Fabrication of A Folic Acid-Modified Arsenic Trioxide Prodrug Liposome and Assessment of its Anti-Hepatocellular Carcinoma Activity

ZHU Jing-Jing,CHEN Xio-Jie,YAO Wen-Dong,WEI Ying-Hui,ZHENG Hng-Sheng,ZHENG Hong-Yue,ZHU Zhi-Hong,WANG Bin-Hui,LI Fn-Zhu*

a.College of Pharmaceutical Sciences,Zhejiang Chinese Medical University,Hangzhou,Zhejiang 310053,China

b.Libraries of Zhejiang Chinese Medical University,Zhejiang Chinese Medical University,Hangzhou,Zhejiang 310053,China

c.The Affiliated Municipal Hospital of Taizhou University,Taizhou,Zhejiang 318000,China

ABSTRACT Objective To reduce the toxicity and side effects of arsenic trioxide (ATO) and provide a new approach for the treatment of primary liver cancer,a folic acid-modified calcium arsenite liposomal “target-controlled” drug delivery system (FA-LP-CaAs)was fabricated using the reverse microemulsion method.Methods A Malvern particle size anal yzer and a transmission electron microscope were employed to determine the particle size,distribution,zeta potential and morphology of FA-LP-CaAs.Further,inductively coupled plasma emission spect rometry was employed to determine the drug loading capacity,entrapment efficiency,and in vitro release behavior of FA-LP-CaAs.To determine its toxicity in human hepatoma cells (HepG2) and human normal hepatocytes (LO2) and its effect on HepG2 cell cycle and apoptosis,the MTT method was used.Laser confocal and flow cytometry were also employed to determine the uptake of FA-LP-CaAs by cells.After establishing a mouse liver cancer model,the in vivo distribution of the drug included in the formulation was investigated using in vivo fluoresc ence.To evaluate the liver cancer targeting and anti-tumor effects of FALP-CaAs in vivo,the distribution of ATO in tissues and changes in tumor volume and body weight after liposomal administration were investigated using hematoxylin-eosin (HE)-stained tumor sections.Results The particle size,zeta potential and PDI of FA-LP-CaAs were ( 122.67 ± 2.18) nm,(12.81 ± 0.75) mV and 0.22 ± 0.01,respectively,while its drug loading capacity was 18.49% ± 1.14%.In vitro experimental results revealed that FA-LP-CaAs had a strong killing effect on HepG2 cells.Further,the cell uptake capacity of this formulation was found to improve.Based on in vivo assessments,FA-LP-CaAs could significantly increase the distribution of ATO in tumor sites and inhibit tumor growth.Conclusions Herein,an FA-LP-CaAs formulation was successfully fabricated.This liposomal drug delivery system had a round appearance,uniform particle size,good polydispersity coefficient,evident “core-shell” structure,high drug loading capacity and pH response,tumor targeted drug delivery and sustained drug release.These findings support further research and the application of ATO as an anti-liver cancer prodrug and provide a new method for the treatment of liver cancer.

Keywords Target-controlled drug delivery system Liver cancer Arsenic trioxide Calcium arsenite Liposomes pH-sensitive

1 Introduction

Hepatocellular carcinoma (HCC) is the fourth most common cause of cancer-related deaths worldwide[1].China is known to have a high incidence of liver cancer.In fact,the number of liver cancer cases in China accounts for more than 50% of the global liver cancer cases[2].Furthermore,the mortality rate of liver cancer is only second to that of lung cancer in China.Currently,the diagnosis and treatment of liver cancer are substandard.The five-year survival rate of liver cancer in China is only 10.1%.Additionally,its prognosis is poor,the survival time of patients is short,and the mortality rate and morbidity rate are extremely high[3].Liver cancer mainly develops following liver cirrhosis caused by a chronic infection of hepatitis B virus and hepatitis C virus,alcoholic injury,and to a lesser extent,genetically determined diseases,such as hemochromatosis[4].Traditional surgical resection cannot completely eradicate the tumor.Further,owing to multidrug resistance and metastasis of liver cancer cells[5],it is difficult to achieve effective results with conventional chemotherapy and molecular targeted therapy for liver cancer[6].

Arsenic,a traditional mineral and Chinese medicine,is bitter,hot and poisonous.Arsenic enters the spleen,lung and liver meridians and can be used to treat many life-threatening diseases,including tumors.The application of arsenic has been recorded as early as inTaiping Shenghui Prescription(Tai Ping Sheng Hui Fang,《太平圣惠方》) andJade Catalpa Medicinal Solution(Yu Qiu Yao Jie,《玉楸藥解》).Based on theCompendium of Materia Medica(Ben Cao Gang Mu,《本草綱目》 ),“arsenic removes asthma,dysentery,rotten meat,blood stasis,scrofula,carbuncle,rotten meat,withered hemorrhoids and insecticides”.Arsenic trioxide (ATO) is the first-line treatment for acute promyelocytic leukemia(APL)[7-11].Studies have shown that ATO can promote cell differentiation,prevent cancer cell migration,and induce cell apoptosis[12],and has a significant inhibitory effect on different solid tumors,such as breast cancer,liver cancer and glioma,with broadspectrum anticancer properties[13,14].However,ATO has a narrow therapeutic window,short elimination half-life,lack of specificity duringin vivodistribution,and strong toxicity to normal tissues,thereby markedly limiting its application as a tumor therapy[15,16].The application of arsenic is derived from the rule of“attacking poison with poison” in traditional Chinese medicine (TCM).According to the medical book,Yellow Emperor’s Inner Classic(Huang Di Nei Jing,《黃帝內(nèi)經(jīng)》),“poison” should be used to cure diseases.This is because without toxicity,drugs cannot cure pain.In TCM,“poison” usually refers to different“pathogens” that damage the human body.These“pathogens” can thus be applied to treat diseases.Accordingly,the classification of pathogens and confirmation of their positive impact are carried out under certain conditions.Life-threatening and positive are not absolute;however,they can be converted to each other.Therefore,fabricating a multifunctional tumor-targeted drug delivery system that increases the amount of ATO entering the tumor,reduces its toxicity and side effects,and enables effective tumor targeting and controlled release of ATO is markedly warranted.

Professor HUANG Leaf ′s team at the University of North Carolina in the United States used the amphiphilic phospholipid,dioleoylphosphatidic acid(DOPA),in complex with calcium and polyvalent metals to synthesize an asymmetric phospholipid bilayer that enabled the nano system to have a stable cell-like structure and tumor targeting ability[17,18].Briefly,CaCl2with siRNA was dispersed in Cyclohexane/Igepal CO-520 solution to form a very well dispersed water-in-oil reverse micro-emulsion.The phosphate part was prepared by Na2HPO4(pH=9.0) in a separate oil phase.Then,DOPA in chloroform was added to the phosphate phase.After mixing the above two solutions for 20 min,ethanol was added to the microemulsion and the mixture was centrifuged at to remove cyclohexane and surfactant.After being extensively washed by ethanol several times,the pellets were dissolved chloroform and stored in a glass vial for further modification.Calcium phosphate (CaP)liposome-based nanoparticles prepared by this method exhibits pH responsiveness[19].In addition,its lipid layer can prevent the precipitation and accumulation of CaP,thereby increasing the stability and biocompatibility of the formulation[20].Similar to CaP,calcium arsenite [Ca3(AsO3)2,CaAs] is a slightly soluble arsenic salt with good degradability and pH sensitivity.CaAs can also dissolve in an acidic environment and form arsenite.Previous studies have utilized the solubility change of arsenite for the fabrication of nanoparticles[21].Although ATO has high loading potential,its biocompatibility is poor,its plasma protein binding rate is high,and the toxicity it induces is severe.Therefore,coating this drug with biocompatibility shells is critical[22].Lipid material possesses high biological safety.As a result,it has been widely used in recent years to synthesize drug delivery systems.Moreover,the use of these materials can effectively improve the biocompatibility of the nano drug delivery system.After the liposome surface is modified with ligands that possess targeting effects,such as folic acid (FA) etc.,human cancer cells overexpress the folate receptor (FR).However,FR has high affinity for folate or folate-drug conjugates (Kd=10?10M)[22-24],causing tumor-specific overexpression of FR.After binding with FA,a depression develops,which can form an inner capsule (endosome) in the cell.Under the action of a proton pump in the inner capsule membrane,the pH value in the capsule decreases from 7 to 4.5 - 5.0,which changes the conformation of the ligand-FR complex and releases the ligand into the cell.FR can then return to the surface of the cell membrane to transport more drugs[25,26],thereby markedly increasing drug enrichment in the target cell[27].

In the present study,the inverse microemulsion method (Figure1) was employed to formulate a liposomal drug delivery system.Briefly,ATO was combined with Ca2+in the form of arsenite;nano-scale CaAs was employed to achieve binding with the anionic phospholipid,DOPA.In addition,a cationic phospholipid was used to serve as the outer layer of the formulation.Other components used to fabricate the drug delivery system include 2-dioleoyl-3-trimethylammonium propane (DOTAP),1,2-disteoylsn-glycero-3-phosphoethanolamine (DSPE),PEGylated disteoyl phosphoethanolamine (DSPE-PEG),and FA-modified DSPE-PEG (DSPE-PEG-FA).The combination of the above components led to the fabrication of an FA-modified CaAs prodrug liposomal “target-controlled” drug delivery system.This drug delivery system could overcome the disadvantage of ATO,evade degradation in organisms,avoid toxicity induced by material accumulation,and provide reference for the design,development and evaluation of therapeutic preparations for HCC.

2 Materials and Methods

2.1 Materials

ATO (99.9% by mass),Suzhou Nord Parson Company;Arsenic standard solution (1 mg/mL),National Center for Analysis and Testing of Nonferrous Metals and Electronic Materials;Calcium Chloride,Tianjin Yongda Reagent Co.,Ltd.;(NH4)2HPO4,Shanghai Yuanye Biotechnology Co.,Ltd.;Cyclohexane,Igepal CO-520,Hoechst 33342,3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT),Sigma-Aldrich USA;DOPA and DOTAP,Avanti Polar Lipids,USA;DSPE-PEG (MW=2000),Lipoid,Germany;DSPE-PEG-FA,Shanghai Advanced Vehile Technology Pharmaceutical Technology Co.,Ltd.;DSPE-PEG-FITC,Shanghai Ponsure Biotechnology Co.,Ltd.;DMEM,RPMI 1640,fetal bovine serum and trypsin,Gibco,USA;Propidium iodide (PI) kit and Annexin V-FITC/PI apoptosis kit,BD Corporation,USA;and Methanol,Honeywell Burdick &Jackson,USA.Other reagents were of analytical grade.

Human hepatoma cells (HepG2) and human normal hepatocytes (LO2) were provided by the Animal Experimental Research Center of Zhejiang Chinese Medicinal University.BALB/c mice were purchased from the Animal Experimental Center of Zhejiang Chinese Medical University (Certificate No.SCXK Shanghai 2017-0005).All animal experiments were carried out in accordance with the animal breeding and use guidelines of Zhejiang Chinese Medical University.

2.2 Preparation of the CaAs liposomal-based nanoparticles

2.2.1 Preparation of the CaAs core using the inverse microemulsion methodBriefly,40 mL of cyclohexane and Igepal CO-520 (70/30,v/v) were placed in a round bottom flask and stirred for 5 min until even mixing was achieved in the oil phase.The mixture was then divided into two parts and stirred continuously.Subsequently,600 μL of 150 mmol/L CaCl2solution and 600 μL of 10 mg/mL sodium arsenite aqueous solution (pH=8.0) were dispersed dropwise into 20 mL of cyclohexane/Igepal CO-520.The two microemulsions were ultrasonicated for 5 min to control particle sizes.After mixing of the two oil phases,500 μL of DOPA (20 mmol/L) chloroform solution was added to the emulsion.Thereafter,the solution was mixed and stirred for 10 min.After 40 mL of absolute ethanol was added,cyclohexane/Igepal CO-520 was removed via centrifugation at 10 000×g for 15 min.After three rounds of washing with ethanol and drying,the CaAs core was obtained and further dissolved in 2 mL of chloroform for later use.

2.2.2 Preparation of LP-CaAs and FA-LP-CaAsThe solution prepared in 2.2.1 (CaAs core),500 L DOTAP (20 mmol/L),250 μL cholesterol (20 mmol/L),20 μL DSPE-PEG (3 mmol/L) and 5 μL DSPE-PEG-FA (3 mmol/L) were combined to prepare the FA-modified CaAs liposome using the thin film dispersion method.Briefly,the material mentioned above and CaAs core were dissolved in chloroform,rotated and evaporated on a film in an eggplant bottle.Thereafter,the organic solvent was removed and the solution was transferred to a vacuum dryer for overnight drying.A volume of 10 mL normal saline was then added and the solution was placed in an air bath shaker at 180 rpm for 20 min.An ultrasonic water bath was carried out for 5 min using an ultrasonic probe to remove the membrane.The obtained product was then centrifuged (10 000×g,5 min),reconstituted with normal saline,and recentrifuged.The above process was repeated five times to remove the phospholipid fragments that did not form a film.Thereafter,FA-LPCaAs was obtained and freeze-dried for later use.To synthesize LP-CaAs,25 μL DSPE-PEG (3 mmol/L)was used instead of 20 μL DSPE-PEG (3 mmol/L) and 5 μL DSPE-PEG-FA (3 mmol/L).

2.3 Characterization of LP-CaAs and FA-LP-CaAs

The sizes of the hydrate particles and the zeta potential of CaAs,LP-CaAs and FA-LP-CaAs were measured by dynamic light scattering (DLS;Malvern Nano-ZS90).The morphological characteristics of CaAs,LP-CaAs and FA-LP-CaAs were obtained using transmission electron microscopy (TEM;H-7 650,Hitachi,Tokyo,Japan).

2.4 Determination of the encapsulation efficiency and drug loading capacity of the formulations

The prepared lipid capsule nanoparticles (LP-CaAs and FA-LP-CaAs) were digested with 50 μL of concentrated nitric acid and 50 μL of perchloric acid,respectively.Thereafter,the nanoparticles were diluted to 10 mL with dilute nitric acid for later use.After 1.5 mL of the solution was centrifuged by ultrafiltration (10 000×g,molecular weight cut-off 10 000),100 μL of the filtrate was digested with 50 μL of concentrated nitric acid and 50 μL of perchloric acid,respectively.Thereafter,a further dilution to 10 mL with dilute nitric acid was performed.After the two solutions were filtered through a 0.22 μm microporous membrane,the content of ATO in the obtained solutions was determined using an Inductively coupled plasma emission spectrum (ICP 6300,Thermo Electron Corporation).The encapsulation efficiency of the formulations was also calculated.

The freeze-dried powder of the CaAs core and CaAs lipid capsule nanoparticle (LP-CaAs and FALP-CaAs) was precisely weighed and dissolved in 10 mL water to formulate the mother liquor;these solutions were placed on standby.After the above solution was filtered through a 0.22 μm microporous membrane,the ATO content in the solution was determined by ICP.Subsequently,the drug loading potential of the formulations was calculated.

The EE (%) and DL (%) of LP-CaAs and FA-LPCaAs were calculated using equations (1) and (2).

2.5 In vitro release assay

Release media with pH values of 7.4,6.8 and 5.5 were selected to investigate the release characteristics of arsenic trioxide solution (ATO-Sol) and the nanopreparations.Briefly,2 mL of ATO-Sol,LP-CaAs and FA-LP-CaAs solution (containing 1.0 mg of As) was precisely measured.In a pre-treated dialysis bag,the samples were collected at 0.02,0.08,0.17,0.25,0.5,1,2,4,6,8,12 and 24 h,respectively;the samples were added at the same temperature,volume and pH.Thereafter,the samples were filtered through a 0.22 μm microporous membrane.Following dilution of the filtrate,the ATO content in the release medium was determined by ICP.The cumulative release rate Qn(%) was calculated using equation (3) and a release curve was generated.

Where Mn,Qn,Cn,V0,Viand Cirepresent the cumulative release at each time point,the cumulative release percentage at each time point,the drug concentration at tn,the volume of the release medium,the volume of the sample removed each time,the drug concentration at ti,and the total drug concentration,respectively.

2.6 In vitro toxicity assessment

The toxicity of the preparation was investigated using the MTT method.HepG2 and LO2 cells in the logarithmic growth phase (8×103cells per well) were seeded in 96-well plate.After incubation at 37 °C for 12 h,the culture medium was removed,and the concentrations of ATO-Sol,LP-CaAs and FA-LP-CaAs were added.After incubation at 37 °C for 48 h,the drug-containing culture medium was removed.After an additional incubation at 37 °C for 4 h,the culture medium was removed.After three rounds of washing with PBS,0.15 mL DMSO was added to each well and the plate was shaken for 15 min.The absorbance value (490 nm) was measured using an enzyme labeling instrument (SpectraMax M2,Molecular Devices,USA).Further,the half maximal inhibitory concentration (IC50) was calculated to determine the drug concentration required to achieve 50%inhibition.All experiments were repeated three times.

2.7 Cellular uptake

2.7.1 Flow cytometryHepG2 and LO2 cells in the logarithmic growth phase were inoculated in 6-well plates at 2×104cells per well,cultured for 24 h,incubated with serum-free medium containing LPCaAs-FITC and FA-LP-CaAs-FITC (containing 5 μmol/L FITC) at 37 °C,washed,and finally digested with PBS (pH 7.4,stored at 4 °C) after 4 h.The average fluorescence intensity was measured and quantified by flow cytometry (Guava Easycyte,Merck Millipore,Germany).All experiments were repeated three times.

2.7.2 Laser confocal experimentThe localization of fluorescent-labeled LP-CaAs-FITC and FA-LP-CaAs-FITC in HepG2 cells was investigated using a laser confocal microscope (Olympus,Tokyo,Japan).Cells were inoculated in a confocal four-compartment dish with 1×105cells per well,cultured for 12 h,and then incubated with serum-free medium containing LPCaAs-FITC and FA-LP-CaAs-FITC (containing 5 μmol/L FITC) at 37 °C for 4 h.The culture solution was discarded and rinsed three times with PBS.Each well was fixed with 0.5 mL 4% (v/v) paraformaldehyde for 10 min.The culture solution was discarded and rinsed three times with PBS.After incubation with 0.1% TritonX-100 for 10 min,the cell membrane was permeated.After the cells were treated with 0.5 mL Hoechst 33343 (12 μg/mL),they were incubated for 10 min.After the medium was then discarded,the cells were rinsed three times with PBS.Phalloidin was diluted 1 000-fold,dyed for 30 min,removed via suction,and rinsed three times with PBS.Phalloidin was then stored away from light and observed under a laser confocal microscope.

2.7.3 Determination of the intracellular ATO concentrationCells in the logarithmic growth period were inoculated in 6-well plates at 5×105/well.After cells adherence,ATO-Sol,LP-CaAs and FA-LP-CaAs were added and allowed to incubate for 6,12 and 24 h.The same volume of the culture solution was employed as a blank control.At the end of the incubation,the cell samples were washed and digested with cold PBS(pH 7.4,stored at 4 °C).After further treatment,the intracellular drug concentration was determined by ICP.

2.8 Cell cycle evaluation

To further examine the mechanism whereby ATO inhibits HepG2 cells and cell growth,the effects of ATO-Sol and the nano drug delivery systems on the cell cycle were determined by flow cytometry.HepG2 cells in the logarithmic growth phase (5×105cells per well) were seeded in each well of 6-well plate.The cells were incubated with serum-free medium containing ATO-Sol,LP-CaAs and FA-LP-CaAs(1 μg/L ATO) at 37 °C for 48 h,digested with trypsin,centrifuged at 800×g for 5 min,immobilized with 70% ethanol solution at 4 °C for 8 h,gently washed with PBS,and then immobilized with DNase-free RNase A at 37 °C for 30 min.Finally,0.1 mL Triton X-100 containing 2 g PI was added and detection was performed using flow cytometry.The experiment was repeated three times.

2.9 Evaluation of apoptosis

To further determine the toxicity of ATO to HepG2 cells,the effects of ATO-Sol and the nano drug delivery systems on cell apoptosis were determined using flow cytometry.Briefly,HepG2 cells in the logarithmic growth phase were harvested,digested with pancreatin,added to cell culture solution,blown,and diluted to 1.5×105cells/mL.The cell suspension (2 mL) was then placed in a 6-well plate,incubated at 37 °C in a 5% CO2incubator for 24 h,treated with ATO-Sol and the nano drug delivery systems (containing 10 μg/mL ATO),and incubated for an additional 48 h.After the culture solution was removed,the cells were washed and then stained according to the manufacturer’s instructions for the Annexin V-FITC/PI apoptosis kit.Apoptosis was then detected by flow cytometry.

2.10 Establishment of HepG2 BALB/c mouse model

To establish the solid tumor-bearing mouse model,HepG2 cells were subcutaneously injected into the lateral neck and back of the forelimbs of BALB/c mice at a rate of 1×107per mouse.Within the three days post-modeling,mice were examined to determine whether a mass bulge developed at the injection site.Mice without an evident mass were not used in the study.

2.11 Fluorescence imaging and tissue distribution experiment in vivo

2.11.1 Fluorescence imaging experiment in vivoTen BALB/c mice bearing HepG2 were randomly divided into two groups:LP-CaAs group and FA-LP-CaAs group.Each tumor-bearing mice was administered 0.2 mL of DiR (5 μg/mg in the preparation) via the tail vein.At 1,2,4,8,12 and 24 h after injection,the nanoparticles were placed in anin vivoimaging apparatus to examine their distribution in different tissues.The excitation wavelength was 748 nm while the emission wavelength was 780 nm.Further,an exposure time of 1 s was employed.

2.11.2 Tissue distribution experimentWhen the tumor size reached 50 - 100 mm3,20 mice were randomly divided into the following four groups:saline,ATO-Sol,LP-CaAs and FA-LP-CaAs groups,with five mice per group.Mice were administered the drug(1 mg/kg) via the tail vein and killed 24 h after cervical dislocation.The heart,liver,spleen,lung,kidney,brain,skin,bone,blood and tumor obtained from mice were weighed and placed in test tubes containing a plug.Concentrated nitric acid and hydrogen peroxide were added to digest the tissues(concentrated nitric acid:hydrogen peroxide=4:1).After standing at room temperature for 24 h,the tissues were fixed with 5% dilute nitric acid in a 10 mL volumetric flask.Before testing,the sample was passed through a 0.22 μm microporous filter membrane.

2.12 Anti-tumor evaluation in vivo

Twenty BALB/c mice were randomly divided into four groups:saline,ATO-Sol,LP-CaAs and FA-LPCaAs groups,with five mice per group.Normal saline,ATO-Sol,LP-CaAs and FA-LP-CaAs (1 mg/kg ATO) were injected into the tail vein of mice on a daily basis.Before mice were administered treatment,their body weight was recorded,and the tumor volume was measured with a Vernier caliper.The tumor volume and tumor growth inhibition rate(IR,%) were calculated using equations (4) and (5).

Where a,b,Wtand Wcrefer to the maximum diameter of the tumor,the minimum diameter of the tumor,the average tumor weight of the treatment group,and the average tumor weight of the control group,respectively.

2.13 Hematoxylin-eosin (HE) staining of the tumor tissue

On day 15,tumor-bearing mice were killed by cervical dislocation.Thereafter,their tumor tissue was removed,weighed,fixed in formalin,embedded in paraffin wax,and sectioned.The sections were stained with HE for tumor cell apoptosis evaluation.

2.14 Statistical analysis

Statistical analysis was carried out using the SPSS 20.0 analysis software.Variance analysis was used for comparisons between multiple groups while independent samplettest was used for comparisons between two groups.Data are expressed as.P<0.05 indicates a significant difference whileP<0.01 indicates extreme significant difference.

3 Results

3.1 Characterization of LP-CaAs and FA-LP-CaAs

As shown in Figure2A and 2B,a uniform particle size was observed for CaAs and FA-LP-CaAs using TEM.Further,an evident “core-shell” structure and an outer lipid layer thickness of approximately 8 nm could be seen.The hydrated particle sizes of CaAs,LP-CaAs and FA-LP-CaAs were (120.20 ± 1.41) nm,(124.83 ± 0.40) nm and (122.67 ± 2.18) nm,respectively,and their dispersion coefficients were approximately 0.21 ? 0.22.The formulations were also identified to have good stability (Figure2C).When LP-CaAs and FA-LP-CaAs were prepared by the “film hydration method”,the zeta potential was reversed from ? (21.93 ± 1.02) mV to (15.32 ± 0.86) mV and(12.81 ± 0.75) mV,respectively (Figure2D).The encapsulation efficiency of LP-CaAs and FA-LP-CaAs was 68.73% ± 4.72% and 68.07% ± 3.88%,respectively,and the drug loading capacity of the CaAs core,LP-CaAs and FA-LP-CaAs was 23.61% ± 3.22%,18.52% ± 1.62%and 18.49% ± 1.14%,respectively.Therefore,the formulations demonstrated good encapsulation efficiency and drug loading capacity.

Thein vitrorelease curves of ATO-Sol,LP-CaAs and FA-LP-CaAs in different pH environments are presented in Figure3.Almost 95% of ATO-Sol was released in the different pH media within 2 ? 4 h.At a pH of 7.4,46.52% ± 1.28% and 42.76% ± 1.06% of LPCaAs and FA-LP-CaAs were respectively released within 24 h,ultimately demonstrating the good stability of these formulations.At a pH of 6.8,73.06% ± 0.76%and 71.05% ± 1.20% of LP-CaAs and FA-LP-CaAs were respectively released within 24 hours.Further,at a pH of 5.5,the release rate of the two formulations was relatively fast,with a cumulative release rate of 91.46% ± 0.66% and 89.25% ± 1.84%,respectively,within 24 h.

3.2 Cytological evaluation of in vitro toxicity

The toxicity results of ATO-Sol,LP-CaAs and FA-LPCaAs in HepG2 and LO2 cells are displayed in Figure4A and 4B.The survival rates of both cells decreased with an increase in drug concentration.The IC50 values of ATO-Sol in HepG2 and LO2 cells were(14.83 ± 0.26) μg/mL and (34.04 ± 1.97) μg/mL,respectively;these values were 1.84- and 1.25-fold greater than those obtained with LP-CaAs (P<0.05).The IC50 values of ATO-Sol in HepG2 and LO2 cells were (14.83 ± 0.26) μg/mL and (34.04 ± 1.97) μg/mL,respectively,while those of FA-LP-CaAs were (6.72 ±0.33) μg/mL and (25.07 ± 1.17) μg/mL,respectively.The IC50 values obtained with FA-LP-CaAs were the lowest of the three formulations and were significantly different from those of LP-CaAs in HepG2 cells(P<0.05) (Figure4C).

Flow cytometry was used to determine the uptake of the FITC-labeled vectors into cells.As shown in Figure4D and 4E,the uptake of LP-CaAs-FITC and FA-LP-CaAs-FITC by HepG2 cells was 90.16% ±3.51% and 98.62% ± 0.49%,respectively,while that in LO2 cells was 22.99% ± 2.04% and 36.87% ± 1.66%,respectively.The uptake of LP-CaAs-FITC and FA-LPCaAs-FITC by HepG2 cells was 3.92- and 2.67-fold greater than that by LO2 cells (P<0.05).Further,the fluorescence intensities of LP-CaAs-FITC and FA-LPCaAs-FITC in HepG2 cells were 4×104and 6×104,respectively,with a difference of 1.5-fold.

Laser confocal microscopy was employed to determine the uptake and intracellular localization of the FITC-labeled vector by HepG2 cells.The carrier labeled with FITC exhibited green fluorescence,the cytoskeleton stained with phalloidin displayed red fluorescence,and the nucleus stained with Hoechst exhibited blue fluorescence.As shown in Figure4G,the nano drug delivery system was mainly distributed in the cytoplasm,and the fluorescence intensity in the nucleus was relatively low.FA-LP-CaAs-FITC was also found to exhibit a stronger fluorescence in cells than LP-CaAs-FITC,indicating that HepG2 exhibited a stronger uptake of FA-LP-CaAs-FITC.Such findings are consistent with those obtained using flow cytometry.The results of intracellular ATO concentration after different periods of incubation are shown in Figure4F.At 6 h after administration,the intracellular ATO concentration in HepG2 cells treated with LP-CaAs was significantly higher than that of cells administered ATO-Sol (P<0.01).In addition,the intracellular ATO concentration in HepG2 cells treated with FA-LP-CaAs was significantly higher than that in cells administered LP-CaAs and ATOSol (P<0.01).

3.3 Evaluation of cell cycle and cell apoptosis

The cell apoptosis results were shown in Figure5A,the proportion of cells in the G2-M phase and S phase following treatment with ATO-Sol,LP-CaAs and FA-LP-CaAs gradually increased while the proportion of cells in the G0-G1 phase (total proportion minus G2-M phase and S phase) gradually decreased (P<0.05).FA-LP-CaAs was also identified to have the smallest proportion of cells in the G0-G1 phase (P<0.05).In fact,the proportion of cells in the G0-G1 phase (total proportion minus the G2-M phase and S phase) was 56.36% ± 1.03%,54.65% ±0.33%,46.535 ± 0.87% and 44.91% ± 0.99% when treated with blank medium,ATO,LP-CaAs,and FALP-CaAs,respectively.

The cell cycle results were shown in Figure5B,the number of apoptotic cells in the ATO group was 31.53% ± 0.90%;this number was significantly lower than that of the LP-CaAs and FA-LP-CaAs groups(45.20% ± 0.34%,51.17% ± 1.41%).Evidently,ATO leads to late apoptosis.

3.4 In vivo fluorescence image and tissue distribution experiment

To evaluate the real-time biological distribution of FA-LP-CaAs,imaging systems were employed forin vivoimaging at different time intervals (Figure6).As shown in Figure6B,LP-CaAs did not display an evident fluorescence aggregation phenomenon at 1 ?8 h;however,fluorescence enhancement was evident at the tumor site at 12 h.FA-LP-CaAs showed no evident fluorescence at 1 h;however,weak fluorescence was observed at 2 h and evident fluorescence could be seen at 4 - 12 h.The tissue distribution results are shown in Figure6A.The group administered FA-LP-CaAs had the highest distribution at the tumor site.In fact,the distribution observed in this group was significantly different from that found in the ATO and LP-CaAs groups (P<0.05).After ATO was prepared into nanoparticles,with CaAs as a prodrug,and encapsulated in liposomes,the distribution of drugs in tumor tissues was significantly increased and the ATO content in the skin was significantly decreased.

3.5 Anti-tumor evaluation in vivo

Thein vivoanti-tumor results are presented in Figure6C.The weight gain observed in the ATO-Sol group was the same as that of other groups at the initial stage (1 - 3 d);however,weight gain began to slow down on the fifth day and was significantly lower than that of the other groups on day 15.As the body weight found in the groups administered LP-CaAs and FA-LP-CaAs was similar to that found in the normal saline group.

Based on the ratio of tumor volume in the treatment group relative to the normal saline group,thein vivoanti-tumor efficacy of each group was evaluated.As shown in Figure6D,on day 15 of treatment,tumor volumes in the normal saline,ATO-Sol,LP-CaAs and FA-LP-CaAs were (1 421.96 ± 392.93) mm3,(941.56 ± 196.01) mm3,(753.97 ± 149.45) mm3and(497.62 ± 234.85) mm3,respectively.The average tumor volume in each treatment group was significantly smaller than that found in the normal saline group (P<0.05).According to the average tumor weight (Figure6E),the tumor inhibition rates of each group were 39.70%,57.31% and 81.17%,respectively.Herein,a negative correlation was thus found between the degree of modification of the nano-preparation and the tumor volume and weight.In fact,as the modification of the nano-preparation improved,the tumor volume and weight were continuously reduced,thereby demonstrating a positive correlation with the rate of inhibition.The tumor volume and inhibition rate of the FA-LP-CaAs group were found to significantly differ from those of the other groups (P<0.05).

3.6 HE staining of the tumor tissue

After 15 days of treatment,paraffin sections were prepared using the tumor tissues collected from each group.The histopathological results obtained after HE staining are shown in Figure6F.The normal saline (negative control) group had dense tumor cells,complete nucleus,small reddish area of cytoplasm,and basically no apoptosis or necrosis of tumor cells.The reddish cytoplasm area in the ATOSol group increased while the density of the nuclei significantly decreased relative to that observed in the normal saline group.Such is indicative of apoptosis and necrosis of some tumor cells.Compared to the ATO-Sol group,the LP-CaAs and FA-LP-CaAs groups possessed more apoptotic and necrotic areas.Further,these groups had a significant decrease in blue-stained nucleus,which indicates that the nucleus had undergone maximum apoptosis while the tumor tissue had undergone severe necrosis and apoptosis.

4 Discussion

4.1 Characterization of LP-CaAs and FA-LP-CaAs

The zeta potential of the formulations was determined with a Malvern particle size analyzer.The change in zeta potential of the formulations could explain the successful encapsulation of cationic phospholipids in the inner core of CaAs.The existence of a lipid layer can prevent direct contact between the drug and the binding protein or protease,which would cause the drug to lose its activity[28].Accordingly,the precipitation and aggregation of the CaAs inner core can be prevented and the stability and biocompatibility of the inner core can be improved.FA-LP-CaAs loading was recognized to be significantly higher than that of some of the ATOloaded nanoparticles,such as the lipid capsule nanoparticles (DL%,approximately 6.7%)[29]and polymer micelles (DL%,approximately 2.4%)[30].The further increase in drug loading may be due to the dual role of the CaAs core (i.e.,a carrier and a drug).Based on thein vitrorelease test,the preparation was found to have good stability.This finding may be due to the potential of calcium ions in the CaAs core to form chemical bonds with the phosphate groups at the end of the phospholipid membrane[31].This formation would promote the subsequent formation of nanoparticles wrapped in a phospholipid monolayer.The addition of the cationic phospholipid,DOTAP and ligand-modified phospholipid enabled the formation of a positively-charged asymmetric phospholipid bimolecular layer,which allowed the nano drug delivery system to possess a stable cell-like structure,reduced cytotoxicity,and enhanced stability[32].The preparation was also demonstrated to exhibit pH sensitivity,be slowly released in a physiological environment (pH=7.4),and possess good stability.However,in a weakly acidic tumor environment (pH=5.5),the preparation could rapidly release the drugs,thereby enabling the treatment of cancer.

4.2 Cytological evaluation in vitro

Based on the results of the cell survival experiments,LP-CaAs and FA-LP-CaAs exhibit stronger antitumor effects than ATO.Such finding may be due to the increased biocompatibility of the preparation and reduced clearance of the reticuloendothelial system,which are both caused by liposome coating[33].Through laser confocal experiments and flow cytometry experiments,the uptake of LP-CaP-FITC and FA-LP-CaP-FITC in HepG2 cells was found to be greater than that in LO2 cells.Such result may be due to the highly selective affinity of FA to the FR overexpressed on the cancer cell surface[34-38].Moreover,compared to tumor tissues,the expression level of FR in normal tissues is markedly reduced,and the drug delivery system modified with FA can be more enriched in HepG2 cells via endocytosis[39-41].In the laser confocal image,several green fluorescent spots were observed in the cell;this may be caused by the internalization of the nanoparticles into the cells and a further concentration of these nanoparticles in organelles,such as lysosomes.The results of cell cycle and cell apoptosis also revealed that FA modification and liposome encapsulation can improve the killing effect of CaAs on target cells.The early stage of apoptosis may be due to the conversion of the ATO-generated CaAs prodrug into + 3-valent arsenite ions and calcium ions in cancer cells following their internalization into cells as nanoparticles[42].When calcium ions and related calcium ion nanoparticles (such as CaP) are broken down into calcium ions,the mitochondria in normal cells and related proteins on the cell membrane will regulate intracellular calcium concentration and will thus result in its excretion[43].However,due to the existence of + 3-valent arsenite ions,some of the related proteins are affected.The simultaneous existence of + 3-valent arsenite ions and calcium ions leads to swelling and dissolution of the mitochondria,which accelerates cell apoptosis[44].Such findings suggest that FA-LP-CaP-FITC exerts a good antitumor effectin vitro.

4.3 In vivo evaluation

To further verify the effectiveness of the preparation,animal experiments were carried out.Thein vivofluorescence imaging analysis revealed that FA-LPCaAs exhibit an evident targeting effect and can improve the aggregation of CaAs at tumor sites.The tissue distribution experiment further verified the significant increase in the distribution of FA-LP-CaAs in tumor tissues.After ATO is synthesized into nanostructures,PEG acts as the wrapped lipid layer.When the hydration capacity is increased,a cluster structure can be formed outside the lipid layer,which can reduce its protein adsorption in physiological or tissue fluid due to steric hindrance.As a result,the formation of protein crowns can be avoided to a certain extent,which reduces the cleaning effect of blood proteins,the uptake of macrophages into the reticuloendothelial system,and the renal clearance rate[45,46].By detecting tumor volume and animal body weight,and generating HE-stained images of the tumor tissue sections viain vivoanti-tumor experiments,FA-LP-CaAs was found to display a stronger killing effect on tumor cells than ATO-Sol and LP-CaAs.Such finding further verifies that liposome encapsulation improves the biocompatibility of the preparation while FA modification causes the preparation to exhibit targeted effect and improves the anti-tumor effect of the formulationin vivo.

5 Conclusions

In the present study,a single-layer phospholipidcoated CaAs core was prepared using the "reverse microemulsion method".During the preparation process,ATO not only served as a pharmacodynamic substance,but also a delivery carrier.CaAs was found to markedly increase drug loading.Furthermore,it is pH-sensitive and can be released in a slightly acidic tumor microenvironment.FA-LP-CaAs was constructed by combining DOTAP,PEG-DSPE,DSPE-PEG-FA and the CaAs core to form an outer phospholipid layer via the thin film dispersion method.As a result,a uniform particle size of approximately 120 - 130 nm was achieved.The nano drug delivery system with a“core-shell” structure,which was generated by wrapping the phospholipid layer on the outer layer,was found to improve the biocompatibility of CaAs and contribute to its sustained release characteristics.The phospholipid bilayer was also found to enhance the affinity between the cell membrane and lipid capsule nanoparticles,thereby increasing the cell uptake of the lipid capsule nanoparticles.The modification of FA occurred outside the lipid layer,which enabled FA-LP-CaAs to exhibit the dual effects of targeting and sustained release.The drug delivery system developed herein can increase the amount of FA-LP-CaAs transported into cells through the specific binding of the ligand receptor,thereby producing strong cytotoxicity.Tissue distribution andin vivofluorescence imaging results revealed that FALP-CaAs exhibited a stronger targeting efficiency than ATO and LP-CaAs.Moreover,histopathological analysis revealed that FA-LP-CaAs displayed a more evident effect on the promotion of necrosis and apoptosis of liver cancer cells relative to normal saline,ATO and LP-CaAs as well as a better anti-liver cancer effectin vivo.

In summary,based on the characteristics of ATO,we constructed a new drug delivery system using CaAs nanoparticles,ATO,and Ca2+,and further encapsulated these components in functional liposomes.As a result,the drug loading and stability of ATO were improved and its pH sensitivity was enhanced.The lipid layer was found to prevent direct contact between ATO and normal tissues.Furthermore,the concentration of the drug in solid tumors was improved,thereby enhancing the anti-liver cancer effect of ATO.The fabrication of FA-LP-CaAs promotes the performance of further research and the application of ATO as an anti-liver cancer drug and provides a new approach for the treatment of HCC.

Acknowledgements

We thank for the funding support from the National Natural Science Foundation of China (No.81873014).

Competing Interests

The authors declare no conflict of interest.