Ha

LQa

b

a

ARRAA

KCPB

1

nisdihPcvd[d

h0

International Journal of Biological Macromolecules 72 (2015) 1391–1401

Contents lists available at ScienceDirect

International Journal of Biological Macromolecules

j ourna l h o mepa ge: www.elsev ier .com/ locate / i jb iomac

yaluronic acid/chitosan nanoparticles for delivery of curcuminoidnd its in vitro evaluation in glioma cells

iu Yanga,1, Shiya Gaoa,1, Sajid Asghara,b,1, Guihua Liua, Jue Songa, Xuan Wanga,ineng Pinga, Can Zhanga, Yanyu Xiaoa,∗

Department of Pharmaceutics, State Key Laboratory of Natural Medicines, China Pharmaceutical University, Nanjing 210009, PR ChinaCollege of Pharmacy, Government College University Faisalabad, Faisalabad, Pakistan

r t i c l e i n f o

rticle history:eceived 9 August 2014eceived in revised form 12 October 2014ccepted 19 October 2014vailable online 27 October 2014

eywords:urcuminoidolyelectrolyte complex nanoparticlesrain gliomas

a b s t r a c t

The aim of this work was to evaluate the potential of polyelectrolyte complex nanoparticles (PENPs)based on hyaluronic acid/chitosan (HA/CS) as carriers for water-insoluble curcuminoid (CUR) and explorein vitro performance against brain glioma cells. PENPs were observed to be affected by the order of addi-tion, mass ratios and initial concentrations of the HA/CS, pH and ionic strength. PENPs remained stableover a temperature range of 5–-55 (C. CUR was successfully encapsulated into the PENPs. CUR-PENPsshowed spherical shape with a mean diameter of 207 nm and positive charge of 25.37 mV. High encap-sulation efficiency (89.9%) and drug loading (6.5%) was achieved. Drug release studies revealed initialburst release of drug from the PENPs up to 4 h followed by sustained release pattern. DSC thermogramsand XRD patterns showed that CUR was encapsulated inside the PENPs in a molecular or amorphous

state. Compared with CUR-solution, CUR-PENPs showed stronger dose dependent cytotoxicity againstC6 glioma cells and higher performance in uptake efficiency in C6 cells. Cellular uptake of CUR-PENPs wasfound to be governed by multi-mechanism in C6 cells, involving active endocytosis, macropinocytosis,clathrin-, caveolae-, and CD44-mediated endocytosis. In conclusion, CUR-PENPs might be a promisingcarrier for therapy of brain gliomas.

© 2014 Elsevier B.V. All rights reserved.

. Introduction

Polyelectrolyte complex nanoparticles (PENPs) are sponta-eously formed when two oppositely charged polyelectrolytes

nteract via the electrostatic attraction upon mixing in an aqueousolution. The simple preparation process, possibility of controlledrug release, protection of the encapsulated payload against chem-

cal degradation, and avoidance of organic solvents, surfactants,eat and vigorous agitation are some of the main advantages ofENPs. However, these PENPs have been mainly investigated asarriers for hydrophilic macromolecules such as peptides, proteins,accines, and DNAs [1–4] and some water-soluble drugs such as

oxorubicin hydrochloride, diclofenac sodium, and salicylic acid5,6]. There have been few studies on loading of water-insolublerugs into PENPs, but with poor encapsulation capacity [7].

∗ Corresponding author. Tel./fax: +86 25 83271079.E-mail address: cpuyanyuxiao@163.com (Y. Xiao).

1 Liu Yang, Shiya Gao and Sajid Asghar contributed equally to this work.

ttp://dx.doi.org/10.1016/j.ijbiomac.2014.10.039141-8130/© 2014 Elsevier B.V. All rights reserved.

Hyaluronic acid (HA), a linear polysaccharide composed ofrepeating disaccharide units of D-glucuronic acid and N-acetyl glu-cosamine linked by � (1,4) and � (1,3) glycosidic bonds, is a highlyefficient targeting molecule for cancer therapy that can bind to theHA receptors (trans-membrane glycoprotein CD44), over expressedin many types of cancer cells [8]. Therefore, HA modified deliv-ery systems can increase drug accumulation specifically in CD44over-expressing cancer cells [9–11].

Chitosan (CS) is a polysaccharide consisting of copolymers ofN-acetyl-d-glucosamine and D-glucosamine units linked by �-(1,4)-glycosidic linkages. Because of the existence of amine groups,CS is positively charged at neutral or acidic pH and is able toform intermolecular complexes with a wide variety of polyanionsincluding poly (galacturonic acid) [12], alginate [13], dextran sul-fate [14], poly(l-malic acid-co-d,l-lactic acid) [15], pectin [16] andHA [1–3].

Curcuminoid (CUR), a naturally active constituent extractedfrom the plants of the Curcuma longa, is valued worldwide as apotential food additive, coloring agent and spice. Moreover, it hasshown a wide range of biological and pharmacological actions [17].

1392 L. Yang et al. / International Journal of Biological Macromolecules 72 (2015) 1391–1401

F lectros

CIUiub[

oPtpaIplc(

dcarcrtc

ig. 1. Hypothetical scheme of PENPs-mediated drug delivery. The formation of eubsequent CD44-mediated drug delivery by PENPs, is shown.

urrently, CUR has been considered by the U.S. National Cancernstitute as a third generation cancer chemopreventive agent [18].nfortunately, clinical advancement of CUR has been hindered by

ts poor solubility in water, instability at neutral-basic pH val-es, oxidation and photodegradation [19]. At present, efforts areeing made to utilize full potential of CUR as an anti-cancer agent20,21].

The main purpose of this research was to study the formationf PENPs based on HA/CS and to evaluate the potential of HA/CSENPs as a carrier for water-insoluble CUR. PENPs were charac-erized for their key features such as size and morphology, zetaotential, drug loading and encapsulation efficiency. Physical statend thermal properties were evaluated by XRD and DSC studies.n vitro drug release studies were conducted to study the releaseattern of CUR from PENPs. Furthermore, in vitro cyto-toxicity, cel-

ular uptake and mechanism of uptake were explored in C6 gliomaells due to high expressions of CD44 in malignant glioma cellsFig. 1) [22].

CUR consists primarily of three bis-�, �-unsaturated-�-iketone hydrophobic polyphenols: curcumin, demethoxycur-umin (DMC), and bisdemethoxycurcumin (BDMC) (Fig. 2). BDMC,mong the components of crude commercial CUR, has beeneported to be the most potent agent for inhibition of cancer

ell invasion [23]. In current study, BDMC was treated as theepresentative component of the CUR in evaluation of the drug con-ents in PENPs and the accumulation of CUR-PENPs in C6 gliomaells.

static complexes between negatively-charged HA and positively-charged CS, and

Fig. 2. The chemical structure of curcuminoid.

logica

2

2

Cf(CdDf

2

sa1wortt(a(ftooCtt

imismsA

2

t(0sNbw(oacap

2

4sZ

L. Yang et al. / International Journal of Bio

. Experimental

.1. Materials

CUR was purchased from Aladdin Reagent Co., Ltd. (Shanghai).S (MW = 50 kDa, degree of deacetylation 90%) was purchased

rom Golden-shell Biochemical Co. Ltd. (Zhejiang, China). HAMW = 120 kDa) was purchased from C.P. Freda Pharmaceuticalo. Ltd. (Shandong, China). 3-(4, 5-dimethylthiazol-2-yl)-2, 5-iphenyltetrazolium bromide (MTT), fetal bovine serum (FBS),MEM/HG medium, penicillin and streptomycin were purchased

rom Thermo Fisher Scientific (USA).

.2. Preparation of blank PENPs

Blank PENPs were prepared by physically mixing HA and CSolutions. In brief, 100 mg of CS was dissolved in 100 mL of 2% (w/v)cetic acid aqueous solution, and 100 mg of HA was dissolved in00 mL of deionized water, and then the CS and HA solutions eachere filtered through a 0.22 �m microporous membrane. The pH

f the HA and CS solutions after the dissolution were 6.50 and 2.75,espectively. The HA solution was added in a dropwise manner tohe CS solution under magnetic stirring (200 rpm) at room tempera-ure at a flow rate of 60 ml/min, thus forming blank HA in CS PENPsHiC PENPs). Additionally, the order in which the solutions weredded was also investigated in order to form blank CS in HA PENPsCiH PENPs). The resultant opalescent preparation was then stirredor 1 h to form uniform nanoparticles. To examine the influence ofhe mass ratio of HA to CS on the physicochemical characteristicsf blank PENPs, ten different mixtures were prepared (mass ratiof HA to CS at 1:9, 2:8, 4:6, 5:5, 6:4, 7:3, 8:2, 9:1, 14:1 and 29:1).hanges in the properties of blank PENPs were also monitored forhe changes in initial concentrations of HA and CS solutions, pH,emperature and ionic strength.

To investigate the interaction between HA and CS, the turbid-ty measurement method was used as reported previously with

inor modifications [24]. Briefly, HA solution (1 mg/mL) was addedn a drop wise manner to the CS solution (1 mg/mL) under gentletirring, and vice versa. Changes in turbidity were estimated byonitoring the absorbance of the mixture at 450 nm by a UV–Vis

pectrophotometer (UV9200 UV/VIS spectrophotometer, China).ll measurements were carried out in triplicate.

.3. Preparation of CUR-PENPs

CUR-PENPs were prepared by dropping anhydrous ethanol con-aining CUR into blank PENPs. 4 mL of the 0.1% (w/v) HA solutiondeionized water) was first dropped into an equal volume of the.1% (w/v) CS solution (2% acetic acid solution) under magnetictirring (200 rpm), and stirred for 5 min to form blank HiC PENPs.ext, 1 mL of 0.25% CUR (anhydrous ethanol) was dropped intolank HiC PENPs, stirred for 15 min, and then the mixed solutionas transferred into a dialysis bag with molecular weight cut off

MWCO) of 12 kDa. The bag was tightened and soaked in 500 mLf water at room temperature for 6 h to remove ethanol and aceticcid. Finally, the dialyzed solution was filtered through a 0.45 �mellulose nitrate membrane and freeze dried until further use. Tovoid the photodegradation of CUR, the preparation process waserformed under light shielding.

.4. Characterization of PENPs

100 �L solution of blank PENPs or CUR-PENPs was diluted to mL with deionized water, and then the particle size, polydisper-ity index (PI), and zeta potential were measured at 25 ◦C by usingetasizer 3000HS (Malvern Instruments Ltd., UK). The pH values

l Macromolecules 72 (2015) 1391–1401 1393

of the samples were measured by using a pH meter (model PHS-2C, Lida equipment mill, Shanghai, China). All measurements werecarried out in triplicates.

Morphology examination of blank PENPs or CUR-PENPs wasperformed by using AFM imaging (SPA3800N, SEIKO, Japan) withtapping model and TEM (H-7650, Hitachi High-Technologies Cor-poration).

The content of CUR (based on BDMC) in the PENPs was deter-mined by Shimadzu LC-10AT HPLC system (Kyoto, Japan). Thestationary phase, �Bondapak C18 column (150 mm × 4.6 mm,5 �m), was kept at 40 ◦C. The mobile phase was a mixture of ace-tonitrile: double distilled water: acetic acid = 50:45:5 (v/v/v). Theflow rate was set at 1.0 mL/min. The injection volume was 20 �Land effluent was monitored at 420 nm. CUR-PENPs was filtratedthrough the 0.45 �m cellulose nitrate membrane. 100 �L of filtratewas dissolved in 10 mL of mobile phase, and then centrifuged at12,000 rpm for 10 min. The drug concentrations before and afterfiltration were determined. The encapsulation efficiency (EE, %)and drug loading (DL, %) of the PENPs were calculated accordingto equations 1 and 2, respectively.

EE(%) = Ce/Ct × 100% (1)

where Ce is the concentration of the drug after filtration and Ct

is the concentration of the drug before filtration.

DL(%) = We/(We + Wf ) × 100% (2)

where We is the amounts of the drug in PENPs and Wf is theamount of the feeding materials.

2.5. XRD

The physical state of CUR in PENPs was assessed using X-raydiffraction (Shimadzu XD-3A X-Ray, Japan).

2.6. DSC

The thermal properties of CUR, the physical mixture of blankPENPs and CUR (the amount of CUR was equal to the loaded drugin CUR-PENPs), blank PENPs, and CUR-PENPs were characterizedusing a Thermal Analyzer-60 WS, DSC-60 (Shimadzu, Kyoto, Japan).

2.7. In vitro release study

The release of CUR from CUR-PENPs was carried out in 5% (w/v)glucose solution containing 1% (w/v) Tween 80 using the dialysisbag diffusion technique according to our previous reported methodwith minor modifications [25]. CUR solution (1.5 mg of CUR dis-solved in 10 mL of polyethylene glycol 400, and then diluted to50 �g/mL with deionized water, based on BDMC) was used as acontrol group. 2 mL of CUR-PENPs or CUR solution (equivalent to100 �g of BDMC) was transferred into a dialysis bag (MWCO of12 kDa), and sealed by dialysis clip. Subsequently, the bag wasimmersed into a beaker containing 50 mL of release medium undermechanical shaking (100 rpm) in a water bath at 37 ± 0.5 ◦C. Atpredetermined intervals, 1.0 mL of the dissolution medium waswithdrawn and replaced with the same amount of pre warmedfresh medium. The withdrawn samples were then centrifuged at12,000 rpm for 10 min and the supernatant was analyzed by theHPLC method described previously. The release experiments werecarried out in triplicate.

2.8. In vitro cytotoxicity assays in C6 glioma cells

C6 glioma cells were kindly provided by Prof. Guo (Depart-ment of pharmacology, China Pharmaceutical University) and

1 logical Macromolecules 72 (2015) 1391–1401

rhtE0f

idi(t

C

wa

2

a5Pfsu2AactwpmJ5(

2

0eeccacwiotatTcae

R

v

394 L. Yang et al. / International Journal of Bio

outinely grown in DMEM/HG medium supplemented by 10%eated-inactivated FBS, 100 U/mL penicillin and 100 �g/mL strep-omycin. Cells were maintained at 37 ◦C in an incubator (Thermolectron Corporation) with 5% CO2. The cells were harvested with.25% trypsin and rinsed. The resulted cell suspension was used forurther experimentation.

Cytotoxicities of blank PENPs and CUR-PENPs were studiedn C6 glioma cells by the 3-(4, 5-dimethylthiazol-2-yl)-2, 5-iphenyltetrazolium bromide (MTT) assay. The UV absorbance

ntensity was measured at 570 nm using the microplate readerThermo Electron Corporation, USA). Each point was performed inriplicate. Cell viability was calculated by the following equation:

ell viability (%) = (ODsample/ODcontrol) × 100% (3)

where ODsample is the absorbance intensity of the cells treatedith blank PENPs, CUR solution or CUR-PENPs; ODcontrol is the

bsorbance intensity of the untreated cells.

.9. Cellular uptake study

C6 glioma cells were seeded in 24-well plates (Costar, IL, USA)t a density of 5 × 105 cells/well. The medium was replaced with

�g/mL (equivalent to BDMC content) of CUR solution or CUR-ENPs diluted with DMEM/HG medium. The cells were incubatedor 1 h, 2 h, 4 h and 6 h, to study cellular uptake kinetics. Totudy the effect of the varying concentrations of CUR on cellularptake, CUR solution or CUR-PENPs at the concentration range of.5–10 �g/mL (based on BDMC) were used and incubated for 4 h.fter the incubation, cells were washed thrice with ice-cold PBS,nd lysed with 0.4 mL of 1% Triton X-100 for 3 min. 20 �L of theell lysate from each well was taken and the total cell protein con-ent was determined using the BCA protein assay kit at 570 nmith the plate reader. The remaining cell lysate was extracted andrecipitated by acetonitrile. The concentration of CUR was deter-ined using HPLC with fluorescence detector (Shimadzu RF-10AXL,

apan) with the excitation and emission wavelengths of 436 nm and18 nm, respectively. The uptake index (UI) was expressed as CURng)/cellular protein (�g).

.10. Uptake mechanism study

C6 glioma cells were pre-incubated at 37 ◦C for 30 min in.5 mL of DMEM/HG medium (free of serum) with the followingndocytic inhibitors [26–28]: (1) sodium azide (0.1%, w/v), a cellnergy metabolism inhibitor; (2) chlorpromazine (10 �g/mL), alathrin-mediated endocytosis inhibitor; (3) nystatin (30 �g/mL), aaveolae-mediated endocytosis inhibitor; (4) amiloride (100 �M),

macropinocytosis inhibitor; (5) HA (0.1%, w/v), an inhibitor toompete cell surface CD44 with CUR-PENPs. All the inhibitors usedere non-toxic to the cells at used concentrations. Following pre-

ncubation, the cells were further treated with either CUR solutionr CUR-PENPs for another 4 h. Subsequently, the cells were washedhrice with PBS and collected according to the method describedbove, and the UI values were calculated. Cells in the presence ofested PENPs but without inhibitor treatment were used as control.o reveal the possible uptake mechanisms of PENPs by C6 gliomaells, the uptake study was also performed at 4 and 37 ◦C. The rel-tive uptake index (RUI) was calculated according to the followingquation

UI = UIS/UIc × 100% (4)

Where UIS is the UI of CUR-solution or CUR-PENPs treated witharious kinds of uptake inhibitors and UIC is the UI of controls.

Fig. 3. The turbidity of PENPs obtained by quickly adding CS solution into HA solu-tion (A); by quickly adding HA solution into CS solution (B).

2.11. Statistical analysis

Statistical data analysis was performed using student’s t-test.Differences were considered to be significant at P < 0.05 and mostsignificant at P < 0.01.

3. Results and discussion

3.1. The formation of blank PENPs based on HA/CS

Generally, PENPs are formed predominantly due to the electro-static attraction between two oppositely charged polysaccharides.The entropy gain associated with the release of counterions is alsoone of the major driving forces for PENPs formation. In addition,both polysaccharides are expected to interact through hydrogenbonds, hydrophobic interactions and other intermolecular forcesto form PENPs [29], and the formation and properties of PENPsdepend on the charge ratio of the anionic-to-cationic species [30].In this study, when the HA solution was dropped into the CS solu-tion, inter- and intramolecular electrostatic attractions occurredbetween the positive charge of NH3

+ group from CS and thenegative charge of the carboxyl group from HA.

3.2. The influencing factors on the formation of PENPs

Studies showed that many factors such as the nature of ionicgroups, initial concentrations of the two polyelectrolytes, the massratio of the two polyelectrolytes, pH and ionic strength of the solu-tion, temperature, and the addition order of polyelectrolytes greatlyaffect the formation and characteristics of PENPs [31].

3.2.1. Effects of the addition order of polyelectrolytes and themass ratio of HA and CS

The formation of PENPs based on HA/CS was studied as a func-tion of the mixing mass ratio of HA and CS ranging from 1:9 to 29:1and the effect of the addition order of polyelectrolytes on the forma-tion of PENPs was also investigated. Five kinds of phenomena wereobserved during the addition of HA into CS solution: clear solu-tion, transparent blue solution, light blue opalescent solution, blueopalescent solution and precipitation. However, only three kinds ofphenomena were observed during the addition of CS into HA solu-tion: light blue opalescent solution, blue opalescent solution and

precipitation.

As shown in Fig. 3, as for the addition of HA into CS solution, theturbidity of the mixed solution heightened along with the increas-ing mass ratio of HA recorded. The absorbance increased sharply

L. Yang et al. / International Journal of Biological Macromolecules 72 (2015) 1391–1401 1395

Table 1Effect of the mass ratio of HA and CS on the physicochemical properties of PENPs when adding HA solution (1 mg/mL) into CS solution (1 mg/mL).

HA:CS (m:m) Mean size (nm) Zeta potential (mV) Solution appearance CS monomersa -HAmonomersb (NetCharged species)

Negative/Positiveratio

1:9 – – Clear 47.7 0.052:8 – – Clear 39.4 0.123:7 166.0 ± 0.7 24.74 ± 0.27 Transparent blue 31.2 0.204:6 113.3 ± 3.7 26.37 ± 1.28 Transparent blue 23.0 0.315:5 130.2 ± 2.1 23.15 ± 0.73 Light blue opalescent 14.8 0.476:4 220.5 ± 3.3 20.19 ± 0.92 Blue opalescent 6.5 0.717:3 1626.1 ± 89.5 −7.93 ± 1.62 Precipitation −1.7 1.108:2 1089.9 ± 53.9 −11.57 ± 2.99 Precipitation −9.9 1.899:1 358.4 ± 11.4 −18.65 ± 2.43 Blue opalescent −18.2 4.2514:1 264.0 ± 6.8 −21.09 ± 0.89 Light blue opalescent −31.3 6.6129:1 238.4 ± 4.1 −21.81 ± 0.27 Light blue opalescent −70.9 13.69

-(1,4) (1,4)

wittit4omomtscItvacs

rsHbwWfotio

oti(

TE

a CS monomers: units of N-acetyl-d-glucosamine and d-glucosamine linked by �b HA monomers: units of d-glucuronic acid and N-acetyl glucosamine linked by �

hen the mass ratio of HA to CS was up to the ratio of 6:4. Furtherncrease in the mass ratio of HA to CS led to the system precipi-ation, at which point the absorbance of the mixed solution begano decrease. The increase of the absorbance of the solution, i.e. thencrease of the turbidity or the decrease of the transmittance ofhe solution, suggests the formation of insoluble PENPs. Because at50 nm, neither HA nor CS absorb light, the particle size and numberf PENPs were the only variables that affected the degree of trans-ittance. As for the addition of CS into HA solution, the absorbance

f the mixed solution also heightened along with the increasingass ratio of HA recorded, and yet, only when the mass ratio of HA

o CS was up to the ratio of 3:7, CiH PENPs with micron particleize were formed, which indicated that tuning of the physico-hemical properties of CiH PENPs would be difficult to control.t is most likely due to the differences in MW of HA and CS, andhe different localization of HA and CS in the PENPs. These obser-ations showed that the addition order of polyelectrolytes wouldffect the properties of PENPs produced. Therefore, the physico-hemical properties of HiC PENPs were studied in the followingtudy.

The relationships between the mass ratios of HA to CS and theesulting PENPs appearance, particle size and zeta potential arehown in Table 1. It was found that only when the mass ratio ofA to CS was lower than 7:3 or higher than 8:2, blank PENPs coulde formed. PENPs with the smallest particle size (113.3 ± 0.6 nm)ere formed when the mass ratio of HA to CS was the ratio of 4:6.hen the HA:CS mass ratio increased, the zeta potential varied

rom positive to negative, thus indicating the emplacement of HAn the PENPs surface. Hence, an increase in size and a decrease inhe surface charge with the increasing mass ratio of HA in PENPsndicate that the size and zeta potential of PENPs were dependentn the composition.

In addition, based on the molecular weight of the repeat unit

f HA and CS, we find that 1 mg of HA contains 2.6 micromoles ofhe HA monomer or carboxylic groups as Mol.wt of HA monomers 379, and 1 mg of CS contains 5.6 micromoles of CS monomerdegree of deacetylation (90%) was considered) or amine groups

able 2Affect of the initial concentrations of HA and CS solutions on particle sizes of blank PENPs

HA concentrationa (mg/mL) CS concentrationa (mg/mL)

0.5 1

0.5 92.4 ± 0.4b 11.0 Aggregation 11.5 613.4 ± 50.8 32.0 473.7 ± 31.4 A

a HA or CS initial concentrations during the process of preparing PENPs.b Mean particle size (nm). Data shown are the mean ± standard deviation, (n = 3).

glycosidic bonds. and � (1,3) glycosidic bonds.

as Mol.wt of CS monomer is 161. Therefore, we can calculate thecharge ratios compared to the mass ratio (Table 1). It is clear thatwhen net charge was highly positive or highly negative, nanopar-ticles with small size were obtained, and when the net charge wasclose to 0 (no of carboxylic groups are equal to number of positivegroups), precipitates or large particles were obtained.

3.2.2. Effect of the initial concentrations of both HA and CSsolutions

As shown in Table 2A, when the initial concentration of HA solu-tion was fixed at 0.5 mg/mL, the particle size of PENPs significantlyincreased with the addition of CS solution with the concentrationin the range of 0.5–2 mg/mL, and PENPs with the smallest particlesize (92.4 ± 0.4 nm) were formed when mixed with 0.5 mg/mL ofCS solution. At the same time, the zeta potential increased alongwith the increase in the initial concentration of CS solution added(Table 2B). In the experiment, we found that when the initial con-centration of HA solution was twice that of CS, zeta potential ofPENPs was close to zero, thus decreasing the repulsion betweenPENPs and leading to PENPs aggregation.

Buehhammer et al. found that when the polymer concentrationwas less than a critical concentration, particle size and the parti-cle size distribution of PENPs had no obvious changes along withthe addition of the polymer. When the polymer concentration wasabove the critical concentration, PENPs became unstable. In orderto prepare a uniformly dispersed system, a low concentration of thepolymer should be selected to form smaller and tighter particles.Nevertheless, even in a dilute solution, PENPs could be stable onlywhen the polyelectrolyte component reached a certain proportion,otherwise, aggregation would occur [32]. Our results showed thatthe particle size of PENPs was really related to the initial concen-tration of the polymer. The experimental results on the stability of

the above PENPs (Table 2C) showed that the some PENPs becamebigger and some even aggregated after 3 days. The PENPs whichshowed least change in diameter were further selected for drugloading.

with volume ratio of HA and CS solutions at 1:1.

.0 1.5 2.0

58.5 ± 5.5 352.4 ± 2.0 436.1 ± 58.840.5 ± 0.8 123.9 ± 0.8 116.9 ± 0.846.5 ± 6.5 198.4 ± 1.8 182.5 ± 3.2ggregation 364.8 ± 7.0 257.8 ± 2.8

1396 L. Yang et al. / International Journal of Biological Macromolecules 72 (2015) 1391–1401

Table 2BEffects of the initial concentration of HA and CS solution on zeta potential of blank PENPs with volume ratio of HA and CS solution at 1:1.

HA concentrationa (mg/mL) CS concentrationa (mg/mL)

0.5 1.0 1.5 2.0

0.5 23.21 ± 3.41 b 25.74 ± 1.51 27.40 ± 2.07 28.00 ± 1.331.0 −1.05 ± 2.13 24.73 ± 0.64 26.71 ± 0.87 26.40 ± 0.261.5 −7.55 ± 0.75 21.75 ± 1.59 25.53 ± 1.09 27.46 ± 1.622.0 −13.99 ± 0.37 3.46 ± 2.88 23.31 ± 1.74 26.25 ± 0.70

).

3

cmPfiatFpfttptoCtfpdaapa

3

tift[ttb5r1

TE

*

a HA or CS initial concentration during the process of preparing PENPs.b Mean zeta potential (mV). Data shown are the mean ± standard deviation, (n = 3

.2.3. Effect of the pHThe pH of the solution determines the degree of ionization of

arboxylic acid groups of the HA and amino groups of the CS, whichight affect the particle size, zeta potential, and the formation of

ENPs [24,33]. The concentrations of both HA and CS solutions werexed at 1 mg/mL. The pH of the CS solution was adjusted with dilutecetic acid or dilute ammonia solution, then the equal volume ofhe HA solution was added into the CS solution to form HiC PENPs.ig. 4A shows that when the pH of the solution was adjusted withinH range of 1.93-2.79, the particle size of HiC PENPs decreasedrom 619.8 to 143.1 nm and zeta potential increased from + 12.54o + 18.33 mV. As the pH was increased from 3.29 to 5.36, the par-icle size of the HiC PENPs significantly increased, and the zetaotential decreased from + 14.21 to −8.14 mV. The observed par-icle size at pH 3.74 (2350.9 nm) was about 6.8 times the sizebserved at pH 3.29 (342.6 nm). The pKa of the amino groups inS is close to 6.5, and the pka of the carboxylic acid in HA is closeo 2.6. Majority of the carboxylic acids in HA were in the molecularorm and most of the amino groups of CS were protonated belowH 2.6, resulting in complexes of high charge density with smalleriameter. However, the carboxylic acids in HA inclined to dissoci-te with a lower protonation tendency of amino groups of CS withn increase in pH from 2.6 to 5.5, which resulted in the formation ofarticles or aggregates in a size range of microns. Our observationsre in good agreement with the previously published findings [16].

.2.4. Effect of the temperatureVarying observations have been reported regarding the effect of

he temperature on the characteristics of PENPs in previous find-ngs. It was reported that the incubation at 50 ◦C for 5.5 h led to theormation of a denser chitosan-DNA complex [34], whereas the par-icle size of xanthan-chitosan nanoparticles increased when heated35]. Stable alginate-chitosan complexes were observed over aemperature range of 4 to 37 ◦C [36]. We observed little effect ofemperature on the particle size and zeta potential of blank PENPs

ased on HA and CS. Fig. 4B shows that at a temperature range of

to 55 ◦C, the particle size and zeta potential of HiC PENPs (massatio of HA to CS was 5:5) remained almost the same even up to0 h of heating.

able 2Cffects of the initial concentration of HA and CS solution on particle sizes of blank PENPs

HA concentrationa (mg/mL) CS concentrationa (mg/mL)

0.5 1

0.5 142.1 ± 7.5 b* 11.0 Aggregation 21.5 Aggregation 52.0 Aggregation A

Formulations are used to further load drug.a HA or CS initial concentration during the process of preparing PENPs.b Mean particle size (nm). Data shown are the mean ± standard deviation, (n = 3).

3.2.5. Effect of the ionic strengthIn order to investigate the effect of ionic strength on PENPs, the

concentrations of the HA and CS solutions were fixed at 1 mg/mL.After the addition of NaCl to the CS solution, equal volume of theHA solution was added into the CS solution to form HiC PENPs. Asshown in Fig. 4C, the particle size of PENPs increased from 123.5to 259.2 nm upon addition of NaCl solution with the ionic strengthin the range of 0–0.17 mmol/L. When ionic strength >0.86 mmol/L,PENPs began to aggregate, which is consistent with the alreadyreported data [24]. The observed phenomenon probably resultsfrom two effects. On one hand, the addition of salt favors the forma-tion and the growth of PENPs, which is indicated by the increase inthe number and particle size of PENPs [37]. On the other hand, theaddition of salt interferes with the electrostatic attraction betweenthe polymer chains of opposite charges, which reduces the interac-tion between the polymer chains and leads to PENPs aggregation[38]. At low salt concentrations, the electrostatic interactions ofPENPs may exceed the screening effect of the salt, however, athigher salt concentrations, the dominant charge screening inducedby salt leads to the substantial reduction of the interaction.

3.3. Preparation of CUR-PENPs

At present, there are two methods used for the preparation ofdrug-loaded PENPs: adsorption method and mixed method. Mixedmethod means the drug is first dissolved in the solution containingone kind of polyelectrolytes, and then the solution containing theoppositely charged polyelectrolyte is added to form drug-loadedPENPs. Adsorption method means after forming blank PENPs, thedrug is added to form drug-loaded PENPs. In preliminary stud-ies, mixed method had been attempted to prepare CUR-PENPs, butmicron-scale unstable CUR-PENPs were obtained (data not shown).Therefore, the adsorption method was chosen to prepare CUR-PENPs.

To investigate the mechanism of the formation of blank PENPs,effect of the pH of solution on the formation of blank PENPs was

discussed, because the pH of the solution determines the degree ofionization of carboxylic acid groups of the HA and amino groupsof the CS. Therefore, in the preparation of blank HiC PENPs, the pHof the CS solution was adjusted with dilute acetic acid or dilute

after 3 days with volume ratio of HA and CS solution at 1:1.

.0 1.5 2.0

40.5 ± 1.0* 369.5 ± 13.7 533.2 ± 12.213.3 ± 1.7* 219.0 ± 9.1* 216.4 ± 8.4*39.2 ± 21.3 288.3 ± 5.7 292.4 ± 6.4ggregation 529.7 ± 16.9 350.5 ± 7.2

L. Yang et al. / International Journal of Biologica

Fig. 4. Effect of pH (A), temperature (B) and concentration of NaCl (C) on the particlesize, zeta potential or polydispersity of blank HiC PENPs (n = 3). * Represents thea

aa

wsmastat

To provide the sink condition, 1% (w/v) tween 80 was added into

ggregation.

mmonia solution, then the equal volume of the HA solution wasdded into the CS solution.

Actually, blank HiC PENPs used in all the other experimentsere prepared through directly mixing the CS solution with the HA

olution. The pH value of blank PENPs suspension was 2.79, whichight cause serious irritation to the blood vessels due to the accept-

ble pH for intravenous preparations in the range of 4 to 9. Fig. 4Ahows that before forming CUR-PENPs, it was not feasible adjust

he pH of CS solution by the addition of dilute acetic acid or dilutemmonia solution to meet the requirement of intravenous injec-ion, which would cause significant changes in the particle size of

l Macromolecules 72 (2015) 1391–1401 1397

PENPs. Finally, we used the dialysis method instead of pH regulatorto adjust the pH of CUR-PENPs in later studies.

Five formulations marked with asterisks in Table 2C were fur-ther studied to load CUR with EE as the evaluation index. Whenthe initial concentrations of both HA and CS solutions at 1 mg/mL,and the volume ratio of HA to CS solution was at 1:1, CUR-PENPshad the highest EE (89.3%) among all labeled formulations (datanot shown). The pre-experiment results showed many factors suchas the drug amount added in blank PENPs, stirring time, and dial-ysis time, had great effects on the physicochemical characteristicsof PENPs. As shown in Table 3, as the drug amount increased from0.75 to 1.5 mg, EE of CUR-PENPs decreased from 90.3% to 82.1%.The stability results showed when 1 mg of the drug was added, thechange of EE of CUR-PENPs was the smallest among all prepara-tions, from 88.9% (the first day) to 82.0% (the fourth day). Therefore,1 mg of CUR was determined as the added amount of drug. Uponprolonging the stirring time, CUR adsorbed on the surface of PENPswas precipitated, leading to the significant reduction of EE from89.6% to 73.1%. The pH value of the dialyzed CUR-PENPs dispersionincreased along with the increase of dialysis time. When dialysistime was set at 6 h, the pH value of CUR-PENPs dispersion was4.13, which was in the allowable range for intravenous injection,and the EE of CUR-PENPs was also the highest (91.4%). Therefore,the stirring time and dialysis time were fixed at 0.25 h and 6 h,respectively.

3.4. Characterization of blank PENPs and CUR-PENPs

The appearance of blank PENPs was uniform and translucentwith a clear blue opalescence (Fig. 5A), and the appearance of CUR-PENPs was uniform and translucent with a clear orange-yellowopalescence (Fig. 5B). The AFM pictures and TEM images furthersuggested that both blank PENPs and CUR-PENPs were almostspherical in shaper without any aggregation (Fig. 5C–F). The otherphysicochemical characteristics of both blank PENPs and CUR-PENPs are shown in Table 4. The results showed that CUR-PENPshad larger particle size and higher PI than blank PENPs, and theaddition of CUR had no effect on the zeta potential of ENPs.

Fig. 5G shows the DSC thermograms of pure CUR, blank PENPs,physical mixture of CUR and blank PENPs, and CUR-PENPs powder.Pure CUR exhibited a sharp endothermal peak at 163.7 ◦C result-ing from melting of CUR crystals [39]. Physical mixture of CUR andblank PENPs clearly exhibited the endothermal peak of CUR. How-ever, DSC of CUR-PENPs did not show the endothermal peak of CUR.It might be explained that the crystallization of CUR was inhibitedby PENPs and CUR might have been in a molecular or amorphousstate in the PENPs.

As shown in Fig. 5H, pure CUR showed a series of characteristicpeaks at 2� angles, indicating the typical crystalline pattern. BlankPENPs exhibited amorphous characteristics and no crystalline peakcould be seen. The diffraction patterns of physical mixture of CURand blank PENPs with apparent peaks were similar to that ofthe pure drug, indicating that only a simple mixing of drug andcarriers did not change the crystallinity of CUR. With regard to CUR-PENPs, XRD showed the similar results to the DSC thermograms andthere was not much difference in the diffraction pattern of blankPENPs and CUR-PENPs, indicating that the addition of CUR had notchanged the nature of PENPs and CUR was entrapped inside thePENPs in a molecular or amorphous form.

3.5. In vitro drug release studies

isotonic glucose solution (5%, w/v). The release studies showed ini-tial burst release (51.60%) of CUR from CUR-PENPs for the first 4 hwhich could be attributed to the adsorbed drug on the surface of

1398 L. Yang et al. / International Journal of Biological Macromolecules 72 (2015) 1391–1401

Table 3Effects of different factors on physicochemical characteristics of CUR-PENPs with 1 mg/mL of the initial concentration of both HA and CS solution and the volume ratio of HAto CS at 1:1. (n = 3).

Factor Size/nm pH EE/%

m(CUR) (mg) 0.75 189.2 ± 2.1 4.18 ± 0.5 90.3 ± 0.81 203.2 ± 2.4 4.12 ± 0.3 88.9 ± 1.41.25 198.1 ± 3.1 4.10 ± 0.2 87.5 ± 0.91.5 185.3 ± 12.4 4.12 ± 0.7 82.1 ± 3.9

Stir time (h) 0.25 204.3 ± 3.3 4.20 ± 2.1 89.6 ± 1.20.5 187.2 ± 13.1 4.30 ± 1.3 85.1 ± 1.71 168.9 ± 21.5 4.21 ± 0.5 74.3 ± 1.72 152.0 ± 17.9 4.22 ± 0.4 73.1 ± 2.8

Dialysis time (h) 2 204.8 ± 2.5 3.82 ± 0.8 89.3 ± 1.74 219.0 ± 4.7 4.08 ± 0.4 88.5 ± 1.26 213.5 ± 2.7 4.13 ± 1.1 91.4 ± 2.0

12 202.7 ± 3.3 4.18 ± 0.3 88.2 ± 3.3

Table 4The physicochemical characterizations of blank PENPs and optimized CUR-PENPs (n = 3).

Size (nm) PI Zeta potential (mV) pH value EE (%) DL (%)

Blank PENPs 130.2 ± 2.13 0.059 ± 0.026 23.15 ± 0.51 2.79 ± 0.13 – –CUR-PENPs 207.0 ± 5.72 0.18 ± 0.017 25.37 ± 0.69 4.13 ± 0.20 89.9 ± 1.2 6.5 ± 0.9

Fig. 5. (A) The appearance of blank PENPs (A) and CUR-PENPs (B); AFM images of blank PENPs (C) and CUR-PENPs(D); TEM images of blank PENPs (E) and CUR-PENPs (F);DSC thermograms of (G) CUR (a), blank PENPs (b), physical mixture of CUR and blank PENPs (c), and CUR-PENPs powder (d); XRD patterns of (H) CUR (a), blank PENPs (b),physical mixture of CUR and blank PENPs (c), and CUR-PENPs powder (d); Cumulative release profiles of (I) CUR-solution and CUR-PENPs into 5% glucose with 1% tween 80(n = 3).

L. Yang et al. / International Journal of Biological Macromolecules 72 (2015) 1391–1401 1399

Fig. 6. (A) In vitro cytotoxicity test of blank PENPs in C6 glioma cells for 24 h, cell viability was measured by MTT-based assay according to various PENPs concentrations; (B)In vitro antitumor effect of CUR-solution and CUR-PENPs in C6 glioma cells for 24 h, cell viability was measured by MTT-based assay according to various CUR concentrations.Data are expressed as mean ± SD (n = 3). * P < 0.05 compared with CUR-solution.

Pdpe6(

3

tPbb

tCttcieiwld

3

itibifAcasAoob

ENPs. After 8 h, the release was steady indicating the sustainedrug release of CUR from PENPs following first order model. Theercentage cumulative drug release after 36 h was 82.77%. How-ver, a more rapid release of CUR from CUR solution was observed,6.93% of CUR were released within 4 h, and 96.20% within 36 hFig. 5I).

.6. Study of the cytotoxicity of PENPs

Fig. 6A displays the percentage of cell viability as a function ofhe concentration of the blank PENPs. Results showed that blankENPs in the studied concentrations have no effects on cell via-ility of C6 glioma cells, suggesting blank PENPs might be goodiocompatible carriers.

Fig. 6B shows the percentage of cell viability as a function ofhe concentration of the CUR. In C6 glioma cell lines, the IC50 ofUR-PENPs and CUR-solution were 14.20 and 28.95 �g/mL, respec-ively. Results showed that for both CUR-solution and CUR-PENPs,he antitumor activity increased concentration-dependently. Foroncentration of CUR below 2.5 �g/mL, no significant differencesn antitumor activity were observed for two formulations. How-ver, compared with CUR-solution, the cytotoxicity of CUR-PENPsncreased significantly (P < 0.05) when the concentration of CUR

as above 5 �g/mL. The increased cytotoxic effect of the drugoaded PENPs could be attributed to the better penetration of therug in the tumor cells due to the carrier.

.7. In vitro cell uptake

Fig. 7A illustrates that both CUR-solution and CUR-PENPs werenternalized in C6 cells as early as in 1 h and the uptake wasime-dependent. Moreover, both CUR-solution and CUR-PENPsnternalization were relatively rapid during the first 4 h of incu-ation followed by a gradual increase reaching a saturation uptake

n 4–6 h, which showed that 4 h of incubation should be sufficientor uptake of both CUR-solution and CUR-PENPs by C6 glioma cells.t almost each time point, CUR-PENPs exhibited higher uptake effi-iency than CUR-solution and uptake efficiency of CUR-PENPs wasbout 1.37 times more than that of CUR-solution at 4 h. Fig. 7Bhows the effect of drug concentration on the cell uptake of CUR.

fter incubation with C6 glioma cells for 4 h at 37 ◦C, the cell uptakef both CUR-solution and CUR-PENPs increased with the increasef the concentrations of CUR, which showed the cell uptake ofoth CUR-solution and CUR-PENPs was concentration-dependent.

Compared with CUR-solution, the cell uptake of CUR-PENPsincreased significantly (P < 0.05) when the concentration of CURwas above 7.5 �g/mL.

As presented in Fig. 7C and D, it could be clearly observed thatcompared with CUR-solution, C6 glioma cells treated with CUR-PENPs exhibited higher fluorescence intensity at the concentrationof 7.5 �g/mL of CUR for 4 h, which was consistent with previousUI data and indicated that CUR-PENPs were more effective in theenhancement of the cell uptake of CUR.

3.8. In vitro cell uptake mechanism study

To identify whether CUR-PENPs are taken up by active endo-cytosis, the cellular uptake of CUR-PENPs in C6 glioma cells wereevaluated at 4 ◦C or in the presence of metabolic inhibitors (sodiumazide) [40]. Compared with the control, cellular uptake of CUR-PENPs decreased significantly when incubated at 4 ◦C or in thepresence of sodium azide at 37 ◦C (P < 0.01) (Fig. 7E). In addition,the UI of CUR-PENPs at 37 ◦C was 6.87 times higher than thoseat 4 ◦C. These results indicate that the internalization processes ofCUR-PENPs were energy- and ATP-dependent.

However, the cellular uptake of CUR-PENPs at 4 ◦C could notbe ignored, which indicated that the uptake process of CUR-PENPs in C6 glioma cells was also governed by other mechanisms.We performed uptake experiments in the presence of endocyto-sis inhibition. As shown in Fig. 7E, after treating cells with freeHA in advance, the uptake amount of CUR-PENPs by C6 gliomacells decreased to 72.05%, demonstrating that the uptake of CUR-PENPs containing HA into C6 glioma cells could be significantlydecreased due to free HA competitively binding with CD44 onthe surface of C6 glioma cells to block the association of CUR-PENPs and cells [41]. In addition, chlorpromazine and nystatinsignificantly inhibited the uptake of CUR-PENPs by 35.40% and18.34%, respectively, indicating that clathrin- and caveolae- medi-ated endocytosis were involved in the internalization of CUR-PENPsin C6 glioma cells. Amiloride slightly decreased the uptake ofCUR-PENPs but had no notable affect. Thus, macropinocytosisseemed to be also involved in the internalization of CUR-PENPsin C6 glioma cells. The key attribute to the macropinocyto-sis of CUR-PENPs is thought to be the non-specific adsorptive

pinocytosis caused by the presence of cationic residues on thesurface of CUR-PENPs, which could enhance attractive electro-static force of nanocarriers with negatively charged cell membrane[42,43].

1400 L. Yang et al. / International Journal of Biological Macromolecules 72 (2015) 1391–1401

Fig. 7. Uptake of CUR-solution and CUR-PENPs in glioma cells at 37 ◦C (A) with 5 �g/mL CUR at different time intervals (n = 3); (B) for different concentration of CUR (n = 3);F tion (Cu ting wC with C

4

ssasagEh

luorescence microscopic images of C6 glioma cells after incubation with CUR-soluptake efficiency of CUR-solution and CUR-PENPs at different temperature and treaUR-solution. # P < 0.05 compared with CUR-PENPs at 37 ◦C, ## P < 0.01 compared

. Conclusion

The mass ratio and the initial concentration of the two oppo-itely charged polyelectrolytes, pH value and ionic strengths of theolution were shown to be parameters that affected particle size,nd zeta potential of blank PENPs. PENPs based on HA/CS wereuccessfully prepared for loading water-insoluble CUR. The drug

mount added in blank PENPs, the stirring time and dialysis timereatly affect the physicochemical characteristics of CUR-PENPs.xperimental results on cells studies showed that CUR-PENPs hadigher cellular uptake in vitro in C6 glioma cells than CUR-solution.

) and CUR-PENPs (D) (equivalent to 7.5 �g/mL of CUR) at 37 ◦C for 4 h; (E) Relativeith various cellular uptake inhibitors in C6 cell line (n = 3). * P < 0.05 compared withUR-PENPs at 37 ◦C.

Therefore, conclusion could be made that CUR-PENPs were moreeffective in inhibiting the proliferation of C6 glioma cells than CUR-solution. However, for the treatment of brain glioma, the vehiclesmust pass through blood-brain barriers in addition to targeting thebrain glioma cells. Therefore, future studies could include a suitableligand (like lactoferrin) for targeting CUR-PENPs to the BBB.

Acknowledgments

This work was supported by the Natural Science Foundationof Jiangsu Province (Program No. BK20130655), the Open Project

logica

PmSge

R

[[

[

[

[

[[[

[

[

[

[

[[

[[[

[

[

[[

[[

[[[[

[[[

L. Yang et al. / International Journal of Bio

rogram of State Key Laboratory of Natural Medicines, China Phar-aceutical University (Program No. SKLNMKF201204) and College

tudents Innovation Project for the R&D of Novel Drugs (Pro-ram No. J1030830) and the National College Students’ innovationntrepreneurial training program (Program No. G13076).

eferences

[1] S. Al-Qadi, M. Alatorre-Meda, P. Taboada, C. Remunán-López, E.M. Zaghloul,Colloid. Surface. B 103 (2012) 615.

[2] H.-D. Lu, H.-Q. Zhao, K. Wang, L.-L. Lv, Int. J. Pharm. 420 (2011) 358–365.[3] Q.-K. Lin, K.-F. Ren, J. Ji, Colloid. Surface. B 74 (2009) 298–303.[4] P. Calvo, C. Remunan-Lopez, J. Vila-Jato, M. Alonso, J. Appl. Polym. Sci. 63 (1997)

125–132.[5] K.A. Janes, M.P. Fresneau, A. Marazuela, A. Fabra, M.a.J. Alonso, J. Control. Release

73 (2001) 255–267.[6] Y. Boonsongrit, A. Mitrevej, B.W. Mueller, Eur. J. Pharm. Biopharm. 62 (2006)

267–274.[7] A. Anitha, V.G. Deepagan, V.V. Divya Rani, D. Menon, S.V. Nair, R. Jayakumar,

Carbohydr. Polym. 84 (2011) 1158–1164.[8] V.M. Platt, F.C. Szoka Jr., Mol. Pharm. 5 (2008) 474–486.[9] I. Rivkin, K. Cohen, J. Koffler, D. Melikhov, D. Peer, R. Margalit, Biomaterials 31

(2010) 7106–7114.10] H. Lee, C.H. Ahn, T.G. Park, Macromol. Biosci. 9 (2009) 336–342.11] M.-Y. Lee, S.-J. Park, K. Park, K.S. Kim, H. Lee, S.K. Hahn, ACS Nano 5 (2011)

6138–6147.12] W. Argüelles-Monal, G. Cabrera, C. Peniche, M. Rinaudo, Polymer 41 (2000)

2373–2378.13] C. Tapia, Z. Escobar, E. Costa, J. Sapag-Hagar, F. Valenzuela, C. Basualto,

a.M. Nella Gai, M. Yazdani-Pedram, Eur. J. Pharm. Biopharm. 57 (2004)65–75.

14] C. Schatz, A. Domard, C. Viton, C. Pichot, T. Delair, Biomacromolecules 5 (2004)1882–1892.

15] K. Barck, M.F. Butler, J. Appl. Polym. Sci. 98 (2005) 1581–1593.16] N.P. Birch, J.D. Schiffman, Langmuir 30 (2014) 3441–3447.17] M. López-Lázaro, Mol. Nutr. Food Res. 52 (2008) S103–S127.

[[[[

l Macromolecules 72 (2015) 1391–1401 1401

18] J.K. Lin, S.Y. Lin-Shiau, Proc. Natl. Sci. Counc. Republic China B Life Sci. 25 (2001)59–66.

19] Y. Xiao, X. Chen, L. Yang, X. Zhu, L. Zou, F. Meng, Q. Ping, J. Agric. Food Chem. 61(2013) 3654–3660.

20] S.S. Dhule, P. Penfornis, J. He, M.R. Harris, T. Terry, V. John, R. Pochampally, Mol.Pharm. 11 (2013) 417–427.

21] T. Harada, L. Giorgio, T.J. Harris, D.-T. Pham, H.T. Ngo, E.F. Need, B.J. Coventry, S.F.Lincoln, C.J. Easton, G. Buchanan, T.W. Kee, Mol. Pharm. 10 (2013) 4481–4490.

22] R. Stern, A.A. Asari, K.N. Sugahara, Eur. J. Cell Biol. 85 (2006) 699–715.23] S. Yodkeeree, W. Chaiwangyen, S. Garbisa, P. Limtrakul, J. Nutr. Biochem. 20

(2009) 87–95.24] S. Shu, X. Zhang, D. Teng, Z. Wang, C. Li, Carbohydr. Res. 344 (2009) 1197–1204.25] Y.X.X. Chen, Y. Chen, Q. Ping, C. Zhang, J. China Pharm. Univer. 43 (2012) 412.26] H. Li, Y. Xiao, J. Niu, X. Chen, Q. Ping, J. Nanosci. Nanotechnol. 11 (2011)

8547–8555.27] Z. Su, Y. Shi, Y. Xiao, M. Sun, Q. Ping, L. Zong, S. Li, J. Niu, A. Huang, W. You, Y.

Chen, X. Chen, J. Fei, J. Tian, Int. J. Pharm. 447 (2013) 281–292.28] R. Mo, X. Jin, N. Li, C. Ju, M. Sun, C. Zhang, Q. Ping, Biomaterials 32 (2011)

4609–4620.29] M. De la Fuente, B. Seijo, M. Alonso, Nanotechnology 19 (2008) 075105.30] J. Baier, J. Koetz, S. Kosmella, B. Tiersch, H. Rehage, J. Phys. Chem. B 111 (2007)

8612–8618.31] S.G. Starodubtsev, A.T. Dembo, K.A. Dembo, Langmuir 20 (2004) 6599–6604.32] H.-M. Buchhammer, M. Mende, M. Oelmann, Colloid. Surface. A 218 (2003)

151–159.33] M.K. Yoo, Y.K. Sung, S.C. Chong, M.L. Young, Polymer 38 (1997) 2759–2765.34] G.M. Signe Danielsen, T. Bjørn, Stokke, Biopolymers 77 (2005) 86.35] B.T.S.G. Maurstad, Biopolymers 74 (2004) 199.36] H.V. Sæther, H.K. Holme, G. Maurstad, O. Smidsrød, B.T. Stokke, Carbohydr.

Polym. 74 (2008) 813–821.37] G. Schneider, G. Decher, Langmuir 24 (2008) 1778–1789.38] H. Dautzenberg, Macromolecules 30 (1997) 7810–7815.39] P. Sanphui, N.R. Goud, U.B.R. Khandavilli, S. Bhanoth, A. Nangia, Chem. Commun.

47 (2011) 5013–5015.40] S. Mukherjee, R.N. Ghosh, F.R. Maxfield, Physiol. Rev. 77 (1997) 759–803.41] S.D. Conner, S.L. Schmid, Nature 422 (2003) 37–44.42] C. He, Y. Hu, L. Yin, C. Tang, C. Yin, Biomaterials 31 (2010) 3657–3666.43] H. Gao, W. Shi, L.B. Freund, Proc. Natl. Acad. Sci. USA 102 (2005) 9469–9474.