 mg DNA/ml, 50 mM

sodium sulfate solution) was added at weight ratio of 6:1 (DGL to DNA) andimmediately vortexed for 30 s. The particle size distribution of DPC/DOX-Trail wasmeasured by a dynamic light scattering detector (Zetasizer, Nano-ZS, Malvern, UK)and transmission electron microscope (Tecnai G2 spirit Biotwin, FEI).

2.5. Cellular uptake and release of DOX from complex and NPs

U87 MG cells were seeded at a density of 5 � 104 cells/well in 6-well plates(Corning-Coaster, Tokyo, Japan). Cells were incubated for 48 h and checked underthe microscope for confluency and morphology. Then cells were incubated with freeDOX, DOX-Trail complex, DP/DOX-Trail and DPC/DOX-Trail at the concentration of1 mg/ml measured by DOX in the presence of serum freemedium for 30 min at 37 �C,respectively. Then cells were digested, centrifuged and washed twice with Hank’ssolution. The fluorescence intensity was analyzed using a flow cytometer (FACSCa-libur, BD Biosciences, Bedford, MA, USA) equipped with an argon ion laser (488 nm)as the excitation source. To evaluate the DOX release from complex and NPs, cellswere cultured in freshmedium containing FBS for another 2 h following the 30 min-incubation. For each flow cytometer analysis, 1 � 104 events were collected.

2.6. Cytotoxicity assay on U87 MG cells

Cytotoxicity assay was evaluated by MTT assay. U87 MG cells were seeded at adensity of 5 �103 cells/well in a 96-well plate. After 48 h incubation, cells were thenwashed twice with Hank’s solution and exposed to different concentrations of DOX,DPC/Trail, DP/DOX-Trail, DPC/DOX-Trail and DPC/DOX-NEG (negative control, DPC/DOX-pGL2) at 37 �C for 2 h in serum free medium. After 2 h of incubation, cellswere washed twice with Hank’s solution and re-fed with 60 mL of DMEM mediumcontaining 10% FBS for 48 h. To assess cell viability, 40 mL of MTT (1.25 mg/ml) so-lution was added into each well and incubated at 37 �C for 2 h. The medium wasremoved and 100 ml of DMSO was added to each well to dissolve the formazancrystals formed by the living cells. Cells without treatment were served as control.The absorbance was read at 570 nm and corrected at 630 nm by dual wavelengthdetection using a Multiskan MK3 microplate reader (Thermo Scientific). Cellviability was calculated as the survival percentage of control.

2.7. Apoptosis detection in vitro

U87 MG cells were seeded at a density of 5 � 104 cells/well in 24-well plates.Cells were incubated for 48 h. Then DOX, DPC/Trail, DP/DOX-Trail, DPC/DOX-Trailand DPC/DOX-NEG at the concentration of 3.16 mM measured by DGL were added tothe cells. The mixture was incubated at 37 �C for 2 h. The cells were washed twicewith Hank’s solution and further cultured in DMEMmedium containing 10% FBS for48 h. Then medium was removed. The cells were incubated with Annexin V-FITC

Apoptosis Detection Kit for 15 min. The fluorescence labeled apoptotic cells wereimaged using a fluorescence microscope.

2.8. In vivo imaging

At 18th day after glioma implantation, DP/DOX-Trail and DPC/DOX-Trail(6:1:0.16, DGL to DNA to DOX, w/w/w) at the dose of 50 mg DNA/mouse wereinjected into the tail vein of model mice, respectively. Images were taken 2 h afterinjection by Cambridge Research & Instrumentation (CRi) in vivo imaging system(CRi, MA, USA). Then mice were anesthetized with diethyl ether and killed bydecapitation and main organs were further visualized. DGL was pre-labeled byNIR783 dye.

2.9. In vivo anti-glioblastoma effect in U87 grafted model

At the 12th, 15th and 18th day after implantation, model mice were adminis-trated of DOX (5 mg/kg), DPC/Trail, DP/DOX-Trail and DPC/DOX-Trail (6:1:0.16, DGLto DNA to DOX, w/w/w) at a dose of 50 mg Trail or/and 8 mg DOX/mouse. At the 21stday, animals were anesthetized with diethyl ether and killed by decapitation. Brainswere removed, fixed in 4% paraformaldehyde for 48 h, placed in 15% sucrose PBSsolution for 24 h until subsidence, then in 30% sucrose for 48 h until subsidence.Afterwards, brains were frozen in OCT embedding medium (Sakura, Torrance, CA,USA) at �80 �C. Frozen sections of 20 mm thickness were prepared with a cryotomeCryostat (Leica, CM 1900, Wetzlar, Germany). Detection of apoptotic cells was per-formed on the basis of TUNEL method using an in situ apoptosis detection kit.Sections were incubated with proteinase K for 15 min at room temperature andendogenous peroxidase was blocked with a solution PBS and 3% H2O2 for 10 min.The slides were then incubated with a solution of digoxigenin-conjugated nucleo-tides and terminal deoxynucleotidyl transferase (TdT) at 30 �C for 60 min. Subse-quently, the anti-digoxigenin antibody was applied and incubated for 30 min atroom temperature. Detection of the antigene antibody link was made throughimmunoperoxydase followed by diaminobenzidine (DAB) chromogenic. The sec-tions were counterstained with hematoxylin, rinsed in distilled water and observedby microscope. The survival of brain tumor-bearing mice was evaluated using thesame course of treatment.

2.10. Statistical analysis

All of the quantitative measurements were repeated three times. The data isexpressed as mean� S.D.. Statistical analysis was performed by One-way Analysis ofVariance (ANOVA) followed by Tukeye Kramer Multiple Comparisons Test usingGraphPad InStat (version 3.06). MTT and Survival data was analyzed by theGraphPad Prism software (version 5.0).

3. Results

3.1. DOX-DNA complex and NPs formation

The DOX-DNA complex was formulated by insertion doxoru-bicin into plasmid DNA and monitored by fluorescence scanningand imaging. The total fluorescence quenchingwas observedwith abinding ratio of doxorubicin to plasmid of 3000:4 (Fig. 1A). Asspeculated, one Dox molecule could occupy 5-6 base pairs ofplasmid DNA. Fluorescence imaging further demonstrated thisintercalation induced fluorescence quenching (Fig. 1B). Dynamiclight scattering showed that the hydrodynamic diameter of DPC/DOX-Trail NPs was 91.28 nm (Fig. 1C). Transmission electron mi-croscope conformed that NPs had a narrow distribution (Fig. 1D).

3.2. Cellular uptake and release of DOX from complex and NPs

U87 MG cells were treated with free DOX, DOX-Trail complex,DP/DOX-Trail and DPC/DOX-Trail NPs for 30 min. Then cellularuptake was examined by flow cytometry (Fig. 2). The fluorescenceintensity in cells treated with free DOX was the highest. Althoughthe fluorescence intensity in cells treated with DP/DOX-Trail andDPC/DOX-Trail NPs showed no significant difference, both NPstreated cells had superior fluorescence intensity than that in DOX-Trail treated cells. After 2 h additional incubation, the fluorescenceof DOXwas examined again. Once DOXwas released from the DOX-Trail complex, the fluorescence quenching was inverted. The fluo-rescence intensity both enhanced in DOX-Trail complex and NPs

B

520 560 600 640 680

0

50

100

150

0:3000

1:3000

2:3000

3:3000

4:3000

Wavelength (nm)

Flu

ore

sc

en

ce

in

te

ns

ity

(a

.u

)

DC

A

In

ten

sity (%

)

0

10

30

20

10 100 1000

Size (d.nm)

Fig. 1. (A) Fluorescence spectra of doxorubicin (3 nmol) with increasing picomoles of plasmid (from top to bottom: 0, 1, 2, 3, and 4). (B) Fluorescence quenching images of plasmidedoxorubicin complex (from left to right: 0, 1, 2, 3, and 4). (C) Dynamic light scattering of DPC/DOX-Trail NPs. (D) Transmission electron microscope of DPC/DOX-Trail NPs.Bar ¼ 100 nm.

J. Li et al. / Biomaterials 34 (2013) 9142e9148 9145

treated cells. The fluorescence intensity in DPC/DOX-Trail NPstreated cells was higher than that in DP/DOX-Trail NPs treated cells.

3.3. Cytotoxicity in vitro

The dominance of combination therapy of DOX and Trail in U87MG cells was investigated by 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay (Fig. 3). After 2 h-

Control

DOX-Trail 0.5 h

DOX-Trail 2 h

DP/DOX-Trail 0.5 h

DP/DOX-Trail 2 h

DPC/DOX-Trail 0.5 h

DPC/DOX-Trail 2 h

DOX 0.5 h

DOX 2 h

100 101 102 103 104

Fig. 2. Uptake of free DOX, DOX-Trail complex, DP/DOX-Trail NPs and DPC/DOX-TrailNPs after a 30-min treatment in U87 MG cells was examined by flow cytometry(solid line). The DOX release from complex and NPs was further examined 2 h aftertreatment (dashed line). Blank cells served as control (black solid line).

treatment followed by 48 h-incubation, the IC50 values of DPC/Trail,DP/DOX-Trail, DPC/DOX-Trail, DPC/DOX-NEG and free DOX was21.98 mM, 2.318 mM,1.645 mM, 3.458 mM and 3.154 mM, respectively. DPC/DOX-Trail showed the highest cytotoxicity, which has an IC50 valueeven lower than that of free DOX. This indicated that the transfectionof Trail elevated the anti-glioblastoma effect of DOX. This enhance-ment was demonstrated by the replacement of Trail by a negativeplasmid DNA (NEG), which resulting in increased IC50 value.

3.4. Apoptosis detection in vitro

Annexin V-FITC could label the membrane phospholipid phos-phatidylserine, which is translocated from the inner to the outerside of cell membranes in early cell apoptosis. PI is a molecular

-2 -1 0 1 2

0

20

40

60

80

100

120DPC/Trail

DP/DOX-Trail

DPC/DOX-Trail

DPC/DOX-NEG

DOX

Log 10 [DGL, µM]

Cell viab

ility (%

)

Fig. 3. Cytotoxicity profile of DPC/Trail, DP/DOX-Trail, DPC/DOX-Trail, DPC/DOX-NEGand DOX for U87 MG cells was determined by MTT. NEG stands for negativeplasmid. Data are expressed as mean � S.D. (n ¼ 4).

Fig. 4. Cellular apoptosis of U87 MG cells induced by DPC/Trail, DP/DOX-Trail, DPC/DOX-Trail, DPC/DOX-NEG and free DOX was examined by fluorescence microscope after 2 h-treatment and 48 h-incubation. Apoptosis cells were labeled with Annexin V-FITC Apoptosis Detection Kit. Green: FITC labeled phosphatidylserine. Red: PI labeled nuclear. Originalmagnification: �100. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

J. Li et al. / Biomaterials 34 (2013) 9142e91489146

probe that can label nucleus through destructed membranes. Cellslabeled only with green fluorescence are at the early state ofapoptosis, and those labeled only with red fluorescence are dead.Cells labeled both with green and red indicating the cells are at thelate state of apoptosis. As shown in Fig. 4, DPC/Trail NPs treatmentled to obvious early apoptosis with little late apoptosis or cell death.While combination treatment caused significantly more lateapoptosis or cell death. Among these, DPC/DOX-Trail NPs treatmentshowed most significant anti-glioblastoma effect.

3.5. In vivo imaging

NIR imaging was performed to evaluate the in vivo targetingefficiency of choline derivate as glioma homing ligand. At 18th day,nude mice were administrated with NIR783 labeled DP/DOX-Trailand DPC/DOX-Trail NPs via tail vein. In vivo images were taken2 h after administration (Fig. 5). Compared to DP/DOX-Trail NPstreatedmice, an obvious accumulation of NIR signal was detected atthe tumor region of mice injected with DPC/DOX-Trail NPs. Ex vivoimaging also exhibited the targeting behavior of choline derivate.

Fig. 5. (A) In vivo imaging of model mice administrated with DP/DOX-Trail (left) and DPC/NIR783. (B) Ex vivo imaging of brain tumors and main organs.

3.6. Apoptosis detection in U87 grafted model

At the 21st day after glioma implantation, TUNEL assay wasperformed to detect the DNA fragmentation in nuclei of apoptotictumor cells (Fig. 6). Although with high cytotoxicity in vitro, DOXinduced only little apoptosis in vivo. An extended apoptosis wasobserved after the combination therapy of Trail and DOX, especiallyin DPC/DOX-Trail treated group.

3.7. In vivo anti-glioblastoma effect in U87 grafted model

The anti-tumor effect of dual targeting and co-delivery systemwas evaluated as the median survival time of model mice. Mediansurvival time of DPC/DOX-Trail, DP/DOX-Trail, DPC/Trail, DOX andsaline treated mice was 47 (p< 0.01, compared to other groups), 37,34.5, 32, 24.5 days, respectively (Fig. 7B). The dominance of tar-geted combination therapy (DPC/DOX-Trail) was also reflected onthe body weight change. The body weight showed a slow decreaseafter the 27th day, while other groups had a rapid decrease(Fig. 7A).

DOX-Trail (right). Images were taken 2 h after administration. NPs were labeled with

Fig. 6. TUNEL assay of frozen sections of brain tumors. Slides were developed with DAB and counterstained with hematoxylin. A: Saline, B: DOX, C: DPC/Trail, D: DP/DOX-Trail, E:DPC/DOX-Trail. Original magnification: �100.

J. Li et al. / Biomaterials 34 (2013) 9142e9148 9147

4. Discussion

Despite recent advances in brain cancer therapy, the prognosisremains poor for patients diagnosed with malignant gliomas. Themedian survival time of glioblastoma patients is less than 15months [16]. Patients have a universal recurrent after the surgeryfollowed by chemotherapy. The effectiveness of single medicationarises from the impermeability across the BBB and drug resistance.Thus, combination medication with BBB permeability may be apromising pathway to treat glioma. The synergistic effect of com-bination of DOX and Trail was reported and a significant thera-peutic beneficial was achieved [11]. In this study, DOX and Trailwere selected as model drugs. The co-delivery system was estab-lished in the following two steps.

In the first step, DOX was interacted into the double-strandedTrail to form a stable complex. The native fluorescence spectrumof doxorubicin was totally quenched at the binding ratio ofdoxorubicin to Trail of 3000:4 (Fig. 1A). The fluorescence imagesalso showed an obvious quenching effect (Fig. 1B). In the secondstep, DOX-Trail complex was condensed by choline derivatemodified DGL. And the cellular uptake of U87 MG cells was eval-uated by flow cytometry (Fig. 2). Compared to free DOX, the

A B

10 20 30 40 50

12

16

20

24

28

Weig

hts (g

)

1

Percen

t su

rviva

l (%

)

Days

Fig. 7. (A) Body weight change of model mice. (B) KaplaneMeier survival curves of mice betreated with Saline, DOX (5 mg/kg), DPC/Trail, DP/DOX-Trail and DPC/DOX-Trail (50 mg DNA aTrail), 37 days (DP/DOX-Trail), 34.5 days (DPC/Trail), 32 days (DOX) and 24.5 days (saline).

cellular uptake of DOX-Trail complex was significantly reduced.Although the interaction of DOX into DNA could improve thepharmacokinetics of DOX, the cellular uptake of complex washindered by the negative charge of plasmid DNA. While,condensation by positively charged DGL led to an increased DOXaccumulation. Unfortunately, no difference was found in cellularuptake of DP/DOX-Trail and DPC/DOX-Trail NPs in the first 30 min.It was reported that DNA-DOX complex was very stable with onlyless than 20% leakage after incubation in PBS 7.4 for 120 h [15].The fluorescence quenching may hind the targeting efficiency ofcholine derivate. We then evaluated the cellular DOX fluorescenceof treated cells 2 h later. A recovery of fluorescence was observedboth in complex treated or NPs treated cells. The fluorescenceintensity of DPC/DOX-Trail treated cells was comparable to that offree DOX treated cells and higher than that of DP/DOX-Trail treatedcells. DOX could escape from the complex or NPs in a acceleratemanner in the intracellular component.

The synergistic effect of the combination of DOX and Trail to-wards U87 MG cells was evaluated by MTT assay and apoptosisdetection in vitro (Figs. 3 and 4). The apoptosis detection showedthat single medication of DPC/Trail induced the significant earlyapoptosis. While less late apoptosis or cell death was detected.

0 20 40 60

0

20

40

60

80

00

Saline

DOX

DPC/Trail

DP/DOX-Trail

DPC/DOX-Trail

Days

aring U87 MG tumors. At 12th, 15th and 18th day after tumor implantation, mice werend/or 8 mg DOX/mouse, n ¼ 12). Median survival for each group was 47 days (DPC/DOX-

J. Li et al. / Biomaterials 34 (2013) 9142e91489148

This was agreed with the high IC50 value of DPC/Trail. Combinationmedication of Trail and DOX had lower IC50 values than free DOXindicating the synergistic effect. This may depend on the apoptosismechanism of each drug. On one hand, TRAIL induces apoptosisvia its cognate receptors (DR4/DR5). DOX increased the expres-sions of DR4 and DR5 thus sensitizing tumor cells to Trail-inducedapoptosis [17]. On the other hand, Bax and Bak deficiency isresponsible for DOX resistance in tumor cells. Proapoptoticmembers Bax and Bak could be activated in Trail-inducedapoptosis pathway. Thus, Trail could in turn enhance DOX sensi-tivity of tumor cells [11].

Blood-tumor barrier (BTB) is another major concern that im-pairs the therapeutic effects in systematic administration. Braintumor vascular pore cutoff size was significantly reduced in cranialmicroenvironment compared with that in peripheral ones [18].Thus, for glioma therapy, one more principle for drug deliverysystem other than long circulation time should be the suitable sizeof NPs. NPs with a diameter around 100 nm might permeate intoglioma more efficiently. Herein, DOX-Trail complex was condensedby DGL at an optimal weight ratio to gain the desired particle size(Fig. 1C, D). Then the in vivo targeting efficiency was evaluated(Fig. 5). None targeting co-delivery system showed less accumu-lation in tumor via EPR effect. Choline derivate modified systemrevealed more accumulation by both EPR effect and dual targeting.Choline transport via choline transporter is in a conformationdependent manner. Here, the targeting mechanismwas hypothesisthat choline derivate with high choline transporter affinity couldfacilitate the NPs anchoring to the choline transporters overex-pressed on tumor cells.

In the evaluation of anti-glioblastoma effect in U87 graftedmodel (Figs. 6 and 7), DPC/DOX-Trail NPs exhibited the highesttherapeutic efficacy for gliomas. The median survival time of freeDOX treated mice is only 32 days. The dosage of free DOX for miceis 5 mg/kg, which will be accompanied by toxicities includingcardiotoxicity and myelosuppression [19]. The low accumulationin tumor and severe side effects made DOX less inefficient forglioma therapy. However, combination therapy with much lowerDOX dosage (8 mg DOX/mouse) had a predominant therapeuticefficacy, specifically for DPC/DOX-Trail NPs treated groups (47days).

5. Conclusions

In summary, a co-delivery system for glioma therapy wasestablished in two steps and achieved enhanced tumor accumula-tion, tumor apoptosis and better therapeutic outcomes. Cholinederivate employed here exhibited as a glioma targeting ligand. Thiseasily established co-delivery system holds great potential forcancer therapy in clinical application.

Acknowledgments

This work was supported by the grant from National BasicResearch Program of China (973 Program, 2013CB932500), NationalNatural Science Foundation of China (81172993), and Program forNew Century Excellent Talents in University.

References

[1] Jain RK, di Tomaso E, Duda DG, Loeffler JS, Sorensen AG, Batchelor TT.Angiogenesis in brain tumours. Nat Rev Neurosci 2007;8:610e22.

[2] Zhan C, Gu B, Xie C, Li J, Liu Y, Lu W. Cyclic RGD conjugated poly(ethyleneglycol)-co-poly(lactic acid) micelle enhances paclitaxel anti-glioblastoma ef-fect. J Control Release 2010;143:136e42.

[3] Rancan F, Todorova A, Hadam S, Papakostas D, Luciani E, Graf C, et al. Stabilityof polylactic acid particles and release of fluorochromes upon topical appli-cation on human skin explants. Eur J Pharm Biopharm 2012;80:76e84.

[4] Ying X, Wen H, Lu WL, Du J, Guo J, Tian W, et al. Dual-targeting daunorubicinliposomes improve the therapeutic efficacy of brain glioma in animals.J Control Release 2010;141:183e92.

[5] Yan H, Wang L, Wang J, Weng X, Lei H, Wang X, et al. Two-order targetedbrain tumor imaging by using an optical/paramagnetic nanoprobe across theblood brain barrier. ACS Nano 2012;6:410e20.

[6] Huang S, Li J, Han L, Liu S, Ma H, Huang R, et al. Dual targeting effect ofAngiopep-2-modified, DNA-loaded nanoparticles for glioma. Biomaterials2011;32:6832e8.

[7] Sun X, Pang Z, Ye H, Qiu B, Guo L, Li J, et al. Co-delivery of pEGFP-hTRAIL andpaclitaxel to brain glioma mediated by an angiopep-conjugated liposome.Biomaterials 2012;33:916e24.

[8] Li J, Zhou L, Ye D, Huang S, Shao K, Huang R, et al. Choline-derivate-modifiednanoparticles for brain-targeting gene delivery. Adv Mater 2011;23:4516e20.

[9] Michel V, Yuan Z, Ramsubir S, Bakovic M. Choline transport for phospholipidsynthesis. Exp Biol Med (Maywood) 2006;231:490e504.

[10] Li J, Huang S, Shao K, Liu Y, An S, Kuang Y, et al. A choline derivate-modifiednanoprobe for glioma diagnosis using MRI. Sci Rep 2013;3:1623.

[11] Liu S, Guo Y, Huang R, Li J, Huang S, Kuang Y, et al. Gene and doxorubicinco-delivery system for targeting therapy of glioma. Biomaterials 2012;33:4907e16.

[12] Zhan C, Wei X, Qian J, Feng L, Zhu J, Lu W. Co-delivery of TRAIL gene enhancesthe anti-glioblastoma effect of paclitaxel in vitro and in vivo. J Control Release2012;160:630e6.

[13] Bagalkot V, Farokhzad OC, Langer R, Jon S. An aptamer-doxorubicin physicalconjugate as a novel targeted drug-delivery platform. Angew Chem Int EdEngl 2006;45:8149e52.

[14] Bagalkot V, Lee IH, Yu MK, Lee E, Park S, Lee JH, et al. A combined chemo-immunotherapy approach using a plasmid-doxorubicin complex. Mol Pharm2009;6:1019e28.

[15] Han L, Huang R, Li J, Liu S, Huang S, Jiang C. Plasmid pORF-hTRAIL anddoxorubicin co-delivery targeting to tumor using peptide-conjugated poly-amidoamine dendrimer. Biomaterials 2011;32:1242e52.

[16] Wen PY, Kesari S. Malignant gliomas in adults. N Engl J Med 2008;359:492e507.

[17] Aroui S, Brahim S, Hamelin J, De Waard M, Breard J, Kenani A. Conjugation ofdoxorubicin to cell penetrating peptides sensitizes human breast MDA-MB231 cancer cells to endogenous TRAIL-induced apoptosis. Apoptosis2009;14:1352e65.

[18] Hobbs SK, Monsky WL, Yuan F, Roberts WG, Griffith L, Torchilin VP, et al.Regulation of transport pathways in tumor vessels: role of tumor type andmicroenvironment. Proc Natl Acad Sci U S A 1998;95:4607e12.

[19] Han L, Huang R, Liu S, Huang S, Jiang C. Peptide-conjugated PAMAM for tar-geted doxorubicin delivery to transferrin receptor overexpressed tumors. MolPharm 2010;7:2156e65.