Mm

Ka

Sb

c

a

ARR1A

KIPMMC

1

dmtataabiamtdoi

tS

(

h0

Colloids and Surfaces B: Biointerfaces 136 (2015) 712–720

Contents lists available at ScienceDirect

Colloids and Surfaces B: Biointerfaces

journa l homepage: www.e lsev ier .com/ locate /co lsur fb

agnetic field activated drug release system based on magnetic PLGAicrospheres for chemo-thermal therapy

un Fang a, Lina Song a, Zhuxiao Gu b, Fang Yang a,c,∗, Yu Zhang a, Ning Gu a,c,∗

State Key Laboratory of Bioelectronics, Jiangsu Key Laboratory for Biomaterials and Devices, School of Biological Science and Medical Engineering,outheast University, Nanjing 210009, ChinaDepartment of Chemical Engineering, University of Florida, Florida 32611, United StatesCollaborative Innovation Center of Suzhou Nano Science and Technology, Suzhou 215123, China

r t i c l e i n f o

rticle history:eceived 16 July 2015eceived in revised form6 September 2015ccepted 10 October 2015

eywords:ron oxide nanoparticles

a b s t r a c t

Controlled drug delivery systems have been extensively investigated for cancer therapy in order to obtainbetter specific targeting and therapeutic efficiency. Herein, we developed doxorubicin-loaded magneticPLGA microspheres (DOX-MMS), in which DOX was encapsulated in the core and high contents (28.3 wt%)of �-Fe2O3 nanoparticles (IOs) were electrostatically assembled on the surface of microsphere to ensurethe high sensitivity to response of an external alternating current magnetic field (ACMF). The IOs inPLGA shell can both induce the heat effect and trigger shell permeability enhancement to release drugswhen DOX-MMs was activated by ACMF. Results show that the cumulative drug release from DOX-MMs

LGAagnetic hyperthermiaagnetic responsive drug release

hemo-thermal therapy

exposed to ACMF for 30 min (21.6%) was significantly higher (approximately 7 times higher) than that notexposed to ACMF (2.8%). The combination of hyperthermia and enhanced DOX release from DOX-MMSis beneficial for in vitro 4T1 breast cancer cell apoptosis as well as effective inhibition of tumor growthin 4T1 tumor xenografts. Therefore, the DOX-MMS can be optimized as powerful delivery system forefficient magnetic responsive drug release and chemo-thermal therapy.

. Introduction

The penetration of anticancer drugs in solid tumors is limitedue to abnormal tumor vasculature, the hypoxic microenviron-ent and the high interstitial fluid pressure in the center of

he tumor [1,2], which may result in the tumor regenerationfter the chemotherapy. In recent years, mild hyperthermia ofumor region (40–45 ◦C) has been increasingly investigated as andjuvant that can effectively sensitize tumors to chemotherapynd radiotherapy as well as induce apoptosis [3,4]. The com-ined treatment modalities provide an emerging approach for

mproving the chemosensitization of cancer cells within tumorsnd the therapeutic efficacy of anticancer agents. In particular,agnetic hyperthermia, resulting from the efficient heat produc-

ion under alternating current magnetic field (ACMF), has been

emonstrated as a promising cancer therapy methodology becausef the relative noninvasive and deep penetration when applied

n vivo [5–8]. Recent studies have demonstrated that magnetic

∗ Corresponding authors at: Southeast University, State Key Laboratory of Bioelec-ronics, Jiangsu Key Laboratory for Biomaterials and Devices, School of Biologicalcience and Medical Engineering, Nanjing 210009, China. Fax: +86 25 83272460.

E-mail addresses: yangfang2080@seu.edu.cn (F. Yang), guning@seu.edu.cnN. Gu).

ttp://dx.doi.org/10.1016/j.colsurfb.2015.10.014927-7765/© 2015 Elsevier B.V. All rights reserved.

© 2015 Elsevier B.V. All rights reserved.

nanoparticles (MNPs) could be loaded with anticancer agents andremotely heated by an external ACMF after localization in tumor [9].Moreover, the combination of hyperthermia and anticancer drugsynergistically enhances therapeutic efficacy. In addition, the factthat appropriate magnetic field causes no adverse effect on biolog-ical tissues serves as a distinctive advantage for noninvasive in vivoapplications.

Poly(lactic-co-glycolic acid) (PLGA), a food and drug administra-tion (FDA) approved biodegradable polymer, has been extensivelyinvestigated in many medical and pharmaceutical fields dueto its good biodegradability and biocompatibility [10–12]. ThePLGA-based drug delivery carriers have shown sustained releasecharacteristics due to degradation and diffusion mechanisms[13–15]. However, the slow degradation of PLGA, the low con-centration drug release may reduce the cytotoxicity of anticancerdrug and result in the multidrug resistance (MDR) [16,17]. There-fore, modulating the drug release kinetics of PLGA-based drugdelivery carries is vital for maximizing their efficacy to tumorstherapy. Until now, there are many approaches to precisely con-trol the release rate by adjusting physicochemical properties of the

polymer, such as molecular weight, structural components (LA/GAratio), hydrophobicity, pH, and glass transition temperature, etc.Among them, some studies have indicated that when the tempera-tures were increased to near the glass transition temperature (Tg) of

es B: B

t[uut

cric(tthSacsmaMMMsaisedh

fht(ispelat

2

2

pLpS(CSDFe

2

mei

K. Fang et al. / Colloids and Surfac

he polymer, the drug release rate would be increased three times18,19]. Based on this theory, a variety of external physical stim-li strategies, such as near-infrared (NIR) light, high intensity focusltrasound (HIFU), ACMF and so on, were used to produce mildemperature to enhance the release rate [20–24].

PLGA microspheres (MS), efficiently loading hydrophilic drug,an be designed with defined size and shell permeability. Yang et al.eported that controlling the magnetic nanoparticle concentrationn the polymer film would result in different controlled drug releaseharacteristics [25]. Therefore, by incorporating MS with MNPsMMS), it opens up a new way to cancer therapy. By assemblinghe MNPs on the surface of MMS, under the ACMF, the increasedemperature around DOX-MMS is beneficial for both localizedyperthermia and enhanced shell permeability to the DOX release.ome reported strategies to design the enhanced synergistic ther-py are to encapsulate both MNPs and drugs within the aqueousore or to load drugs in the core and to embed MNPs in the polymerhell of MMS [26–28]. However, high MNPs in the polymer shellay interfere with the integrity of shell to cause non-specific leak-

ge of the encapsulated drug [29]. Recently, we fabricate a novelMS with MNPs coated surface, and find that the MNPs-coatedMS preserve better magnetic properties than MNPs-embeddedMS with the same MNPs concentration [30,31]. Moreover, in

ome cases MNPs on surfaces could crosslink to form higher hier-rchy aggregates, in which the localized magnetic hyperthermias conducted more efficiently. Thereby, the MNPs coated on theurface of MMS can directly interact with surrounding tissues tonhance the therapeutic effect. So we predict that the elaborateesign of the shell structure to optimize drug release with magneticyperthermia may be suitable for further applications.

To test this hypothesis, a magnetic responsive microsphere plat-orm was developed as combinational action of chemotherapy andyperthermia as shown in Fig. 1. The PLGA microspheres con-ained DOX drug solution in the core and the �-Fe2O3 nanoparticlesIOs) were adsorbed on the surface of microspheres by electrostaticncorporation. The aims of this work were to verify two hypothe-es: (1) the increase of local temperature at the vicinity of the IOsromoted permeation of the polymer shell when ACMF was appliedxternally, which enhanced drug diffusion rate and actuated a localarge-dosage drug release. (2) Optimum therapeutic effects can bechieved by the combination of chemotherapy and hyperthermiaherapy.

. Materials and methods

.1. Materials

Poly(lactic-co-glycolic acid) (50:50) (PLGA, MW = 10,000) wasurchased from Jinan Dai Gang biological technology Co.,td. (Jinan, China). Poly(vinyl alcohol) (PVA, MW = 31,000) andolyethylenimine (PEI, 25 kDa branched) were obtained fromigma–Aldrich (Shanghai, China). Doxorubicin hydrochlorideDOX, ≥98%) was purchased from Dalian Melone Pharmaceuticalo., Ltd. (Dalian, China). Methylene dichloride (DCM) was fromhanghai Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China).imercaptosuccinic Acid coated �-Fe2O3 nanoparticles (IOs, �-e2O3@DMSA) were prepared as previously described [32]. In thexperiments, all other materials were of analytical grade.

.2. The fabrication of DOX-MMS

DOX-loaded PLGA microspheres (DOX-MS) were prepared by aodified water-in-oil-in-water (W/O/W) double emulsion-solvent

vaporation method [26,33]. Briefly, PLGA (300 mg) were dissolvedn DCM (20 mL). The DOX solution (2 mg/mL) was emulsified into

iointerfaces 136 (2015) 712–720 713

the PLGA solution using a probe sonicator (Sonics & Materials,Newtown, CT, USA) at an amplitude setting of 40% for 100 s inan ice bath. Then, the primary W/O emulsion was poured intoa PVA solution (3%) and homogenized at 6000 rpm for 30 min inan ice bath by using a disperser T-25 digital ULTRA-TURRAX®

(IKA®, Germany). Afterwards, the produced W/O/W emulsion wastransferred into PVA solution (1%) with stirring overnight at roomtemperature to make almost all the DCM evaporated off and thecapsules harden. The solidified microspheres were collected bycentrifugation (3000 rpm for 30 min), and washed three times withdeionized (DI) water to remove residual and unencapsulated freeDOX.

The magnetic PLGA microspheres (DOX-MMS) were fabricatedby electrostatic deposition with the help of positively charged PEIas an interlayer. The DOX-MS suspension (5 mL) was added in PEIsolution (35 mL, 1.0 mg/mL) in a 50 mL vial, shaked and mixed forabout 2 h, and then centrifuged at 3000 × g for 30 min. The super-natant was discarded and excessive non-adsorbed PEI was washedthree times with DI water. The achieved microspheres suspension(35 mL) was mixed with 5 mL of IOs, and the mixture was incu-bated for 2 h. The resulting mixture was centrifuged for 30 min at3000 rpm to eliminate unbound IOs.

2.3. Characterization of DOX-MMS

The morphology of the microspheres was evaluated by usingfield-emission scanning electron microscopy (SEM, FEI Sirion-200,USA) and transmission electron microscope (TEM, JEOL, JEM-2000EX, Japan). The mean size distribution of the microparticleswas measured on a Microtrac S3500 particle size analyzer (USA).Suspensions of microspheres (1 × 105/mL in DI water) were used asanalyzed samples. Glass transition temperatures (Tg) of MMS wereperformed on a differential scanning calorimeter (DSC, NETZSCHSTA449C, Germany) from 10 ◦C to 300 ◦C at a rate of 10 ◦C/min undera constant flow of nitrogen gas. Total IOs content of the particles wasdetermined by thermo-gravimetric analysis (TGA) from 40 ◦C to700 ◦C at 10 ◦C/min under a constant flow of nitrogen gas. Addition-ally, the iron concentration in the aqueous dispersion of the MMSwas determined by the 1, 10-phenanthroline colorimetric method[34]. The magnetization properties were further investigated usinga vibrating sample magnetometer (VSM, LakeShore 7407, USA) inthe field H range of ±5000 Oe at room temperature.

The loading and encapsulation efficiency of DOX in themicrospheres were determined in triplicate using ultraviolet vis-ible (UV–vis) spectrophotometry. Briefly, freeze–dried DOX-MMS(5 mg) were completely dissolved in dimethyl sulfoxide (DMSO,10 mL) and then the supernatant was collected with centrifuga-tion and magnetic separation. The concentration of DOX in thesupernatant was evaluated using an UV–vis spectrophotometer(UV-3600, SHIMADZU, Japan) at 480 nm. Empty MMS were used asa blank test. The drug concentration was calculated from a standardcurve established previously. The drug loading and encapsula-tion efficiency were calculated according to the following equation[12,35]:

Encapsulation efficiency(%)

= the total DOX amount in the DOX-MMSthe gross weight of the feedingDOX

× 100%

Loading efficiency(%) = the total DOX amount in the DOX-MMSthe gross weight of the DOX-MMS

× 100%

714 K. Fang et al. / Colloids and Surfaces B: Biointerfaces 136 (2015) 712–720

ase ba

2

0Atrw3

F(

Fig. 1. Schematic demonstration of magnetic-induced drug rele

.4. Magnetic responsive drug release behavior

The moderate ratio frequency heating machine (Shuangpin SPG-6-III) (ACMF, 390 kHz) was applied to investigate the drug release.

fiber-optic thermometry device was introduced to measure the

emperature. In order to assess the magnetic responsive drugelease behavior, the DOX-MMS solution (5 mL, at [Fe] = 1.6 mg/mL)as subjected to alternating current magnetic fields (ACMF,

90 kHz). The various times (10, 20 and 30 min) were used for the

ig. 2. Preparation of PLGA hollow microspheres decorated with IOs (MMS) and their feB–D) and MMS (E–G).

sed magnetic PLGA hollow microspheres in response to ACMF.

ACMF treatment. The particles treated with ACMF were immedi-ately separated by centrifugation (12,000 rpm, 30 min at 4 ◦C). Theoptical imaging was obtained using Maestro In-Vivo Optical Imag-ing System (Caliper Life Sciences, MA) at the selected time points.The images were acquired and analyzed using Maestro 2.4 software

(Caliper Life Sciences, MA).To obtain the drug release profiles, theDOX-MMS solution treated with ACMF transferred to dialysis bags(molecular weight cutoff 8 kDa) and immersed in 30 mL of PBS at25 ◦C under moderate shaking. The buffer medium (1 mL) was taken

ature. (A) Schematic showing the preparation of MMS. SEM and TEM image of MS

K. Fang et al. / Colloids and Surfaces B: Biointerfaces 136 (2015) 712–720 715

F grapho , the i

owmmoj

2

Bom1ih9a

Fac

ig. 3. The characterization of MMS. (A) VSM curve of MMS, the inset shows photof MMS. (C) Time-dependent temperature increase curve of MMS exposed to ACMF

ut at the given time interval and an equal volume of fresh mediumas compensated. The release amount of DOX in each sample waseasured via UV–vis spectrophotometry at 480 nm. Then, the accu-ulative ratios of the released DOX were calculated as a function

f time. Control groups were performed under the same conditionust without ACMF treatment.

.5. In vitro chemo-thermal therapy experiments and cytotoxicity

The mouse breast cancer cell line 4T1 was provided by the Cellank of Shanghai Institutes of Biological Sciences, Chinese Academyf Sciences (Shanghai, China) and cultured in RPMI 1640 supple-ented with 10% fetal bovine serum, 100 �g/mL penicillin and

00 �g/mL streptomycin. The cells were maintained in a humid-

fied incubator at 37 ◦C with 5% CO2. For the in vitro magneticyperthermia treatment experiments, 4T1 cells were grown up to0% confluency in a glass sample vial. The medium was removednd replaced the culture medium with DOX-MMS dispersed in

ig. 4. (A) The supernatant DOX solutions released from DOX-MMS after ACMF treatmentnd fluorescence intensities of were quantified. (B) DOX release profiles of DOX-MMS werharacterized by SEM after 24 h incubation without (C) and with (D) ACMF.

s of MMS in the water before and after applying an external magnet. (B) TGA curvenset figure shows the DSC of MMS.

RPMI-1640 media. Then, the cells were exposed to a magnetichyperthermia setup at frequency of 390 kHz (Shuangpin SPG-06-III) for 10 min, 20 min and 30 min, respectively. The procedure wasalso performed for DOX-MMS with various DOX concentrations toevaluate their combined effect for cancer cell. In addition, 4T1 cellswere treated using MMS (without DOX) under ACMF exposure tocompare. Cells with PBS were used as a control. The 4T1 cells treatedwith ACMF were centrifuged and washed three times with PBS, fol-lowed by incubation for 24 h. The cell viability was measured by theMTT assay. The absorbance was measured by a microplate reader(model 680, Bio-RAD) at the wavelength of 490 nm. Results wereexpressed as the percentage of the metabolic activity of treatedcells relative to untreated cells.

The treated cells were stained with fluorescein diacetate (FDA)

and propidium iodide (PI) for 30 min to differentiate live/dead cell,which were observed using a Living Cell Time Lapse ObservationSystem (CellR̂, Olympus, Germany).

at different times treated was detected using a fluorescence optical imaging systeme described at the treated time of 0, 10, 20 and 30 min, respectively. DOX-MMS was

7 es B: B

2

immeSppcm

2

dtDtmtitaact0asw

T

R

wt

paAAsfiwtitt

wt

3

3

FtsR

16 K. Fang et al. / Colloids and Surfac

.6. Animal protocol

Female BALB/c nude mice were purchased from the Exper-mental Animal Center of Yangzhou University and were

aintained under specific pathogen-free (SPF) conditions. All ani-al procedures were performed in compliance with the animal

xperimentation guidelines of the Animal Research Ethics Board ofoutheast University. For solid tumor model, 4T1 cells (5 × 105 cellser mouse) were administered by subcutaneous injection into theroximal thigh region of the BALB/c nude mice. Experiments wereonducted after the tumor reached an average volume of approxi-ately 80 mm3.

.7. In vivo chemo-thermal therapy

To investigate the anti-tumor efficiency, the mice were ran-omly divided into four groups (n = 8 per group). Group 1: micereated with PBS buffer; Group 2: mice injected intratumorally withOX-MMS; Group 3: mice exposed to ACMF for 30 min after intra-

umorally administration with MMS (containing no DOX). Group 4:ice exposed to ACMF for 30 min after intratumorally administra-

ion with DOX-MMS. The Group 2 and Group 4 were intratumorallynjected with single dose at the beginning (day 0). The adminis-rated concentrations were consistent in terms of the total DOXnd Fe content (100 �L: 2.8 mg/mL of IOs, 3.0 mg/kg of DOX). Thenesthetized mice (Group 3 and Group 4) were placed into a water-ooled magnetic induction coil in which ACMF was applied to theumor region for 30 min after injection (390 kHz, Shuangpin SPG-6-III). Each mouse was injected once during the experiment. Afterdministration, tumor growth was monitored every other day. Theize of the tumor was measured by a caliper. The tumor volumeas calculated by using Eq. (1):

umor volume = (Tumor length) × (Tumor width)2

2(1)

Relative tumor volumes were calculated as Eq. (2):

elative tumor volumes = Vt

Vc(2)

here Vc was the mean tumor volume of the control group, Vt washe mean tumor volumes of the treated groups.

Mice were sacrificed by cervical vertebra dislocation at 7th dayost-injection. Subsequently, tumors in each group were immedi-tely excised and weighed, followed by fixed, embedded in paraffin.fter being embedded, ultra-thin sections (50 nm) were obtained.poptotic cells in tumor sections were visualized by the TUNELtaining according to the manufacturer’s protocol. Five equal-sizedelds were randomly chosen, and the percentage of apoptotic cellsas calculated as a ratio of the apoptotic cell number to the total

umor cell number in each field. The micro-vessel density (MVD)n the tumor sections was quantified using CD31 staining methodo evaluate vascular change around tumor tissue after differentreatment.

All data are expressed as mean ± SD. Statistical analysis of dataere done using Student’s t-test, and P < 0.05 was considered sta-

istical significance.

. Results and discussion

.1. Preparation and characterization of DOX-MMS

The technical route of MMS preparation is demonstrated in

ig. 2A. The PLGA hollow microspheres (MS) without IOs modifica-ion exhibited a smooth and spherical morphology (Fig. 2B), and theurface morphology of the IOs modified MMS was rough (Fig. 2E).epresentative SEM images revealed that the microsphere surface

iointerfaces 136 (2015) 712–720

was successfully decorated by the IOs. TEM images of MS and MMSare shown in Fig. 2C and F, which revealed the hollow structure.In Fig. 2G, the IOs shell was also clearly observed on surface of theMMS. Beneficial for the hollow structure, the quantities of encap-sulated DOX in DOX-MMS were 6.2%, and the drug entrapmentefficiency was (85.1 ± 2.1)%. The average sizes were measured tobe 2.2 �m and 2.4 �m for blank MS and DOX-MMs, respectively(Fig. S1).

The zeta potential of MS increased from −26.0 ± 4.5 mV (neg-ative) to 49.0 ± 6.3 mV (positive) after the addition of PEI, whichfurther demonstrated the adsorption of the IOs onto the surface ofthe MMS during the preparation process of MMS (Fig. S2). Our pre-vious report indicated the IOs, functionalized with DMSA, exhibiteda high negative charge [32]. When the IOs were adsorbed onto thesurface of the PEI-coated MS, zeta potential is −21.0 ± 4.8 mV, indi-cating the successful deposition of PEI and IOs onto the surface ofMS templates.

3.2. Magnetic characteristic of DOX-MMS

The results of magnetic performance of MMS showed that therewas no remanent magnetization observed in this curve and the sat-uration magnetization was 11.2 emu/g (Fig. 3A). MMS could rapidlyaggregate toward the side of the cuvette nearest to the magnet, asdisplayed in inset Fig. 3A. The rate of weight loss was augmentedwith the increase of temperature (Fig. 3B). Due to the completethermo-decomposition of the organic polymer, the lost weight rateof MMS at 600 ◦C was 71.7%, left with a 28.3% residual of IOs.

The magneto-thermal effect was measured to evaluate thehyperthermia potentiality of the platforms under an ACMF at thefrequency of 390 KHz (Fig. 3C). The MMS solution showed a rapidtemperature rise with the increase of irradiation time. The sam-ples can be easily heated to above 42 ◦C after 15 min, which issufficiently higher than the Tg of MMS (40.9 ◦C) (Fig. 3C inset).The phase transition from the glassy state to the rubbery statecan provide a significant increase of the molecular motion. Fur-thermore, the curves also indicate the corresponding temperatureshave increased up to roughly 45 ◦C after 30 min. Such temperaturecould lead to an irreversible damage to tumor cells and ultimatelydeath.

3.3. Magnetic triggering drug release

The magnetic responsive release of DOX-MMS was observedunder ACMF at different time. The supernatant solutions containedDOX released from DOX-MMS were collected by centrifugationafter ACMF treatment, and the magnetic stimulation effect of DOX-MMS on the fluorescence change of DOX was observed by usingthe commercially available optical imaging system (Fig. 4A). Aquantitative investigation based on fluorescence intensity was alsoconducted by measuring the fluorescence intensity of the opti-cal image. The results indicated that the fluorescence intensityincreases with prolonged exposure time under ACMF. After expos-ing to ACMF for 30 min, the fluorescence intensity was 6 timeshigher than that not exposed to ACMF.

Fig. 4B shows the characteristics of the DOX release from DOX-MMS with or without ACMF treatment. The DOX-MMS displayedlow drug release because the IOs coating layer restricted outwardDOX diffusion. The DOX only released less than 11.4% in PBS buffer,and no burst release was observed. Furthermore, the accumulativerelease of DOX from MS was about 18.5% and slightly higher thanthat decorated with IOs in PBS buffer (Fig. S3). It is interesting to

find that the cumulative release of DOX from DOX-MMS was signif-icantly faster with ACMF than without ACMF. It was observed thatthe release profile of DOX-MMS after a shorted initial burst releasefollowed by a slower diffusion release. The magnetic responsive

K. Fang et al. / Colloids and Surfaces B: Biointerfaces 136 (2015) 712–720 717

Fig. 5. (A) Fluorescence images of 4T1 cells co-stained with FDA (green, live cells) and PI (red, dead cells) treated with PBS (A), DOX-MMS (B) and DOX-MMS + ACMF (C),r CMF foe e refera

rle(3f1Aawooawoce

dmtroshtwresmr

3

c

espectively. (D) The viability of 4T1 cells exposure to DOX-MMS with or without Axposure to DOX-MMS exposed to ACMF for various times. (For interpretation of thrticle.)

elease characteristic would depend on exposed time. The cumu-ative drug release from DOX-MMS was increased from 2.8% (notxposed to ACMF) to 8.7% (exposed to ACMF for 10 min), 11.7%exposed to ACMF for 20 min) and 21.6% (exposed to ACMF for0 min), respectively. After 24 h, the cumulative release increasedrom 11.4% (not exposed to ACMF) to 28.1% (exposed to ACMF for0 min), 49.5% (exposed to ACMF for 20 min) and 79.5% (exposed toCMF for 30 min), respectively. The different release process withnd without ACMF is related with the shell surface permeability,hich demonstrated by SEM results in Fig. 4C and D. SEM images

f the DOX-MMS exposed to ACMF for 30 min showed the presencef broken particles with pores on the surface (Fig. 4D), and the leak-ge of IOs was also observed, which suggested that the DOX-MMSere destroyed after ACMF treatment. In contrast, the morphol-

gy of DOX-MMS without exposure to ACMF did not significantlyhange and remained relatively structural integrity throughout thentire study (Fig. 4C).

Two possible factors may determine the magnetic-responsiverug release behavior on the PLGA shell of DOX-MMS. One is theagnetically induced heat energy on the shell of the DOX-MMS due

o magnetic energy dissipation from IOs through Néel and Brownelaxation under an ACMF. It is believed that the temperature risef the IOs coating layer was initially much larger than that of theurrounding [36,37]. Such localized and focused intensive heat, orot spots, caused a destruction of the PLGA shell. The second ishe increased mobility of the polymer chains and drug molecules,hen the PLGA underwent a thermal transition from a glassy to a

ubbery state at elevated temperature (T > Tg). Such a state of highernergy would render them more susceptible to increased chaincission and result in increased PLGA hydrolysis [18,19]. These twoechanisms were found to work synergistically to manipulate drug

elease through tuning ACMF treatment.

.4. In vitro chemo-thermal therapy of DOX-MMS

We investigated the possibility of utilizing DOX-MMS for in vitrohemo-thermal therapy. Fig. 5A–C shows the resultant fluorescent

r as a function of DOX concentration after 24 h incubation. (E) The viability of cellsences to color in this figure legend, the reader is referred to the web version of this

microscopy images of cells treatment with PBS buffer, DOX-MMS(DOX, 20 �g/mL) and DOX-MMS + ACMF (DOX, 20 �g/mL, treatedfor 30 min) treatment. To directly observe the magnetic hyper-thermia efficacy, the treated cells were co-stained with FDA andPI to differentiate the live (green) and dead (red) cells. It was foundthat all the cells incubated with DOX-MMS were almost scaldedto death after exposure to ACMF, with the evidence that only redfluorescence emission was present (Fig. 5C). In contrast, the redfluorescence emission was shown by the merge presence of DOX-MMS, and the cell viability was almost not influenced due to thelow doxorubicin dose released (Fig. 5B).

For more quantitative evaluation of cytotoxicity, 4T1 cells wereincubated with different concentrations of DOX-MMS appliedACMF for 30 min. They were incubated for 24 h after treated byACMF, and then their viability was measured using MTT assay(Fig. 5D). There was a significant difference of the cell-killing effectbetween DOX-MMS + ACMF and DOX-MMS. The cell viability wasslightly reduced with the increase of the DOX concentration, indi-cating excellent biocompatibility and low toxicity of DOX-MMSwithout ACMF (Fig. 5D), which attributed to the protective effect ofIOs coating layer outside of DOX-MMS, delaying DOX release fromDOX-MMS. When 4T1 cells were incubated with DOX-MMS (DOX,20 �g/mL), 81.6% of the cells were viable. However, when the cellswere treated with DOX-MMS (DOX, 20 �g/mL) exposed to ACMFfor 30 min, the cell viability was reduced to 10%, as evidenced bythe fact that MMS combined with ACMF are able to enhance therelease of DOX and increase cytotoxicity of DOX. Besides, the ACMFtreatment time also play an important role in inducing heat anddrug release. Fig. 5E showed the effect of ACMF treatment time onthe viability of 4T1 cells treated with DOX-MMS. With the increaseof time exposed to ACMF, the cell viability was decreased. Afterexposed to ACMF from 10 to 30 min, the viability of 4T1 cells treatedwith DOX-MMS was decreased from 60% to 10%. When DOX-MMS

were exposed to ACMF, the induced local hyperthermia promotedpermeation of the PLGA polymer shell, which increased the drugdiffusion rate and actuated a local large drug concentration. Thelocal hyperthermia further enhanced cytotoxicity of DOX.

718 K. Fang et al. / Colloids and Surfaces B: Biointerfaces 136 (2015) 712–720

Fig. 6. (A) 4T1 tumor treated with PBS (left), MMS (containing no DOX) with ACMF (middle) and DOX-MMS with ACMF (right) after 7 days. (B) Tumor therapy using MMSwith or without DOX and with or without ACMF treatment for 30 min. (C) the typical photographs of excised tumors from mice on the 7 day. (D) Mouse mass as a functionof days post treatment.

Fig. 7. (A) Tumor tissue sections in each group staining with H&E (Upper), TUNEL assay (Middle), CD31 immunohistochemical staining (Underneath). (B) The apoptotic indexin each group. (C) Quantitative assay of the microvessel density (MVD) in each group.

es B: B

3

fmatsgamrsM(ottatttasmme

3

tttsaMt

wathetieGTtGcAmptlM

vetius4

[[

[[

[[

[

K. Fang et al. / Colloids and Surfac

.5. In vivo chemo-thermal therapy of DOX-MMS

The in vivo antitumor activity of DOX-MMS under ACMF wasurther demonstrated using the 4T1 tumor model on BALB/c nude

ice. 7 days after treatment, both MMS (containing no DOX)nd DOX-MMS treated with ACMF showed treatment effect athe injected site with a dark gray color, and severe necrotic tis-ue also occurred (Fig. 6A). The relative tumor volume for eachroup was assessed for the therapeutic effect (Fig. 6B). There was

significant difference in tumor growth rates among the treat-ent groups over time. The tumors in Group 1 (control group)

apidly increased by 6.2 ± 0.8-fold. The Group 3 (MMS + ACMF)howed better antitumor effect compared with Group 2 (DOX-MS) (3.4 ± 0.6 versus 4.0 ± 0.8-folds). Remarkably, the Group 4

DOX-MMS + ACMF) showed the highest inhibition compared tother groups (3.4 ± 0.6 folds), which indicated that the fusionreatment led to a significant benefit, which was consistent withhe in vitro therapeutic efficiency. It was noteworthy that a largemount of DOX-MMS still remained within the tumor tissue afterhe hyperthermia treatment (Fig. 6C). These results demonstratedhat the combination of DOX-MMS and hyperthermia could effec-ively kill cancer cells in vivo, this synergistic effect was probablyscribed to hyperthermia enhanced cytotoxicity of DOX. At theame time, body weight loss was not significantly observed in theice among all the groups (Fig. 6D), which indicated that the treat-ent with a single-dose did not result in significant acute side

ffect.

.6. Histological assay

Histological and immunohistochemical examinations of tumorissues from each group were performed to further assess the anti-umor effect of the MMS. Hematoxylin and eosin (H&E) staining ofumor slices were performed (Fig. 7). A small amount of necro-is was shown in the tumor treated with DOX-MMS (Group 2)nd MMS + ACMF (Group 3). However, tumors treated with DOX-MS + ACMF (Group 4) had visible necrosis compared to the control

umors (Group 1) (Fig. 7A).The effect of DOX-MMS on apoptosis in the tumor tissues

as evidently confirmed by TUNEL staining, in which necrotic orpoptotic cells were stained brown. It was found that the apop-otic cancer cells of Group 4 (DOX-MMS + ACMF) were significantlyigher than that of the other treatment groups. The quantitativevaluation of apoptosis cells in 5 random fields from 5 differentumors in each group was counted manually, and the apoptoticndex is shown in Fig. 7B. Control tumors (Group 1) showed the low-st degree of apoptosis (6.08%). Tumors in Group 2 (DOX-MMS) androup 3 (MMS + ACMF) had some apoptotic cells (14.8% and 39.2%).he tumor in Group 4 (DOX-MMS + ACMF) significantly showedhe most apoptotic cells (65.4%), which was 4.4-fold higher thanroup 2. Angiogenesis was estimated by the immunohistochemi-al anti-CD31 staining of the tumor tissue from mice in each group.ngiogenesis plays an important role in both tumor growth andetastasis, and intratumor MVD has been accepted as an inde-

endent prognostic factor. As shown in Fig. 7C, the MVD of tumorissues from Group 4 (DOX-MMS + ACMF, 11.3) was significantlyower compared to that of, Group 1 (Control, 45.3), Group 2 (DOX-

MS, 36.3) or Group 3 (MMS + ACMF, 23.3).The above results further demonstrated that DOX-MMS acti-

ated by ACMF could provide a significantly enhanced therapeuticfficacy in tumor. Two main factors may be involved to explainhe mechanism of the highly efficient therapeutic effect. One factor

s the localized magnetic hyperthermia treatment with DOX-MMSnder an ACMF. In the presence of ACMF, the IOs coated shell canerve as hyperthermia agents. When magnetic heating to above2 ◦C, the localized magnetic hyperthermia causes tumor cell apo-

[

[[

iointerfaces 136 (2015) 712–720 719

ptosis because tumor cells are more susceptible to hyperthermia[3]. Another factor is the triggered DOX release from DOX-MMSby magnetic hyperthermia. The accelerated drug release increasedthe intracellular concentration of DOX within tumor tissues dur-ing the heating process and enhanced cytotoxicity [4]. In addition,the DOX-MMS still remained within the tumor tissues after thehyperthermia treatment. The DOX released from DOX-MMS canpenetrate into the tissue and kill the remaining live tumor cells,which reduced angiogenesis in the tumor. All these contributed toincreased apoptosis and reduced angiogenesis, ultimately resultingin tumor growth inhibition.

4. Conclusions

In summary, we have successfully demonstrated a magneticresponsive drug delivery system based on PLGA hollow micro-spheres, which decorated with IOs and encapsulated doxorubicin inthe core. When exposed to ACMF for 30 min, the localized heat pro-duced by IOs through Néel and Brown processes, promoted the poreformation in PLGA shell, which enabled to enhance drug releasefrom DOX-MMS approximately 6 times. The combination of DOX-MMS and ACMF demonstrated excellent efficiency of tumor growthinhibition in vivo. By taking advantage of this magnetic-induceddrug release behavior, this platform can be applied as a synergisticplatform for chemo-thermal therapy.

Acknowledgments

This work was financially supported by National Basic ResearchProgram of China (No. 2011CB933503, 2013CB733804), NationalNatural Science Foundation of China (No. 31370019, 61127002),National Natural Science Foundation of China for Key Projectof International Cooperation (61420106012). Foundation for theAuthor of National Excellent Doctoral Dissertation of PR China (No.201259), the Fundamental Research Funds for the Central Univer-sities.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.colsurfb.2015.10.014.

References

[1] C.H. Heldin, K. Rubin, K. Pietras, Nat. Rev. Cancer 4 (2004) 806–813.[2] R.-C. Ji, Cancer Lett. 346 (2014) 6–16.[3] M. Torres-Lugo, C. Rinaldi, Nanomedicine 8 (2013) 1689–1707.[4] R.D. Issels, Eur. J. Cancer 44 (2008) 2546–2554.[5] J. Jaber, E. Mohsen, Colloid. Surf. B 102 (2013) 265–272.[6] D. Ramimoghadam, S. Bagheri, S.B. Abd Hamid, Colloid Surf. B 133 (2015)

388–411.[7] J.H. Young, M.T. Wang, I.A. Brezovich, Electron Lett. 16 (1980) 358–359.[8] R. Hergt, S. Dutz, J. Magn. Magn. Mater. 311 (2007) 187–192.[9] K. Hayashi, M. Nakamura, H. Miki, S. Ozaki, M. Abe, T. Matsumoto, W.

Sakamoto, T. Yogo, K. Ishimura, Theranostics 4 (2014) 834–844.10] M.S. Shive, J.M. Anderson, Adv. Drug Deliv. Rev. 28 (1997) 5–24.11] F. Danhier, E. Ansorena, J.M. Silva, R. Coco, A. Le Breton, V. Preat, J. Control.

Release 161 (2012) 505–522.12] F. Li, J. Sun, H. Zhu, X. Wen, C. Lin, D. Shi, Colloid Surf. B 88 (2011) 58–62.13] A. Giteau, M.C. Venier-Julienne, A. Aubert-Pouessel, J.P. Benoit, Int. J. Pharm.

350 (2008) 14–26.14] A.N.F. Versypt, D.W. Pack, R.D. Braatz, J. Control. Release 165 (2013) 29–37.15] S. Fredenberg, M. Wahlgren, M. Reslow, A. Axelsson, Int. J. Pharm. 415 (2011)

34–52.16] K.T. Oh, H.J. Baik, A.H. Lee, Y.T. Oh, Y.S. Youn, E.S. Lee, Int. J. Mol. Sci. 10 (2009)

3776–3792.

17] L. Solorio, A.M. Olear, J.I. Hamilton, R.B. Patel, A.C. Beiswenger, J.E. Wallace,

H.Y. Zhou, A.A. Exner, Theranostics 2 (2012) 1064–1077.18] B.S. Zolnik, P.E. Leary, D.J. Burgess, J. Control. Release 112 (2006) 293–300.19] N. Faisant, J. Akiki, F. Siepmann, J.P. Benoit, J. Siepmann, Int. J. Pharm. 314

(2006) 189–197.

7 es B: B

[

[

[

[

[

[

[

[

[

[

[

[

[

[[[

W. Xu, L. Zhang, Q. Guo, Int. J. Pharm. 7 (2012) 1697–1708.

20 K. Fang et al. / Colloids and Surfac

20] M. Mathiyazhakan, Y. Yang, Y. Liu, C. Zhu, Q. Liu, C.-D. Ohl, K.C. Tam, Y. Gao, C.Xu, Colloid Surf. B 126 (2015) 569–574.

21] C.C. Niu, Z.G. Wang, G.M. Lu, T.M. Krupka, Y. Sun, Y.F. You, W.X. Song, H.T. Ran,P. Li, Y.Y. Zheng, Biomaterials 34 (2013) 2307–2317.

22] H. Park, J. Yang, J. Lee, S. Haam, I.H. Choi, K.H. Yoo, ACS Nano 3 (2009)2919–2926.

23] J. Majeed, L. Pradhan, R.S. Ningthoujam, R.K. Vatsa, D. Bahadur, A.K. Tyagi,Colloid Surf. B 122 (2014) 396–403.

24] S.D. Kong, M. Sartor, C.M. Hu, W. Zhang, L. Zhang, S. Jin, Acta Biomater. 9(2013) 5447–5452.

25] F. Yang, X.X. Zhang, L.N. Song, H.T. Cui, J.N. Myers, T.T. Bai, Y. Zhou, Z. Chen, N.Gu, ACS Appl. Mater. Interface 7 (2015) 9410–9419.

26] F. Yang, P. Chen, W. He, N. Gu, X.Z. Zhang, K. Fang, Y. Zhang, J.F. Sun, J.Y. Tong,Small 6 (2010) 1300–1305.

27] W.L. Chiang, C.J. Ke, Z.X. Liao, S.Y. Chen, F.R. Chen, C.Y. Tsai, Y.N. Xia, H.W.Sung, Small 8 (2012) 3584–3588.

28] E. Amstad, E. Reimhult, Nanomedicine (Uk) 7 (2012) 145–164.

[

[

iointerfaces 136 (2015) 712–720

29] F. Yang, Y.X. Li, Z.P. Chen, Y. Zhang, J.R. Wu, N. Gu, Biomaterials 30 (2009)3882–3890.

30] W. He, F. Yang, Y.H. Wu, S. Wen, P. Chen, Y. Zhang, N. Gu, Mater. Lett. 68(2012) 64–67.

31] L. Duan, F. Yang, L. Song, K. Fang, J. Tian, Y. Liang, M. Li, N. Xu, Z. Chen, Y.Zhang, N. Gu, Soft Matter 11 (2015) 5492–5500.

32] S. Zhang, Z.P. Bian, C.R. Gu, Y. Zhang, S.Y. He, N. Gu, J.N. Zhang, Colloid Surf. B55 (2007) 143–148.

33] K. Fang, F. Yang, Q.Y. Zhang, T.Z. Zhang, N. Gu, Mater. Lett. 117 (2014) 86–89.34] A.E. Harvey, J.A. Smart, E.S. Amis, Anal. Chem. 27 (1955) 26–29.35] Y. Jia, M. Yuan, H. Yuan, X. Huang, X. Sui, X. Cui, F. Tang, J. Peng, J. Chen, S. Lu,

36] S.H. Hu, Y.Y. Chen, T.C. Liu, T.H. Tung, D.M. Liu, S.Y. Chen, Chem. Commun. 47(2011) 1776–1778.

37] H.Y. Huang, S.H. Hu, C.S. Chian, S.Y. Chen, H.Y. Lai, Y.Y. Chen, J. Mater. Chem.22 (2012) 8566–8573.