lable at ScienceDirect

Biomaterials 116 (2017) 130e144

Contents lists avai

Biomaterials

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

Combination therapy with doxorubicin-loaded galactosylatedpoly(ethyleneglycol)-lithocholic acid to suppress the tumor growth inan orthotopic mouse model of liver cancer

Bijay Singh a, d, 1, Yoonjeong Jang b, 1, Sushila Maharjan a, Hyeon-Jeong Kim b,Ah Young Lee b, Sanghwa Kim b, Nomundelger Gankhuyag b, Myeon-Sik Yang c,Yun-Jaie Choi a, Myung-Haing Cho b, *, Chong-Su Cho a, **

a Department of Agricultural Biotechnology, Research Institute for Agriculture and Life Sciences, Seoul National University, Seoul, 08826, South Koreab Research Institute for Veterinary Science, College of Veterinary Medicine, Seoul National University, Seoul, 08826, South Koreac Laboratory of Veterinary Pathology, College of Veterinary Medicine, Chonbuk National University, Jeonju, 54896, South Koread Research Institute for Bioscience and Biotechnology, Kathmandu, Nepal

a r t i c l e i n f o

Article history:Received 28 July 2016Received in revised form12 November 2016Accepted 24 November 2016Available online 24 November 2016

Keywords:CancerCombination therapyDoxorubicinNanoparticlesOrthotopic mouse modelTargeted drug delivery

* Corresponding author.** Corresponding author.

E-mail addresses: mchotox@snu.ac.kr (M.-H. Cho)1 These authors contributed equally to this work.

http://dx.doi.org/10.1016/j.biomaterials.2016.11.0400142-9612/© 2016 Elsevier Ltd. All rights reserved.

a b s t r a c t

Despite advances in technology, neither conventional anti-cancer drugs nor current nanoparticle (NP)drugs have gained substantial success in cancer treatment. While conventional chemotherapy drugs haveseveral limitations such as low potency, poor in vivo stability and limited bioavailability, non-specifictargeting of NP drugs diminishes their potency at actual target sites. In addition, the development ofdrug resistance to anti-cancer drugs is another challenging problem. To overcome these limitations, weaimed to develop a polymer-drug conjugate, which functions as an active NP drug and drug carrier both,to deliver a chemotherapeutic drug for combination therapy. Accordingly, we made targeting NP carrierof lithocholic acid-poly(ethylene glycol)-lactobionic acid (LPL) loading doxorubicin (Dox) to produce Dox/LPL NPs. The cellular uptake of Dox/LPL NPs was relatively higher in human liver cancer cell line (SK-HEP-1) due to galactose ligand-asialoglycoprotein receptor interaction. Consequently, the cellular uptake ofDox/LPL NPs led to massive cell death of SK-HEP-1 cells by two different mechanisms, particularlyapoptotic activity by LPL and mitotic catastrophe by Dox. Most importantly, Dox/LPL NPs, whenadministered to orthotopic xenograft model of liver cancer, greatly reduced proliferation, invasion,migration, and angiogenesis of liver tumor in vivo. Thus, this study exemplifies the superiority of com-bination therapy over individual NP drug or conventional small molecule drug for cancer therapy.Overall, we present a promising approach of combinatorial therapy to inhibit the hepatic tumor growthand metastasis in the orthotopic xenograft model mice, thus representing an effective weapon for cancertreatment.

© 2016 Elsevier Ltd. All rights reserved.

1. Introduction

Cancer is one of the leading cause of deaths worldwide claimingmillions of lives each year with an estimated 600,000 deathsattributable to liver cancer. Despite advances in technology,chemotherapy has been remained as the standard of treatment formost of the cancers. However, conventional chemotherapy drugs

, chocs@snu.ac.kr (C.-S. Cho).

have several limitations such as low potency, poor in vivo stabilityand limited bioavailability [1]. It is because chemotherapy drugs areusually non-specific small molecules that simply diffuse anddistribute freely throughout the body diminishing their potency atactual target sites. In addition, non-specific targeting of chemo-therapy drugs can damage rapidly proliferating normal cells [2].Hence, the development of innovative anti-cancer drugs which aremuch more effective and selective to the cancer cells are requiredfor the specific treatment of cancers.

Polymeric nanoparticles (NPs) have attracted significant interestover the last two decades to overcome the limitations of conven-tional drug delivery system [1,3]. The NPs offer great advantages

B. Singh et al. / Biomaterials 116 (2017) 130e144 131

over conventional drugs because they can be manipulated intodiverse shapes and sizes using a variety of polymeric materialsincorporating lipids, metals or chemical and biological substances[4e6]. Besides, NPs can be tailored to achieve site-specific activetargeting, in contrast to passive targeting of conventional drugs, toenhance the intracellular concentration of drugs in cancer cellswhile avoiding toxicity to normal cells [7]. Currently, the develop-ment of drug resistance to anti-cancer drugs is major challengingproblem [8,9]. One way to avoid the development of resistantcancer cells is to give chemotherapy drugs in combination thatpresumably circumvents the possibility of development of drugresistance by acting on multiple molecular mechanisms [10]. Theother way is to deliver chemotherapy drugs in groups by using theemerging field of NP-based drug delivery platforms. Themajority ofNPs for combinatorial drug delivery is primarily based on lipo-somes, polymeric micelles, dendrimers and polymer-drug conju-gates. However, polymeric NP drug delivery platform has beenrarely explored to use chemotherapy drug with polymer-drug forcombination therapy.

Usually, polymeric conjugates of drug (prodrug or permanentdrug) are built by attaching a small molecule anticancer drug ortherapeutic agent to a polymer, with either a labile or a permanentlinker. While the prodrug is made with labile linkage with drug foreasy hydrolysis to release active drug, permanent drug denotes thedrug conjugate with polymer with stable linkage that has thera-peutic activity without breakdown of the linkage. These polymer-drug conjugates have several advantages over their low molecu-lar weight precursors. The main advantages include water solubi-lity, higher stability, prolonged circulation half-life, lowerimmunogenicity and specific targeting to tissues or cells [11].Recently, lithocholic acid (LCA) has been reported to induceapoptotic cell death to cancer cells, while sparing normal cells [12].Although LCA appears a promising small molecule drug for thetreatment of cancers, it faces many challenges for in vivo use due toits low solubility, short circulating half-life, and rapid clearancefrom the body by renal filtration. In order to eliminate these typesof defects of LCA, we recently conjugated LCA with poly(ethyleneglycol)(PEG) as a permanent drug and the product exhibited higherapoptotic activity than LCA itself [13]. Since the product self-assembles to NP, it could be used as an efficient carrier to deliverconventional anti-cancer drugs for combination chemotherapy.

Liver is one of the most important organs in the body thatmetabolizes nutrients, detoxifies harmful substances, producesvital enzymes and secretes bile acids to perform many vital func-tions in the body. Liver is composed of parenchymal tissues mainlycomprising of hepatocytes which consist of numerous cell surfacereceptors (asialoglycoproteins) that readily binds with carbohy-drates such as galactose [14]. Receptor-mediated drug targeting isof major interest in the field of drug delivery systems where drugscan be delivered to the specific cells or tissues through specificinteraction between ligand and receptor on the cells [15e17].Owing to specific interaction between galactose and asialoglyco-protein receptors on liver cells, a number of NPs modified withgalactose have been utilized for liver-targeted drug delivery sys-tems [18e20]. The results from these experiments suggest that thesurface characteristics of NPs greatly contribute to targeted-drugdelivery.

In this study, we used a polymer-drug conjugate, which func-tions both as an active NP drug and a drug carrier, to deliver achemotherapeutic drug for combination therapy in order to sup-press hepatic tumor growth and spontaneous metastasis in anorthotopic mouse model of liver cancer. In addition, we studied theunderlying molecular mechanisms of combination drugs in thesuppression of proliferation, invasion, migration, and angiogenesisof liver tumor in vivo.

2. Materials and methods

2.1. Materials

Lithocholic acid (LCA), lactobionic acid (LBA), N,N0-dicyclohex-ylcarbodiimide (DCC), N-hydroxysuccinimide (NHS), dimethylsulf-oxide (DMSO), doxorubicin (Dox) hydrochloride, sodium dodecylsulfate (SDS), Fluorescein isothiocyanate (FITC) isomer I, 40,6-diamidino-2-phenylindole (DAPI) dilactate and triethylaminewere purchased from Sigma-Aldrich (St. Louis, MO,USA). Diaminopoly (ethylene glycol) 6000 Da (PEG) was purchased from SunBio(Anyang, Korea).

Antibodies of cyclin dependent kinase 4 (CDK4) and pro-caspase-3 (Casp-3) were purchased from Cell Signaling (Danvers,MA, USA) while antibodies of proliferating cell nuclear antigen(PCNA), vascular endothelial growth factor (VEGF), cyclin B1, cyclindependent kinase 2 (CDK2), cytochrome c and actin were fromSanta Cruz Biotechnology (Santa Cruz, CA, USA). Fetal bovine serum(FBS), penicillin/streptomycin were purchased from Invitrogen(Carlsbad, CA, USA). All other chemicals were purchased fromSigma-Aldrich unless otherwise specifically stated.

2.2. Synthesis and preparation of NPs

LPL was synthesized as previously described [21]. Briefly, LBA(2 mM) was activated in the presence of DCC (4 mM) and NHS(4mM) in DMSO (10ml) at room temperature for 18 h. The LBAwasfurther reacted with diamino PEG (20 mM) for 4 h. The reactionproduct was initially dialyzed against DMSO with a dialysis mem-brane (Spectra/Por MWCO: 1000) for 24 h, and finally dialyzedagainst water at 4 �C overnight. The product, LBA-PEG (LP), wasfreeze-dried and conjugated with LCA under following condition.First, LCA (50 mM) was activated in the presence of DCC (100 mM)and NHS (100 mM) in DMSO (10 ml) with at room temperature for18 h. The LCA was further conjugated with LP (10 mM) for 4 h. Theproduct, LPL, was first dialyzed against DMSO with a dialysismembrane (MWCO: 1000) for 24 h, and finally dialyzed againstwater at 4 �C overnight to obtain LPL NPs which were furtherwashed with water, freeze-dried and stored at �80 �C until use.Similarly, PEG-LCA (PL) was synthesized and PL NPs were preparedas described above. FITC-labeled LPL or PL was prepared as follows.Twenty milligrams of LPL or PL and 1 mg of FITC was dissolved in1 ml of DMSO and stirred for 4 h in dark at room temperature. Thereaction product was dialyzed in water at 4 �C for 24 h in dark toobtain FITC-labeled LPL or PL NPs which were further freeze-driedand stored at �80 �C until use. Similarly, Dox/LPL NPs were pre-pared by following method. A mixture of Dox (5 mg) with trie-thylamine (2.5 ml) was stirredwith LPL (20mg) in DMSO (5 ml). Themixture was dialyzed against water with a dialysis membrane(MWCO: 1000) at 4 �C for 6 h to obtain Dox/LPL NPs which werefurther washed with water, freeze-dried and stored at �80 �C untiluse.

2.3. Characterization of NPs

The size and surface zeta potential of the NPs were assessedusing a dynamic light scattering spectrophotometer (DLS 7000,Otsuka Electronics, Osaka, Japan) and electrophoretic light scat-tering spectrometer (ELS 8000, Otsuka Electronics) by dispersingthe NPs (1 mg/ml) in water. The morphology of NPs was visualizedusing field-emission scanning electron microscope (FE-SEM)(SIGMA, Carl Zeiss Microscopy GmbH, Jena, Germany). For SEM, thedry NPs were dispersed on a grid with electric tape. A thin film ofplatinumwas then sputtered onto the NPs on the grid using a highvacuum coater Leica EM ACE600. The coated samples were

B. Singh et al. / Biomaterials 116 (2017) 130e144132

observed and imaged with SEM.

2.4. Drug loading content and drug release test of NPs

The drug encapsulation efficiency was defined as the weightpercentage of Dox encapsulated in NPs relative to Dox initiallyadded for encapsulation while the drug loading content wasdefined as the weight percentage of Dox in Dox-loaded NPs. Astandard curve of Dox was plotted by measuring the absorbance ofDox at 480 nm by UVeVIS spectrophotometer (Tecan Infinite® 200PRO series, Tecan Group Ltd. Mannedorf, Switzerland) to calculatethe encapsulation efficiency and loading content. The release ofDox from Dox/PL and Dox/LPL NPs was performed by dialysismethod. Briefly, 10 mg of each NP sample was dispersed in 1 mL ofPBS buffer (pH 7.4) or acetate buffer (pH 5.5) including 0.05 wt%SDS and transferred to Slide-A-Lyzer MINI Dialysis Unit, 2000MWCO (Thermo Scientific, Logan, UT, USA). The dialysis unit wasinserted in a microcentrifuge tube containing the same PBS buffer.The tube was inserted in a float, placed on a water beaker andstirred with 100 rpm at 37 �C. The amount of Dox released in agiven time of interval was observed at 480 nm by UV-VIS spec-troscopy. The release experiment was carried out in triplicates. Therelease of Dox was quantified using the standard curve plot.

2.5. Cytotoxicity of NPs

Human liver cancer cell line SK-HEP-1 (Korean Cell Line Bank,Seoul, Korea) was cultured in complete DMEM/high glucose me-dium supplemented with 10% fetal bovine serum (FBS), 100 U/mlpenicillin and 100 mg/ml streptomycin at 37 �C in a humidified 5%CO2 incubator. Cytotoxicity of NPs in SK-HEP-1 cells was deter-mined by a methylthiazolyldiphenyl tetrazolium bromide (MTT)assay. The cells were seeded in 24-well plates at an initial density of10 � 104 cells/well in 1 ml of growth medium for 24 h. The cellswere treated with the different amount of NPs in Gibco Opti-MEMreduced serum medium (Thermo Scientific, USA) for 4 h andreplaced by 0.5 ml fresh growth media. The cells were furtherincubated for 24 h and then treated with MTT reagent for 3 h. Thecell viability was determined by measuring the absorbance at570 nm using multimode microplate reader (Tecan Infinite® 200PRO series).

2.6. Cellular uptake of NPs and subsequent detection of Dox

SK-HEP-1 cells were seeded in confocal dish at an initial densityof 20 � 104 cells/well in 2 ml of growth medium for 24 h. Todetermine the cell specificity of the NPs, the cells were treated with25 mM concentration of FITC-NPs in Opti-MEMmedium for 4 h. Thecells were washed, stained with DAPI and fixed with formalin priorto observation by confocal laser scanning microscope (CLSM,LSM710, Carl Zeiss Microscopy GmbH, Germany). To observe theinhibitory effect of galactose on the receptor-mediated endocytosisof LPL-core NPs, SK-HEP-1 cells were cultured as described before.Various concentration of galactose (0, 0.1, 1, 10 mM) was treated tothe cells prior to the treatment with 25 mM concentration of Dox/LPL NPs in Opti-MEM medium for 2 h. After incubation with theNPs, the cells were collected and the amount of Dox was detectedby flow cytometry in a FACSCalibur™ flow cytometer (BD Bio-sciences, Franklin Lakes, NJ, USA). The amount of Dox detection isdirectly proportional to the uptake of Dox/LPL NPs.

On the other hand, to observe the cellular uptake of NPs andsubsequent release of Dox inside cells, the cells were treated withdifferent amount of Dox/LPL NPs in Opti-MEMmedium for 4 h. Thecells were replaced by fresh growth media and further incubatedfor 18 h. The cells were washed, stained and fixed as above to

observe through CLSM. Quantitative analysis of the degree of up-take of NPs was done by using ImageJ v1.36b software (http://rsb.info.nih.gov/ij/).

2.7. Flow cytometric detection of apoptosis

To analyze apoptosis induced by NPs, SK-HEP-1 cells wereseeded in 6-well plates at an initial density of 20 � 104 cells/well in2 ml of growth medium for 24 h. The cells were treated with theoptimized concentration of NPs in Opti-MEM medium for 4 h. Thecells were replaced by fresh growth media and further incubatedfor 18 h. The cells were collected and stainedwith Ezway annexinV-FITC apoptosis detection kit (Komabiotech, Seoul, Korea) accordingto manufacturer's instructions. The incidence of apoptosis in cellcultures was analyzed by flow cytometry.

2.8. Establishment of orthotopic xenograft mouse model of livercancer

The laboratory animal studywas approved by the Seoul NationalUniversity Institutional Animal Care and Use Committee (IACUC)(SNU-150112-3-1). 7-week-old male nude mice (BALB/c, nu/nu)were purchased from Central Lab Animal Co. (Seoul, Korea). Theywere housed under a 12 h light/dark cycle with temperature of23 ± 2 �C, a relative humidity level of 50 ± 20%, and fed with foodand water ad libitum. After 1 week of acclimation, the mice weretaken surgery for orthotopic tumor graft and randomly divided into5 groups (4mice/group). To establish orthotopic xenograft model ofliver cancer, the mice were anesthetized using alfaxalone (60 mg/kg; Alfaxan, Jurox) and xylazine (35 mg/kg; Rompun, Bayer). Afterapplying 70% ethanol and povidone-iodine to incision site for sur-gical scrub and laparotomy, 3 � 106 SK-HEP-1 cells were injectedinto the liver. The incision was closed and applied with antibioticointment (Fig. S1). Eighteen days after the injection, the tumordevelopment in liver was examined.

2.9. Administration of NPs in vivo and histological examination

After confirmation of tumor development in mice, the micewere administered with a dose of 200 mM of Dox/LPL NPs, 200 mMof LPL NPs, 140 mM of LCA (equivalent amount of LCA content inLPL), 9.6 nM of Dox (equivalent amount of Dox content in Dox/LPL)in a volume of 0.2 ml saline via intraperitoneal (IP) injection for 4weeks with a total of 7 doses at the interval of 4 days (Scheme 1).Three days after the last dose of the treatment, the mice wereanesthetized for necropsy. Blood samples were collected from theabdominal vein and sera were obtained by centrifugation at14,000 rpm for 50 min and stored at �70 �C until use. Analyses ofserum aspartate aminotransferase (AST), and alanine aminotrans-ferase (ALT) levels from the sera were performed by Neodin Vet-erinary Science Institute (Seoul, Korea).

Livers were isolated from each group of mice and the tumor sizewas measured with caliper and the volume was calculated as fol-lows: tumor volume [mm3] ¼ (length [mm]) � (width[mm])2 � 0.5. For the histological analysis, the liver tissues werefixed in 10% neutral buffered formalin, paraffin-processed, andsectioned. The sections were deparaffinized in xylene for 5 mintwice and rehydrated with 100%, 95%, 90%, 80%, 70%, and 50%ethanol for 2 min each. Then, they were stained with hematoxylinand eosin, mounted with Permount mounting medium (FisherScientific, Fair Lawn, NJ, USA) and observed under a lightmicroscope.

B. Singh et al. / Biomaterials 116 (2017) 130e144 133

2.10. Western blot analyses

Livers were homogenized with lysis buffer and the proteinconcentration of each tissue lysate was measured using Pierce™BCA Protein Assay Kit (Thermo Scientific, USA). Equal amounts(25 mg) of each sample were loaded and separated by 10e12% so-dium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Then the protein-loaded gel was transferred to a nitrocel-lulose membrane using iBlot® 2 Gel Transfer Device (Life Technol-ogies, Carlsbad, CA, USA). The protein expression analyses wereperformed with antibodies of PCNA, VEGF, cyclin B1, CDK2, CDK4,Casp-3, cytochrome c and actin. After incubating with corre-sponding secondary antibodies conjugated with horseradishperoxide, the Western bands were detected by using Imageanalyzer ATTO EZ Capture MG (ATTO Corp., Tokyo, Japan) anddensitometric analyses of the bands were performed with CSanalyzer 3.0 program (ATTO Corp.).

2.11. Statistical analyses

Unless otherwise stated, independent experiments were run atleast in triplicate (n ¼ 3). The results were expressed asmean ± standard deviation (SD). The statistical significance of thedifferences was calculated by one-way analysis of variance(ANOVA) or two-tailed unpaired Student's t-test using GraphpadPrism 5 Software (San Diego, CA, USA). Statistical significance isdenoted by *P < 0.05, **P < 0.01, and ***P < 0.001.

3. Results

3.1. Preparation and characterization of NPs

LBA and LCA were successively conjugated at the two ends ofdiamino PEG to produce LPL as described inmaterials andmethods.Both LBA and LCA were activated with DCC/NHS to link covalentlywith the amino groups of PEG. By similar reaction, PEG-LCA (PL)was also synthesized. The successful synthesis of LPL and PL wasconfirmed by proton nuclear magnetic resonance (1H NMR). Thecorresponding methylene peaks of LCA, PEG and LBA are shown inthe 1H NMR spectrum of LPL (Fig. 1A) and PL (Fig. 1B). The peaks inthe spectrum were authenticated by comparing with the corre-sponding peaks of 1H NMR spectrum of original compound of LCA(Fig. S2), LBA (Fig. S3) and PEG (Fig. S4). The conjugation of LCA andLBA in LPL was further confirmed by Fourier transform infraredspectroscopy (FTIR). FT-IR spectrum with the peaks correspondingto O-H bending vibration (3405 cm�1) of LBA, C-H stretching vi-bration (2928 cm�1) of LCA, C-O-C stretching vibration (1101cm�1) of PEG and C]O bending vibration (1107 cm�1) of amideindicated the chemical bonds of LPL (Fig. 2A). Similar bands

Day 1 18 42

Injection of SK-Hep1 cells to establish orthotopic liver tumor

Treatment of drugs by IP injection for four weeks

Necropsy and collection of samples

Scheme 1. Schematic representation of in vivo experiment schedule. SK-HEP-1 cellswere injected into the liver of mice to develop orthotopic liver tumor. Eighteen daysafter the tumor development, each group of mice received a total of 7 doses of drugs atthe interval of 4 days. Three days after the last dose treatment, mice were sacrificed tocollect liver tissues and blood samples.

appeared on FT-IR spectrum of PL (Fig. 2B). The mole% of conju-gated molecules in the polymer was calculated by the integralvalues of NMR spectra. Thus, LBA and LCA content in LPL was 9 and69 mol%, respectively. Similarly, the LCA content in PL was 64 mol%.

Dox/LPL or Dox/PL NPs were prepared by dialysis method.Briefly, a mixture of Dox and polymer at the weight ratio of 1:4 inDMSO was put in a dialysis tubing and dialyzed against water toproduce NPs through self-assembly. The NPs thus formed werefreeze-dried to obtain solid NPs for further characterization. Theobtained NPs of Dox/LPL and Dox/PL were both solid powders butorange-red in color in contrast to their polymer precursors whichwere solid white powders (Fig. 3A).When suspended inwater, bothDox/LPL and Dox/PL NPs formed homogenous suspension similar totheir polymer precursors. However, the suspensions of NPs withLPL backbones exhibited slightly higher turbidity than that of NPswith PL backbones (Fig. 3B).

Particularly, DLS is an important tool for determining the poly-dispersity index (PDI) or variability in particle size. Hence, DLSmeasures the agglomeration state of nanoparticles in suspension. Ifthe particles are unagglomerated, the size measured by DLS is oftensimilar or slightly larger than the SEM sizewith low PDI value. If theparticles are agglomerated, the size measured by DLS is muchlarger than the SEM size with high PDI value. Hence, any solutionthat contains larger particles will dominate the light scatteringsignal and mask the presence of the smaller particles during DLSmeasurement. Similarly, the hydrodynamic sizes of Dox/LPL, Dox/PL, LPL and PL NPs are consistent with the PDI values of the sameparticles. The larger the sizes of the particles, the higher the PDIvalues. As a result, the PDI of Dox/LPL, Dox/PL, LPL and PL NPs, whenmeasured by DLS, was approximately 0.538, 0.383, 0.254 and 0.244,respectively.

Interestingly, the shapes of the NPs with LPL backbone differedfrom that of PL NPs when observed through SEM (Fig. 4). While PLNPs had spherical shapes, LPL NPs had non-spherical shapes, pre-dominantly nanodiscs that resemble with red blood cells (RBCs).After loading Dox, the shapes of Dox/PL NPs were slightly modifiedwith increase in the volume of the spheres whereas Dox/LPL NPswere preferably RBC-like nanodiscs with greater size than that ofLPL NPs. The exact sizes of the NPs were measured by DLS (Fig. 5).The mean hydrodynamic diameter of Dox/LPL, Dox/PL, LPL and PLNPs was approximately 916, 866, 355 and 310 nm, respectively.Because the diameter of particles calculated by DLS is hydrody-namic which considers the particles as spherical regardless of theirshape and approximate size, it is common that the hydrodynamicsize of NPs is typically larger than the actual one as that observed bySEM. Consistently, the largest particle sizes of Dox/LPL NPs were~500 nm when observed by SEM (Fig. S5). It is, therefore, weconsidered the particles formed by LPL were nanoparticles, notmicelles, as LPL produced nanoparticles due to bulky group of LCA.The surface zeta potentials of the NPs were given in Table 1.

3.2. Release test of Dox from NPs

Initially, the drug encapsulation efficiency and loading contentof Dox/LPL and Dox/PL NPs was calculated. The encapsulation ef-ficiency of Dox/LPL and Dox/PL NPs was 95.6 ± 0.9% and 92.1 ± 3%,respectively. The Dox content in Dox/LPL and Dox/PL NPs wasapproximately 3.9% and 4.2%, respectively. The release of Dox fromDox/PL and Dox/LPL NPs was performed in PBS buffer (pH 7.4) andacetate buffer (pH 5.5). The release of Dox from these NPs wasnegligible even after 5 days of incubation at both pH conditions. Itshowed that these NPs were highly stable due to the tight packingof Dox into the self-assembled LPL NPs. Further, the test was per-formed with the same buffers including 0.05 wt% SDS to disas-semble LPL NPs. The release of Dox from Dox/LPL NPs were around

Fig. 1. 1H NMR spectra of LPL (A) and PL (B). The characteristic peaks of LBA, PEG and LCA are depicted.

B. Singh et al. / Biomaterials 116 (2017) 130e144134

50% in PBS buffer (pH 7.4) within 2 h while the release wasapproximately 60% in acetate buffer (pH 5.5) in 2 h (Fig. 6). Therelease of Dox from Dox/PL NPs was 44% and 57% in PBS buffer (pH7.4) and acetate buffer (pH 5.5), respectively in 2 h. Although theDox release from NPs increased slightly with time after 2 h of in-cubation, the release of Dox from NPs was pH dependent. Therelease of Dox from NPs was also tested in the presence of serumusing Opti-MEMmedium. Addition of serummedium in the buffersdid not alter the release of Dox from NPs.

3.3. Cell specificity and cytotoxicity of NPs in vitro

To determine the cell specificity of NPs, SK-HEP-1 humanendothelial cell line derived from adenocarcinoma of the liver wasused. The cell line was cultured and maintained as described inmaterials and methods. FITC-labeled NPs with liver-specific ligand(LPL) or without ligand (PL) at 25 mM concentrationwere treated toSK-HEP-1 cells to observe the cell-specificity of these NPs by CLSM(Fig. 7A). Quantitative analysis of CLSM images exhibited that thecellular uptake of FITC-LPL NPs was significantly higher than theuptake of FITC-PL NPs (Fig. 7B). The results showed the markeddifference between the cellular uptake of NPs with or withoutligand indicating the receptor-mediated endocytosis of NPs. Toobserve the ligand-receptor interaction on endocytosis, a compet-itive binding experiment was performed where Dox/LPL NPs weretreated with SK-HEP-1 cells in the presence of different concen-tration of galactose, and the uptake of NPs inside cells was analyzedby flow cytometry. The results demonstrated that the higher theconcentration of galactose treatment to the cells, the lower theuptake of Dox/LPL NPs inside cells (Fig. 7C). Since the pre-treatment

of galactose hindered the cells to uptake the LPL-core NPs, it wasobvious that the endocytosis of Dox/LPL NPs occurred throughgalactose-asialoglycoprotein interaction. As a result, the cellviability of SK-HEP-1 cells with the treatment of 25 mM concen-tration of Dox/LPL and Dox/PL NPs was 20% and 80%, respectively(Fig. 8A). Remarkably, Dox/LPL NPs exhibited significantly highercytotoxicity than Dox/PL NPs due to higher internalization of Dox/LPL NPs through receptor-mediated endocytosis that subsequentlyresulted in higher cytotoxic effect. Consistently, the cell viability ofHepG2 liver cancer cells treated with LPL and PL at 50 mM con-centration was 20% and 65%, respectively due to the targetingproperties of LPL nanoparticles through asialoglycoprotein re-ceptors [13]. The cell viability decreased drastically with increaseconcentration of Dox/PL NPs used while high concentration of Dox/LPL NPs treatment showed less changes in the cell viability as theviable cells were very low even at the lowest concentration of Dox/LPL NPs used.

To further compare the dose-dependent effects of NPs withsmall molecule drugs, various concentrations of Dox, LCA, LPL andDox/LPL NPs were treated to the cells. Since LPL is the core of theNPs, the concentration used for in vitro experiments was on thebasis of LPL. Thus, the amount of free Dox and LCA used wereequivalent to the amount present in Dox/LPL and LPL, respectively.When 100 mM of LPL or Dox/LPL was treated to the cells, 70 mM ofLCA was used since LPL contains 70 mol% of LCA. Similarly, 4.8 nMof Dox was used when 100 mM of Dox/LPL was treated to the cells.When the cells were treated with 50 mM concentration of LPL orDox/LPL, the cell viability of SK-HEP-1 cells was approximately 50%and 90% with LCA and Dox treatment, respectively while the cellviability of SK-HEP-1 cells treated with LPL and Dox/LPL NPs was

Fig. 2. FTIR spectra of LPL (A) and PL (B). The characteristic IR peaks of LBA, PEG and LCA are indicated by arrowheads.

B. Singh et al. / Biomaterials 116 (2017) 130e144 135

Fig. 3. Physical properties of NPs in solid state (A) and aqueous suspension (B). PL (I), LPL (II), Dox/PL (III) and Dox/LPL (IV).

Fig. 4. SEM images of NPs. Dox/LPL (A), Dox/PL (B), LPL (C) and PL (D). Scale Bar ¼ 1 mm.

B. Singh et al. / Biomaterials 116 (2017) 130e144136

20% and 15%, respectively (Fig. 8B). Although NPs and smallmolecule drugs exerted cytotoxic effects in dose-dependentmanner, the efficacy of NPs seemed much higher than that ofsmall molecule drugs. Previously, the cytotoxicity of LPL NPs onnormal LO2 liver cells and HepG2 liver cancer cells was compared[13]. At 50 mM concentration of LPL treatment, the cell viability ofLO2 and HepG2 cells and was 75% and 20%, respectively. The resultsindicated that LPL NPs selectively killed cancer cells, while sparingnormal cells. In addition, the present data apparently indicated thatDox/LPL NPs had higher efficacy than LPL NPs.

3.4. Delivery of Dox by NPs and mechanistic study of cytotoxicityin vitro

To investigate the delivery of Dox by NPs inside cells in vitro, SK-HEP-1 cells were cultured as described before. Various concentra-tion of Dox/LPL NPs (5, 25 and 50 mM) in minimal medium weretreated to the cells for 4 h and further incubated with growthmedium for 18 h. The cellular uptake of NPs was observed by CLSM

(Fig. 9A). When the cells were treated with 5 mM concentration ofDox/PL NPs, Dox was sparsely detected inside the cells possibly dueto low amount of NPs uptake. When the cells were treated withhigher concentration of NPs, the accumulation of Dox around nu-cleus was clearly visible due to high binding affinity of Dox withDNA. At 50 mM concentration of NPs treatment, Dox was denselydetected inside the nucleus (Fig. 9B). These results clearly illus-trated that Dox/LPL NPs efficiently released Dox inside the cells.Altogether, the accumulation of Dox inside the cells was directlyproportional to the concentration of NPs treated to the cells.

To elucidate the mechanism of cytotoxicity induced by NPsdelivery, Annexin V-FITC/PI assay was performed. SK-HEP-1 cellswere treated with 50 mM concentration of Dox, LCA, LPL or Dox/LPLNPs as described above. The cells were then stained with AnnexinV-FITC/PI and analyzed by fluorescence-activated cell sorting(FACS). The FACS data demonstrated that the apoptosis induced byDox was very low while the majority of LCA-treated cells followedapoptotic pathway (Fig. 10A). Similarly, the cell death occurred byLPL NPs primarily followed apoptotic pathway, and the induction of

Fig. 5. Size distribution of NPs. The mean hydrodynamic diameters of Dox/LPL (A), Dox/PL (B), LPL (C) and PL (D) were measured by DLS.

Table 1Characterization of nanoparticles.

Nanoparticles Sizea [nm] Zeta potentialb [mV] Polydispersity indexa

Dox/LPL 915.9 ± 136.1 �9.345 ± 0.3 0.538Dox/PL 866.6 ± 215.7 �9.855 ± 0.1 0.383LPL 355.7 ± 39.7 �8.215 ± 1.2 0.254PL 310.6 ± 39.0 �7.765 ± 0.3 0.244

a Values determined by DLS.b Values determined by ELS.

0 0.5 1 2 4 8 12 24

20

40

60

80

100Dox/LPL (pH 7.4)

Dox/PL (pH 7.4)

Dox/LPL (pH 5.5)

Dox/PL (pH 5.5)

Time (h)

)%(

esaeleR

xoD

Fig. 6. Release profile of Dox from NPs at different pH.

B. Singh et al. / Biomaterials 116 (2017) 130e144 137

apoptosis by LPL NPs was greater than only LCA treatment(Fig. 10B). Importantly, the flow cytometric data indicated that Dox/LPL NPs induced maximum apoptotic cell death compared withother groups resulting from combined effect of Dox and LPL.

3.5. Effect of NPs on tumor growth in liver in vivo

Orthotopic xenograft mouse models of liver cancer weredeveloped to investigate the in vivo efficacy of NPs. The bodyweight of mice in each group were observed at the interval of 3e4days after injecting SK-HEP-1 cells in liver to drug treatment period.

There was no significant loss in body weight of the mice in anygroup after the treatment of drugs or NPs (Fig. S6). Following thetreatments of NPs in mice as described in materials and methods,

Fig. 7. Cell specificity of NPs. CLSM images of cell specificity of FITC-NPs treated at 25 mM concentration (A), Normalized FITC-NPs uptake (B) and Competitive inhibition of galactoseon cellular uptake of Dox/LPL NPs in SK-HEP-1 cells. Data were represented as mean ± SD (n ¼ 3, *p < 0.05, **p < 0.01 versus control).

DOX LCA LPL DOX/LPL0

20

40

60

80

100 25 �M 50 �M100 �M

*****

)%( ytilibai v lle

C

25 μM 50 μM 100 μM0

20

40

60

80

100DOX/PLDOX/LPL

*

***

)%( ytilibaiv lle

C

A B

Fig. 8. Cell specificity and cytotoxicity of NPs. Cell specificity of NPs with or without ligand (A), and cytotoxicity of NPs at different concentration (B). SK-HEP-1 cells were treatedwith Dox, LCA, LPL and Dox/LPL for 4 h and cell viability of the cells was analyzed by MTT assay. The amounts of Dox and LCA used were equivalent to their content in Dox/LPL andLPL, respectively. Data were represented as mean ± SD (n ¼ 3, *p < 0.05, **p < 0.01, ***p < 0.001 versus Dox).

B. Singh et al. / Biomaterials 116 (2017) 130e144138

liver was isolated from each mouse to observe the tumor conditionand intrahepatic metastasis (Fig. 11A). Morphological analysis ofliver revealed that the number of tumor nodules (Fig. 11B) andtumor volume (Fig. 11C) were relatively smaller in both LPL andDox/LPL NPs-treated groups compared to control group. Particu-larly, the tumor volume prominently decreased in Dox/LPL NPs-treated group. When compared between small molecule drugs,the tumor number was lower in Dox treatment group, while tumorvolume was comparatively smaller in LCA treatment group. On thecontrary, the liver tumor volume of Dox-treated group was similarto control group and the average number of tumor nodules in liverof LCA-treated group was relatively higher than that of control

group. To assess the general state of the liver, blood sera from thedrugs or NPs-treated mice were collected and analyses of alanineaminotransferase (ALT) and aspartate aminotransferase (AST)levels were performed. The results exhibited the presence of lowerquantity of ALT (Fig. 11D) and AST (Fig. 11E) in NPs or drugs-treatedmice than saline-treated group. Consistently, Dox/LPL NPs-treatedgroup had the lowest levels of both ALT and AST activities indi-cating the normal state of livers on these treatment group. Toexamine tumor margins or metastases, the liver tissues weresectioned and stained for histology observation. When the sectionsof liver tissues were observed under microscope, the margins oftumor metastases were clearly visualized in the images (Fig. 11F).

Fig. 9. Fluorescent images of Dox released inside cells. SK-HEP-1 cells were treated with Dox/LPL NPs at various concentration for 4 h and the localization of Dox inside cells wasobserved by CLSM (A) and normalized Dox uptake (B). Data were represented as mean ± SD (n ¼ 3, *p < 0.05, **p < 0.01, ***p < 0.001 versus control).

Fig. 10. Apoptosis detection by Annexin V-FITC analysis. SK-HEP-1 cells were treated with Dox, LCA, LPL and Dox/LPL at concentration of 50 mM for 4 h. The apoptotic cells weredepicted in FACS data (A) and the percentage of total cell populations undergoing apoptosis were measured (B). Data were represented as mean ± SD (n ¼ 3, *p < 0.05, **p < 0.01,***p < 0.001 versus control).

B. Singh et al. / Biomaterials 116 (2017) 130e144 139

3.6. Effect of NPs on cell proliferation and angiogenesis in liverin vivo

To determine the anti-tumor effect of NPs in orthotopic

xenograft model of liver cancer in vivo, liver tissues were homog-enized, proteins were isolated and Western blots were performedto detect the tumor proliferation and angiogenesis markers(Fig. 12A). The protein expression levels of PCNA and VEGF were

Fig. 11. Effects of NPs on orthotopic liver tumor in mice. Orthotopic tumor and metastatic tumor nodules after the treatment (A), the red circles indicate the tumor nodules on theliver. The number of tumor nodules in liver indicating intrahepatic metastasis (B). Calculated liver tumor volume (C). Serum analyses of liver damage indicators, ALT (D) and AST (E).Histological examination of the liver tumors using Hematoxylin & Eosin staining (F), the arrows indicate the margin of tumor in liver tissue section. Data were represented asmean ± SD (n ¼ 4, *p < 0.05 versus saline; #p < 0.05 versus LCA). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of thisarticle.)

B. Singh et al. / Biomaterials 116 (2017) 130e144140

comparatively smaller in NPs-treated groups than Dox or LCA-treated groups. Notably, the expression levels of PCNA and VEGFwere remarkably low in Dox/LPL NPs-treated groups. Densito-metric analyses of Western blots also confirmed that the treatmentwith Dox/LPL NPs in the mouse models of liver cancer significantlydecreased the level of PCNA and VEGF expression (Fig. 12B).

3.7. Effect of NPs on cell cycle in liver in vivo

To determine the effect of NPs on cell cycle of liver tumor in vivo,Western blot was performed to detect the expression levels of cellcycle-related proteins such as cyclin B1, CDK2, and CDK4 (Fig. 13A).As in previous results, the drugs treated with NPs formulations inorthotopic mouse models of liver cancer had low cell cycle activity.

Specifically, the treatment of Dox/LPL NPs in the mouse modelssignificantly suppressed CDK2 and CDK4 expressions inhibitingrapid cell cycle of tumor growth (Fig. 13B).

3.8. Effect of NPs on apoptotic protein expression in liver in vivo

To determine the mechanism of anti-tumor effect of NPs in vivo,the protein expression levels of apoptosis-related proteins, cyto-chrome c and Casp-3, in the liver tissues were observed byWesternblot (Fig. 14A). The expression levels of these proteins weresignificantly higher in LPL and Dox/LPL NPs-treated groups thansaline, Dox or LCA-treated groups (Fig. 14B). Interestingly, theexpression of Casp-3 in LCA-treated group was relatively higherthan Dox-treated groupwhile cytochrome c expressionwas slightly

B. Singh et al. / Biomaterials 116 (2017) 130e144 141

higher in case of Dox-treated group than LCA-treated group.

4. Discussion

Nanoparticle drug has received much attention in cancer ther-apy following the first clinical approval of Doxil (PEGylatedliposome-encapsulated Dox) by the US Food and Drug Adminis-tration (FDA) in 1995 [22]. Over the past two decades, a number ofNPs have been designed and developed for cancer drug delivery[23]. However, the treatment of cancers with NPs, in practice, hasbeen confronted with limited success. It is understood that severalbiological barriers hinder the successful delivery of NPs in vivo, butthe prime causes of hindrance are associated with the physicalcharacteristics of NPs such as shape, size and surface propertieswhich greatly contribute to non-specific distribution resultinginadequate delivery of therapeutic drugs at cancer sites [24].Therefore, rational design of NPs is critical to address the biologicalbarriers to successful cancer drug delivery.

Traditionally, chemotherapy drugs have been used as the primechoice of treatment for most of the cancers, although the actualefficacy of drugs rely on the accumulation of the drugs to a specificsite. The extensive use of chemotherapy drugs is primarily due totheir ability to arrest the normal functions of a cell by inhibitingreplication or inducing apoptosis [25]. Conversely, a number ofinherent nature of these small molecule drugs hinder their realefficiency in cancer chemotherapy. These limitations, however,have opened new avenue of anti-cancer drug delivery using com-bination chemotherapy. In fact, chemotherapy drugs exert greater

Saline

LCADox

LPL

Dox/LPL

0.00

0.02

0.04

0.06

0.08

0.10

**#&

*

**

VEGF

nitcA/FGEV

Saline

LCADox

LPL

Dox/LPL

0

1

2

3

***##&

*****

*

#

PCNA

nitcA /ANC P

VEGF

PCNA

Actin

Saline LCA Dox Dox/LPLLPLA

B

Fig. 12. Effects of NPs on proliferation and angiogenesis in orthotopic liver tumor inmice. The protein levels of PCNA and VEGF in the liver tissue lysates were determinedbyWestern blot. Only representative bands of one mouse from each group were shownin the graph (A). To compare the levels of protein expression, densitometric analyses ofWestern blot bands were performed using four orthotopic liver tumor mice in eachgroup. Actin was used as a control (B). Data were represented as mean ± SD (*p < 0.05,**p < 0.01, ***p < 0.001 versus saline; #p < 0.05, ##p < 0.01 versus LCA; &p < 0.05versus Dox).

efficiency when given in combination. The rationale for combina-tion chemotherapy is to use the combination of two or more drugswith different modes of actions, thereby reducing the risk ofdeveloping resistant cancer cells. The common combinations ofchemotherapy drugs include Dox, paclitaxel, cisplatin and cyclo-phosphamide which have been extensively used to treat variouscancer types [26e28]. Recently, a new small molecule drug LCA hasbeen introduced in scientific literatures that induces apoptotic celldeath specifically to cancer cells [12]. However, LCA encountersseveral hurdles for in vivo use due to its low solubility and rapidelimination from systemic circulation. However, these types ofdefects of small molecule drugs have been ameliorated by conju-gation with PEG to enhance the solubility as well as molecularweight of the drugs for greater retention in body fluids [29e31].

In this study, we developed a polymeric NP, a self-assembledproduct of LPL, to deliver Dox (anti-cancer drug) for effective andselective delivery to liver. LPL consists of LCA (anti-cancer drug) andLBA (targeting ligand) at the two ends of PEG. LCA is a secondarybile acid product that is formed by the action of intestinal microbialenzymes in the gut and metabolized in liver [32]. LBA is a disac-charide formed from gluconic acid and galactose which has highbinding affinity with asialoglycoprotein receptors on hepatocytes ofliver. Dox is a small molecule drug belongs to anthracyclines thatforms a ternary complex with topoisomerase-II and DNA to impedecellular replication triggering cell death [33]. Depending upon thedoses, Dox activates different regulatory mechanisms to induceeither apoptosis or cell death through mitotic catastrophe [34].Mitotic catastrophe is governed by numerous molecular players, inparticular, cell-cycle-specific kinases (such as cyclin B1-dependentkinase CDK1) and cell-cycle checkpoint proteins [35]. On the con-trary, LCA mainly triggers cell death through apoptosis which isprimarily dependent on the activation of Casp-3 and -8 in cells [36].

The main goal of this study was to develop a polymeric NP withdual properties, in fact, to use a NP drug as a drug carrier (Scheme2). Previously, we developed PL and LPL as active polymer-drugsboth of which self-assembled to NPs (300e400 nm) whenobserved by DLS and TEM measurements. Surprisingly, when weobserved the same NPs by SEM, PL NPs had spherical shapes whileLPL NPs had non-spherical shapes that resemble with RBCs.Consistently, Dox/PL NPs were spherical whereas Dox/LPL NPs weremostly RBC-like nanodiscs. The distinct shape added an anotheradvantage to Dox/LPL NPs as drug carrier. The importance of NPshape in cancer drug delivery has recently been recognized and it isnow acknowledged that the shape of NP plays a key role in drugcirculation time, biodistribution, cellular uptake and even targetingin NP drug delivery [37,38]. Although recent advances in nano-technology have enabled the modification of NPs with variousshapes and sizes, most current NPs designed for anti-cancer drugsare formed in the spherical shape [39,40]. In contrast, bacteria andviruses have evolved in non-spherical shapes enabling them toevade an immune response [41]. Hence, there is a growingawareness and interest that NPs with non-spherical shapes, thatmimic natural biological systems such as RBCs, might displaydistinctly beneficial properties over identical nanospheres [42,43].In fact, the unique morphological properties of RBCs offer them agreat advantage to routinely pass through blood vessels travelingthroughout the body.

Cells constantly internalize molecules from external environ-ment to various intracellular compartments through endocytosis.The efficiency of this process is increased when the receptors ofcells have high binding affinity with external molecules for inter-nalization. The process of internalization through receptor-ligandinteraction is called receptor-mediated endocytosis. Therefore,this route has been exploited to enhance the efficiency of drugdelivery into specific target cell types, including the

CDK2

Cyclin B1

Actin

CDK4

Saline LCA Dox Dox/LPLLPL

Saline

LCADox

LPL

Dox/LPL

0

1

2

3

4

5

**$

Cyclin B1

n itcA/1Bnilcy

C

Saline

LCADox

LPL

Dox/LPL

0.0

0.2

0.4

0.6

0.8

*$$

&&&

CDK2

nitc A/2KD

C

Saline

LCADox

LPL

Dox/LPL

0.0

0.5

1.0

1.5

2.0

2.5

$ &&#

#$

CDK4

nitcA/4

KD

C

A

B

Fig. 13. Effects of NPs on cell cycle in orthotopic liver tumor in mice. The protein expression levels of cyclin B1, CDK2 and CDK4 in the liver tissue lysates were determined byWestern blot (A). To compare the levels of protein expression, densitometric analyses of Western blot bands were performed using three orthotopic liver tumor mice in each group.Actin was used as a control (B). Data were represented as mean ± SD (*p < 0.05, **p < 0.01 versus saline; #p < 0.05 versus LCA; $p < 0.05, $$p < 0.01 versus Dox; &&p < 0.01,&&&p < 0.001 versus LPL).

B. Singh et al. / Biomaterials 116 (2017) 130e144142

asialoglycoprotein for hepatocyte targeting [44,45]. Despitecontradiction in the expression of asialoglycoprotein at low level inextrahepatic cells such as intestine and kidney [46], many authorsused lactobionic acid for targeted delivery of anticancer drug tohepatocarcinoma cells because of higher expression of asialogly-coprotein on the surface of hepatocytes [47]. To verify the conceptof receptor-ligand interaction for targeted delivery of NPs, wedesigned and synthesized LPL and PL, with and without ligand,respectively. We further labeled LPL and PL with FITC. When theseFITC-NPs were treated to asialoglycoprotein expressing SK-HEP-1 cells, the cellular uptake of FITC-LPL NPs by SK-HEP-1 cells wasremarkably higher in comparison with the uptake of FITC-PL NPs.We further encapsulated Dox in LPL and PL NPs to produce Dox/LPLNPs and Dox/PL NPs, respectively. Again, the cytotoxicity inducedby Dox/LPL NPs was much higher than that of Dox/PL NPs. The highcytotoxic effect of Dox/LPL NPs to SK-HEP-1 cells was attributed tohigh cellular uptake of these NPs by receptor-mediated endocy-tosis. These results are consistent with our previous study of LPLNPs which exhibited higher cytotoxicity than PL NPs in anotherhuman liver cancer cell line HepG2 [13]. We had also observedin vivo biodistribution profile of LPL NPs and confirmed the tar-geting ability of LPL NPs to liver in the study. It is now evident fromour previous and current findings that the ligand can make markeddifference in endocytosis of NPs. Therefore, the properties of Dox/LPL NPs are deemed to reduce the therapeutic dose as well as off-target effects when applied for cancer therapy in vivo. Hence, wedid not use Dox/PL NPs for further in vivo experiment.

Prior to in vivo experiment, we investigated the mechanism ofDox/LPL NPs-induced cell death of SK-HEP-1 cells. We first inter-rogated and analyzed the rate of Dox delivery by Dox/LPL NPs in SK-

HEP-1 cells using confocal fluorescence microscopy. The confocalmicroscopy images clearly demonstrated the increased accumula-tion of Dox inside the Dox/LPL NPs-treated cells in concentration-dependent manner. The images also highlighted the efficientrelease of Dox from the Dox/LPL NPs inside the cells. The highamount of Dox delivered inside cells is attributed to the enhancedinternalization of Dox/LPL NPs through endocytosis which subse-quently resulted in high cytotoxicity. The high cytotoxic effect ofDox/LPL NPs led us to investigate the mode of action of cytotoxicity.Hence, we used flow cytometry to distinguish the cell populationstending to apoptosis induced by Dox/LPL NPs, Dox and LCA indi-vidually. The FACS data demonstrated a clear difference in celldeath induced by Dox and LCA. While the majority of cell deathinduced by LCA followed apoptotic pathway, only a minor popu-lation of cell death occurred through apoptosis with Dox treatment.As in case of LCA, apoptosis was the main cause of cell deathoccurred with LPL NPs treatment. Accordingly, Dox/LPL NPs causedcell death jointly through apoptosis and mitotic catastrophe. Be-sides, we previously observed that the cytotoxicity of LPL or LCAwas significantly higher in cancer cells than normal cells [13]. Dueto the fact that Dox/LPL NPs had similar properties with LPL NPs, wespeculated that Dox/LPL NPs would exert low toxicity to normalcells in vivo.

To further examine pharmacological effects of Dox/LPL NPsin vivo, we established orthotopic xenograft mouse model of livercancer. Orthotopic tumor models are deemed more clinically rele-vant and comprehensive predictive models of drug test than con-ventional subcutaneous xenograft tumor models because tumorcells are directly implanted into the organ of origin in orthotopicmodels to imitate and reflect the original condition of tumor

Casp-3

Cyt c

Actin

Saline LCA Dox Dox/LPLLPL

Saline

LCADox

LPL

Dox/LPL

0

10

20

30

**##$$

*#

Cytochrome c

t yC

cnitcA/

A

B

Saline

LCADox

LPL

Dox/LPL

0.0

0.5

1.0

1.5

**##$$

**###$$$

Caspase-3nitcA/3-psa

C

Fig. 14. Effects of NPs on the expression of apoptotic proteins in orthotopic liver tumorin mice. The protein expression levels of cytochrome c (Cyt c) and Casp-3 in the livertissue lysates were determined by Western blot (A). To compare the levels of proteinexpression, densitometric analyses of Western blot bands were performed using fourorthotopic liver tumor mice in each group. Actin was used as a control (B). Data wererepresented as mean ± SD (*p < 0.05, **p < 0.01 versus saline; #p < 0.05, ##p < 0.01,###p < 0.001 versus LCA; $$p < 0.01, $$$p < 0.001 versus Dox).

Receptor

Liver cancer cell

Mitotic catastropheApoptosis

Dead cancer cell

Ligand

Nanoparticle drug(also carrier)Anti-cancer drug

Scheme 2. Schematic diagram of drug-loaded NP and its action for cancer cell killing.

B. Singh et al. / Biomaterials 116 (2017) 130e144 143

growth, tumor distribution and metastases. Hence, we treated theNPs or drugs to orthotopic tumor models through IP route for fourweeks. After the treatment, there were a noticeable difference innumber of tumor nodules and tumor volume in the liver of Dox/LPLNPs-treated mice compared to control group. In agreement withthe in vitro results of cytotoxicity, Dox/LPL NPs greatly reduced thetumor volume as well as the liver metastasis. The tumor volumeand metastasis in liver varied between Dox and LCA treatment,

probably due to their inherent mode of actions. We further testedthe general function of liver using blood serum samples from thedrugs-treated mice. The healthy state of livers in Dox/LPL NPs-treated mice also verified the effective delivery of the drugs to liver.

We further analyzed tumor markers at protein level to detectcancer activity in vivo. Particularly, we detected the expressionlevels of PCNA and VEGF related with tumor proliferation andangiogenesis in liver tissues of drugs-treated mice. While PCNA iscritical for cell survival, proliferation and cell cycle progression [48],VEGF is important factor of angiogenesis in liver cancer that isnecessary to stimulate blood vessel growth for tumor and metas-tasis [49]. These proteins can be down-regulated or over-expressedin tumor tissues depending upon the effectiveness of treatment.The lowest levels of PCNA and VEGF in Dox/LPL NPs-treated groupindicated that Dox/LPL NPs have the highest anti-proliferating andangiogenetic effects on liver tumor and hence resulted in greaterefficacy in liver cancer treatment.

Generally, cancer alters cell cycle of mitosis leading to unre-strained proliferation and increased expression of cell cycle regu-latory proteins, such as cyclins and CDKs [50]. Therefore, wedetected the protein levels of key cell cycle-related proteins CDK4,CDK2, and cyclin B1 in liver tissues after treatment. We observedthat the treatment with Dox/LPL NPs substantially suppressed cellcycle activities in the liver tissues compared with saline, LCA, orDox treatment. Besides, Dox/LPL NPs-treated groups significantlyreduced CDK2 and CDK4 expressions arresting the rapid cell cycleof tumor growth. Taken together, these results were also consistentwith small tumor volume in livers in Dox/LPL NPs-treated mice.

Normally, cell proliferation is offset by cell death through anintricate cellular process of apoptosis in which the cell usesspecialized cellular machinery to eliminate itself. Deregulation inapoptotic cell death machinery is a hallmark of cancer [51]. Mostanti-cancer drugs currently used in cancer therapy exploit theintact apoptotic machinery to trigger cancer cell death. Cytochromec and Casp-3 are the fundamental regulators of apoptosis. Releaseof cytochrome c from mitochondria is a central event in apoptoticsignaling to activate a caspase cascade leading to cell death [52].The upregulated cytochrome c and Casp-3 in LPL and Dox/LPL NPs-treated groups revealed that LPL and Dox/LPL NPs could potentlyinduce apoptosis in liver cancer to inhibit tumor growth. In addi-tion, the results were consistent with the in vitro cell deathmechanism of LPL and Dox/LPL NPs.

In conclusion, we have successfully developed an efficient NPdelivery system for liver-targeted combination cancer therapy andillustrated to suppress liver tumor growth in an orthotopic xeno-graft model of liver cancer. This study further exemplifies the su-periority of combination therapy over individual NP drug orconventional small molecule drug for cancer therapy. It is antici-pated that the combination therapy with different mode of actionsof drugs could be a promising therapeutic strategy toward specificcancer treatment.

Acknowledgements

This researchwas supported by National Research Foundation ofKorea Grant (NRF-2014R1A1A2007163) funded by Ministry of Sci-ence, ICT and Future Planning (MSIF). Yoonjeong Jang, Hyeon-JeongKim, Ah Young Lee and Sushila Maharjan are supported by BrainKorea 21 Program for Leading Universities & Students (BK21 PLUS),Republic of Korea. We acknowledge National Center for Inter-University Research Facilities (Seoul National University) for FACSanalysis and we are grateful to Yong Jae Kim to operate CLSM. Wealso acknowledge the National Instrumentation Center for Envi-ronmental Management (Seoul National University) for permissionto take NMR, FTIR, DLS, ELS and SEM measurements.

B. Singh et al. / Biomaterials 116 (2017) 130e144144

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.biomaterials.2016.11.040.

References

[1] K. Cho, X. Wang, S. Nie, D.M. Shin, Therapeutic nanoparticles for drug deliveryin cancer, Clin. Cancer Res. 14 (5) (2008) 1310e1316.

[2] M. Chidambaram, R. Manavalan, K. Kathiresan, Nanotherapeutics to overcomeconventional cancer chemotherapy limitations, J. Pharm. Pharm. Sci. 14 (1)(2011) 67e77.

[3] A.Z. Wang, R. Langer, O.C. Farokhzad, Nanoparticle delivery of cancer drugs,Annu. Rev. Med. 63 (2012) 185e198.

[4] R.H. Mu�Eller, K. Ma�Eder, S. Gohla, Solid lipid nanoparticles (SLN) for controlleddrug deliveryea review of the state of the art, Eur. J. Pharm. Biopharm. 50 (1)(2000) 161e177.

[5] M. Arruebo, R. Fern�andez-Pacheco, M.R. Ibarra, J. Santamaría, Magneticnanoparticles for drug delivery, Nano Today 2 (3) (2007) 22e32.

[6] S.D. Steichen, M. Caldorera-Moore, N.A. Peppas, A review of current nano-particle and targeting moieties for the delivery of cancer therapeutics, Eur. J.Pharm. Sci. 48 (3) (2013) 416e427.

[7] K.T. Nguyen, Targeted nanoparticles for cancer therapy: promises and chal-lenges, J. Nanomedicine Nanotechnol. 2011 (2012).

[8] Y.A. Luqmani, Mechanisms of drug resistance in cancer chemotherapy, Med-ical principles and practice, Int. J. Kuwait Univ. Health Sci. Centre 14 (Suppl 1)(2005) 35e48.

[9] M.M. Gottesman, Mechanisms of cancer drug resistance, Annu. Rev. Med. 53(1) (2002) 615e627.

[10] J.Y. Choi, R.K. Thapa, C.S. Yong, J.O. Kim, Nanoparticle-based combination drugdelivery systems for synergistic cancer treatment, J. Pharm. Investig. (2016)1e15.

[11] J. Khandare, T. Minko, Polymeredrug conjugates: progress in polymeric pro-drugs, Prog. Polym. Sci. 31 (4) (2006) 359e397.

[12] A.A. Goldberg, A. Beach, G.F. Davies, T.A. Harkness, A. Leblanc, V.I. Titorenko,Lithocholic bile acid selectively kills neuroblastoma cells, while sparingnormal neuronal cells, Oncotarget 2 (10) (2011) 761e782.

[13] N. Gankhuyag, B. Singh, S. Maharjan, Y.J. Choi, C.S. Cho, M.H. Cho, Gal-actosylated poly (Ethyleneglycol)-Lithocholic acid selectively kills hepatomacells, while sparing normal liver cells, Macromol. Biosci. 15 (6) (2015)777e787.

[14] G.Y. Wu, C.H. Wu, Receptor-mediated delivery of foreign genes to hepato-cytes, Adv. Drug Deliv. Rev. 29 (3) (1998) 243e248.

[15] J. Sudimack, R.J. Lee, Targeted drug delivery via the folate receptor, Adv. DrugDeliv. Rev. 41 (2) (2000) 147e162.

[16] Z.M. Qian, H. Li, H. Sun, K. Ho, Targeted drug delivery via the transferrinreceptor-mediated endocytosis pathway, Pharmacol. Rev. 54 (4) (2002)561e587.

[17] E. Wagner, D. Curiel, M. Cotten, Delivery of drugs, proteins and genes into cellsusing transferrin as a ligand for receptor-mediated endocytosis, Adv. DrugDeliv. Rev. 14 (1) (1994) 113e135.

[18] C.-S. Cho, A. Kobayashi, R. Takei, T. Ishihara, A. Maruyama, T. Akaike, Receptor-mediated cell modulator delivery to hepatocyte using nanoparticles coatedwith carbohydrate-carrying polymers, Biomaterials 22 (1) (2000) 45e51.

[19] H.-L. Jiang, J.-T. Kwon, E.-M. Kim, Y.-K. Kim, R. Arote, D. Jere, H.-J. Jeong, M.-K. Jang, J.-W. Nah, C.-X. Xu, Galactosylated poly (ethylene glycol)-chitosan-graft-polyethylenimine as a gene carrier for hepatocyte-targeting, J. Control.Release 131 (2) (2008) 150e157.

[20] Y. Yang, Z. Zhang, L. Chen, W. Gu, Y. Li, Galactosylated poly(2-(2-amino-ethyoxy)ethoxy)phosphazene/DNA complex nanoparticles: in vitro andin vivo evaluation for gene delivery, Biomacromolecules 11 (4) (2010)927e933.

[21] N. Gankhuyag, B. Singh, S. Maharjan, Y.J. Choi, C.S. Cho, M.H. Cho, Gal-actosylated poly(ethyleneglycol)-lithocholic Acid selectively kills hepatomacells, while sparing normal liver cells, Macromol. Biosci. 15 (6) (2015)777e787.

[22] J.S. James, DOXIL approved for KS, AIDS Treat. News (no 236) (1995) 6.[23] R.A. Petros, J.M. DeSimone, Strategies in the design of nanoparticles for

therapeutic applications, Nat. Rev. Drug Discov. 9 (8) (2010) 615e627.[24] E. Blanco, H. Shen, M. Ferrari, Principles of nanoparticle design for overcoming

biological barriers to drug delivery, Nat. Biotechnol. 33 (9) (2015) 941e951.[25] S.-S. Feng, S. Chien, Chemotherapeutic engineering: application and further

development of chemical engineering principles for chemotherapy of cancerand other diseases, Chem. Eng. Sci. 58 (18) (2003) 4087e4114.

[26] R.B. Weiss, The anthracyclines: will we ever find a better doxorubicin?Seminars Oncol. 19 (6) (1992) 670e686.

[27] C.M. Spencer, D. Faulds, Paclitaxel, Drugs 48 (5) (1994) 794e847.[28] D. Gately, S. Howell, Cellular accumulation of the anticancer agent cisplatin: a

review, Br. J. Cancer 67 (6) (1993) 1171.[29] A. Pendri, C.W. Gilbert, S. Soundararajan, D. Bolikal, R.G.L. Shorr,

R.B. Greenwald, PEG modified anticancer drugs: synthesis and biological ac-tivity, J. Bioact. Compat. Polym. 11 (2) (1996) 122e134.

[30] D. Vllasaliu, R. Fowler, S. Stolnik, PEGylated nanomedicines: recent progressand remaining concerns, Expert Opin. Drug Deliv. 11 (1) (2014) 139e154.

[31] Y. Cu, W.M. Saltzman, Controlled surface modification with poly(ethylene)glycol enhances diffusion of PLGA nanoparticles in human cervical mucus,Mol. Pharm. 6 (1) (2009) 173e181.

[32] J.K. Nicholson, E. Holmes, J. Kinross, R. Burcelin, G. Gibson, W. Jia, S. Pettersson,Host-gut microbiota metabolic interactions, Science 336 (6086) (2012)1262e1267.

[33] L.H. Hurley, DNA and its associated processes as targets for cancer therapy,Nat. Rev. Cancer 2 (3) (2002) 188e200.

[34] Y.W. Eom, M.A. Kim, S.S. Park, M.J. Goo, H.J. Kwon, S. Sohn, W.H. Kim, G. Yoon,K.S. Choi, Two distinct modes of cell death induced by doxorubicin: apoptosisand cell death through mitotic catastrophe accompanied by senescence-likephenotype, Oncogene 24 (30) (2005) 4765e4777.

[35] M. Castedo, J.-L. Perfettini, T. Roumier, K. Andreau, R. Medema, G. Kroemer,Cell death by mitotic catastrophe: a molecular definition, Oncogene 23 (16)(2004) 2825e2837.

[36] A.A. Goldberg, V.I. Titorenko, A. Beach, J.T. Sanderson, Bile acids induceapoptosis selectively in androgen-dependent and-independent prostate can-cer cells, PeerJ 1 (2013) e122.

[37] J.A. Champion, Y.K. Katare, S. Mitragotri, Particle shape: a new designparameter for micro- and nanoscale drug delivery carriers, Journal ofcontrolled release, Off. J. Control. Release Soc. 121 (1e2) (2007) 3e9.

[38] S. Venkataraman, J.L. Hedrick, Z.Y. Ong, C. Yang, P.L. Ee, P.T. Hammond,Y.Y. Yang, The effects of polymeric nanostructure shape on drug delivery, Adv.Drug Deliv. Rev. 63 (14e15) (2011) 1228e1246.

[39] E.A. Simone, T.D. Dziubla, V.R. Muzykantov, Polymeric carriers: role of ge-ometry in drug delivery, Expert Opin. Drug Deliv. 5 (12) (2008) 1283e1300.

[40] N. Doshi, S. Mitragotri, Macrophages recognize size and shape of their targets,PloS One 5 (4) (2010) e10051.

[41] S.S. Justice, D.A. Hunstad, L. Cegelski, S.J. Hultgren, Morphological plasticity asa bacterial survival strategy, Nat. Rev. Microbiol. 6 (2) (2008) 162e168.

[42] N.P. Truong, M.R. Whittaker, C.W. Mak, T.P. Davis, The importance of nano-particle shape in cancer drug delivery, Expert Opin. Drug Deliv. 12 (1) (2015)129e142.

[43] R.H. Fang, C.M. Hu, L. Zhang, Nanoparticles disguised as red blood cells toevade the immune system, Expert Opin. Biol. Ther. 12 (4) (2012) 385e389.

[44] J.C. Perales, T. Ferkol, H. Beegen, O.D. Ratnoff, R.W. Hanson, Gene transferin vivo: sustained expression and regulation of genes introduced into the liverby receptor-targeted uptake, Proc. Natl. Acad. Sci. 91 (9) (1994) 4086e4090.

[45] J.H. Kim, Y.K. Kim, M.T. Arash, S.H. Hong, J.H. Lee, B.N. Kang, Y.B. Bang, C.S. Cho,D.Y. Yu, H.L. Jiang, M.H. Cho, Galactosylation of chitosan-graft-spermine as agene carrier for hepatocyte targeting in vitro and in vivo, J. Nanosci. Nano-technol. 12 (7) (2012) 5178e5184.

[46] Y.Y. Seow, M.G. Tan, K.T. Woo, Expression of a functional asialoglycoproteinreceptor in human renal proximal tubular epithelial cells, Nephron 91 (3)(2002) 431e438.

[47] G. Chen, D. Li, J. Li, X. Cao, J. Wang, X. Shi, R. Guo, Targeted doxorubicin de-livery to hepatocarcinoma cells by lactobionic acid-modified laponite nano-disks, New J. Chem. 39 (4) (2015) 2847e2855.

[48] T. Paunesku, S. Mittal, M. Proti�c, J. Oryhon, S.V. Korolev, A. Joachimiak,G.E. Woloschak, Proliferating cell nuclear antigen (PCNA): ringmaster of thegenome, Int. J. Radiat. Biol. 77 (10) (2001) 1007e1021.

[49] C. Stroescu, A. Dragnea, B. Ivanov, C. Pechianu, V. Herlea, O. Sgarbura,A. Popescu, I. Popescu, Expression of p53, Bcl-2, VEGF, Ki67 and PCNA andprognostic significance in hepatocellular carcinoma, J. Gastrointestin Liver Dis.17 (4) (2008) 411e417.

[50] J. Kota, R.R. Chivukula, K.A. O'Donnell, E.A. Wentzel, C.L. Montgomery, H.-W. Hwang, T.-C. Chang, P. Vivekanandan, M. Torbenson, K.R. Clark, Thera-peutic microRNA delivery suppresses tumorigenesis in a murine liver cancermodel, Cell 137 (6) (2009) 1005e1017.

[51] G. Pistritto, D. Trisciuoglio, C. Ceci, A. Garufi, G. D'Orazi, Apoptosis as anti-cancer mechanism: function and dysfunction of its modulators and targetedtherapeutic strategies, Aging 8 (4) (2016) 603e619.

[52] J.C. Goldstein, C. Munoz-Pinedo, J.E. Ricci, S.R. Adams, A. Kelekar, M. Schuler,R.Y. Tsien, D.R. Green, Cytochrome c is released in a single step duringapoptosis, Cell Death Differ. 12 (5) (2005) 453e462.