lable at ScienceDirect

Biomaterials 32 (2011) 5880e5888

Contents lists avai

Biomaterials

journal homepage: www.elsevier .com/locate/biomateria ls

Aggregation-enhanced fluorescence in PEGylated phospholipid nanomicellesfor in vivo imaging

Dan Wang a,b,1, Jun Qian a,1, Sailing He a,b,*, Jin Sun Park c, Kwang-Sup Lee c, Sihai Han d, Ying Mu d

aCentre for Optical and Electromagnetic Research, State Key Laboratory of Modern Optical Instrumentations, JORCEP (KTH-ZJU Joint Center of Photonics), Zhejiang University,Hangzhou 310058, PR Chinab ZJU-SCNU Joint Research Center of Photonics, South China Normal University (SCNU), 510006 Guangzhou, PR ChinacDepartment of Advanced Materials, Hannam University, Daejeon 305-811, Republic of KoreadResearch Center for Analytical Instrumentation, Institute of Cyber-Systems and Control, State Key Lab. of Industrial Control Technology, Zhejiang University (ZJU),Hangzhou 310058, PR China

a r t i c l e i n f o

Article history:Received 18 April 2011Accepted 26 April 2011Available online 20 May 2011

Keywords:Aggregation-enhanced fluorescencePhospholipid-PEG nanomicellesBright nanoprobesIn vivo imaging

* Corresponding author. Centre for Optical and Elejiang University, PR China. Department of ElectromInstitute of Technology (KTH), Sweden. Tel.: þ86 488206512.

E-mail address: sailing@kth.se (S. He).1 D. Wang and J. Qian contributed equally to this w

0142-9612/$ e see front matter � 2011 Elsevier Ltd.doi:10.1016/j.biomaterials.2011.04.080

a b s t r a c t

We report polymeric nanomicelles doped with organic fluorophores (StCN, (Z)-2,3-bis[4-(N-4-(diphe-nylamino)styryl)phenyl]-acrylonitrile), which have the property of aggregation-enhanced fluorescence.The fluorescent nanomicelles have two unique features: (1) They give much brighter fluorescenceemission than mono-fluorophores. (2) The nanomicelles with amphiphilic copolymers [e.g.,phospholipids-PEG (polyethylene glycol)] make the encapsulated fluorophores more stable in variousbio-environments and easy for further conjugation with bio-molecules. After chemical and opticalcharacterization, these fluorescent nanomicelles are utilized as efficient optical probes for in vivo sentinellymph node (SLN) mapping of mice. The StCN-encapsulated nanomicelles, as well as their bioconjugateswith arginine-glycine-aspartic acid (RGD) peptides, are used to target subcutaneously xenograftedtumors in mice, and in vivo fluorescence images demonstrate the potential to use PEGylated phospho-lipid nanomicelles with aggregation-enhanced fluorescence as bright nanoprobes for in vivo diagnosis oftumors.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

In recent years, photoluminescence based bioimaging formsamajor thrust of bio-photonics [1].Much focus has been given to thedevelopment of efficient, inexpensive, stable, and tunable exoge-nous optical agents in biological systems, such as quantumdots/rods[2e6], silica nanoparticles [7e9],metal nanoparticles [10,11], carbonnanomaterials [12,13] and up-converting nanophosphores [14e17].Although organic fluorophores exhibit remarkably high photo-luminescence quantum efficiencies, their applications in the area ofbiological imaging are still limiteddue to their intrinsic hydrophobicproperty and instability in bio-environments. To overcome theseproblems, some approaches (e.g., phospholipid nanomicellesencapsulation) have been adopted.

ctromagnetic Research, Zhe-agnetic Engineering, Royal

6 8 7908465; fax: þ86 571

ork.

All rights reserved.

Phospholipid-PEG nanomicelles can be utilized to encapsulatevarious fluorophores and drugs, and have many advantages in bio-applications: (1) They possess no cytotoxicity; (2) The preparationprocess of phospholipid-PEG nanomicelles is much simpler thanthat of other nanocarriers, such as silica nanoparticles [7e9] andgold nanoparticles [10,11]; (3) A large hydrophobic core in thenanomicelle, which arises due to the presence of long acyl chains ofphospholipids, can facilitate the loading of high concentrations ofhydrophobic molecules per micelle; (4) The long PEG chains innanomicelles can improve the long-time circulation of nano-particles in an animal body and help to avoid capture/degradationby reticuloendothelial systems (RES), and this is very important forin vivo animal experiments [18e20]. Self-assembled constructionsof phospholipid-PEG nanomicelles have beenwidely utilized in theapplications of drug delivery [21,22] and in vivo animal imaging[23] in the past several years.

However, there is still an obstacle for fluorophore doped nano-micelles. Most commonly used organic fluorophores suffer from anaggregation-induced fluorescence quenching phenomenon andtheir fluorescence emission decreases when they are highly loadedin phospholipid-PEG nanomicelles. Thus, it would be very useful if

D. Wang et al. / Biomaterials 32 (2011) 5880e5888 5881

certain fluorescent materials, which have aggregation-inducedenhanced emission properties, could be encapsulated inphospholipid-PEG nanomicelles and applied in fluorescence bio-imaging. Recently, some groups have successfully synthesized these“special” fluorescent materials and demonstrated that they couldexhibit enhanced emission when aggregated in certain solvents, orfabricated into solid films [24,25]. Furthermore, some other groupsalso began to use these fluorescent materials for various in vitro cellimaging experiments [7,26e28]. However, to the best of ourknowledge, aggregation-enhanced emission dyes have still not beenapplied in in vivo animal imaging.

In this paper, (Z)-2,3-bis[4-(N-4-(diphenylamino)styryl)phenyl]acrylonitrile (StCN), a DeAeD type stilbene derivative with cyanogroup (as fluorescence acceptor group) in the center and vinyl andtriphenylamine (as fluorescence donor group) in both ends, wassynthesized as previously reported [29]. The intermolecularorientation is interrupted by the bulky cyano group during theformation of aggregation, which prevents the parallel overlap offluorophores from becoming face-to-face oriented H-aggrega-tions, but facilitates the formation of head-to-tail orientedJ-aggregations [30,31]. Subsequently, StCN-encapsulated phos-pholipid-PEG (StCN@PEG) nanomicelles with a diameter ofw20 nm were prepared, and excellent chemical stability of thenanomicelles was demonstrated. The absorption and emissionproperties of StCN@PEGnanomicelleswith various loading densitieswere characterized with an absorbance spectrophotometer andphotoluminescence spectroscopy, and no noticeable aggregation-induced fluorescence quenching or emission peak-wavelength shiftwas observed.We then used the StCN@PEG nanomicelles for in vitrostain of HeLa cells, and their biological uptake by tumor cells wasconfirmed with fluorescence microscopy. To demonstrate thepotential of StCN@PEG nanomicelles as bright fluorescent probes forin vivo animal imaging, we first used them for sentinel lymph node(SLN) mapping, which is a key process in SLN biopsy (SLNB) forcancer staging and surgery [32]. Subsequently, we developedafluorescence imagingprotocol basedonStCN@PEGnanomicelles ascontrast probes for in vivo diagnosis of tumors of mice. As is wellknown, nanoparticles have a property to be preferentially taken upby malignant tissues (e.g., tumors) due to the “enhanced perme-ability and retention” (EPR) effect [33,34]. We demonstrated thepassive targeting of tumors using StCN@PEG nanomicelles as fluo-rescent agents in nude mice, which bear subcutaneous lung tumorxenografts. Furthermore, StCN@PEG nanomicelles bioconjugatedwith arginine-glycine-aspartic acid (RGD) peptides were used forin vivo tumor targeting. RGD peptides, which have high bindingaffinity to the avb3 integrin receptor [16,35e39], are promising newtools for imaging of tumors. The integrin avb3 is a type of cell-surfacereceptors that are overexpressed at the endothelium of growingblood vessels (vasculature) associated with tumor growth (angio-genesis) [40,41]. It plays an important role in angiogenesis and tumorcell metastasis, and is currently being evaluated as a target for newdiagnosis and therapeutic treatment of tumors in vivo [35,36]. Ourexperimental results revealed that RGD peptide-conjugatedStCN@PEG nanomicelles provided higher targeting efficiency to thesubcutaneous lung tumor xenografts of mice than StCN@PEGnanomicelles without RGD peptides.

2. Experimental section

2.1. Materials and instruments

(Z)-2,3-bis[4-(N-4-(diphenylamino)styryl)phenyl]-acrylonitrile(StCN) was synthesized by the Department of Advanced Materials,Hannam University, Daejeon, Korea. The detailed information,including preparation and spectroscopic characterizations, were

described in Ref. [29]. 1,2-distearoyl-sn-glycero-3-phosphoe-thanolamine-N-[methoxy(polyethylene glycol)-5000] (mPEG-DSPE-5000) and 1,2-distearoyl-sn-glycero-3-phosphoethanol-amine-N-[maleimide[poly(ethylene glycol)] (DSPE-PEG-Maleimi-de-3400) were purchased from Creative PEGWorks, Inc.Chloroform, 3,30-diethylthiadicarbocyanine iodide (DTDC), NileRed, hydrochloric acid and sodium hydroxide were obtained fromthe Chemical Reagent Department of Zhejiang University. RGDpeptides (PCI-3686-PI) were purchased from Peptides Interna-tional, Inc. Cell-culture products, unless otherwise mentioned,were purchased from Gibco. All the reagents, which were notspecially pointed out, were analytical grade, and deionized (DI)water was used in all the experimental procedures.

Transmission electron microscopy (TEM) images were taken bya JEOL JEM-1230 transmission electron microscope operating at160 kV in bright-field mode. A Shimadzu 2550 UVevis scanningspectrophotometer and a HITACHI F-2500 fluorescence spectro-photometer were used to measure the absorption and photo-luminescence (PL) spectra of samples.

2.2. Preparation of StCN@PEG nanomicelles

Typically, 1.5 mL of StCN solutions in chloroform (1 mg/mL) wereinjected into 100 mLmPEG-DSPE solutions in chloroform (10mg/mL).Themixture solutionwas thenevaporatedanddriedundervacuumina rotary evaporator at 70 �C. Next, 2.5 mL of DI water was added intothe solid lipidic mass obtained, and the solution was sonicated for2 min. After that, an optically clear suspension containing StCN@PEGwas prepared. As excess mPEG-DSPE molecules were used toencapsulate StCN molecules, the small loss of StCN in the experi-mental process could be negligible. The loading density (StCN/[StCNþmPEG-DSPE] inwt%) of StCN in StCN@PEG nanomicelleswascalculatedby theweight ratioof thematerialsput into the reaction. Byvarying the added quantities of StCN and mPEG-DSPE solution,nanomicelles with different StCN loading densities (20 wt%, 40 wt%,60 wt%) were prepared by the same method (Table S1).

2.3. Preparation of RGD-conjugated StCN@PEG nanomicelles

First, maleimide-functionalized StCN@PEG nanomicelles wereprepared bymixing 1.5 mL StCN solutions (1 mg/mL in chloroform),80 mL mPEG-DSPE solutions (10 mg/mL in chloroform) and 10 mLDSPE-PEG-maleimide solutions (10mg/mL in chloroform) together.After evaporated and dried under vacuum, 2.5 mL PBS (pH ¼ 7.4,10 mM) solutions were added into the obtained lipidic film, and thesolution was sonicated for 2 min to obtain StCN@PEG-maleimidenanomicelles. The StCN@PEG-maleimide nanomicelles were thenconjugated with thiolated RGD through specific thiolemaleimidereactions. Briefly, 1 mL PBS solution of StCN@PEG-maleimidenanomicelles was mixed with 0.3 mL thiolated RGD peptide solu-tion (24 mg/mL) and incubated for 2 h in room temperature. Theresulting dispersion was further centrifugated at 12,000 rpm ina 0.2 mm membrane filter for 15 min to remove the excessunreacted RGD molecules and the pellets blocked on themembrane (mainly containing bioconjugates) were redispersed in1 mL of PBS (pH ¼ 7.4, 10 mM) and kept at 4 �C for further use.

2.4. Release kinetics and chemical stability analyses

For release kinetics studies, 1 mg StCN@PEG (60 wt% of StCNloading) nanomicelles were incubated with 1 mL Tween-20 solu-tions (1% in DI water) at 40 �C. After a certain time, the sample wasspin-filtered using microfuge membrane-filter (NANOSEP 100KOMEGA, Pall Corporation, USA) at 12,000 rpm for 15 min (spinfil-tration). The filtrated solution, which passed through the

D. Wang et al. / Biomaterials 32 (2011) 5880e58885882

membrane, was collected. Its extinction spectrum, as well as that ofthe original aqueous solution of StCN@PEG nanomicelles, wasmeasured. The ratio of the two extinction peak intensities could beused to characterize the release percentage of StCN from StCN@PEGnanomicells. For chemical stability tests, the PL spectra of 100 mgStCN@PEG nanomicelles incubated in 1 mL solutions (pH valuesfrom 4e10) of PBS and serum were monitored. The chemicalstabilities were evaluated according to the changes of PL intensity.

2.5. In vitro studies with tumor cells

HeLa cells were cultured in a Dulbecco’s minimum essentialmedia (DMEM/f12) with 10% fetal bovine serum (FBS), 1% penicillin,and 1% amphotericinB. For fluorescence imaging, cells were treatedseparatedly with nothing, 200 mL solutions of PEG (400 mg/mL inwater), 200 mL solutions of StCN@PEG nanomicelles (60wt% of StCNloading, 1 mg/mL in water), and then incubated at 37 �C with 5%CO2 for 2 h. Thereafter, all the cell samples were gently washedthrice with PBS (pH ¼ 7.4, 10 mM) and directly imaged with a fluo-rescence microscope underblue broadband light excitation.

2.6. In vivo imaging studies

All in vivo experiments were performed in compliance withZhejiang University Animal Study Committee’s requirements for

Fig. 1. Synthesis and characterization of StCN@PEG nanomicelles. (a) A schematic illustratiStCN@PEG nanomicelles. (c) UVevis absorption spectra of StCN in chloroform and StCN@PEnanomicelles in water solution. Inset: Pictures of StCN in chloroform (left) and StCN@PEG narespectively. The StCN in both vials are with the same masses of 0.2 mg.

the care and use of laboratory animals in research. 18e21 g malenude mice from the Animal Experimentation Center of ZhejiangUniversity were used for animal imaging studies. Every time beforeimaging, the mice were anaesthetized with intraperitoneal injec-tion of 0.5% pentobarbital sodium. To investigate the SLN mappingof StCN@PEG nanomicelles in mice, samples (60 wt% of StCN@PEG,1 mg/mL in 10 mM PBS, 10 mL solutions per mouse) were intrader-mally injected into the right forepaw pad of a group of nude mice.As a control experiment, the other group of nude mice wereintradermally injected with DSPE-mPEG solutions (0.4 mg/mL in10 mM PBS, 10 mL solutions per mouse) on their right forepaw pads.The in vivo fluorescence imaging of the experimental group and thecontrol group was performed immediately after sample injection,by utilizing a Maestro in vivo optical imaging system (CRI, Inc.Woburn, MA). The in vivo imaging system consists of an opticalhead, an optical coupler, a cooled scientific-grade monochromeCCD camera, and an image acquisition/analysis software. A liquidcrystal tunable filter was automatically tuned with 10 nm incre-ments from 500 nm to 750 nm while the camera captured theimage at each wavelength with a constant exposure time (500 ms).The resulting images were then used to create the unmixed imagesof the mouse with both auto-fluorescence and fluorescence signals.To set up tumor models, A549 cells (human lung cancer cell lines,3 � 106 cells in 0.3 mL 10 mM PBS, pH ¼ 7.4) were implantedsubcutaneously in the scapular region of 18e21 g male nude mice.

on for the preparation of StCN@PEG nanomicelles. (b) A representative TEM image ofG nanomicelles in water solution. (d) PL spectra of StCN in chloroform and StCN@PEGnomicelles in water (right) under visible light and ultraviolet light excitation (325 nm),

D. Wang et al. / Biomaterials 32 (2011) 5880e5888 5883

Tumor growth was monitored until a palpable tumor size wasobserved. Tumor-bearing mice were intravenously injected withStCN@PEG nanomicelles (with 60 wt% loading density, 1 mg/mL inPBS, 100 mL solutions per mouse) and RGD peptide-conjugatedStCN@PEG nanomicelles (with 60 wt% loading density, 1 mg/mLin PBS, 100 mL solutions per mouse). mPEG-DSPE solutions (0.4 mg/mL in PBS, 100 mL solutions per mouse) were also intravenouslyinjected to tumor-bearing mice, acting as control. All the mice wereimaged with the in vivo imaging system, and the imaging resultswere analyzed with the software package provided by CRI Inc.

3. Results and discussion

3.1. Synthesis and characterization of StCN@PEG nanomicelles

Hydrophobic StCN molecules were encapsulated in amphi-philic PEGylated phospholipid based on a well-establishedmethod [23]. The major steps of StCN@PEG nanomicellessynthesis and the chemical structures of StCN and mPEG-DSPE areshown in Fig. 1a. mPEG-DSPE is a type of well-established

Fig. 2. Stability studies of StCN@PEG nanomicelles. (a) Release kinetics studies of StCN@PEGof the two absorbance peak intensities of the filtrate solution containing Tween-20 nanomchanges at various time points (0e12 h), indicating that the release of StCN from StCN@PEG nof the StCN@PEG nanomicelles treated with PBS, serum, and pH 4 to 10 solutions. The PL ininset show the corresponding fluorescence images of StCN@PEG nanomicelles dispersed in vlight with a peak wavelength of 455 nm.

surfactant with stable properties in aqueous solutions, whichhas been widely used in bio-applications [18e23]. The presence ofPEG chains not only enabled the water solubility of StCN@PEGnanomicelles, but also provided a feasible bioconjugation methoddue to the existence of functional groups (e.g., COOHe, NH2e, orSHe) on its molecule end. Fig. 1b shows a representative TEMimage of the StCN@PEG nanomicelles with 60 wt% loading density,and the nanomicelles exhibited roughly spherical shapes and werewell dispersed. The StCN@PEG nanomicelles also appeareduniform in size and had an average diameter of less than 30 nm,which is small enough to minimize any disturbance of normalcellular physiology [42]. A digital camera was used to observe thesolutions of StCN in chloroform and StCN@PEG nanomicelles in DIwater under visible light and ultraviolet light (325 nm) excitationin a darkroom. The photographs were presented in the insets ofFig. 1c and d. As the StCN in both vials is of the same masses(0.2 mg), both solutions exhibited similar yellow colors undervisible light (shown in Fig. 1c). After the excitation of ultravioletlight (325 nm), StCN chloroform solution emitted orange fluo-rescence while aqueous solution of StCN@PEG nanomicelles

nanomicelles (60 wt% of StCN loading) in 1% Tween-20 suspension at 40 �C. The ratiosicelles and the original aqueous solution of StCN@PEG nanomicelles exhibit nearly noanomicelles is not severe and independent of incubation time. (b) Stability comparisontensities of mPEG-DSPE solutions in water were also measured as control. The panelsarious solutions and the control solutions. The excitation source was a blue broadband

D. Wang et al. / Biomaterials 32 (2011) 5880e58885884

emitted reddish fluorescence (shown in Fig. 1d). The two solutionswere then characterized using a UVevis scanning spectropho-tometer and a fluorescence spectrophotometer, respectively. Asboth the absorbance and PL intensities of mPEG-DSPE in waterwere almost “none” (Fig. S1), we can conclude that the absorbanceand emission of StCN@PEG nanomicelles in water were mainlycontributed from the encapsulated StCN molecules. As indicatedby the results shown in Fig. 1c and d, the absorbance spectrum ofStCN@PEG nanomicelles in DI water was similar to that of StCN inchloroform, with a maximum absorbance wavelength at 435 nm.However, compared to StCN chloroform solution, the fluorescencepeak-wavelength of the nanomicelles solution (in water) red-shifted from 554 nm to 577 nm (the excitation wavelength was435 nm). We attributed the red-shift of the fluorescence peak-wavelength to the nanomicelle-encapsulation with PEGylatedphospholipid and/or the difference of the solvents, sinceStCN@PEG nanomicelles were “dispersed” in water while StCNmolecules were “dissolved” in chloroform.

Fig. 3. Aggregation-enhanced fluorescence analysis of StCN@PEG nanomicelles. (a) UVevisdensities (20 wt%, 40 wt%, 60 wt%). (b) Normalized fluorescence intensities of hydrophobicloading densities (20 wt%, 40 wt%, 60 wt%), respectively.

3.2. Stability analyses of StCN@PEG nanomicelles

Since the stability of nanomicelles is of high importance foroptical imaging, we tested the leakage of the encapsulated StCNfrom StCN@PEG nanomicelles in Tween-20 solution. As StCN dyecan’t dissolve (or disperse) in water, if it leaked from nanomicellesand formed aggregates, the aqueous solutions would turn turbid. Inour experiments, the Tween-20 solutions containing StCN@PEGnanomicelles were always optically clear, and thus the possibilitythat aggregates were formed by leaked dye could be excluded.According to the release kinetics study, if the nanomicellesstructure is not stable, Tween-20 solution would breakphospholipid-PEG nanomicelles and extract StCN molecules fromthem, forming smaller StCN@Tween-20 nanomicelles in theaqueous solution [23]. Under high speed centrifugation, microfugemembrane-filters would allow Tween-20 nanomicelles to passthrough the membrane with the phospholipid-PEG nanomicellesarrested by the membrane. According to Beer’s Law, the optical

absorption and PL spectra of the StCN@PEG nanomicelles with various StCN loadingdyes (StCN, DTDC and Nile Red) doped in phospholipid-PEG nanomicelles with various

Fig. 4. In vitro images of HeLa cells treated with nothing (a), mPEG-DSPE (b),StCN@PEG nanomicelles (c) for 2 h at 37 �C. The scale bar is 50 mm.

D. Wang et al. / Biomaterials 32 (2011) 5880e5888 5885

absorbance of a chromophore in a transparent solvent varies line-arly with the concentration of the chromophore, if the pathlengthof light in the chromophore sample is fixed. In our experiment, thefiltrated solutions containing StCN@Tween-20 micelles werecollected and dissolved in a certain volume of DI water for absor-bance measurement by a UVevis scanning spectrophotometer.Measurements were carried out at various time points and theabsorbance intensity at 435 nm was used to calibrate the StCNconcentration. The release percentage of StCN from StCN@PEGnanomicelles was calculated by the concentration ratio of StCN inthe filtrated solution and StCN in the original solution of StCN@PEGnanomicelles. As shown in Fig. 2a, the release percentages of StCNfrom StCN@PEG nanomicelles were less than 5% after incubatedwith 1% of the Tween-20 surfactant at 40 �C during the entire timeof the experiment (even after 12 h), indicating that no severerelease of StCN from StCN@PEG nanomicelles occurred in aqueoussolutions.

Furthermore, as a kind of efficient fluorescent probes for bio-logical applications, StCN@PEG nanomicelles should be stable invarious bio-environments including a wide range of pH values.Otherwise, monodispersed nanoparticles may be transformed intoclusters composed of largenumbers of particles,which are too big tobe taken up by cells, and cannot circulate smoothly in the vessels forin vivo experimental applications. To confirm the biological andchemical stabilities of StCN@PEG nanomicelles, we systematicallystudied the fluorescence intensity changes of these nanomicellesunder different treatments (e.g., PBS, serum, and pH 4 to 10 solu-tions) for 12 h. The solutions are selected with consideration to themost basic environments in thehumanbody: PBS is a buffer solutioncommonly used in biological research; serum is used to simulate thebio-environments of in vivo experiments; pH ¼ 4 and pH ¼ 10 arealmost the limit values of pH in the humanbody. As shown in Fig. 2b,the fluorescence peak intensities of StCN@PEG nanomicelleschanged by less than10% in all experimental conditions (normalizedby the fluorescence peak intensity at pH ¼ 7), indicating that theStCN@PEG nanomicelles was chemically stable in those solutions,which is very positive for various bio-applications.

3.3. Aggregation-enhanced fluorescence properties of StCN@PEGnanomicelles

Nanomicelles solutions with various StCN loading densities(20 wt%, 40 wt%, 60 wt%) were optically characterized in order toinvestigate the aggregation-induced properties of StCN@PEGnanomicelles. For an accurate quantitative comparison of photo-luminescence emission, absorption intensities of the excitationwavelength for all the samples were kept the same, by utilizingdifferent amounts of nanomicelles to encapsulate the same amountof StCN molecules (Table S2). The total PL intensity increasedsignificantly without any distinct wavelength shift in emission peak(Fig. 3a), as the StCN loading density increased. Furthermore, twocommon fluorophores (DTTC and Nile Red, with absorbance and PLspectra shown in Fig. S2) were selected, and the PL intensities ofDTDC and Nile Red encapsulated phospholipid-PEG nanomicelleswith various loading densities weremonitored for comparisonwiththose of StCN@PEG nanomicelles. As shown in Fig. 3b, the PLintensities (normalized by the PL intensity in the condition ofloading density ¼ 20 wt%) of DTDC/Nile Red doped nanomicellesdecreased significantly as the loading density of DTDC/Nile Red innanomicelles increased, which was due to the aggregation-inducedfluorescence quenching. However, the behavior of StCN dopednanomicelles was quite opposite to those of the two common dye-doped nanomicelles. It can be concluded that StCN@PEG nano-micelles exhibit an aggregation-enhanced fluorescence emissioneffect, due to the unique chemical structure of StCN molecules.

3.4. StCN@PEG nanomicelles for in vitro cell imaging

Due to the properties of ultra small size, highly monodispersity,excellent chemical stability and aggregation-enhancedfluorescenceemission, StCN@PEG nanomicelles were utilized for in vitro opticalbioimaging of cells. Fig. 4c shows the in vitro images of HeLa cellstreated with StCN@PEG nanomicelles. The bright-field imagesshowed that the morphologies of all the cells (treated withStCN@PEG nanomicelles) kept very well, and they were viable afterthe sample treatment, indicating StCN@PEG nanomicelles did notcause any toxicity to the cells. The fluorescence images illustratedthat the fluorescence of StCN could be clearly observed from theHeLa cells, indicating that the StCN@PEG nanomicelles were effec-tively taken up by cells. However, control cells A (without anytreatment, Fig. 4a) andcontrol cellsB (treatedonlywithmPEG-DSPE,Fig. 4b) showed no fluorescence.

3.5. StCN@PEG nanomicelles for in vivo SLN mapping and tumortargeting of mice

To assess the potential of StCN@PEG nanomicelles for bio-imaging in living animals, we first investigatedwhether they can beused for fluorescence mapping of SLN (sentinel lymph node) ina mouse model. SLN is the first group of lymph nodes receivingmetastatic cancer cells by direct lymphatic drainage from a primarytumor. Accurate identification and biopsy of SLN can enable clini-cians to focus on certain lymph nodes and perform more detailedtracking of cancer cell diffusion. SLN imaging utilizing fluorophoresas labeling agents has become a research areawith intense interest.The aggregation-enhanced fluorescence emission of StCN, as wellas the excellent chemical stability of StCN@PEG nanomicelles,motivate us to investigate their applicability in in vivo SLNmapping.In our experiment, StCN@PEG nanomicelles were intradermallyinjected into the right forepaw pads of nudemice. AMaestro opticalimaging system was then used to record the in vivo fluorescenceimaging immediately after sample injection. The diffusion andaccumulation process of nanomicelles at the SLN over time wasshown in Fig. 5b. After injected, nanomicelles diffused rapidly from

Fig. 5. Pseudo-color fluorescence images of mice with mPEG-DSPE (a) and StCN@PEG nanomicelles (b) injected into its right paw at various time points (5, 20, 40 and 60 min).Arrows indicate the SLN sites. (c) Fluorescence spectra obtained from the SLN (red) and skin (green) of the mice. (d) Variations of fluorescence peak intensities in the SLN sites of themice, which were injected with StCN@PEG nanomicelles. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

D. Wang et al. / Biomaterials 32 (2011) 5880e58885886

the injection site into the lymphatics. 5 min later, fluorescencesignals were detected at an axillary node, giving a clear SLNmapping of the mouse (shown in Fig. 5a). In contrast, mice injectedwith mPEG-DSPE solutions (Fig. 5a) showed no signal in the area ofSLN. Fluorescence spectra acquired from the SLN and the skin of themouse were shown in Fig. 5c. As time went by, nanomicellesgradually migrated from SLN, and the fluorescence signal intensityat the SLN decreased (as shown in Fig. 5d). Furthermore, a mousewas put down and its SLN was taken out and dissected, thereafterthe in vivo SLN mapping of StCN@PEG nanomicelles was confirmedaccording to ex vivo imaging result (Fig. S3).

3.6. StCN@PEG nanomicelles for tumor targeting of mice

To verify the applicability of StCN@PEG nanomicelles for in vivotumor targeting, we injected them intravenously into nude micebearing subcutaneous lung tumor xenografts. PEG molecules canimprove the long-time circulation of nanomicelles in the animalbody, and due to the EPR effect in the tumor tissues, nanomicellescould be preferentially taken up by tumors. Hence, we couldanticipate that StCN@PEG nanomicelles could passively targettumors of mice. Furthermore, in order to obtain high efficiency oftumor targeting by using nanomicelles, we conjugated StCN@PEG

nanomicelles with a kind of triplet peptide called RGD (arginine-glycine-aspartic acid peptides). Triplet peptide RGD can specificallytarget (bind to) avb3 integrins overexpressed on the tumor endo-thelium, which can potentially play a critical role for the diagnosisof developing tumors in vivo via noninvasive optical imaging. In ourexperiment, mice bearing subcutaneous lung tumor xenograftswere intravenously injected separately with mPEG-DSPE (ascontrol), StCN@PEG nanomicelles and RGD peptide-conjugatedStCN@PEG nanomicelles. In vivo fluorescence imaging of tumor-bearing nude mice was carried out at various time points post-injection and spectrally unmixed using the Maestro imagingsoftware. Fig. 6 shows the representative whole body in vivo opticalimaging results of mice injected with StCN@PEG nanomicelles (topof the images) and RGD peptide-conjugated StCN@PEG nano-micelles (bottom of the images). Spectral signatures from the tumorsites and the auto-fluorescence of skin sites were also respectivelypresented in Fig. 6. It is obvious that the fluorescence spectra ofsignals were consistent with the fluorescence spectra of StCN and itcould easily be differentiated from the auto-fluorescence of skinsites. For control mice, no fluorescence contrast could be observedfrom tumors and surrounding skin after injection of mPEG-DSPE(Fig. S4). For the mice injected with StCN@PEG nanomicelles,there were fluorescence signals from the tumors at 48 h post-

Fig. 6. In vivo imaging of mice bearing subcutaneous lung tumor xenografts, injected with StCN@PEG nanomicelles (top) and RGD peptides-conjugated StCN@PEG nanomicelles(bottom). The concentrations of two StCN@PEG nanomicelles samples were the same (with 60 wt% loading density, 1 mg/mL in PBS, 100 mL solutions per mouse). Spectral profiles in(dei) were used to unmix images.

D. Wang et al. / Biomaterials 32 (2011) 5880e5888 5887

injection (Fig. 6f and l top), while the images taken 1 h and 24 hpost-injection only showed bright signals at the injection sites oftails (Fig. 6d, e, j, k, top). These results showed that the accumula-tion of the StCN@PEG nanomicelles in the tumor sites needs longtime (e.g., 48 h post-injection) to take place, which could beattributed to the slow process of the EPR effect [33,34]. For the miceinjected with RGD peptide-conjugated StCN@PEG nanomicelles,there were no signals in the tumor sites 1 h post-injection (Fig. 6dand j bottom), but intense fluorescence signals could be observed inthe tumor sites at 24 h post-injection (Fig. 6e and k bottom), as wellas at 48 h post-injection (Fig. 6f and l bottom). The accumulation ofthe RGD peptide-conjugated StCN@PEG nanomicelles in the tumorsites was much faster than StCN@PEG nanomicelles, and onepossible explanation is: the high binding affinity of RGD peptides tothe avb3 integrin receptor contributed more than the EPR effect totumor targeting of nanomicelles, making the RGD peptide-conjugated StCN@PEG nanomicelles exhibit higher efficiency oftargeting to the subcutaneous lung tumor xenografts in mice. Ourexperiment results have suggested that bioconjugated/non-bioconjugated StCN@PEG nanomicelles of aggregation-enhancedfluorescence can be used as bright nanoprobes for potentialapplications in in vivo tumor targeting and diagnoses.

4. Conclusions

We have synthesized phospholipid-PEG nanomicelles (<30 nm)loaded with (Z)-2,3-bis[4-(N-4-(diphenylamino)styryl)phenyl]-acrylonitrile (StCN). The as-prepared StCN@PEG nanomicelles were

characterized and exhibited ultra small size, high monodispersity,excellent chemical stability and aggregation-enhanced fluores-cence emission. In vitro fluorescence microscopy and in vivo SLNmapping using StCN@PEG nanomicelles (as bright luminescentlabels) were carried out. Moreover, we have demonstrated thatStCN@PEG nanomicelles (with and without RGD-conjugated) couldbe used as an excellent tool for tumor targeting in live animals.

Acknowledgments

This work is partially supported by the National Natural ScienceFoundation of China (No. 61008052 and 60688401), the SwedishFoundation for Strategic Research (SSF), AOARD, the FundamentalResearch Funds for the Central Universities and the China Post-doctoral Science Foundation (No. 20090461394). We are alsograteful toNational Basic Research Programof China 2007CB714503and Innovation Method Fund of China 2008IM040800.

Appendix. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.biomaterials.2011.04.080.

References

[1] Prasad PN. Introduction to biophotonics. New York: Wiley-Interscience; 2004.[2] Bruchez M, Moronne M, Gin P, Weiss S, Alivisatos AP. Semiconductor nano-

crystals as fluorescent biological labels. Science 1998;281(5385):2013e6.

D. Wang et al. / Biomaterials 32 (2011) 5880e58885888

[3] Chan WCW, Nie SM. Quantum dot bioconjugates for ultrasensitive non-isotopic detection. Science 1998;281(5385):2016e8.

[4] Jamieson T, Bakhshi R, Petrova D, Pocock R, Imani M, Seifalian AM. Biologicalapplications of quantum dots. Biomaterials 2007;28(31):4717e32.

[5] Zhang P, Liu WG. ZnO QD@PMAA-co-PDMAEMA nonviral vector for plasmidDNA delivery and bioimaging. Biomaterials 2010;31(11):3087e94.

[6] Yong KT, Qian J, Roy I, Lee HH, Bergey EJ, Tramposch KM, et al. Quantum rodbioconjugates as targeted probes for confocal and two-photon fluorescenceimaging of cancer cells. Nano Lett 2007;7(3):761e5.

[7] Kim S, Ohulchanskyy TY, Pudavar HE, Pandey RK, Prasad PN. Organicallymodified silica nanoparticles co-encapsulating photosensitizing drug andaggregation-enhanced two-photon absorbing fluorescent dye aggregatesfor two-photon photodynamic therapy. J Am Chem Soc 2007;129(9):2669e75.

[8] He XX, Nie HL, Wang KM, Tan WH, Wu X, Zhang PF. In vivo study of bio-distribution and urinary excretion of surface-modified silica nanoparticles.Anal Chem 2008;80(24):9597e603.

[9] Yang PP, Quan ZW, Hou ZY, Li CX, Kang XJ, Cheng ZY, et al. A magnetic,luminescent and mesoporous coreeshell structured composite material asdrug carrier. Biomaterials 2009;30(27):4786e95.

[10] Zhan QQ, Qian J, Li X, He SL. A study of mesoporous silica-encapsulated goldnanorods as enhanced light scattering probes for cancer cell imaging. Nano-technology 2010;21(5):055704e12.

[11] Qian J, Jiang L, Cai FH, Wang D, He SL. Fluorescence-surface enhanced Ramanscattering co-functionalized gold nanorods as near-infrared probes for purelyoptical in vivo imaging. Biomaterials 2011;32(6):1602e10.

[12] Liu Z, Robinson JT, Sun XM, Dai HJ. PEGylated nanographene oxide for deliveryof water-insoluble cancer drugs. J Am Chem Soc 2008;130(33):10876e7.

[13] Yang K, Zhang S, Zhang GX, Sun XM, Lee ST, Liu Z. Graphene in mice: ultrahighin vivo tumor uptake and efficient photothermal therapy. Nano Lett 2010;10(9):3318e23.

[14] Nyk M, Kumar R, Ohulchanskyy TY, Bergey EJ, Prasad PN. High contrast in vitroand in vivo photoluminescence bioimaging using near infrared to nearinfrared up-conversion in TM3þ and Yb3þ doped fluoride nanophosphors.Nano Lett 2008;8(11):3834e8.

[15] Xiong LQ, Chen ZG, Yu MX, Li FY, Liu C, Huang CH. Synthesis, characterization,and in vivo targeted imaging of amine-functionalized rare-earth up-converting nanophosphors. Biomaterials 2009;30(29):5592e600.

[16] Xiong LQ, Yang TS, Yang Y, Xu CJ, Li FY. Long-term in vivo biodistributionimaging and toxicity of polyacrylic acid-coated upconversion nanophosphors.Biomaterials 2010;31(27):7078e85.

[17] Zhou J, Sun Y, Du XX, Xiong LQ, Hu H, Li FY. Dual-modality in vivo imagingusing rare-earth nanocrystals with near-infrared to near-infrared (NIR-to-NIR) upconversion luminescence and magnetic resonance properties.Biomaterials 2010;31(12):3287e95.

[18] Dubertret B, Skourides P, Norris DJ, Noireaux V, Brivanlou AH, Libchaber A. InVivo imaging of quantum dots encapsulated in phospholipid micelles. Science2002;298(5599):1759e62.

[19] Ashok B, Arleth L, Hjelm RP, Rubinstein I, Önyuksel H. In vitro characterizationof PEGylated phospholipid micelles for improved drug solubilization: effectsof PEG chain length and PC incorporation. J Pharm Sci 2004;93(10):2476e87.

[20] Joralemon MJ, McRae S, Emrick T. PEGylated polymers for medicine: fromconjugation to self-assembled systems. Chem Commun 2010;46(9):1377e93.

[21] Perkins WR, Ahmad I, Li X, Hirsh DJ, Masters GR, Fecko CJ, et al. Novel ther-apeutic nano-particles (lipocores): trapping poorly water soluble compounds.Int J Pharm 2000;200(1):27e39.

[22] Sandström MC, Johansson E, Edwards K. Structure of mixed micelles formedin PEG-lipid/lipid dispersions. Langmuir 2007;23(8):4192e8.

[23] Kumar R, Ohulchanskyy TY, Roy I, Gupta SK, Borek C, ThompsonME, et al. Near-infrared phosphorescent polymeric nanomicelles: efficient optical probes fortumor imaging and detection. ACS Appl Mater Interfaces 2009;1(7):1474e81.

[24] Tong H, Dong YQ, Haußler M, Hong YN, Lam JWY, Sung HHY, et al. Molecularpacking and aggregation-induced emission of 4-dicyanomethylene-2,6-dis-tyryl-4H-pyran derivatives. Chem Phys Lett 2006;428(4e6):326e30.

[25] Dong YQ, Lam JWY, Qin AJ, Sun JX, Liu JZ, Li Z, et al. Aggregation-induced andcrystallization-enhanced emissions of 1,2-diphenyl-3,4-bis(diphenyl-methylene)-1-cyclobutene. Chem Commun 2007;31:3255e7.

[26] Kim S, Zheng Q, He GS, Bharali DJ, Pudavar HE, Baev A, et al. Aggregation-enhanced fluorescence and two-photon absorption in nanoaggregates ofa 9,10-Bis[40-(400-aminostyryl)styryl]anthracene derivative. Adv Funct Mater2006;16(18):2317e23.

[27] Wu WC, Chen CY, Tian YQ, Jang SH, Hong YN, Liu Y, et al. Enhancement ofaggregation-induced emission in dye-encapsulating polymeric micelles forbioimaging. Adv Funct Mater 2010;20(9):1413e23.

[28] Faisal M, Hong YN, Liu JZ, Yu Y, Lam JWY, Qin AJ, et al. Fabrication of fluo-rescent silica nanoparticles hybridized with AIE luminogens and explorationof their applications as nanobiosensors in intracellular imaging. Chem Eur J2010;16(14):4266e72.

[29] Park JS, Kim RH, Cho NS, Shim HK, Lee KS. Enhanced emission and two-photonabsorption cross-section by nanoaggregation of a cyano-substituted stilbenederivative. J Nanosci Nanotechnol 2008;8(9):4793e6.

[30] Lippert E, Rettig W, Bona�ci�c-Koutecký V, Heisel FF, Miehé JA. Photophysics ofinternal twisting. Adv Chem Phys 1987;68:1e173.

[31] An BK, Kwon SK, Jung SD, Park SY. Enhanced emission and its switching influorescent organic nanoparticles. J Am Chem Soc 2002;124(48):14410e5.

[32] Jakub JW, Pendas S, Reintgen DS. Current status of sentinel lymph nodemapping and biopsy: facts and controversies. Oncologist 2003;8(1):59e68.

[33] Maeda H, Wu J, Sawa T, Matsumura Y, Hori KJ. Tumor vascular permeabilityand the EPR effect in macromolecular therapeutics: a review. J Control Release2000;65(1e2):271e84.

[34] Lukyanov AN, Gao ZG, Mazzola L, Torchilin VP. Polyethylene glycol-diacyllipidmicelles demonstrate increased accumulation in subcutaneous tumors inmice. Pharm Res 2002;19(10):1424e9.

[35] HaubnerR,WeberWA, BeerAJ, Vabuliene E, ReimD, SarbiaM, et al. Noninvasivevisualization of the activated avb3 integrin in cancer Patients by positronemission tomography and [18F]Galacto-RGD. PLoS Med 2005;2(3):244e52.

[36] Chen X, Conti PS, Moats RA. In vivo near-infrared fluorescence imaging ofintegrin avb3 in brain tumor xenografts. Cancer Res 2004;64(21):8009e14.

[37] Cai WB, Shin DW, Chen K, Gheysens O, Cao QZ, Wang SX, et al. Peptide-Labeled near-infrared quantum dots for imaging tumor vasculature in livingsubjects. Nano Lett 2006;6(4):669e76.

[38] Kalininaa S, Gliemannb H, López-Garcíac M, Petershansb A, Auernheimerc J,Schimmeld T, et al. Isothiocyanate-functionalized RGD peptides for tailoringcell-adhesive surface patterns. Biomaterials 2008;29(20):3004e13.

[39] Zako T, Nagata H, Terada N, Utsumi A, Sakono M, Yohda M, et al. Cyclic RGDpeptide-labeled upconversion nanophosphors for tumor cell-targetedimaging. Biochem Biophys Res Commun 2009;381(1):54e8.

[40] Cai W, Gambhir SS, Chen X. Multimodality tumor imaging targeting integrinavb3. BioTechniques 2005;39(6 suppl):S14e25.

[41] Hu R, Yong KT, Roy I, Ding H, Law WC, Cai HX, et al. Functionalized near-infrared quantum dots for in vivo tumor vasculature imaging. Nanotech-nology 2010;21(14):145105.

[42] Kim S, Pudavar HE, Bonoiu A. Aggregation-enhanced fluorescence in organi-cally modified silica nanoparticles: a novel approach toward high-signal-output nanoprobes for two-photon fluorescence bioimaging. Adv Mater2007;19(22):3791e5.