ot

an

ghaolecnghay, Sh

Article history:Received 20 August 2014Accepted 19 October 2014

tubes and graphene [10,11], all of which show strong opticalabsorbance in the near-infrared (NIR) tissue optical transparencywindow. The previous results demonstrated that photoabsorber-based therapy might be a promising approach for cancer therapy,

d enhance survivalby photothermalene, and so on), isitably limit futuredable or slowlyin bodily organs,matory cytokinenicant to explorethe photothermal

Fe3O4 NPs) havecellent magnetic,

biocompatible and potentially non-toxic properties [19,20]. Fe3O4NPs, which can be manipulated and controlled by an externalmagnetic eld, are additional important materials and have beenemployed in many areas, including biology, pharmaceuticals anddiagnostics. Moreover, iron is a nutrient and readily metabolized bycellular regulation using the transferrin pathway. Thus, Fe3O4 NPsare easily degradable and passes in and out of cells across theplasma membrane [21]. For the acknowledged advantages of Fe3O4NPs, they are extremely suitable for in vivo applications. In addition,

* Corresponding author. Tel.: 86 21 65642385.E-mail address: wlyang@fudan.edu.cn (W. Yang).

1

Contents lists availab

Biomat

ev

Biomaterials 39 (2015) 67e74Both authors contributed equally to this work.1. Introduction

Nanomaterials have been extensively used in biomedicalresearch during the past twenty years [1e4]. Especially in recentyears, more and more studies focused on photo-absorbing nano-agents, and photothermal therapy (PTT) employing photoabsorberscan convert optical energy into thermal energy to kill cancer cellswithout affecting healthy tissues [5,6]. Compared with radio-therapy, chemotherapy and surgical management, PTT is lessinvasive, controllable and highly efcient. Our and other researchgroups have developed a large number of nanomaterials as PTTagents, such as gold-based nanomaterials [5,7e9], carbon nano-

which could effectively reduce tumor growth an[12e15]. Even so, the potential toxicity inducedagents (especially for carbon nanotubes or graphstill an unresolved debate [16,17], which will inevclinic applications of PTT. These non-degradegradable nanoparticles easily accumulateresulting in increased oxidative stress, inamproduction and cell death [18]. Therefore, it is siga bio-safety and biodegradable photoabsorber forablation of cancer with NIR irradiation.

Magnetic iron oxides nanoparticles (namelyreceived tremendous attention for their exAvailable online 15 November 2014

Keywords:Magnetic nanoparticle clustersFe3O4TumorPhotothermal therapyhttp://dx.doi.org/10.1016/j.biomaterials.2014.10.0640142-9612/ 2014 Elsevier Ltd. All rights reserved.a b s t r a c t

In this study, the photothermal effect of magnetic nanoparticle clusters was rstly reported for thephotothermal ablation of tumors both in vitro in cellular systems but also in vivo study. Compared withindividual magnetic Fe3O4 nanoparticles (NPs), clustered Fe3O4 NPs can result in a signicant increase inthe near-infrared (NIR) absorption. Upon NIR irradiation at 808 nm, clustered Fe3O4 NPs inducing highertemperature were more cytotoxic against A549 cells than individual Fe3O4 NPs. We then performedin vivo photothermal therapy (PTT) studies and observed a promising tumor treatment. Compared withPBS and individual magnetic Fe3O4 NPs by NIR irradiation, the clustered Fe3O4 NPs treatment showed ahigher therapeutic efcacy. The treatment effects of clustered Fe3O4 NPs with different time of NIRillumination were also evaluated. The result indicated that a sustained high temperature generated byNIR laser with long irradiation time was more effective in killing tumor cells. Furthermore, histologicalanalysis of H&E staining and TUNEL immunohistological assay were further employed for antitumorefcacy assessment of PTT against A549 tumors.

2014 Elsevier Ltd. All rights reserved.a r t i c l e i n f oMagnetic nanoparticle clusters for photnear-infrared irradiation

Shun Shen a, 1, Sheng Wang a, c, 1, Rui Zheng b, XiaoyWuli Yang b, *

a School of Pharmacy & Key Laboratory of Smart Drug Delivery, Fudan University, Shanb State Key Laboratory of Molecular Engineering of Polymers & Department of Macromc Pancreatic Disease Institute, Department of Pancreatic Surgery, Huashan Hospital, Shad Department of Integrative Oncology, Shanghai Cancer Center, Department of OncologShanghai 200032, China

journal homepage: www.elshermal therapy with

Zhu d, Xinguo Jiang a, Deliang Fu c,

i 201203, Chinaular Science, Fudan University, Shanghai 200433, Chinai Medical College, Fudan University, Shanghai 200040, Chinaanghai Medical University, Fudan University,

le at ScienceDirect

erials

ier .com/locate/biomater ia ls

teriaFe3O4 NPs are exceptional materials for hyperthermia treatment oftumors [22,23]. Under an alternating magnetic eld (AMF), mag-netic hyperthermia of Fe3O4 NPs is produced via dipole relaxation,which can be employed to destroy tumor cells because these cellsare more sensitive to temperatures in excess of ca. 41 C than theirnormal counterparts [24]. Although magnetic hyperthermia hasbeen clinically used [25e28], the technique requires high currentand voltage due to large air volume within the applied eld inwhich energy cannot be easily focused [29].

Very recently, NIR light induced photothermal effect for Fe3O4NPs has been studied and it exhibits good photothermal convertingefciency. For example, Chu et al. applied individual Fe3O4 NPs withhigh concentration for the photothermal ablation of cancer by NIRlaser irradiation [29]. Considering extra-high-dose magnetitemight generate potential toxicity to the body, the number of usableelements should also be severely limited. Thus, Fe3O4 NPs used forPTT must be improved, mainly involving two routes. One is tomodify NIR light-absorbing materials onto the surface of magneticNPs, another is to concentrate the magnetic NPs. Newly, it has beenshown that the clustering of magnetic NPs induces a signicantincrease in the magnetic moment, and consequently, the clusteredmagnetic NPs have a much higher relatively high saturationmagnetization and specic absorption rate (SAR) than individualmagnetic NPs [30e33]. Hayashi et al. have utilized the uniquemeritof magnetite clusters for the magnetic hyperthermia therapy oftumors by AMF [32,33]. More signicantly, previous studies havealso indicated that metallic NPs clusters can induce a red-shift inthe light absorption spectra [34,35], which enhances the lightabsorbance in the NIR region, expanding new application elds formetallic NPs clusters to be utilized as photosensitizers in the NIRPTT. Therefore, numerous papers have been reported the photo-thermal ablation of tumors by utilizing aggregation-inducedenhanced photothermo (AIEP) of metallic NPs clusters [36,37],such as gold NPs aggregation. Until now, to the best of ourknowledge, the papers on the study of photothermal effect ofmagnetite clusters were rarely reported, let alone for the use ofmagnetite clusters with NIR irradiation to destroy tumor in vivo.

Herein, we present the synthesis of magnetic colloidal nano-crystal clusters that effectively produce heat in response to harm-less NIR irradiation. The biomedical parameters were not onlyassessed in vitro in cellular systems but also in vivo study. The re-sults proved that the photothermal effects of clustered magneticNPs for the treatment of A549 (Human lung adenocarcinomaepithelial cell line) tumor were better than those obtained by singlecrystal magnetic NPs.

2. Materials and methods

2.1. Materials

Iron (III) chloride (FeCl3$6H2O), iron (II) chloride (FeCl2$4H2O), trisodium citratedehydrate (C8H5Na3O7$2H2O), sodium acetate anhydrous (NaOAc), ethylene glycol(EG), acetone, sodium hydroxide (NaOH), and nitric acid (HNO3) were purchasedfrom Shanghai Chemical Reagents Company (China). MTT (3-(4, 5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay, and other biological reagents werepurchased from Invitrogen Corp. Dulbecco's modied Eagle medium (DMEM), Fetalbovine serum (FBS), Penicillin-Streptomycin solution and Trypsin-EDTA solutionwere purchased from Gibco (Tulsa, OK, USA). All the other chemicals were ofanalytical grade, and puried water was produced by a Millipore water puricationsystem.

2.2. Synthesis of individual magnetic NPs

The individualmagnetic NPs were prepared by chemical coprecipitationmethod[38]. In a typical recipe, FeCl3$6H2O (10.81 g, 0.04 mol) was dissolved in 50 mLdeionized water. The resulting solution was transferred into a three-necked ask(250 mL), which was equipped with a mechanical stirrer, a dropping funnel, and anitrogen inlet. Then 3.98 g of FeCl2$4H2O (0.02 mol) was also dissolved in another

S. Shen et al. / Bioma6850 mL deionized water, and the resulting solution was transferred into the above-mentioned three-necked ask. The as-prepared mixture was then stirred at roomtemperature in nitrogen atmosphere, and NaOH solution (10 M, 50 mL) was thenslowly dropped into the ask from the dropping funnel in 1 h. After that, themixturewas continuously stirred at room temperature for 1 h, which was then stirred at90 C for another 2 h and cooled to room temperature under continuous stirring. Thewhole process was in the nitrogen atmosphere. The products were washed withdeionized water three times, collected with the help of a magnet. After that, nitricacid solution (2 M, 100 mL) was added into the above products, which were thenwashed with deionized water several times and collected with the magnet, until thepH value of the supernatant is neutral. Then 100 mL (0.3 M) of trisodium citrate wasadded into the resulting products, the mixture was stirred for 30 min at 90 C in thenitrogen atmosphere. Then it was cooled to room temperature, transferred into abeaker and collected with the magnet, and 20 mL of deionized water and a lot ofacetonewere added into the beaker towash the products twice. Finally, the obtainedproducts were dispersed in 50 mL of deionized water, and stirred at 80 C for a longtime to remove acetone.

2.3. Synthesis of clustered magnetic NPs

The clustered magnetic NPs (magnetic particles) were prepared by a modiedsolvothermal reaction [39]. Briey, FeCl3$6H2O (1.028 g), C8H5Na3O7$2H2O(0.24 g), and NaOAc (1.2 g) were rst dissolved in 20 mL of EG under vigorousstirring for 0.5 h. The resulting solution was then transferred into a Teon-linedstainless-steel autoclave with a capacity of 50 mL. The autoclave was sealed andheated at 200 C for 10 h. Then it was cooled to room temperature. The as-prepared black products were washed with ethanol and deionized water severaltimes, and collected with a magnet. The nal products were dispersed in 10 mL ofethanol for the further use.

2.4. Characterization

Ultravioletevisible (UVevis) spectra were performed using a PerkineElmerLambda 750 spectrophotometer. Transmission electron microscopy (TEM) imageswere obtained on a Tecnai G2 20 TWIN transmission electron microscope. Magneticcharacterization was carried out with a vibrating sample magnetometer (VSM) on aModel 6000 physical propertymeasurement system (Quantum, USA) at 300 K. X-raydiffraction (XRD) measurements were recorded on a X'pert PRO diffractometer todetermine the composition of Fe3O4 particles. All the diffraction peaks in the XRDpatterns were indexed and assigned to the typical cubic structure of Fe3O4 (JCPDS75-1609).

2.5. Cell lines and culture conditions

A549 cells obtained from Chinese Academy of Sciences Cells Bank, Shanghai,China, were routinely cultured in RPMI-1640 cell medium supplemented with 10%FBS, 100 U/mL penicillin and 100 mg/mL streptomycin, at 37 C in 5% CO2 and 95% airatmosphere with >95% humidity. All experiments were performed on cells in thelogarithmic phase of growth.

2.6. NIR-heating effect of magnetite NPs in solution

The individual magnetic NPs or clustering of magnetic NPs stock solution at1 mg/mL was diluted to the different concentrations (10e80 mg/mL) and 200 mL ofaliquots were deposited into wells of a 48-well cell culture plate. Wells were illu-minated by an 808-nm continuous-wave NIR laser (Changchun New IndustriesOptoelectronics Technology, Changchun, China; uence: 5 W/cm2, spot size: 5 mm)with the different exposure time from 60 to 180 s. Pre- and postillumination tem-peratures were taken by thermocouple.

2.7. Cell viability assay

A549 cells were seeded in 96-well plates with a density of 104 cells/well, andallowed to adhere for 24 h prior to assay. The cells were exposed to the clusteredmagnetic NPs or individual magnetic NPs with the same concentration of 50 mg/mL,respectively. The cells were or were not irradiated by NIR laser light at a powerdensity of 5 W/cm2 with different illumination time from 60 to 180 s. After the cellswere incubated at 37 C for another 24 h, they were incubated with 0.5 mg/mL MTTin DMEM for 4 h in dark and then mixed with dimethyl sulfoxide after the super-natant was removed. The OD value at 570 nmwas read using the microplate reader(Synergy TM2, BIO-TEK Instruments Inc. USA). Cell viability was determined by thepercentage of OD value of the study group over the control group. Afterward, cellswere co-stained by a mixture of Calcein AM and PI solution for 20 min. The sampleswere subjected to observe using a ZEISS LSM710 live cell confocal laser imagingSystem (Carl Zeiss, German).

2.8. Flow cytometry analysis

To investigate the photothermal ablation of A549 cells by clustered Fe3O4 NPswithout or with the 808 nm laser irradiation using ow cytometry, clusteredFe3O4 NPs dispersion (100 mg/mL in PBS solution) was added to a 12-well cellculture plate containing A549 cells with a density of 5 105 cells/well in RPMI-

ls 39 (2015) 67e741640 medium supplemented with 10% FBS and 1% penicillinestreptomycin, thenthe A549 cells were incubated for 4 h at 37 C in 5% CO2 and 95% air atmosphere

with >95% humidity. Then, a group of cell solution was exposed to an 808 nmlaser at a power intensity of 5 W/cm2 for 180 s. After the laser irradiation treated,A549 cells were stained with Annexin-V-FITC and PI (Becton Dickinson, Moun-tain View, CA, USA), and analyzed by BD FACSAria I ow cytometry with exci-tation at 488 nm. Fluorescent emission of FITC was measured at 515e545 nm andthat of DNA-PI complexes at 564e606 nm. Compensation was used wherevernecessary.

2.9. In vivo photothermal treatment

Male Balb/c mice, aging 4e5 weeks and weighing 18.8 1.9 g, were supplied bythe Department of Experimental Animals, Fudan University (Shanghai, China), andmaintained under standard housing conditions. All animal experiments were car-ried out in accordance with guidelines evaluated and approved by the ethics com-mittee of Fudan University. Through a subcutaneous injection of 2 106 cellssuspended in 100 mL PBS into the ank region, the mice bearing A549 tumor weresuccessfully established. The longest dimension and the shortest dimension of tu-mors were monitored by digital calipers. Once tumors were measured ~5.0 mm inlongest dimension, mice were randomized into 4 treatment groups (n 5 pergroup): PBS with laser treated (Group I), individual Fe3O4 NPs with laser treated(Group II), clustered Fe3O4 NPs with laser treated (Group III and IV for different NIRillumination time). Each type of dispersion (2 mg/mL, injection volume of 25 mL) wasintratumorally injected into mice. Two hours after injection, the 808-nmcontinuous-wave NIR laser was applied to irradiate tumors under a power in-tensity of 5 W/cm2 (Group I, II and IV for 180 s, Group III for 120 s). The temperaturewas recorded using an infrared camera thermographic system (InfraTec, Vari-oCAMhr research, German). After treatment, all mice were returned to animalhousing and tumor volumes were tracked every 2 days by digital caliper measure-ments. The tumor volume was calculated by the formula V 1/2(L$W2), where L islength (longest dimension) andW is width (shortest dimension). At 19th day, all theanimals were euthanized, and the tumors were dissected and weighed. According toour previous reports [40e43], some histological examinations included hematoxylinand eosin (H&E) stains and TUNEL assays were carried out for the assessment oftherapy efcacy.

2.10. The retention of the magnetic NPs in the tumor sites

According to the previous literature [29], the retention of the Fe3O4 NPs in thetumor sites was also studied. Once tumors were measured ~5.0 mm in longestdimension, mice were randomized into 2 groups (n 9 per group): individual Fe3O4NPs with laser treated (Group I), clustered Fe3O4 NPs with laser treated (Group II).Each type of dispersion (2 mg/mL, injection volume of 25 mL) was intratumorallyinjected intomice. Two hours after injection, the 808-nm continuous-wave NIR laserwas applied to irradiate tumors under a power intensity of 5 W/cm2 for 180 s. Theywere sacriced on day 1, 5 and 19 after injection, respectively. Three mice for onegroup were randomly selected and sacriced at each time point and the tumorswere extracted immediately after. The tumors were weighed and washed withdistilled water. Afterward, the tissue was mixed with 0.5 mL of aqua regia andincubated at room temperature. Two days later, the samples were centrifuged andthe upper clear solutions were collected and adjusted with distilled water. Theconcentrations of iron atoms in the solutions were performed by ame atomic ab-sorption spectroscopy (Hitachi, Japan). An iron hollow cathode lamp was used as alight source. The excitation current, slit width and absorption wavelength of ironwere set at 15 mA, 0.2 nm and 248.3 nm, respectively.

2.11. Statistical analysis

Unpaired student's t test was used for between two-group comparison and one-way ANOVAwith Fisher's LSD for multiple-group analysis. A probability (P) less than0.05 was considered statistically signicant. Results were expressed asmean standard deviation (SD) unless otherwise indicated.

3. Results and discussion

3.1. Characterization of magnetic NPs

A TEM image shows the dimension of our obtained individualFe3O4 NPs is about 15 nm (Fig. 1a), while that of the clustered Fe3O4

S. Shen et al. / Biomaterials 39 (2015) 67e74 69Fig. 1. TEM images of (a) individual magnetic Fe3O4 NPs and (b) clustered magnetic Fe3O4tization loops of individual and clustered magnetic Fe3O4 NPs.NPs. (c) The XRD pattern of individual and clustered magnetic Fe3O4 NPs. (d) Magne-

NPs has a nearly uniform size of ca. 225 nm with spherical shape(Fig. 1b). A TEM image of the clustered Fe3O4 NPs at highermagnication (Fig. S1) illustrates that the magnetite con-globulations are loose clusters, and the particles are composed ofnanocrystals with the size of about 5e10 nm. According to thepreviously reported literature [44], these nanocrystals are con-nected with each other by amorphous matrix as a bridge. Thesamples of individual and clustered Fe3O4 NPs both show similartypical XRD patterns of magnetite (JCPDS no. 75-1609; Fig. 1c),which indicates that the nanocrystalline structure of the clusteredFe3O4 NPs is the same as individual Fe3O4 NPs. The magneticproperty characterization (Fig. 1d) shows that both individual andclustered Fe3O4 NPs have superparamagnetic property and highmagnetization with saturation value of 58.69 emu/g for the indi-vidual Fe3O4 NPs, and 63.70 emu/g for the clustered Fe3O4 NPs,respectively.

3.2. Photothermal effect of magnetic NPs

Fig. 2a illustrates the UVevis spectra of the individual andclustered Fe3O4 NPs dispersed in deionized water. The individual

shorter NIR irradiation time in our study, which was benecial forthe safe application in vivo. In addition, the NIR-heating effects ofthe individual and clustered Fe3O4 NPs were both time andconcentration-dependent. The NPs with higher concentrationperformed markedly better than that of low concentration. Like-wise, the longer the laser irradiation time, the higher the solutiontemperature is. The results indicate that optoelectronic excitationof clustered Fe3O4 NPs by NIR irradiation can occur rapidly, and theextra energies are efciently transferred into molecular vibrationmodes to generate signicant amounts of thermal energy, whichcould be useful as photothermal mediators.

In order to validate the feasibility of the Fe3O4 NPs in biomedi-cine, the nanomaterials of individual or clustered Fe3O4 NPs wereadministered to A549 cells in 96-well plates, respectively, and eachgroup was subdivided into groups of with and without NIR laserexposure. The cell counting Kit-8 (CCK-8) assay was employed forthe quantitative evaluation of cell viability [45]. The viability ofuntreated cells was assumed to be 100%. The cell viability remainedabout 92.63% when they were incubated with individual or clus-tered Fe3O4 NPs at higher concentration of 500 mg/mL for 24 h(Fig. S2, Supporting information), which indicated our magnetic

Fe3

S. Shen et al. / Biomaterials 39 (2015) 67e7470Fe3O4 NPs have a very weak absorption in the NIR spectroscopyfrom600 to 1000 nm. Surprisingly, the clustered Fe3O4 NPs showanobvious absorption band containing a maximum at ca. 420 nm.Moreover, comparedwith individual Fe3O4 NPs, the clustered Fe3O4NPs also resulted in a signicant ~3.6 fold increase in the NIR ab-sorption at 808 nm for the aggregation of nanoparticles. Thus, theclustered Fe3O4 NPs that have a stronger absorption in the NIRspectroscopy may be conducive to destroying tumors in vivo as aneffective thermal generator due to its in-depth tissue penetration.The photothermal effects of the PBS solutions of the individual andclustered Fe3O4 NPs were rstly examined and compared in vitro byNIR laser irradiation (l 808 nm, 5 W/cm2). As seen from Fig. 2b,the clustered Fe3O4 NPs were more efcient than the individualFe3O4 NPs in inducing a temperature increase with the same con-centration and exposure time. For example, the temperature raisedby the clustered Fe3O4 NPs solution could reach 51.4 1.2 C withthe NPs concentration of 50 mg/mL and NIR exposure time of 180 s,while that of the individual Fe3O4 NPs only reached 42.0 1.5 C.Compared with the previous report using individual Fe3O4 NPswith high concentration [29], the higher temperature was easilyacquirable by clustered Fe3O4 NPs with lower concentration and

Fig. 2. (a) UVevis spectra of aqueous dispersion of individual and clustered magnetic

magnetic Fe3O4 NPs with the concentration of 0e80 mg/mL were examined ex vitro by ithermocouple.nanomaterials had no inherent toxicity. Subsequently, themagneticnanomaterials were applied for the photothermal ablation of can-cer cells. The cytotoxic effects by NIR irradiation were shown inFig. 3a. About 8.9%, 33.5% and 72.8% of cells were killed by clusteredFe3O4 NPs at the concentration of 50 mg/mL with a power density of5 W/cm2 for different illumination time of 60, 120 and 180 s,respectively. And only 0.8%, 3.5% and 14.5% of cells were killed byindividual Fe3O4 NPs. In addition, direct irradiation of the cellsremained close to 100%, exhibiting no effect on cell viability. This ismainly because the low light absorption by natural endogenouscytochromes of these cells caused minimal temperature elevationaccounts for their high survival [5]. The CCK-8 experimental resultindicated that clustered Fe3O4 NPs inducing higher temperaturewere more cytotoxic against A549 cells than individual Fe3O4 NPsby NIR irradiation. These results also proved that cancer cells couldbe effectively killed by clustered Fe3O4 NPs with lower concentra-tion and shorter NIR irradiation time compared with previousreport [29]. Fluorescence images of calcein AM (green, live cells)and PI (red, dead cells) co-stained cells after PTT were furtherconrmed the effectiveness of PTT using clustered Fe3O4 NPs(Fig. 3b). A large number of killed A549 cells were observed by

O4 NPs. (b) The photothermal effects of the PBS solutions of individual and clustered2rradiating NIR (l 808 nm, 5 W/cm ) for 60e180 s. Temperature was measured by

Fig. 3. (a) Relative viabilities of PBS, individual and clustered magnetic Fe3O4 NPs with the concentration of 50 mg/mL treated A549 cells without or with NIR laser irradiation(l 808 nm, 5 W/cm2) for 60e180 s. (b) Confocal uorescence images of calcein AM (green, live cells) and propidium iodide (red, dead cells) co-stained A549 cells treated byclustered magnetic Fe3O4 NPs at the concentration of 50 mg/mL with NIR laser irradiation (l 808 nm, 5 W/cm2) for (1) 0 s, (2) 60 s, (3) 120 s, (4) 180 s. (c) Flow cytometry graphs ofA549 cells treated by clustered magnetic Fe3O4 NPs at the concentration of 50 mg/mL without or with NIR laser irradiation (l 808 nm, 5 W/cm2) for 180 s. The treated A549 cellswere double stained by Annexin-V-FITC/PI and analyzed by ow cytometry. Quadrant I doesn't representatively correspond to live or damaged cells; Quadrant II representativelycorresponds to the population of dying cells. The membranes of these cells had been compromised so PI could penetrate and then hybridize with their DNAs; Quadrant III representsthe population of live cells. Since the live cells have intact cell membranes, PI could not stain their DNAs; The population of dead cells, or stained DNAs, exhibits a high PI and smallforward scattering signal as shown in quadrant IV. These are free DNAs, no longer enclosed within a membrane. The analysis data reveal that cells targeted by the clusteredmagnetic Fe3O4 NPs combined with NIR lighting display irreversible damage, as shown by the larger population in quadrant IV, where the cellular membrane is broken to such anextent that the cell can no longer function nor recover from the damage [46].

S. Shen et al. / Biomaterials 39 (2015) 67e74 71

clustered Fe3O4 NPs exposed to laser illumination with a powerdensity of 5W/cm2 for 180 s. Once the irradiation timewas reducedto 120 s, most of cells remained survival. In addition, cells incubatedwith clustered Fe3O4 NPs for irradiation time of 60 s or without NIRlaser treatment both showed limited damage, which could beattributable to insufciently lethal temperature rise. The uores-cence imaging result was consistent with the CCK-8 result.

To make clear the cell death mode after photothermal treat-ment, an Annexin-V-FITC/PI method was conducted by owcytometry. Fig. 3c showed the ow cytometry graphs of the cells byclustered Fe3O4 NPs with and without laser irradiation. Annexin-V-FITC emission signal was plotted on the x-axis, while PI emissionsignal was plotted on the y-axis. The graphs quadrants weredivided into four quadrants. The quantities of living cells, apoptoticcells and necrotic cells were determined by the percentage ofAnnexin V/PI, Annexin V/PI and Annexin V/PI. Almost noapoptosis or necrosis cells were observed in the group of no NIRtreatment. After laser irradiation, the apoptosis rate of cells reached67.7%, while the necrosis rate was only 6.4%. The ow cytometrydata revealed that cells treated by clustered Fe3O4 NPs displayed

Fig. 4 showed the infrared thermal images of tumor surface inGroup I, II and IV mice. In Group IV, the maximum temperature ofthe tumor surface rapidly increased 21.6 C to reach 55.9 Cattributing to the clustered Fe3O4 NPs with a good NIR-heatingproperty. In addition, as can be seen from the gure, the increasein NIR irradiation time from 120 s to 180 s could not signicantlyincrease the temperature of tumor surface, showing the tempera-ture reached a balance after two minutes of NIR irradiation. At thattime, the NIR laser didn't stop working, and kept irradiating for oneminute to maintain a high temperature in the tumor site to effec-tively kill A549 cells. In Group II, the maximum temperature of thetumor surface reached 50.3 C, which was inferior to Group IV dueto the weak photothermal effects. In contrast, less temperaturechanged in Group I by PBS with NIR treated, and increased only4.2 C. Furthermore, in order to investigate the effect of PTT usingclustered Fe3O4 NPs with different NIR irradiation time, the 808-nm continuous-wave NIR laser was employed to irradiate tumorsunder a power intensity of 5 W/cm2 for 120 s in Group III.Compared with Group IV, the duration time of high temperature inGroup III reduced one minute. As we know, high temperature can

S. Shen et al. / Biomaterials 39 (2015) 67e7472irreversible damage upon photothermal treatment, and theircellular membranes were broken to such an extent that the cellscould no longer function nor recover from the damage. Therefore,the mechanism of in vitro PTT is mainly triggered by cell apoptosis,not necrosis.

3.3. Photothermal ablation of tumor in vivo

Encouraged by the effective photothermal outcome in vitro, wenext performed a pilot in vivo photothermal treatment study. Inthis work, A549 tumor-bearing mice were established by subcu-taneous injection of 2 106 A549 cells suspended in RMPI-1640medium into the ank region. Once the tumors were measuredca. 5.0 mm in longest dimension, mice were randomized into fourtreatment groups (n 5 per group) mentioned in the experimentalsection. There was no statistical difference among group meantumor volumes at the onset of treatment (P > 0.05). Twenty-fourhours before irradiation, mice were then injected intratumorallywith 25 mL of treatment solution containing 50 mg of individual orclustered Fe3O4 NPs in Group II, III and IV, leaving light gray scars onthe original tumor sites. The tumors were illuminated with an 808-nm NIR laser (5 W/cm2; spot size, 5 mm) for 120 s in Group III andfor 180 s in Group II, IV. No mice died during the course of therapy.Fig. 4. Infrared thermal images of PBS, individual and clustered magnetic Fe3O4 NPs with th(l 808 nm, 5 W/cm2) for 0e180 s.lead to cell apoptosis, even instantaneous coagulative necrosis andirreversible cell death [47]. Tumor sizes after diverse treatmentswere measured every two days (Fig. 5a). The tumors of Group I, IIand III grew rapidly. There was no statistically signicant differencein nal tumor size (P > 0.05), indicating that tumor growth was notaffected by PBS or individual Fe3O4 NPs with NIR irradiation for180 s, and clustered Fe3O4 NPs with NIR irradiation time of 120 s.On the contrary, a statistically signicant of tumor growth wasobserved in mice treated with clustered Fe3O4 NPs plus NIR laserfor illumination time of 180 s (P < 0.05). Effects were most pro-nounced in the Group IV, where mean tumor volume at 19 dayswas reduced from 955.3 mm3 in controls to 222.8 mm3

(P 0.0139). At 19th days, mice were euthanized, tumors wereexcised and weighted (Fig. 5). The weights of tumors for Group I, II,III, IV and V were 0.5311 0.0599 g, 0.4369 0.0452 g,0.4072 0.0479 g and 0.2297 0.0333 g, respectively (Fig. 5d).Obviously, the antitumor efciency of Group IV was particularlyprominent and was superior to all the other groups (P < 0.01),showing an inhibition rate of 56.75%, 47.43% and 43.59% comparedwith Group I, II and III. However, there was no statistically signi-cant difference in Group III compared with Group I, II, indicatingthat a lack of sustained high temperature would hardly effectivelykill the tumor cells.e concentration of 100 mg/mL injected A549 tumor sample under NIR laser irradiation

easth Ns supfter

teriaFig. 5. (a) Tumor growth curves of different groups after treatment. Tumor sizes were mmagnetic Fe3O4 NPs with NIR irradiation for 180 s and clustered magnetic Fe3O4 NPs wimagnetic Fe3O4 NPs with NIR irradiation for 180 s was particularly prominent and watherapy every 2 days. Means and standard errors are shown. (c) Photograph of tumors a(d) Tumor weights of each group. n.s. no signicant difference, *P < 0.05, **P < 0.01.S. Shen et al. / BiomaAccordingly, histological analysis of H&E staining illustratedthat the tumors in test groups exhibited a scarlike structure con-taining numerous collagen bundles (Fig. S3), especially for GroupIV. However, a lot of tumor cells were observed in Group I, II and III.The antitumor efcacy of PTT against A549 tumor was also evalu-ated by the TUNEL immunohistological assay (Fig. S4), whichexclusively detected apoptosis in tumor tissues. TUNEL signals(brown uorescence) were observed in Group II, III and IV. Tumortissues treated with clustered Fe3O4 NPs plus laser irradiation for180 s showedmore TUNEL signals than did tumors treatedwith PBSand individual Fe3O4 NPs plus laser irradiation for 180 s, andclustered Fe3O4 NPs plus laser irradiation for 120 s, indicating thatphotothermal ablation of clustered Fe3O4 NPs with long time NIRtreatment was more efcient for tumor therapy. Last but not theleast, to investigate the metabolism of the Fe3O4 NPs after therapy,the iron atoms captured in the tumors were determined by a ameatomic absorption spectroscopy. The analytical results exhibitedthat 98.7%, 77.8%, 53.5% of individual Fe3O4 NPs or 99.1%, 72.3%,46.7% of clustered Fe3O4 NPs in the tumors for 1, 5 and 19 days post-injection with NIR irradiation with 180 s (Fig. S5). The analyticalresult showed that our Fe3O4 NPs could be slowly cleared from thetumors after therapy. Besides, the high Fe3O4 NPs uptake in the rstfew days is particularly important for photothermal cancer treat-ment if repeated NIR irradiation therapy was needed.

4. Conclusions

The individual and clustered Fe3O4 NPs with the same nano-crystalline structure were successfully synthesized. The clusteredFe3O4 NPs were more efcient than the individual Fe3O4 NPs ininducing a temperature increase. Signicantly greater cell killingwas detected when A549 cells incubated with clustered Fe3O4 NPsured every 2 days. Means and standard errors are shown. The groups of PBS, individualIR irradiation for 120 s were statistically identical. The antitumor efciency of clusterederior to all the other groups (P < 0.05). (b) Body weight changes were recorded afterexcision from PBS, individual and clustered magnetic Fe3O4 NPs under NIR irradiation.ls 39 (2015) 67e74 73were irradiated with NIR illumination. The ow cytometry analysisdemonstrated that the mechanism of in vitro PTT was mainlytriggered by cell apoptosis, not necrosis. Using an A549 tumormodel, a higher therapeutic efcacy was obtained by the NIR-induced hyperthermia of the clustered Fe3O4 NPs. We believethat our study has the potential in the future clinical application ofcancer therapies for the clustered Fe3O4 NPs with good photo-thermal effect and excellent biological safety.

Acknowledgment

This work was supported by the National Natural ScienceFoundation of China (grant, No. 51273047, 81301974 and81102847). W. Yang thanks Prof. Hao Zhang (Jilin University) for thediscussion about the properties of Fe3O4.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.biomaterials.2014.10.064.

References

[1] Karakoti AS, Das S, Thevuthasan S, Seal S. PEGylated inorganic nanoparticles.Angew Chem Int Ed 2011;50:1980e94.

[2] Yang K, Hu L, Ma X, Ye S, Cheng L, Shi X, et al. Multimodal imaging guidedphotothermal therapy using functionalized graphene nanosheets anchoredwith magnetic nanoparticles. Adv Mater 2012;24:1868e72.

[3] Zhao F, Zhao Y, Liu Y, Chang XL, Chen CY, Zhao YL. Cellular uptake, intracel-lular trafcking, and cytotoxicity of nanomaterials. Small 2011;7:1322e37.

[4] Liu HY, Chen D, Li LL, Liu TL, Tan LF, Wu XL, et al. Multifunctional goldnanoshells on silica nanorattles: a platform for the combination of photo-thermal therapy and chemotherapy with low systemic toxicity. Angew ChemInt Ed 2011;50:891e5.

[5] Shen S, Tang HY, Zhang XT, Ren JF, Pang ZQ, Wang DG, et al. Targeting mes-oporous silica-encapsulated gold nanorods for chemo-photothermal therapywith near-infrared radiation. Biomaterials 2013;34:3150e8.

[6] Shen S, Ren JF, Zhu XY, Pang ZQ, Lu XH, Deng CH, et al. Monodisperse mag-netites anchored onto carbon nanotubes: a platform for cell imaging, mag-netic manipulation and enhanced photothermal treatment of tumors. J MaterChem B 2013;1:1939e46.

[7] Okuno T, Kato S, Hatakeyama Y, Okajima J, Maruyama S, Sakamoto M, et al.Photothermal therapy of tumors in lymph nodes using gold nanorods andnear-infrared laser light. J Control Release 2013;172:879e84.

[8] Tsai MF, Chang SH, Cheng FY, Shanmugam V, Cheng YS, Su CH, et al. Aunanorod design as light-absorber in the rst and second biological near-infrared windows for in vivo photothermal therapy. ACS Nano 2013;7:

[28] Johannsen M, Gneveckow U, Taymoorian K, Thiesen B, Waldofner N, Scholz R,et al. Morbidity and quality of life during thermotherapy using magneticnanoparticles in locally recurrent prostate cancer: results of a prospectivephase I trial. Int J Hyperth 2007;23:315e23.

[29] Chu MQ, Shao YX, Peng JL, Dai XY, Li HK, Wu QS, et al. Near-infrared laser lightmediated cancer therapy by photothermal effect of Fe3O4 magnetic nano-particles. Biomaterials 2013;34:4078e88.

[30] Lartigue L, Hugounenq P, Alloyeau D, Clarke SP, Levy M, Bacri JC, et al.Cooperative organization in iron oxide multi-core nanoparticles potentiatestheir efciency as heating mediators and MRI contrast agents. ACS Nano2012;6:10935e49.

[31] Qiu P, Jensen C, Charity N, Towner R, Mao C. Oil phase evaporation-inducedself-assembly of hydrophobic nanoparticles into spherical clusters withcontrolled surface chemistry in an oil-in-water dispersion and comparison of

S. Shen et al. / Biomaterials 39 (2015) 67e74745330e42.[9] Bagley AF, Hill S, Rogers GS, Bhatia SN. Plasmonic photothermal heating of

intraperitoneal tumors through the use of an implanted near-infrared source.ACS Nano 2013;7:8089e97.

[10] Robinson JT, Welsher K, Tabakman SM, Sherlock SP, Wang H, Luong R, et al.High performance in vivo near-IR (>1 mum) imaging and photothermalcancer therapy with carbon nanotubes. Nano Res 2010;3:779e93.

[11] Markovic ZM, Harhaji-Trajkovic LM, Todorovic-Markovic BM, Kepic DP,Arsikin KM, Jovanovic SP, et al. In vitro comparison of the photothermalanticancer activity of graphene nanoparticles and carbon nanotubes. Bio-materials 2011;32:1121e9.

[12] Zhou ZG, Sun YA, Shen JC, Wei J, Yu C, Kong B, et al. Iron/iron oxide core/shellnanoparticles for magnetic targeting MRI and near-infraredphotothermaltherapy. Biomaterials 2014;35:7470e8.

[13] Li XJ, Takashima M, Yuba E, Harada A, Kono K. PEGylated PAMAM dendrimer-doxorubicin conjugate-hybridized gold nanorod for combinedphotothermal-chemotherapy. Biomaterials 2014;35:6576e84.

[14] Tao Y, Ju EG, Liu Z, Dong K, Ren JS, Qu XG. Engineered, self-assembled near-infrared photothermal agents for combined tumor immunotherapy andchemo-photothermal therapy. Biomaterials 2014;35:6646e56.

[15] Bai J, Liu YW, Jiang XE. Multifunctional PEG-GO/CuS nanocomposites for near-infrared chemo-photothermal therapy. Biomaterials 2014;35:5805e13.

[16] Zhu MT, Nie GJ, Meng H, Xia T, Nel A, Zhao YL. Physicochemical propertiesdetermine nanomaterial cellular uptake, transport, and fate. Acc Chem Res2013;46:622e31.

[17] Nystrom AM, Fadeel B. Safety assessment of nanomaterials: implications fornanomedicine. J Control Release 2012;161:403e8.

[18] Nel A, Xia T, Meng H, Wang X, Lin SJ, Ji ZX, et al. Nanomaterial toxicity testingin the 21st century: use of a predictive toxicological approach and high-throughput screening. Acc Chem Res 2013;46:607e21.

[19] Li WP, Liao PY, Su CH, Yeh CS. Formation of oligonucleotide-gated silica shell-coated Fe3O4-Au coreeshell nanotrisoctahedra for magnetically targeted andnear-infrared light-responsive theranostic platform. J Am Chem Soc2014;136:10062e75.

[20] Torres-Lugo M, Rinaldi C. Thermal potentiation of chemotherapy by magneticnanoparticles. Nanomedicine 2013;8:1689e707.

[21] Dormer KJ, Awasthi V, Galbraith W, Kopke RD, Chen K, Wassel R. Magneti-cally-targeted, technetium 99m-labeled nanoparticles to the inner ear.J Biomed Nanotechnol 2008;4:174e84.

[22] Gogoi M, Sarma HD, Bahadur D, Banerjee R. Biphasic magnetic nanoparticles-nanovesicle hybrids for chemotherapy and self-controlled hyperthermia.Nanomedicine 2014;9:955e70.

[23] Shi DL, Cho HS, Chen Y, Xu H, Gu HC, Lian J. Adv Mater 2009;21:2170e3.[24] Storm FK, Harrison WH, Elliott RS, Morton DL. Cancer Res 1979;39:2245e51.[25] Maier-Hauff K, Ulrich F, Nestler D, Niehoff H, Wust P, Thiesen B, et al. Efcacy

and safety of intratumoral thermotherapy using magnetic iron-oxide nano-particles combined with external beam radiotherapy on patients withrecurrent glioblastoma multiforme. J Neurooncol 2011;103:317e24.

[26] Maier-Hauff K, Rothe R, Scholz R, Gneveckow U, Wust P, Thiesen B, et al.Intracranial thermotherapy using magnetic nanoparticles combined withexternal beam radiotherapy: results of a feasibility study on patients withglioblastoma multiforme. J Neurooncol 2007;81:53e60.

[27] Johannsen M, Gneveckow U, Thiesen B, Taymoorian K, Cho CH, Waldofner N,et al. Thermotherapy of prostate cancer using magnetic nanoparticles: feasi-bility, imaging, and three-dimensional temperature distribution. Eur Urol2007;52:1653e61.behaviours of individual and clustered iron oxide nanoparticles. J Am ChemSoc 2010;132:17724e32.

[32] Hayashi K, Nakamura M, Sakamoto W, Yogo T, Miki H, Ozaki S, et al. Super-paramagnetic nanoparticle clusters for cancer theranostics combining mag-netic resonance imaging and hyperthermia treatment. Theranostics 2013;3:366e76.

[33] Hayashi K, Nakamura M, Miki H, Ozaki S, Abe M, Matsumoto T, et al.Magnetically responsive smart nanoparticles for cancer treatment with acombination of magnetic hyperthermia and remote-control drug release.Theranostics 2013;4:834e44.

[34] Qin W, Lohrman J, Ren SQ. Magnetic and optoelectronic properties of goldnanocluster-thiophene assembly. Angew Chem Int Ed 2014;53:7316e9.

[35] Ni WX, Qiu YM, Li M, Zheng J, Sun RWY, Zhan SZ, et al. Metallophilicity-drivendynamic aggregation of a phosphorescent gold(I)-ailver(I) cluster prepared byaolution-based and mechanochemical approaches. J Am Chem Soc 2014;136:9532e5.

[36] Zhang XD, Chen J, Luo ZT, Wu D, Shen X, Song SS, et al. Enhanced tumoraccumulation of sub-2 nm gold nanoclusters for cancer radiation therapy. AdvHealth Mater 2014;3:133e41.

[37] Retnakumari A, Jayasimhan J, Chandran P, Menon D, Nair S, Mony U, et al.CD33 monoclonal antibody conjugated Au cluster nano-bioprobe for targetedow-cytometric detection of acute myeloid leukaemia. Nanotechnology2011;22:285102.

[38] Deng YH, Yang WL, Wang CC, Fu SK. A novel approach for preparation ofthermoresponsive polymer magnetic microspheres with coreeshell structure.Adv Mater 2003;15:1729e32.

[39] Ma W, Xu S, Li JM, Guo J, Lin Y, Wang CC. Hydrophilic dual-responsivemagnetite/PMAA core/shell microspheres with high magnetic susceptibilityand pH sensitivity via distillation-precipitation polymerization. J Polym SciPart A Polym Chem 2011;49:2725e33.

[40] Ren JF, Shen S, Wang DG, Xi Z, Guo LR, Pang ZQ, et al. The targeted delivery ofanticancer drugs to brain glioma by PEGylated oxidized multi-walled carbonnanotubes modied with angiopep-2. Biomaterials 2012;33:3324e33.

[41] Zhang B, Sun X, Mei H, Wang Y, Liao Z, Chen J, et al. LDLR-mediated peptide-22-conjugated nanoparticles for dual-targeting therapy of brain glioma. Bio-materials 2013;34:9171e82.

[42] Zhang B, Shen S, Liao Z, Shi W, Wang Y, Zhao J, et al. Targeting bronectins ofglioma extracellular matrix by CLT1 peptide-conjugated nanoparticles. Bio-materials 2014;35:4088e98.

[43] Zhang B, Wang H, Liao Z, Wang Y, Hu Y, Yang J, et al. EGFP-EGF1-conjugatednanoparticles for targeting both neovascular and glioma cells in therapy ofbrain glioma. Biomaterials 2014;35:4133e45.

[44] Liu J, Sun Z, Deng Y, Zou Y, Li C, Guo X, et al. Highly water-dispersiblebiocompatible magnetite particles with low cytotoxicity stabilized by citrategroups. Angew Chem Int Ed 2009;48:5875e9.

[45] Wang S, Zhang Q, Luo XF, Li J, He H, Yang F, et al. Magnetic graphene-basednanotheranostic agent for dual-modality mapping guided photothermaltherapy in regional lymph nodal metastasis of pancreatic cancer. Biomaterials2014;35:9473e83.

[46] Au L, Zheng D, Zhou F, Li ZY, Li XD, Xia YN. A quantitative study on thephotothermal effect of immuno gold nanocages targeted to breast cancercells. ACS Nano 2008;2:1645e52.

[47] Zhou YF. Noninvasive treatment of breast cancer using high-intensity focusedultrasound. J Med Imaging Health Informatics 2013;3:141e56.

Magnetic nanoparticle clusters for photothermal therapy with near-infrared irradiation1. Introduction2. Materials and methods2.1. Materials2.2. Synthesis of individual magnetic NPs2.3. Synthesis of clustered magnetic NPs2.4. Characterization2.5. Cell lines and culture conditions2.6. NIR-heating effect of magnetite NPs in solution2.7. Cell viability assay2.8. Flow cytometry analysis2.9. In vivo photothermal treatment2.10. The retention of the magnetic NPs in the tumor sites2.11. Statistical analysis

3. Results and discussion3.1. Characterization of magnetic NPs3.2. Photothermal effect of magnetic NPs3.3. Photothermal ablation of tumor in vivo

4. ConclusionsAcknowledgmentAppendix A. Supplementary dataReferences