lable at ScienceDirect

Biomaterials 124 (2017) 35e46

Contents lists avai

Biomaterials

journal homepage: www.elsevier .com/locate/biomater ia ls

Shape-controlled fabrication of magnetite silver hybrid nanoparticleswith high performance magnetic hyperthermia

Qi Ding a, b, Dongfang Liu c, Dawei Guo d, Fang Yang a, b, Xingyun Pang a, b, Renchao Che e,Naizhen Zhou a, b, Jun Xie a, b, Jianfei Sun a, b, Zhihai Huang a, b, Ning Gu a, b, *

a State Key Laboratory of Bioelectronics, Jiangsu Key Laboratory for Biomaterials and Devices, School of Biological Science and Medical Engineering,Southeast University, Nanjing 210096, PR Chinab Collaborative Innovation Center of Suzhou Nano-Science and Technology, Suzhou, 215123, PR Chinac Jiangsu Key Laboratory of Molecular and Functional Imaging, Department of Radiology, Medical School, Zhongda Hospital, Southeast University, Nanjing210009, PR Chinad Laboratory of Veterinary Pharmacology and Toxicology, College of Veterinary Medicine, Nanjing Agricultural University, Nanjing 210095, PR Chinae Laboratory of Advanced Materials, Department of Materials Science, Collaborative Innovation Center of Chemistry for Energy Materials, Fudan University,Shanghai 200438, PR China

a r t i c l e i n f o

Article history:Received 24 August 2016Received in revised form10 January 2017Accepted 28 January 2017Available online 31 January 2017

Keywords:Magnetic nanoparticlesSilverNanocompositesShape controlHyperthermia

* Corresponding author. State Key Laboratory ofLaboratory for Biomaterials and Devices, School of BiEngineering, Southeast University, 2 Sipailou, Nanjing

E-mail address: guning@seu.edu.cn (N. Gu).

http://dx.doi.org/10.1016/j.biomaterials.2017.01.0430142-9612/© 2017 Elsevier Ltd. All rights reserved.

a b s t r a c t

Superparamagnetic Fe3O4 nanoparticles (NPs)-based hyperthermia is a promising non-invasive approachfor cancer therapy. However, the heat transfer efficiency of Fe3O4 NPs is relative low, which hinders theirpractical clinical applications. Therefore, it is promising to improve the magnetic hyperthermia efficiencyby exploring the higher performance magnetic NPs-based hybrid nanostructures. In the current study, itpresents a straightforward in situ reduction method for the shape-controlled preparation of magnetite(Fe3O4) silver (Ag) hybrid NPs designed as magnetic hyperthermia heat mediators. The magnetite silverhybrid NPs with core-shell (Fe3O4@Ag) or heteromer (Fe3O4-Ag) structures exhibited a higher biocom-patibility with SMMC-7721 cells and L02 cells than the individual Ag NPs. Importantly, in the magnetichyperthermia, with the exposure to alternating current magnetic field, the Fe3O4@Ag and Fe3O4-Aghybrid NPs indicated much better tumor suppression effect against SMMC-7721 cells than the individualFe3O4 NPs in vitro and in vivo. These results demonstrate that the hybridisation of Fe3O4 and Ag NPs couldgreatly enhance the magnetic hyperthermia efficiency of Fe3O4 NPs. Therefore, the Fe3O4@Ag and Fe3O4-Ag hybrid NPs can be used to be as high performance magnetic hyperthermia mediators based on asimple and effective preparation approach.

© 2017 Elsevier Ltd. All rights reserved.

1. Introduction

Colloidal hybrid nanoparticles (NPs), which include multiplecomponents in a single nanosized object, are attractive forbiomedical applications because novel important properties andmultiple or enhanced functions can be generated due to the syn-ergy of the individual component in the hybrid NPs [1e4]. Amongthe multitude of investigations to date, superparamagnetic NPs-based hybrid NPs have been considered as an important family ofmultifunctional or enhanced functional nanocomposites [5]. These

Bioelectronics, Jiangsu Keyological Science and Medical210096, PR China.

types of NPs have shown great application potential in variousbiomedical fields such as drug delivery, magnetic resonance im-aging and hyperthermia treatment [6]. Especially, based on the factthat magnetic NPs such as iron oxide NPs can generate heat whenan alternating current magnetic field (ACMF) is applied to them[7,8], the magnetic hyperthermia is increasingly becoming one ofthe novel noninvasive approaches for cancer therapy. Cancer cells,more sensitive to temperature (�40 �C) than healthy cells [9,10],can be selectively killed [11] by the magnetically induced heating,which can damage the biological integrity of the cell membraneand cytoskeleton [12]. However, the clinical trials has demon-strated that a considerably high concentration of iron oxide NPs[13,14] is needed for the efficient cancer thermal treatment becauseof the relatively poor heat transfer efficiency of current Fe3O4 NPs[15], which hinders their practical applications. In this regard, the

Q. Ding et al. / Biomaterials 124 (2017) 35e4636

development of high-performance hyperthermiamediators such asmagnetic NPs attached therapeutic agents is of great importance tothe practical applications of magnetic NPs-based hyperthermia[7,16,17].

Noble metallic silver (Ag) NPs have been recognized as a po-tential anti-cancer or antibacterial agent and extensively studied inthe biomedical fields [18e20]. Recently, several recent studies havedemonstrated that Ag NPs can enhance the efficiency of hyper-thermia [13,21e24]. For example, Wang et al. [25] reported that AgNPs improved the thermo-sensitivity of C6 cells. Liu et al. [13]studied the influence of a mixture of Fe3O4 and Ag NPs on mag-netic hyperthermia, and found that the presence of Ag NPsenhanced the cancer cell death from magnetic heat both in vitroand in vivo. However, the heat generated by Fe3O4 NPs can onlytransmit very short distances and induce highly focused heating[15], and thus the separation of Fe3O4 and Ag NPs may limit thesensitivity of Ag NPs to Fe3O4 NPs-mediated hyperthermia.

It is promising to attain the lattice-mismatched heterojunctionsby tailor-made chemical combining Fe3O4 and Ag NPs into a singleunit; this would integrate the properties of the two NPs as well asenhance the functions between them at nanoscale level [26e29].Various approaches have been reported for the preparation ofFe3O4-based magnetite silver hybrids with various morphologies,including centric or eccentric core@shells [30e32], phase-segregated core-satellites [33e35] and heteromer architectures[36,37]. However, the study on shape control of themagnetite silverhybrid NPs is still scarce [38e40].

In this study, we present a simple and effective in situ reductionstrategy to prepare the magnetite silver hybrid NPs. It designed thepre-existing carboxyl-rich Fe3O4 NPs to carry inorganic Ag bondingjunctions [2] in order to overcome the high interfacial energybarrier. Based on this design strategy, the shape of magnetite silverhybrid NPs can be tuned easily by adjusting the concentration ofthe intermediate product Fe3O4-COOAg NPs to produce core-shellFe3O4@Ag or heteromer Fe3O4-Ag structures after in situ reduc-tion. The cytotoxicity of the magnetite silver hybrid NPs is muchlower than the individual Ag NPs. Compared to the individual Fe3O4NPs, both Fe3O4@Ag and Fe3O4-Ag NPs exhibited greatly enhancedhyperthermia effects in vitro and in vivo. To the best of ourknowledge, so far, this is the first case in which magnetite silverhybrid NPs were applied to in vitro and in vivo magnetic hyper-thermia, thereby providing a novel and effective approach toimprove the hyperthermia efficiency.

2. Materials and methods

2.1. Preparation of magnetite silver hybrid NPs

2.1.1. Synthesis of monodisperse hydrophobic Fe3O4-OA NPsThe Fe3O4-OA NPs (OA: oleic acid) were prepared by the classical

thermal decomposition method [41,42]. In a typical chemicalpreparation, iron chloride (FeCl3$6H2O, 2.16 g) and sodium oleate(7.3 g) were dissolved in a mixture solvent consisting of ethanol(16 mL), pure water (12 mL) and hexane (28 mL). The mixture wasstirred for 4 h at 70 �C. Then the upper organic layer containing theiron-oleate complex was washed three times with pure water in aseparating funnel, and subsequently, hexane was evaporated off.The as-prepared iron-oleate complex was collected and driedovernight in a drying oven at 40 �C under vacuum. Then, the iron-oleate complex (3.4 g) and oleic acid (500 mL) were dissolved in thehigh boiling 1-octadecene solvent (20 mL). With the protection ofnitrogen, the solution was programmed heated from the roomtemperature to 320 �C at 3.3 �C/min, and held isothermally for30 min at 120 �C (open to air so as to remove trace water) and320 �C, respectively. When the synthetic procedurewas completed,

the resulting mixture was cooled, and the Fe3O4-OA NPs wereprecipitated by ethanol (50 mL) and separated by centrifugation.The NPs were washed 3 times with 1:3 hexane and acetone, andcollected by centrifugation, and finally dried under vacuum. Theobtained Fe3O4-OA NPs were in a black solid form.

2.1.2. Synthesis of hydrophilic Fe3O4-PAA NPsFe3O4-PAA NPs (PAA: Poly(acrylic acid)) were prepared

following a previously reported method [43]. In a typical process,under an inert nitrogen atmosphere, diethylene glycol (DGE, 35mL)mixed with PAA (2 g) was heated to 120 �C with strongly stirringwith the injection of the prepared Fe3O4-OA NPs toluene solution(100 mg/mL). Then the mixture was heated to 210 �C under refluxfor 1 h. After the mixture was cooled down to room temperature, abrown-black magnetic powder was collected by the addition ofdilute hydrochloric acid andmagnetic separation. The precipitate ofFe3O4-PAA NPs was washed several times with pure water, andthen well dispersed in pure water by ionizing the carboxyl groupswith triethylamine to reserve. Fe3O4-PAA NPs were purified by ul-trafiltration with an Ultra centrifugal filter unit (Millipore, 100,000NMWL).

2.1.3. Synthesis of core-shell magnetite silver hybrid NPs (Fe3O4 @Ag)

Aqueous solutions of Fe3O4-PAA NPs (1.5 mg/mL, 13.33 mL) andsilver nitrate (AgNO3, 40 mg/mL, 10 mL) were mixed and shakenvigorously on amicro-oscillator for 20min. Amassive precipitate ofFe3O4-COOAg NPs was produced and collected by magnetic sepa-ration. The precipitate was ultrasonically redissolved in an aqueoussolution of sodium citrate (Na3Cit, 10.6 mmoL/L, 16 mL). After 12 hstirring and membrane filtration (220 nm), a saturated solution ofFe3O4-COOAg NPs was obtained. Then the saturated solution ofFe3O4-COOAg NPs (5.55 mg Ag/mL, 15 mL) was stirred and addedwith aqueous solution of freshly prepared sodium borohydride(NaBH4, 26 mmol/L, 100 mL) in every 10 s to a total volume of 1 mL,and stirring was continued for 30 min. The resulting mixture waspurified by ultrafiltration to yield Fe3O4@Ag.

2.1.4. Synthesis of heteromer magnetite silver hybrid NPs (Fe3O4-Ag)

A saturated solution of Fe3O4-COOAg NPs (5.55 mg Ag/mL,15 mL) was diluted with an aqueous solution of Na3Cit (10.6 mmol/L) to prepare an unsaturated solution of Fe3O4-COOAg NPs (4.37 mgAg/mL). Then an aqueous solution of freshly prepared NaBH4(26 mmol/L, 100 mL) was added in every 10 s to a total volume of1 mL under stirring. The solution was stirred for 30 min, and pu-rified by ultrafiltration to yield Fe3O4-Ag.

2.2. Experimental setup and measurements for magnetichyperthermia

The magnetically induced heating efficiency of Fe3O4-PAA,Fe3O4@Ag and Fe3O4-Ag NPs was measured and recorded with acommercial system (illustrated in Fig. S1) with a moderate radiofrequency (390 KHz) heating machine equipped with an inductivecopper coil of 3 turns, an inner diameter of 31 mm, an outerdiameter of 41mm (Shuangping SPG-06-II, China); and a fiber optictemperature sensor (UMI8 universal multichannel, FISO Technol-ogies Inc., Quebec, Canada) fitted with a pre-calibrated fiber optictemperature probe (FISO, model FOT-L-SD-C1-F1-M2-R1-ST). Theprepared NPs at a concentration of 2 mg Fe/mL were placed insidethe copper coil of the applied ACMF (390 KHz, 18 A). Specific ab-sorption rate (SAR) was used to quantify the heating efficiencygenerated per unit gram of Fe per unit time [56], expressed asEquation (1):

Q. Ding et al. / Biomaterials 124 (2017) 35e46 37

SAR ¼ CdTdt

ms

mm(1)

where C is the specific heat capacity of the solution (here, the heatcapacity of the solvent, Cwater ¼ 4.185 J/g$�C), dT/dt is the initialslope of the temperature-time curve, ms is the mass of the sus-pension, and mm is the mass of the Fe content in the suspension.

2.3. Cell culture and animals

Human hepatoma cell line SMMC-7721 and human normalhepatocyte cell line L02 were purchased from KeyGen BiologyTechnology Company (Nanjing, China). SMMC-7721 cells andL02 cells were cultured in RPMI 1640 medium (KeyGen Biology,China) supplemented with 10% fetal bovine serum (FBS, Sijiqing,Hangzhou, China) and incubated at 37 �C in a 5% CO2 incubator(HERA cell 150, Thermo Electron Corporation).

BALB/c nude mice (female, aged 4 weeks) were purchased fromthe Comparative Medicine Center of Yangzhou University (Yangz-hou, China). All animal studies were carried out according to theGuideline for Animal Experimentation with the approval of theAnimal Care Committee of Southeast University, Nanjing, China.

2.4. In vitro cytotoxicity test

The in vitro cytotoxicity of Fe3O4-PAA, Fe3O4@Ag and Fe3O4-AgNPs in SMMC-7721 cells and L02 cells was assessed using an MTT(3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyl-tetrazolium bromide,KeyGen Biology, China) assay. The cells were seeded in 96-wellplates (Corning, USA) at a density of 5 � 103 cells per well andincubated inmedium (100 mL) for 24 h to allow cell attachment. Themedium was then replaced with another medium (100 mL) con-taining Fe3O4-PAA, Fe3O4@Ag and Fe3O4-Ag NPs at a final concen-tration of 0e60 mg Fe/mL per well, and the cells were cultured for24 h. After the addition of MTT (50 mL, 1 mg/mL) per well, the cellswere incubated for another 4 h. Formazan was dissolved indimethyl-sulfoxide (DMSO, 150 mL) and the absorbance wasmeasured at 490 nm using a microplate reader (Bio-Rad 680, USA).The relative cell viability was based on optical density (OD) andwascalculated as (ODtreated � ODbackground)/(ODcontrol � ODbackground) � 100%. ODtreated was taken from the cellstreated with the prepared NPs, ODcontrol was from the untreatedcontrol cells, and ODbackground was from the empty wells. Theviability of the control cells was 100%. All data are presented as themean ± standard deviation of triplicate experiments. The in vitrocytotoxicity test of Ag NPs is the same as above.

2.5. In vitro magnetic hyperthermia experiment

SMMC-7721 cells (5 � 104) were grown in sterile glass Petridishes (specially made, 22 mm in diameter and 30mm in height) ina 5% CO2 incubator at 37 �C for 24 h. After removing the old me-dium, cells were incubated with 1 mL medium containing Fe3O4-PAA, Fe3O4@Ag and Fe3O4-Ag NPs at a final concentration of 15, 30and 60 mg Fe/mL in duplicate, which was representative of twogroups as follows: non-hyperthermia groups: Fe3O4-PAA,Fe3O4@Ag and Fe3O4-Ag NPs group; hyperthermia groups: Fe3O4-PAA plus ACMF, Fe3O4@Ag plus ACMF and Fe3O4-Ag NPs plus ACMFgroup. Non-hyperthermia groups were incubated for 12 h. Hyper-thermia groups were incubated for 12 h, and then subjected toACMF (390 kHz, 18 A) for 20 min. Cell inhibition was measuredusing an MTT assay. MTT (250 mL, 1 mg/mL) was added to each Petridish and incubated with cells for 4 h. After removing the super-natant, DMSO (750 mL) per dish was added to dissolve the

formazan. DMSO (150 mL) containing formazanwas transferred to a96-well plate prior to the measurement of absorbance at 490 nm bya microplate reader. The calculation of relative cell inhibition wascarried out as [1 � (ODtreated � ODbackground)/(ODcontrol � ODbackground)] � 100%. ODtreated was taken from thecells treated with the prepared NPs and ACMF, ODcontrol was fromthe control cells untreated with NPs and ACMF, and ODbackgroundwas from the empty wells. The inhibition of the control cells wasassumed to be 0%. All data analysis was based on three independentexperiments.

2.6. In vivo magnetic hyperthermia experiment

The tumor models were established by subcutaneous injectionof SMMC-7721 cells (1 � 107 cells/mouse) into the right flanks ofnude mice. When the volume of the tumors reached about300 mm3, the mice were randomly divided into eight groups(n ¼ 6/group): (1) saline group (negative control), (2) saline plusACMF group, (3) Fe3O4-PAA NPs group, (4) Fe3O4-Ag NPs group, (5)Fe3O4@Ag NPs group, (6) Fe3O4-PAA NPs plus ACMF group, (7)Fe3O4-Ag NPs plus ACMF group, (8) Fe3O4@Ag NPs plus ACMFgroup. Following a multipoint injection strategy, moving clockwiseat the 3, 6, 9 and 12 o'clock points, the mice were intratumorallyinjected with aqueous solution of Fe3O4-PAA, Fe3O4-Ag, Fe3O4@AgNPs and saline, respectively, on the 1st and 4th day. The singleinjection does of all of the Fe3O4-based NPs was normalized to be15 mg Fe/Kg body weight. Particularly for hyperthermia groups of6e8, the tumors were given to ACMF treatments at 2 h after eachinjection. In seven day follow-up experiment, the tumors in groups2 and 6e8 were exposed to ACMF for 20 min, repeating 6 times(days 1e6) at 24 h intervals (in vivo therapeutic protocol diagramsee Fig. 7a). Tumor-bearing mice were placed into the copper coil ofthe applied ACMF (390 KHz, 18 A) using a specially made poly-propylene supporter so that tumors were positioned in the middleof the coil interior space possessing the highest field density ofACMF (Fig. S2). The temperature of the tumors was monitored withan infrared thermometer (FLUKE thermal imager, Ti32 model). Inaddition, the mice were weighed by electronic balance and theirtumor growth were measured by vernier caliper on an alternateday. The tumor volumes were calculated as V ¼ d2 � D/2 (whered and D represent the shortest and longest diameter of the tumor inmm, respectively). The relative tumor volume was calculated as Vx

e V0, where Vx is the tumor volume on day x post treatment and V0

is the initial tumor volume. The tumor volume inhibition wascalculated (1 �‾Vtreated group/‾Vcontrol group) � 100%, where‾V is themean tumor volume.

2.7. Statistical analysis

Values were expressed as mean ± SDs. The data were analyzedusing the unpaired t-test, and a P value of <0.05 was considered asstatistically significant difference.

3. Results and discussion

3.1. Synthesis and characterization of magnetite silver hybrid NPs

The synthesis of magnetite silver hybrid NPs is depicted sche-matically in Scheme 1. First, the excellent superparamagnetic andmonodispersed Fe3O4-OA NPs core were prepared by the classicalthermal decomposition of iron oleate in 1-octadecene. Short-chainPAA with abundant free carboxyl groups was used to replace of OAon Fe3O4 NPs surface to achieve carboxyl-rich Fe3O4-PAA NPs viathe ligand exchange reaction. The hydrophilic carboxyl-rich Fe3O4-PAA NPs then provided the abundant anchor sites of Ag source on

Scheme 1. Schematic representation of the formation process of Fe3O4@Ag and Fe3O4-Ag hybrid NPs.

Q. Ding et al. / Biomaterials 124 (2017) 35e4638

their surfaces. Then, because of a strong affinity of ionized carboxylgroup to Agþ ion [44], the abundant Agþ ions were bonded to thesurface of Fe3O4 NPs by excess of AgNO3 added resulting in theFe3O4-COOAg NPs precipitate [45]. This precipitate was redissolvedin different amounts of Na3Cit solution to yield different concen-tration Fe3O4-COOAg NPs solutions. During this process, it wasfound that the concentration of Ag in these solutions may be a keyfactor for the shape control of magnetite silver hybrid NPs. After insitu reduction with NaBH4, core-shell Fe3O4@Ag hybrid NPs wereobtained at a concentration of 5.55 mg Ag/mL (the saturated con-centration of the Fe3O4-COOAg NPs solution), and heteromerFe3O4-Ag hybrid NPs were obtained at a concentration of 4.35 mgAg/mL.

It is well-recognized that the high density of carboxyl groups onmagnetic NPs surface [43,46] can help create the formation of thecore-shell [47] or core-satellite [27,48] Fe3O4@Ag hybrid structure.Encouraged by the existing precipitation-dissolution behavior ofthe precipitate of Fe3O4-COOAg NPs (Fig. S3), the possible mecha-nism of magnetite silver hybrid NPs shape control may be

associated with the ionization equilibrium of the Fe3O4-COOAg NPssolution shown in Equation (2) [49].

Fe3O4 � COO�AgþðsÞ# Fe3O4 � COO�AgþðaqÞ # Fe3O4 � COO�ðaqÞ

þ AgþðaqÞ(2)

According to the Le Chatelier's principle, the equilibrium movesto the right when the solution is diluted, and moves to the leftwhen the solution is concentrated. The smaller the concentration ofFe3O4-COOAg NPs (aq), the greater the dissociation degree of[-COO-] and [Agþ] (the proving experiment see S1.4). We hypoth-esized that controlling the dissociation degree of [-COO-] and [Agþ]could rebuild site-differential surface distributions of Ag source onthe surface of Fe3O4-COOAg NPs (aq). When in a saturated Fe3O4-COOAg NPs solution (5.55 mg Ag/mL), the minimized dissociationof [-COO-] and [Agþ] was kept; and most of Agþ ions were boundedon the carboxyl groups of the Fe3O4-COOAg NPs (aq) surface. Thus,

Q. Ding et al. / Biomaterials 124 (2017) 35e46 39

the core-shell Fe3O4@Ag hybrid NPs were formed after in situreduction (Scheme 1). When the Fe3O4@COOAg NPs solution wasdiluted to 4.35 mg/mL, the dissociation degree of [-COO-] and [Agþ]was increased. Under this condition, only partial Agþ ions were stillbounded on the carboxyl groups of the Fe3O4-COOAg NPs (aq)surface to generate the heteromer Fe3O4-Ag hybrid NPs after in situreduction (Scheme 1).

The morphology of magnetite silver hybrid NPs was character-ized by transmission electron microscope (TEM). Fe3O4@Ag hybridNPs (Fig. 1b) are spherical and uniform with a diameter of10.41 ± 0.95 nm, which is similar to 10.1 ± 0.98 nm of Fe3O4-PAANPs (Fig. 1a, the size distribution histograms see Fig. S4), indicatingthat the Ag shell is very thin. The High-resolution TEM (HRTEM)image provides detailed structural information of the core-shellstructure (Fig. 1b). Both the lattice fringe spacing of ~0.253 nmand ~0.237 nm, corresponding to the (311) planes of the cubicspinel structured Fe3O4 (JCPDS card no. 19-0629) and (111) planesin the face-centered cubic Ag (JCPDS card no. 04-0783) respectively,can be observed in a single Fe3O4@Ag hybrid NP (Fig. 1d), con-firming the coexistence of Ag and Fe3O4 crystals in a single hybridNP. Ag generally appears darker than Fe3O4 in TEM images becauseit possesses a high electron density compared to Fe3O4 [50], andthus it is not easy to confirm the core-shell structure of Fe3O4@AgNPs by TEM images. To further confirm the core-shell structure ofFe3O4@Ag hybrid NPs, a single Fe3O4@Ag NP was further analyzed

Fig. 1. TEM images of Fe3O4-PAA (a), Fe3O4@Ag (b) and Fe3O4-Ag (c) NPs; HRTEM images of F(g) NPs; EDS patterns of Fe3O4@Ag (h) and Fe3O4-Ag (i) NPs (the inserts are the scanning a

by high-angle annular dark-field scanning transmission electronmicroscopy (HAADF-STEM). As shown in Fig. 1f, the NP was muchbrighter than the Fe3O4 portions of the heteromers shown in Fig.1g.Ag appears brighter than Fe3O4 because of the higher intensity ofscattered electrons in the HAADF-STEM image [50], which dem-onstrates the existence of an Ag shell in the Fe3O4@Ag NP. Theenergy dispersive X-ray spectroscopy (EDS) analysis of a singleFe3O4@Ag NP further confirmed the coexistence of iron and Agelements in the hybrid NP (Fig. 1h).

Themorphology of the heteromer Fe3O4-Ag hybrid NPs was alsocharacterized by TEM, HRTEM, HAADF-STEM and EDS. As shown inthe TEM image (Fig. 1c), the structure of a single heteromer NP wascomposed of one NP of ~10 nm bearing 1e3 NPs of 2e5 nm. Fromthe HRTEM image of the heteromer NP (Fig. 1e), it can be seen thatthe lattice fringe spacings of the NP of ~10 nmwere ~0.253 nm and~0.297 nm, corresponding to the (311) and (220) planes of Fe3O4(JCPDS card no. 19-0629). The smaller NPs attached to the Fe3O4had a lattice fringe spacing of ~0.237 nm, corresponding to the (111)plane of fcc Ag (JCPDS card no. 04-0783). As expected, in theHAADF-STEM image (Fig. 1g), the Ag portions of the heteromer NPswere much brighter than the Fe3O4 portions. In addition, EDSanalysis of a single heteromer NP confirmed the coexistence of ironand Ag elements in the heteromer (Fig. 1i). These results demon-strate that the two types of magnetite silver hybrid NPs are core-shell and heteromer structures, respectively. The constituents of

e3O4@Ag (d) and Fe3O4-Ag (e) NPs; HAADF-STEM images of Fe3O4@Ag (f) and Fe3O4-Agreas corresponding to the EDS spectra).

Q. Ding et al. / Biomaterials 124 (2017) 35e4640

the two types of hybrid NPs were further qualified by ICP-MS. Themolar ratio of Fe to Ag in Fe3O4@Ag is 1:0.22 for Fe3O4@Ag hybridNPs, and Fe to Ag in Fe3O4-Ag is 1:0.17 for Fe3O4-Ag hybrid NPs.

The crystal structures of magnetite silver hybrid NPs werecharacterized by powder X-ray diffraction (XRD). As shown inFig. 2a, the XRD spectra of Fe3O4@Ag and Fe3O4-Ag hybrid NPs werevery similar. In both spectra, the reflections of 30.22�, 35.46�,43.20�, 57.28� and 62.74� (marked with asterisks) can be attributedto the (220), (311), (400), (511) and (440) planes of cubic-phaseFe3O4 (JCPDS Card, No. 19-629, see Fig. S5), and the reflections of38.14� and 44.42� (marked with rhombi) can be attributed to the(111) and (200) planes of the fcc Ag (JCPDS Card, No. 4-783, seeFig. S6). This demonstrates that both Fe3O4@Ag and Fe3O4-Ag NPsare composed of crystalline Fe3O4 and Ag.

The optical properties of magnetite silver hybrid NPs werecharacterized using UVevis absorption spectra (Fig. 2b). BothFe3O4@Ag and Fe3O4-Ag hybrid NPs showed intense peaks at407 nm and 416 nm, respectively, which can be attributed to thesurface plasmon resonance (SPR) effect of nano Ag. The SPR peaksof Fe3O4@Ag and Fe3O4-Ag NPs were red-shifted to different de-grees compared to that of Ag NPs, which is due to the localdielectric effect according to classical Mie theory [51]. Similar red-

Fig. 2. (a) X-ray diffraction patterns of Fe3O4@Ag and Fe3O4-Ag hybrid NPs. (b) UVevisabsorption spectra of the aqueous solutions of Fe3O4-PAA, Fe3O4@Ag, Fe3O4-Ag and AgNPs.

shifts have also been observed previously [39,51].

3.2. Magnetically induced heating characteristics of magnetitesilver hybrid NPs in ACMF

VSM was used to evaluate the magnetic properties of these twotypes of magnetite silver hybrid NPs for future magnetic hyper-thermia. Results shown in Fig. 3a indicate that both Fe3O4@Ag andFe3O4-Ag hybrid NPs exhibited superparamagnetic propertiessimilar to Fe3O4-PAA NPs. The saturation magnetizations (Ms) ofFe3O4-PAA, Fe3O4-Ag and Fe3O4@Ag NPs were 82.4, 78.1 and 75.1emu/g Fe, respectively. Although the saturation magnetization ofboth Fe3O4-Ag and Fe3O4@Ag NPs has a little bit decrease, thepresence of Ag in the hybrid NPs had no significant effect on themagnetic properties of Fe3O4.

To further investigate the magnetically induced heating prop-erties of magnetite silver hybrid NPs, the heating efficiency ofFe3O4-PAA, Fe3O4-Ag and Fe3O4@Ag NPs were studied. SAR wasused to quantify the heating efficiency generated per unit gram ofFe per unit time. Three samples, 1 mL each of Fe3O4-PAA, Fe3O4-Agand Fe3O4@Ag NPs aqueous solutions at a concentration of 2mg/mLFe, were placed inside an ACMF with a moderate radio frequency of

Fig. 3. (a) Hysteresis loops of Fe3O4-PAA, Fe3O4-Ag and Fe3O4@Ag NPs. (b)Temperature-time curves of aqueous solutions of Fe3O4-PAA, Fe3O4-Ag, Fe3O4@Ag NPsand pure water with Fe concentration of 2 mg/mL as a function of the ACMF exposuretime.

Q. Ding et al. / Biomaterials 124 (2017) 35e46 41

390 KHz and a current of 18 A. As shown in Fig. 3b, the tempera-tures of all samples rose approximately 12 �C after 600 s ACMFtreatment; and the SAR values were 87, 83 and 76 W/g for Fe3O4-PAA, Fe3O4-Ag and Fe3O4@Ag NPs, respectively. The SAR result in-dicates that the existence of Ag in the two latter hybrid NPs had nosignificant influence on their magnetically induced heating.Moreover, the pure water exposed to the applied ACMF showedonly a negligible temperature increase, indicating that the tem-perature rise of the samples can be attributed to the magneticallyinduced heating effect of the NPs.

3.3. Evaluation of in vitro cytotoxicity

To use magnetite silver hybrid NPs in living systems, it isimportant to assess their stability and biocompatibility. Dynamiclight scattering (DLS, Zeta Plus Particle Size Analyzer, Brookhaven,US) was used to determine the hydrodynamic diameters of Fe3O4-PAA, Fe3O4-Ag and Fe3O4@Ag NPs in aqueous solution, biologicalmedium and biological medium suffering from ACMF, whichexhibited the high stability level (Table S1). The cytotoxicity of eachtype of NPs with various Fe concentrations was evaluated againstSMMC-7721 cells and L02 cells using a MTT assay. As shown inFig. 4a and 4d, after treatment with Fe3O4@Ag, Fe3O4-Ag or Fe3O4-PAA NPs for 24 h, the cell viability was above 90% at Fe concen-trations up to 60 mg/mL, indicating that magnetite silver hybrid NPshave no obvious cytotoxicity at the tested concentrations. Ascomparison, we also tested the cytotoxicity of the individual Ag NPsand found that its toxicity was evident at Ag concentrations greaterthan 10.35 mg/mL (Fig. 4b, 4c, 4e and 4f). In contrast, Fe3O4-Ag andFe3O4@Ag hybrid NPs showed no significant cytotoxicity at Agconcentrations up to 24.83 mg/mL. This demonstrates that the

Fig. 4. In vitro cell viability of SMMC-7721 cells and L02 cells treated with Fe3O4-PAA, Fe3O4

Fe3O4-PAA, Fe3O4-Ag and Fe3O4@Ag NPs at different Fe concentrations; (b) SMMC-7721 ce(*P < 0.05, **P < 0.01 versus Fe3O4-Ag NPs at the same Ag concentration); (c) SMMC-7721 c(**P < 0.01, ***P < 0.001 versus Fe3O4@Ag NPs at the same Ag concentration).

incorporation of Ag into magnetite silver hybrid NPs reduces thetoxicity of Ag.

3.4. In vitro magnetic hyperthermia for cancer cells

Next, the in vitro use of the preparedmagnetite silver hybrid NPsfor magnetic hyperthermia in SMMC-7721 cells was evaluated. Tomaximize the heat generation of the magnetic NPs and minimizethe background heating effect of the eddy current induced byACMF, the applied ACMF was set at a moderate 390 KHz frequencyand a current of 18 A (Optimization shown in Fig. S7) [52,53]. Cellinhibition was examined by MTT assay after 12 h incubation with aseries of doses of Fe3O4@Ag, Fe3O4-Ag and Fe3O4-PAA NPs, followedwith or without a 20-min exposure to ACMF. Fig. 5 shows that thecell inhibition rates were 51%, 75% and 83% for Fe3O4@AgNPs plus ACMF treatment and 46%, 72% and 75% for Fe3O4-AgNPs plus ACMF treatment at non-cytotoxic concentrations of 15, 30and 60 mg Fe/mL, respectively, which were significantly higher thanthe rates for Fe3O4-PAA NPs plus ACMF treatment at the corre-sponding concentrations. It's believed that both Fe3O4@Ag NPs andFe3O4-Ag NPs have higher inhibition ability for SMMC-7721 cellsthan Fe3O4-PAA NPs; magnetic hyperthermia efficiency was greatlyenhanced by using Fe3O4@Ag and Fe3O4-Ag NPs compared toFe3O4-PAA NPs.

The results demonstrate that the efficiency of in vitro magnetichyperthermia can be greatly enhanced in the presence of Agincorporated into magnetite silver hybrid NPs. The higher inhibi-tion may come from silver ions released from the surface ofFe3O4@Ag and Fe3O4-Ag NPs. Silver ions released from the particlessurface are considered to be the principal mechanism of Ag NPstoxicity [54]. On the one hand, silver ions readily bind to sulfur- and

-Ag, Fe3O4@Ag and Ag NPs for 24 h (a) SMMC-7721 cells and (d) L02 cells treated withlls and (e) L02 cells treated with Fe3O4-Ag and Ag NPs at different Ag concentrationsells and (f) L02 cells treated with Fe3O4@Ag and Ag NPs at different Ag concentrations

Fig. 5. In vitro cell inhibition of SMMC-7721 cells treated with Fe3O4-PAA, Fe3O4-Agand Fe3O4@Ag NPs at various Fe concentrations for 12 h incubation, followed with orwithout a 20-min exposure to ACMF. *P < 0.05, **P < 0.01 versus cells treated withFe3O4-PAA plus ACMF NPs at the same Fe conditions.

Q. Ding et al. / Biomaterials 124 (2017) 35e4642

phosphorus-containing biomolecules such as proteins and DNA,thereby potentially causing cell damage [55]. On the other hand,silver ions could also interfere with mitochondrial activity andinduce apoptosis [20]. Several investigations have reported thatAgþ ions are released from Ag NPs in aerobic aqueous environ-ments [56]. The dissolution of Ag NPs to silver ions is an oxidationprocess. Upon the internalization of Ag NPs by cells, the rate ofintracellular oxidation of Ag NPs is higher compared with that inwater due to higher intracellular oxygen level [57,58]. Also, Ag NPscould be further dissolved to silver ions in the lysosomes due to thelower pH environment [59]. TEM images in Fig. 6 indicate thecellular uptake and localization of magnetite silver hybrid NPs inthe absence and presence of ACMF. The results showed that all theNPs could indeed be internalized and localized in the cytoplasmand lysosomes with and without ACMF treatment. The localization

Fig. 6. Representative TEM images of SMMC-7721 cells treated with Fe3O4-PAA (a), Fe3O4-AFe3O4-Ag (e) and Fe3O4@Ag (f) NPs for 12 h followed by a 20-min ACMF treatment.

in acid lysosome would enhance the release of silver ion from thehybrid NPs. In addition, Heat, including photothermal [22,23,60]and magnetically induced heating [13,25], could trigger therelease of Agþ ions. We further compared in vitro ionic silverreleased from the two hybrid NPs in aqueous solution with andwithout exposure to ACMF. A negligible (0.04%) increase in Agþ ionrelease from Ag NPs was observed with ACMF treatment comparedto without ACMF. However, 1.2% and 3.1% increases in Agþ ionrelease from Fe3O4@Ag and Fe3O4-AgNPs were observed withACMF compared to without ACMF, indicating a boosting of silverions release. Taken together, magnetic hyperthermia efficiency wassignificantly enhanced using Fe3O4@Ag and Fe3O4-Ag hybrid NPscompared to Fe3O4-PAA NPs at a series of non-cytotoxic concen-trations, suggesting a possibly synergistic effect of magneticallyinduced heating and heat-assisted silver ions release. The inhibi-tion rate of the Fe3O4@Ag was slightly higher than that of theFe3O4-Ag, which may be due to the comparatively high Ag contentin the core-shell structure.

3.5. In vivo tumor magnetic hyperthermia for tumors

The in vivo magnetic hyperthermia performance of themagnetite silver hybrid NPs was further investigated to assess theirapplication potentials in cancer treatment. In the 7-day experi-ment, the two injections with saline, Fe3O4-PAA, Fe3O4-Ag andFe3O4@Ag NPs aqueous solution (single dose 15 mg Fe/kg bodyweight) were administrated to saline group 1 (negative control);saline plus ACMF group 2; non-hyperthermia groups of Fe3O4-PAANPs group 3, Fe3O4-Ag NPs group 4, Fe3O4@Ag NPs group 5; hy-perthermia groups of Fe3O4-PAA NPs plus ACMF group 6, Fe3O4-AgNPs plus ACMF group 7, Fe3O4@Ag NPs plus ACMF group 8,respectively, on the 1st and 4th day. And six ACMF treatments(390 KHz, 18 A, 20 min) were administrated to groups 2 and 6e8 at24 h intervals on days 1e6. The intratumoral distribution of injec-ted Fe3O4-based NPs were monitored by T2 and T2*-weighted MRI.The clearMR signal attenuation demonstrated that the injected NPsdiffused to the whole tumor regions at 2 h post injections (Fig. S8).

g (b) and Fe3O4@Ag (c) NPs for 12 h, and SMMC-7721 cells treated with Fe3O4-PAA (d),

Fig. 7. In vivo antitumor effect of subcutaneous SMMC-7721 tumor-bearing mice in the course of a 7-day follow-up experiment (the two intratumoral injections and six hyper-thermia treatments): (1) saline group, (2) saline plus ACMF group, (3) Fe3O4-PAA NPs group, (4) Fe3O4-Ag NPs group, (5) Fe3O4@Ag NPs group, (6) Fe3O4-PAA NPs plus ACMF group,(7) Fe3O4-Ag NPs plus ACMF group, (8) Fe3O4@Ag NPs plus ACMF group. (a) In vivo protocol diagram of time course of magnetic hyperthermia treatment. (b) Tumor growth behaviorand (c) body weight changes of mice in different treatment groups as a function of time. yP < 0.05 and ‡P < 0.0001 versus group 1, §P < 0.0001 versus corresponding non-hyperthermia group treated with the same NPs, #P < 0.001 versus group 6.

Q. Ding et al. / Biomaterials 124 (2017) 35e46 43

The tumor temperature of the hyperthermia groups 6e8 wasdirectly detected by the infrared imaging (FLUKE thermal imager,Ti32 model) after ACMF treatment. The thermal images displayedthat the average temperatures of tumor surface could reached40e43 �C (Table S2, a representative thermal image shown inFig. S9). Since the temperature deviation between tumor interiorand surface is usually ±2 �C [61], the temperature in our experi-ment is in the range of valid hyperthermia temperature, whichwere able to induce the apoptosis of tumor cells and inflict nodamage to normal tissue [12]. After 7-day experiment, the evolu-tion of tumor growth is presented in Fig. 7b for all the treatedgroups 1e8. The tumor volume inhibition rate is 65% for group 7and 67% for group 8 respectively, which is significantly higher than47% of group 6, 14% of group 5, 10% of group 4, 5% of group 3 and 3%of group 2. Even for comparison of non-hyperthermia groups 4e5and group 1, Fe3O4-Ag and Fe3O4@Ag treatments exhibited signif-icant inhibitions (P < 0.05) of tumor growth, which suggested thatFe3O4-Ag and Fe3O4@Ag hybrid NPs themselves have a certain in-hibition effect due to the cytotoxicity derived from the presence ofAg in magnetite silver hybrid NPs; even for comparison of hyper-thermia groups 6e8 and non-hyperthermia groups 3e5 treatedwith corresponding NPs, ACMF treatments exhibited extremelysignificantly higher inhibitions (P < 0.0001) of tumor growth,which suggested that magnetically induced heating can greatlyenhance tumor growth inhibition effect. It is further noticed thatthere was no significant difference of tumor growth inhibition inthe non-hyperthermia groups 4e5 compared to group 3, but anextremely significant difference (P < 0.001) in the hyperthermiagroups 7e8 compared to group 6 under the condition of no obvioushyperthermia temperature difference. Considering statistical dif-ference degree, it is believed that the combination of magnetitesilver hybrid NPs and ACMF treatments can take a synergistically

enhanced inhibition effect consistent with the results of in vitromagnetic hyperthermia. The tumor inhibition effects can also beobserved intuitively from the typical tumor photographs recordedfrom each treated group during treatments (Fig. S10). In addition,there is no noticeable mortality and body weight discrepancies(Fig. 7c) between non-hyperthermia groups and hyperthermiagroups compared with the saline group, indicating a well-tolerateddose level of the magnetite silver hybrid NPs.

To further assess the tumor therapy efficiency, all mice weresacrificed after treatments on the 7th day for H&E and TUNELstaining analysis (Fig. 8). Histopathology indicated that the com-bination of magnetite silver hybrid NPs and ACMF treatmentsinduced massive complete necrosis and fibrosis symptoms of tu-mor tissues in hyperthermia groups 7e8, which were far moreserious than that in non-hyperthermia groups 4e5 and hyper-thermia group 6. This can further proof the excellent magnetichyperthermia efficiency of magnetite silver hybrid NPs on thedestruction of tumor cells. No obvious histopathological abnor-malities or lesions were observed in the main organs (heart, liver,spleen and kidney) in all the groups, implying either the three NPsor magnetically induced heating to cancer therapy with low acutetoxicity or little risk. Very small amounts accumulation of Fe3O4-PAA, Fe3O4-Ag and Fe3O4@Ag NPs were found in the liver andspleen using Prussian blue and nuclear fast red double staining(Fig. S11) in the corresponding treated groups. TUNEL assay wasperformed to evaluate whether tumor growth inhibition by treat-ment materials and magnetically induced heating was related tocell apoptosis. It was assuredly observed that the tumors treatedwith combination of magnetite silver hybrid NPs and ACMF treat-ments exhibited the extensive regions and high populations ofapoptotic cells (brown cell) in hyperthermia groups 7 and 8. Ac-cording to statistics, either the apoptosis rates 61.29± 14.38% for

Fig. 8. Histological analysis of mouse organs and tumors from groups 1e8 on the 7th day after treatments, using H&E staining and TUNEL staining. Groups 1e8: (1) saline group, (2)saline plus ACMF group, (3) Fe3O4-PAA NPs group, (4) Fe3O4-Ag NPs group, (5) Fe3O4@Ag NPs group, (6) Fe3O4-PAA NPs plus ACMF group, (7) Fe3O4-Ag NPs plus ACMF group, (8)Fe3O4@Ag NPs plus ACMF group. Images were taken at 400� magnification.

Q. Ding et al. / Biomaterials 124 (2017) 35e4644

group 7 or 63.26± 16.88% for group 8 were significantly higher thanthe rates 41.29 ± 10.3% for group 6, 15.69 ± 8.85% for group 5,12.21 ± 5.45% for group 4, 7.91 ± 5.2% for group 3, 4.25 ± 2.23% forgroup 2, 2.17 ± 1.98% for group 1 (Table S3). The high apoptosis levelin hyperthermia groups 7e8 may explain the greatly reduction intumor size in these two groups.

4. Conclusion

In conclusion, an in situ reduction strategy to fabricate thelattice-mismatched Fe3O4 and Ag into a single nano-unit wasdeveloped. Using this method, the morphology of the magnetitesilver hybrid NPs can be easily tuned to a core-shell (Fe3O4@Ag) orheteromer (Fe3O4-Ag) structure, which proved to a higherbiocompatibility with SMMC-7721 cells and L02 cells than the in-dividual Ag NPs. Both Fe3O4-Ag and Fe3O4@Ag hybrid NPs exhibited

the significantly enhanced in vitro and in vivo tumor therapeuticeffect compared to the individual Fe3O4 NPs, which may beattributed to the synergistic effect of magnetically induced heatingand heat-assisted silver ions release. Therefore, the preparedmagnetite silver hybrid NPs provide the potential better candidatesfor tumor magnetic hyperthermia treatment.

Conflict of interest

These authors declare no conflict of interest.

Acknowledgements

This work was supported by the grants from the National KeyBasic Research Program of China [2011CB933503, 2013CB932901],the National Natural Science Foundation of China for Key Project of

Q. Ding et al. / Biomaterials 124 (2017) 35e46 45

International Cooperation [61420106012], the National NaturalScience Foundation of China [31370019, 51201034, 11274066], theNatural Science Foundation of Jiangsu Province in China[BK20140716], and the Special Project on Development of NationalKey Scientific Instruments and Equipment of China[2011YQ03013403].

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.biomaterials.2017.01.043.

References

[1] P.D. Cozzoli, T. Pellegrino, L. Manna, Synthesis, properties and perspectives ofhybrid nanocrystal structures, Chem. Soc. Rev. 35 (2006) 1195e1208.

[2] L. Carbone, P.D. Cozzoli, Colloidal heterostructured nanocrystals: synthesisand growth mechanisms, Nano Today 5 (2010) 449e493.

[3] R. Costi, A.E. Saunders, U. Banin, Colloidal hybrid nanostructures: a new typeof functional materials, Angew. Chem. Int. Ed. 49 (2010) 4878e4897.

[4] C. Sanchez, P. Belleville, M. Popall, L. Nicole, Applications of advanced hybridorganic-inorganic nanomaterials: from laboratory to market, Chem. Soc. Rev.40 (2011) 696e753.

[5] R. Hao, R. Xing, Z. Xu, Y. Hou, S. Gao, S. Sun, Synthesis, functionalization, andbiomedical applications of multifunctional magnetic nanoparticles, Adv.Mater 22 (2010) 2729e2742.

[6] A.-H. Lu, E.L. Salabas, F. Schüth, Magnetic nanoparticles: synthesis, protection,functionalization, and application, Angew. Chem. Int. Ed. 46 (2007)1222e1244.

[7] S. Ota, N. Yamazaki, A. Tomitaka, T. Yamada, Y. Takemura, Hyperthermia usingantibody-conjugated magnetic nanoparticles and its enhanced effect withcryptotanshinone, Nanomaterials 4 (2014) 319e330.

[8] L. Lartigue, C. Innocenti, T. Kalaivani, A. Awwad, M.d.M. Sanchez Duque,Y. Guari, et al., Water-dispersible sugar-coated iron oxide nanoparticles. anevaluation of their relaxometric and magnetic hyperthermia properties, J. Am.Chem. Soc. 133 (2011) 10459e10472.

[9] P. Guardia, R. Di Corato, L. Lartigue, C. Wilhelm, A. Espinosa, M. Garcia-Her-nandez, et al., Water-soluble iron oxide nanocubes with high values of specificabsorption rate for cancer cell hyperthermia treatment, ACS Nano 6 (2012)3080e3091.

[10] J.L. Roti Roti, Cellular responses to hyperthermia (40e46�C): cell killing andmolecular events, Int. J. Hyperth. 24 (2008) 3e15.

[11] T. Sadhukha, T.S. Wiedmann, J. Panyam, Inhalable magnetic nanoparticles fortargeted hyperthermia in lung cancer therapy, Biomaterials 34 (2013)5163e5171.

[12] M. Lin, J. Huang, J. Zhang, L. Wang, W. Xiao, H. Yu, et al., The therapeutic effectof PEI-Mn0.5Zn0.5Fe2O4 nanoparticles/pEgr1-HSV-TK/GCV associated withradiation and magnet-induced heating on hepatoma,, Nanoscale 5 (2013)991e1000.

[13] L. Liu, F. Ni, J. Zhang, X. Jiang, X. Lu, Z. Guo, et al., Silver nanocrystals sensitizemagnetic-nanoparticle-mediated thermo-induced killing of cancer cells, ActaBiochim. Biophys. Sin. 43 (2011) 316e323.

[14] B. Thiesen, A. Jordan, Clinical applications of magnetic nanoparticles for hy-perthermia, Int. J. Hyperth. 24 (2008) 467e474.

[15] A. Riedinger, P. Guardia, A. Curcio, M.A. Garcia, R. Cingolani, L. Manna, et al.,Subnanometer local temperature probing and remotely controlled drugrelease based on azo-functionalized iron oxide nanoparticles, Nano Lett. 13(2013) 2399e2406.

[16] A. Hervault, N.T.K. Thanh, Magnetic nanoparticle-based therapeutic agents forthermo-chemotherapy treatment of cancer, Nanoscale 6 (2014)11553e11573.

[17] K. Hayashi, M. Nakamura, H. Miki, S. Ozaki, M. Abe, T. Matsumoto, et al.,Magnetically responsive smart nanoparticles for cancer treatment with acombination of magnetic hyperthermia and remote-control drug release,Theranostics 4 (2014) 834e844.

[18] L.S. Nair, C.T. Laurencin, Silver nanoparticles: synthesis and therapeutic ap-plications, J. Biomed. Nanotechnol. 3 (2007) 301e316.

[19] M. Mahmoudi, V. Serpooshan, Silver-coated engineered magnetic nano-particles are promising for the success in the fight against antibacterialresistance threat, ACS Nano 6 (2012) 2656e2664.

[20] D. Guo, L. Zhu, Z. Huang, H. Zhou, Y. Ge, W. Ma, et al., Anti-leukemia activity ofPVP-coated silver nanoparticles via generation of reactive oxygen species andrelease of silver ions, Biomaterials 34 (2013) 7884e7894.

[21] H. Jiang, C. Wang, Z. Guo, Z. Wang, L. Liu, Silver nanocrystals mediated com-bination therapy of radiation with magnetic hyperthermia on glioma cells,J. Nanosci. Nanotechnol. 12 (2012) 8276e8281.

[22] K.-W. Hu, T.-M. Liu, K.-Y. Chung, K.-S. Huang, C.-T. Hsieh, C.-K. Sun, et al.,Efficient near-IR hyperthermia and intense nonlinear optical imaging contraston the gold nanorod-in-shell nanostructures, J. Am. Chem. Soc. 131 (2009)14186e14187.

[23] R. Di Corato, D. Palumberi, R. Marotta, M. Scotto, S. Carregal-Romero,P. Rivera_Gil, et al., Magnetic nanobeads decorated with silver nanoparticlesas cytotoxic agents and photothermal probes, Small 8 (2012) 2731e2742.

[24] H. Jang, Y.-K. Kim, H. Huh, D.-H. Min, Facile synthesis and intraparticle self-catalytic oxidation of dextran-coated hollow AueAg nanoshell and its appli-cation for chemo-thermotherapy, ACS Nano 8 (2014) 467e475.

[25] R. Wang, C. Chen, W. Yang, S. Shi, C. Wang, J. Chen, Enhancement effect ofcytotoxicity response of silver nanoparticles combined with thermotherapyon c6 rat glioma cells, J. Nanosci. Nanotechnol. 13 (2013) 3851e3854.

[26] J. Jiang, H. Gu, H. Shao, E. Devlin, G.C. Papaefthymiou, J.Y. Ying, BifunctionalFe3O4eAg heterodimer nanoparticles for two-photon fluorescence imagingand magnetic manipulation, Adv. Mater 20 (2008) 4403e4407.

[27] J. Chen, Z. Guo, H.-B. Wang, M. Gong, X.-K. Kong, P. Xia, et al., MultifunctionalFe3O4@C@Ag hybrid nanoparticles as dual modal imaging probes and near-infrared light-responsive drug delivery platform, Biomaterials 34 (2013)571e581.

[28] P. Dallas, J. Tucek, D. Jancik, M. Kolar, A. Panacek, R. Zboril, Magneticallycontrollable silver nanocomposite with multifunctional phosphotriazine ma-trix and high antimicrobial activity, Adv. Funct. Mater 20 (2010) 2347e2354.

[29] J.M. Hodges, A.J. Biacchi, R.E. Schaak, Ternary hybrid nanoparticle isomers:directing the nucleation of Ag on PteFe3O4 using a solid-state protectinggroup, ACS Nano 8 (2013) 1047e1055.

[30] J. Shen, Y. Zhu, X. Yang, J. Zong, C. Li, Multifunctional Fe3O4@Ag/SiO2/Aucoreeshell microspheres as a novel SERS-activity label via long-range plas-mon coupling, Langmuir 29 (2012) 690e695.

[31] Z. Xu, Y. Hou, S. Sun, Magnetic core/shell Fe3O4/Au and Fe3O4/Au/Ag nano-particles with tunable plasmonic properties, J. Am. Chem. Soc. 129 (2007)8698e8699.

[32] C. Yuen, Q. Liu, Optimization of Fe3O4@Ag nanoshells in magnetic field-enriched surface-enhanced resonance Raman scattering for malaria diag-nosis, Analyst 138 (2013) 6494e6500.

[33] B.-H. Jun, M.S. Noh, J. Kim, G. Kim, H. Kang, M.-S. Kim, et al., Multifunctionalsilver-embedded magnetic nanoparticles as sers nanoprobes and their ap-plications, Small 6 (2010) 119e125.

[34] X.X. Han, A.M. Schmidt, G. Marten, A. Fischer, I.M. Weidinger, P. Hildebrandt,Magnetic silver hybrid nanoparticles for surface-enhanced resonance Ramanspectroscopic detection and decontamination of small toxic molecules, ACSNano 7 (2013) 3212e3220.

[35] T. Zheng, Q. Zhang, S. Feng, J.-J. Zhu, Q. Wang, H. Wang, Robust nonenzymatichybrid nanoelectrocatalysts for signal amplification toward ultrasensitiveelectrochemical cytosensing, J. Am. Chem. Soc. 136 (2014) 2288e2291.

[36] H. Gu, Z. Yang, J. Gao, C.K. Chang, B. Xu, Heterodimers of Nanoparticles: for-mation at a liquid�liquid interface and particle-specific surface modificationby functional molecules, J. Am. Chem. Soc. 127 (2005) 34e35.

[37] M.R. Buck, J.F. Bondi, R.E. Schaak, A total-synthesis framework for the con-struction of high-order colloidal hybrid nanoparticles,, Nat. Chem. 4 (2012)37e44.

[38] J. Huang, Y. Sun, S. Huang, K. Yu, Q. Zhao, F. Peng, et al., Crystal engineeringand SERS properties of Ag-Fe3O4 nanohybrids: from heterodimer to core-shellnanostructures, J. Mater. Chem. 21 (2011) 17930e17937.

[39] Y. Zhai, L. Han, P. Wang, G. Li, W. Ren, L. Liu, et al., Superparamagnetic plas-monic nanohybrids: shape-controlled synthesis, tem-induced structure evo-lution, and efficient sunlight-driven inactivation of bacteria, ACS Nano 5(2011) 8562e8570.

[40] Y. Chen, N. Gao, J. Jiang, Surface matters: enhanced bactericidal property ofcoreeshell AgeFe2O3 nanostructures to their heteromer counterparts fromone-pot synthesis, Small 9 (2013) 3242e3246.

[41] J. Park, K. An, Y. Hwang, J.G. Park, H.J. Noh, J.Y. Kim, et al., Ultra-large-scalesyntheses of monodisperse nanocrystals, Nat. Mater 3 (2004) 891e895.

[42] Z. Chen, R. Xu, Y. Zhang, N. Gu, Effects of proteins from culture medium onsurface property of silanes- functionalized magnetic nanoparticles, NanoscaleRes. Lett. 4 (2008) 204e209.

[43] T.R. Zhang, J.P. Ge, Y.X. Hu, Y.D. Yin, A general approach for transferring hy-drophobic nanocrystals into water, Nano Lett. 7 (2007) 3203e3207.

[44] Z. Shen, H. Duan, H. Frey, Water-soluble fluorescent ag nanoclusters obtainedfrom multiarm star poly(acrylic acid) as “Molecular Hydrogel” templates, Adv.Mater 19 (2007) 349e352.

[45] M.D. Judd, B.A. Plunkett, M.I. Pope, The thermal decomposition of calcium,sodium, silver and copper(II) acetates, J. Therm. Anal. 6 (1974) 555e563.

[46] D. Liu, W. Wu, J. Ling, S. Wen, N. Gu, X. Zhang, Effective PEGylation of ironoxide nanoparticles for high performance in vivo cancer imaging, Adv. Funct.Mater 21 (2011) 1498e1504.

[47] P. Joshi, Y. Zhou, T.O. Ahmadov, P. Zhang, Quantitative SERS-based detectionusing Ag-Fe3O4 nanocomposites with an internal reference, J. Mater. Chem. C2 (2014) 9964e9968.

[48] B. Liu, W. Zhang, F. Yang, H. Feng, X. Yang, Facile method for synthesis ofFe3O4@polymer microspheres and their application as magnetic support forloading metal nanoparticles, J. Phys. Chem. C 115 (2011) 15875e15884.

[49] R.W. Ramette, The nature of dissolved silver acetate, J. Chem. Educ. 43 (1966)299.

[50] H. Yu, M. Chen, P.M. Rice, S.X. Wang, R.L. White, S.H. Sun, Dumbbell-likebifunctional Au-Fe3O4 nanoparticles, Nano Lett. 5 (2005) 379e382.

[51] Y. Chen, N. Gao, J. Jiang, Surface matters: enhanced bactericidal property ofcore-shell Ag-Fe2O3 nanostructures to their heteromer counterparts fromone-pot synthesis, Small 9 (2013) 3242e3246.

Q. Ding et al. / Biomaterials 124 (2017) 35e4646

[52] S. Dutz, R. Hergt, Magnetic nanoparticle heating and heat transfer on amicroscale: Basic principles, realities and physical limitations of hyperthermiafor tumour therapy, Int. J. Hyperth. 29 (2013) 790e800.

[53] D. Maity, P. Chandrasekharan, C.T. Yang, K.H. Chuang, B. Shuter, J.M. Xue, et al.,Facile synthesis of water-stable magnetite nanoparticles for clinical MRI andmagnetic hyperthermia applications, Nanomedicine 5 (2010) 1571e1584.

[54] A. Ivask, A. ElBadawy, C. Kaweeteerawat, D. Boren, H. Fischer, Z. Ji, et al.,Toxicity mechanisms in escherichia coli vary for silver nanoparticles and differfrom ionic silver, ACS Nano 8 (2013) 374e386.

[55] Z.-m. Xiu, Q.-b. Zhang, H.L. Puppala, V.L. Colvin, P.J.J. Alvarez, Negligibleparticle-specific antibacterial activity of silver nanoparticles, Nano Lett. 12(2012) 4271e4275.

[56] R.D. Glover, J.M. Miller, J.E. Hutchison, Generation of metal nanoparticles fromsilver and copper objects: nanoparticle dynamics on surfaces and potentialsources of nanoparticles in the environment, ACS Nano 5 (2011) 8950e8957.

[57] S.W.P. Wijnhoven, W.J.G.M. Peijnenburg, C.A. Herberts, W.I. Hagens,

A.G. Oomen, E.H.W. Heugens, et al., Nano-silver e a review of available dataand knowledge gaps in human and environmental risk assessment, Nano-toxicology 3 (2009) 109e138.

[58] X. Zhao, T. Toyooka, Y. Ibuki, Synergistic bactericidal effect by combinedexposure to Ag nanoparticles and UVA, Sci. Total Environ. 458e460 (2013)54e62.

[59] W. He, Y.-T. Zhou, W.G. Wamer, M.D. Boudreau, J.-J. Yin, Mechanisms of thepH dependent generation of hydroxyl radicals and oxygen induced by Agnanoparticles, Biomaterials 33 (2012) 7547e7555.

[60] K.C.L. Black, T.S. Sileika, J. Yi, R. Zhang, J.G. Rivera, P.B. Messersmith, Bacterialkilling by light-triggered release of silver from biomimetic metal nanorods,Small 10 (2014) 169e178.

[61] J. Xie, Y. Zhang, C. Yan, L. Song, S. Wen, F. Zang, et al., High-performancePEGylated MneZn ferrite nanocrystals as a passive-targeted agent formagnetically induced cancer theranostics, Biomaterials 35 (2014) 9126e9136.